Sie sind auf Seite 1von 430

Die Reine Rechtslehre

auf dem Prüfstand


Hans Kelsen’s Pure Theory of Law:
Conceptions and Misconceptions
Tagung der Deutschen Sektion der Internationalen Vereinigung
für Rechts- und Sozialphilosophie vom 27.–29. September 2018
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

in Freiburg im Breisgau

Herausgegeben von Matthias Jestaedt, Ralf Poscher


und Jörg Kammerhofer

Archiv für Rechts- und Sozialphilosophie Beiheft 163

Franz Steiner Verlag

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


archiv für rechts- und sozialphilosophie
archives for philosophy of law
and social philosophy
archives de philosophie du droit
et de philosophie sociale
archivo de filosofía jurídica y social

Herausgegeben von der Internationalen Vereinigung für Rechts-


und Sozialphilosophie (IVR)
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Redaktion: Dr. Annette Brockmöller, LL. M.

beiheft 163

Franz Steiner Verlag


DIE REINE RECHTSLEHRE
AUF DEM PRÜFSTAND
HANS KELSEN'S PURE THEORY
OF LAW: CONCEPTIONS AND
MISCONCEPTIONS
Tagung der Deutschen Sektion der Internationalen
Vereinigung für Rechts- und Sozialphilosophie
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

vom 27.–29. September 2018 in Freiburg im Breisgau

Herausgegeben von Matthias Jestaedt, Ralf Poscher


und Jörg Kammerhofer

Franz Steiner Verlag

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Umschlagbild: Justitia, Landgericht Ulm


Quelle: shutterstock.com / Georg_89

Bibliografische Information der Deutschen Nationalbibliothek:


Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen
Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über
<http://dnb.d-nb.de> abrufbar.

Dieses Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt.


Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes
ist unzulässig und strafbar.
© Franz Steiner Verlag, Stuttgart 2020
Layout, Satz und Herstellung durch den Verlag
Druck: Beltz Grafische Betriebe, Bad Langensalza
Gedruckt auf säurefreiem, alterungsbeständigem Papier.
Printed in Germany.
ISBN 978-3-515-12568-0 (Print)
ISBN 978-3-515-12579-6 (E-Book)

Franz Steiner Verlag


Inhaltsverzeichnis / Table of Contents

Vorwort. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

I. Die Grundlagen der Reinen Rechtslehre / The Foundations of the Pure


Theory of Law

JOHN GARDNER †
Normativity (in Kelsen and otherwise) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

ROBERT ALEXY
Hans Kelsen’s Legal Theory in the System of Non-Positivism . . . . . . . . . . . . . . . . . . . . 31

ALEXANDER SOMEK
The Demystification Impasse
Legal Positivism Divided Against Itself 45

CHRISTOPH KLETZER
Decentralisation and the Limits of Jurisprudence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

CARSTEN HEIDEMANN
Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen* . . . . . . . . . . . . . . . . . . . . . . . . 81

MAXIMILIAN KIENER
Fictionalising Kelsen’s Pure Theory of Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

II. Rechtstheorie / The Mechanics of Law

MATHIEU CARPENTIER
Kelsen on Derogation and Normative Conflicts
An Essay in Critical Reconstruction 125

Franz Steiner Verlag


6 Inhaltsverzeichnis / Table of Contents

MATHEUS PELEGRINO DA SILVA


Suspension als eine Art von Derogation oder als weitere Funktion
der Rechtsnorm? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

THOMAS HOCHMANN
Welche Rolle für die Rechtswissenschaft?
Zu einer Debatte innerhalb der Wiener rechtstheoretischen Schule 161

RODRIGO GARCIA CADORE


Alternativermächtigung vs. Fehlerkalkül
Wie geht das Recht mit Fehlern um? 177

BENEDIKT PIRKER
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Kelsen meets Cognitive Science


The Pure Theory of Law, Interpretation, and Modern Cognitive Pragmatics 203

III. Die Anwendung der Reinen Rechtslehre auf Rechtsordnungen /


Applying the Pure Theory of Law to Legal Orders

LENA FOLJANTY
Hans Kelsen, das Privatrecht und die Demokratie . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227

FERNANDO MENEZES
Hans Kelsen’s Ideas on Public and Private Law and their Applicability
to a Critical Analysis of the Theory of Administrative Law . . . . . . . . . . . . . . . . . . . . . . 245

D. A. JEREMY TELMAN
Problems of Translation and Interpretation
A Kelsenian Commentary on Positivist Originalism 257

TOMASZ WIDŁAK
Kelsen’s Monism and the Structure of Global Law
On the Relevance of a Kelsenian Account for the Polycentric International Law 275

ANNE KÜHLER
Constitutional Pluralism and the Problem of the Hierarchy of Norms beyond
National Law . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291

Franz Steiner Verlag


Inhaltsverzeichnis / Table of Contents 7

IV. Ideengeschichtlicher Kontext / Kelsen in Context

FREDERICK SCHAUER
Fuller and Kelsen – Fuller on Kelsen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

URSZULA KOSIELI Ń SKA-GRABOWSKA


The Impact of Hans Kelsen’s Pure Theory on Alf Ross’s Ideas
A Case of Scientific Plagiarism? 319

MARIO G. LOSANO
Bobbios „Bekehrung“ zur Reinen Rechtslehre
Die Turiner Schule und die Rezeption Hans Kelsens in Italien 339
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

STANLEY L. PAULSON
Did Walter Jellinek Invent Hans Kelsen’s Basic Norm? . . . . . . . . . . . . . . . . . . . . . . . . . . 351

MARIJAN PAVČNIK
Rechtliche Natur des Staates nach Leonid Pitamic
The Legal Nature of the State according to Leonid Pitamic 361

ULRICH WAGRANDL
Kelsen was no Relativist
Reading Hans Kelsen in the Light of Isaiah Berlin’s Value Pluralism 373

FEDERICO LIJOI
The Democratic Value of Law
Hans Kelsen on the Theory and Praxis of Relativism 393

REUT YAEL PAZ / MAXIMILIAN WAGNER


Salvaging Scientific Socialism?
Hans Kelsen’s Attachments and Detachments to Austro-Marxism 413

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


Vorwort

MATHIAS JESTAEDT / RALF POSCHER /


JÖRG KAMMERHOFER
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Die Reine Rechtslehre Hans Kelsens stellt einen spezifischen Zugang zu Recht und
Normativität dar; sie liefert eine eigene Grammatik und eine besondere Nomenklatur,
mit deren Hilfe man gewisse rechtswissenschaftliche Begriffe, Konzepte und Proble-
me als wesentlich und andere als irreführend für die Hauptaufgaben der (rechts)wis-
senschaftlichen Beschäftigung mit dem Recht ausweisen kann. Grundlegender noch
erhebt sie den Anspruch, eine Rechtswissenschaftstheorie zu sein, d. h. eine Theorie
darüber, was Rechtswissenschaft ist bzw. sein kann oder sein sollte. Dieser Ansatz zu
Bedeutung und Aufgabe einer selbständigen Rechtswissenschaft bleibt freilich oft im-
plizit und wird in der einschlägigen Literatur eher selten adressiert.
Auch nach über einhundert Jahren weckt die Reine Rechtslehre die Aufmerksam-
keit der Rechtswissenschaft – und das in ebenso heftigem Zuspruch wie Widerspruch.
Dabei treten Missverständnisse und Ausblendungen zutage  – bei Kelsens Lehren,
bei den Versuchen, sich dessen Lehren anzueignen, und bei den Versuchen, diese zu
widerlegen. Die internationale Community setzt sich nach wie vor mit bestimmten
Schriften und Konzepten Hans Kelsens recht intensiv auseinander. Darüber geraten
andere Wesenszüge und Argumentationslinien der Reinen Rechtslehre gar nicht erst
ins Blickfeld. Und während Kelsens auch auf Englisch verfügbare Werke international
vergleichsweise intensiv diskutiert werden, gilt Gleiches nicht für die Werke anderer
wegbereitender Mitglieder der „Wiener rechtstheoretischen Schule“. Just die Debatte
um die Reine Rechtslehre in und zwischen den unterschiedlichen rechtswissenschaft-
lichen Kulturen und Traditionen ist besonders spannend.
Es war also wert, sich im Rahmen der Deutschen Sektion der IVR mit Kelsen und
der Reinen Rechtslehre zu beschäftigen und den Versuch zu unternehmen, in einem
offenen Diskurs die rechtskulturellen, sprachlichen und disziplinären Disjunktionen
kenntlich zu machen und möglichst zu überbrücken, um eine freimütige internationa-
le und intradisziplinäre Debatte um Hans Kelsen und die Reine Rechtslehre zu eröff-
nen. Mit der Tagung der Deutschen Sektion der IVR am 27.–29.  September 2018 an

Franz Steiner Verlag


10 Vorwort

der Albert-Ludwigs-Universität Freiburg i. Br. zum Thema „Die Reine Rechtslehre auf
dem Prüfstand / Hans Kelsen’s Pure Theory of Law: Conceptions and Misconceptions“
wurde also nichts weniger gewagt, als der heutigen Relevanz der Thesen der Reinen
Rechtslehre in einem international und intradisziplinär reflektierten Rahmen nachzu-
gehen.
Einer der Gründe für diese besondere Spannungslage eröffnet sich bereits bei ein-
er bloß kursorischen Kontrastierung von Leben und Werk von Kelsen mit Herbert
Hart. Harts Geburtsort Harrogate in North Yorkshire und sein Studien-, Arbeits- und
Sterbeort Oxford1 liegen keine 250 km Luftlinie auseinander, das ist in etwa die Entfer-
nung vom Tagungsort Freiburg nach Frankfurt. Seine akademische Karriere hat sich
exklusiv an der Universität Oxford abgespielt; er hat sich und seine Lehren niemals in
anderen Sprachen, gar in anderen Kulturkreisen, verständlich machen müssen. Vita
und Oeuvre sind also britisch, eigentlich rein englisch, und verlassen diesen Orbit
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

nicht. Auch ist die Rezeptionsgeschichte des Hartschen Werkes ohne den Aufstieg des
Englischen zur wissenschaftlichen lingua franca und der damit einhergehenden Pers-
pektivenverschiebung nicht vollends zu erklären.
Bei Kelsen sieht das völlig anders aus; ihm widerfährt ein Schicksal, welches das
20. Jahrhundert für ideologiekritische Protagonisten der wissenschaftlichen Moderne
in Deutschland und Österreich vielfach bereitgehalten hat – das der „Vertriebene[n]
Vernunft“.2 Kelsen sieht sich im Österreich der 1920er und im Deutschland der frü-
hen 1930er Jahre immer wieder angefochten, angefeindet, unerwünscht und schließ-
lich vertrieben. Von Wien nach Köln, nach Genf, Prag, Harvard und schließlich nach
Berkeley muss Kelsen seinen beruflich-wissenschaftlichen Standort wechseln und er
und seine Familie auch den sozialen Neuanfang wagen. Sein Geburtsort Prag ist vom
Sterbeort Orinda (bei Berkeley) denn auch nicht wie bei Hart 250 km, sondern rund
9.400 km Luftlinie auseinander. Er findet „des Wandermüden letzte Ruhestätte“ am
sprichwörtlich anderen Ende der Welt.3
Kelsen lehrt und publiziert zunächst auf Deutsch, sodann auf Französisch, schließ-
lich auf Englisch. Die fast 18.000 Seiten von Originalarbeiten Kelsens sind zu mehr als
50 % auf Deutsch erschienen, rund 6.000 Seiten liegen auf Englisch vor, einer Sprache,
die Kelsen im Alter von knapp 60 Jahren als seine Alltags- und Berufssprache zu erler-
nen gezwungen war und die er bis zuletzt nicht mit der Leichtigkeit und Selbstverständ-
lichkeit, Virtuosität und Nuanciertheit seiner Muttersprache beherrschte. Mit gewisser

1 Vgl. Nicola Lacey, A Life of H. L. A. Hart. The Nightmare and the Noble Dream, Oxford 2004.
2 Friedrich Stadler (Hrsg.), Vertriebene Vernunft I. Emigration und Exil österreichischer Wissenschaft
1930–1940, Wien 1987; Friedrich Stadler (Hrsg.), Vertriebene Vernunft II. Emigration und Exil österreichi-
scher Wissenschaft 1930–1940, Wien 1988.
3 Die nach wie vor maßgebliche Biographie Kelsens ist: Rudolf Aladár Métall, Hand Kelsen. Leben und
Werk, Wien 1969; siehe auch Hans Kelsen, Selbstdarstellung (1927), in: HKW 1, S. 19–27; Hans Kelsen, Auto-
biographie (1947), in: HKW 1, S. 29–91. Thomas Olechowski, Geschäftsführer des Hans Kelsen-Instituts,
Wien, wird in Kürze eine umfassende Biographie Kelsens im Verlag Mohr Siebeck vorlegen.

Franz Steiner Verlag


Vorwort 11

Zuspitzung könnte man sagen, dass der Nationalsozialismus Kelsen nicht nur seine Hei-
mat, sondern auch seine Sprache geraubt hat, ihn zu nicht unwichtigen Teilen sprach-
los gemacht und im Diskurs-Niemandsland ausgesetzt hat. In den Vereinigten Staaten,
dem Land, das ihm großzügig eine neue Heimat bot, hatte er in des Wortes doppelter
Bedeutung Mühe, sich verständlich zu machen.4 Kelsens Vita und Œuvre bezeugen
eine Odyssee, die sozusagen die Begleitgeschichte und den Resonanzboden abgibt für
die Reine Rechtslehre und die ihre Rezeption und Diskussion bis heute beeinflusst.
Man möchte angesichts dieses Befundes meinen, dass die Rezeptionsgeschichte
der Reinen Rechtslehre in Deutschland, wo im Gegensatz zum US-amerikanischen
Exil doch der sprachlich unmittelbarste und mithin einfachste Zugang zur Reinen
Rechtslehre herrscht,5 besonders günstig verlaufen ist. Doch mitnichten: Hierzulan-
de begreift man jetzt erst in Ansätzen, welcher Schatz seit vielen Jahrzehnten vor den
eigenen Füßen schlummert. Die Reine Rechtslehre wird in Deutschland dramatisch
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

anders rezipiert als im Gros der übrigen Länder. Kelsen ist auf dem Höhepunkt seines
wissenschaftlichen Schaffens und seiner Anerkennung als public intellectual, als der
sogenannte Weimarer Methoden- und Richtungsstreit ausbricht und er neben Rudolf
Smend, Carl Schmitt und Hermann Heller zum Weimarer Quartett zählt. Doch schon
bald, nämlich bereits im April 1933, wird er hors du discours gesetzt. Seine Wiener
Schule wird, ohne dass sie sich davon je erholen sollte, zerschlagen. Bis 1945 verhängt
die regimehörige Rechtswissenschaft eine damnatio memoriae über den „Juden Kel-
sen“ und dessen Werk.
Die mit der bedingungslosen Kapitulation NS-Deutschlands schlagende Stunde
Null ist aber nicht die Stunde, in der Kelsen rehabilitiert wird und seine Ideen mit
offenen Armen Aufnahme finden. Die Naturrechtsrenaissance, flankiert von der Rad-
bruchschen Positivismuslegende,6 erübrigt, so ist man sich einig, eine Auseinander-
setzung mit Kelsen und seinem Œuvre. Bis weit in die 1960er Jahre bleibt Kelsen hors
du discours. Sowohl die Erstauflage der „Reinen Rechtslehre“, die Kelsen 1934 in Genf
vollendet, als auch die Zweitauflage 1960, gehören streng genommen zur Gattung der
Exil-Literatur, sind also Außenseiter-Schriften im deutschsprachigen Wissenschafts-
diskurs. Ab Mitte der 1980er Jahre scheint sich das Blatt zu wenden, und rund zehn
Jahre später beginnt eine Art Kelsen-Renaissance in Deutschland. Nun fängt die
Staatsrechtslehre langsam an, sich das bislang ausgeschlagene Erbe nach und nach an-
zueignen. Eingeübte vorurteilsgetragene Abwehrreflexe sind zwar immer noch viru-
lent, doch überwiegen heute die wissenschaftliche Neugierde und der Wunsch nach

4 Vgl. Jeremy Telman, The Reception of Hans Kelsen’s Legal Theory in the United States: A Sociological
Model, in: L’Observateur des Nations Unis 24 (2008), S. 299 ff.; Jeremy Telman, A Path Not Taken: Hans
Kelsen’s Pure Theory of Law in the Land of the Legal Realists, in: Robert Walter / Clemens Jabloner / Klaus
Zeleny (Hrsg.), Hans Kelsen anderswo – Hans Kelsen abroad, Wien 2010, S. 353–376.
5 Für Österreich gilt Besonderes.
6 Gustav Radbruch, Gesetzliches Unrecht und übergesetzliches Recht, in: Süddeutsche Juristen-Zeitung 1
(1946) 105–108 (107).

Franz Steiner Verlag


12 Vorwort

sachlicher Auseinandersetzung mit den vielfach als frisch und unverbraucht in ihrem
Provokationsgehalt empfundenen Thesen Kelsens.
Im Gegensatz dazu gilt Kelsen im globalen Maßstab als Mastermind der Rechtstheo-
rie; viele Schatzsucher aus Lateinamerika, aus Südostasien, aus Südeuropa, um nur drei
besonders Kelsen-affine Weltregionen zu nennen, sind unterwegs; sie sehen sich mit der
Sprachbarriere konfrontiert, wenn sie den als solchen erkannten Schatz heben wollen,
und sind bereit, diese zu überwinden. Die (zweite Auflage der) Reine(n) Rechtsleh-
re gilt in einer ganzen Reihe von Ländern zumeist des romanischen Sprachkreises als
Pflichtlektüre im ersten Studienjahr, gehört also zu den kanonischen Schriften. Nicht
zuletzt waren es die ausländischen Kelsen-Schüler zu Wiener Zeiten, die sich als eifrige
Exporteure und Verbreiter der Reinen Rechtslehre in ihre Heimatländer und darüber
hinaus betätigten, etwa Leonidas Pitamic, Alf Ross, Charles Eisenmann, Luis Recaséns
Siches oder Luís Legaz y Lacambra. Daran haben sich reiche rechtsphilosophische und
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

rechtstheoretische Folgediskurse angeschlossen, wie zum Beispiel der argentinische, der


mit den Namen von Carlos Eduardo Alchourrón und Eugenio Bulygin verbunden ist.
Doch sosehr der Status von Kelsen und der Reinen Rechtslehre in einer Vielzahl
von Ländern außer Frage steht, so sehr ergeben sich Rezeptionshindernisse, die mit
dem Faktum zusammenhängen, dass nahezu alle wichtigen Schriften der Reinen
Rechtslehre im Original auf Deutsch verfasst worden sind.7 Das Arbeiten mit Über-
setzungen – mögen sie auch so akkurat und bestechend sein wie jene der Erstauflage
der Reinen Rechtslehre durch Bonnie Litschewski Paulson und Stanley Paulson oder
jene der Zweitauflage durch Mario Losano8 – bringt unvermeidlicher Weise Probleme
mit sich. Diese wachsen um ein Vielfaches, wenn und da eine große Zahl auch sehr
wichtiger Schriften insbesondere nicht ins Englische übersetzt sind.
Und wer des Deutschen nicht mächtig ist, dem wird sich kaum erschließen, welcher
Reichtum und welche Heterogenität sich mit der Reinen Rechtslehre verbindet, die
eben nicht nur das isoliert-exklusive Werk des einsamen Genies ist, sondern Ausdruck
und Folge einer ungemein kraftvollen, unerschrockenen und fruchtbaren kollektiven
Anstrengung von rund 40 jungen Männern und Frauen, die Kelsen als „Jungösterrei-
chische Schule der Rechtstheorie“9 um sich geschart hatte. Um Kelsens Gedankenge-

7 Hans Kelsen, Hauptprobleme der Staatsrechtslehre entwickelt aus der Lehre vom Rechtssatze, Tübingen
1911 = HKW 2, S. 21–878; Hans Kelsen, Allgemeine Staatslehre, Berlin 1925; Hans Kelsen, Reine Rechtslehre.
Einführung in die rechtswissenschaftliche Problematik, Leipzig und Wien 1934; Hans Kelsen, Reine Rechts-
lehre, 2. Aufl., Wien 1960; Hans Kelsen, Allgemeine Theorie der Normen, Wien 1979.
8 Hans Kelsen, Introduction to the Problems of Legal Theory (übersetzt von Bonnie Litschewski Paulson
und Stanley L. Paulson), Oxford 1992; Hans Kelsen, La dottrina pura del diritto (übersetzt von Mario Losa-
no), Turin 1966.
9 Die Bezeichnung dürfte auf Alfred Verdroß, Bernhard Stark: Die Analyse des Rechts (1916), in: Schmol-
lers Jahrbuch für Gesetzgebung, Verwaltung und Volkswirtschaft im Deutschen Reiche 41 (1917), S. 477–
479 (478, 479), und Bernhard Starck, Die jungösterreichische Schule der Rechtswissenschaft und die natur-
wissenschaftliche Methode, in: Juristische Blätter 47 (1918), S. 301–304, zurückgehen.

Franz Steiner Verlag


Vorwort 13

bäude in seinem Reichtum, in seinen Prägungen, aber auch seinen Spannungen wirk-
lich zu ergründen, kommt man nicht nur nicht an den beiden Hauptschülern, Adolf
Julius Merkl und Alfred Verdroß, vorbei, auch die Schriften des Hauptes der Brünner
Schule, František Weyr, oder von jüngeren Schülern wie beispielsweise Fritz Sander,
Felix Kaufmann, Fritz Schreier, Julius Kraft oder Josef L. Kunz, müssen studiert wer-
den  – solange und soweit aber belastbare Übersetzungen fehlen, kommt man hier
ohne Deutsch nicht weit. Zugänglichkeit und Zugangswunsch fallen folglich ausein-
ander: In Deutschland die Möglichkeit, aber nicht (oder nur in geringem Maße) das
Interesse – global das Interesse, aber nur eingeschränkt die Zugänglichkeit. Der vorlie-
gende Band will genau hier einsetzen und erblickt in diesem Paradoxon eine Chance
wechselseitiger Hilfe, um einen großen Schritt vorwärts zu machen.
Und das ist mit den 24 Beiträgen dieses Bandes, die allesamt auf Vorträgen bei der
IVR-Tagung basieren, gelungen: Es schreiben Wissenschaftler aus dem In- und Aus-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

land, große Namen – an den tragisch während der Drucklegung verstorbenen John
Gardner sei in diesem Zusammenhang besonders erinnert und damit eines Großen des
Faches gedacht – wie Nachwuchstalente, diejenigen, welche (vorranging) auf Deutsch
und im deutschen Rechts- und Kulturkreis arbeiten und publizieren, diejenigen aus
anderen Kulturkreisen sowie diejenigen, deren Vita selbst das Kelsensche Wanderle-
ben zwischen den Kulturen (wenn auch heute weniger brutalen Zwängen geschuldet)
widerspiegelt. Die Beiträge sind denn auch heterogen – kritische, neutral analytische
wie zustimmend die Reine Rechtslehre anwendende Stimmen finden sich gleicher-
maßen; die Zugänge sind philosophischer, (rechts)theoretischer, ideengeschichtlicher
oder rechtsdogmatischer Natur; die Rechtskreise und -gebiete reichen vom deutschen
Privatrecht über das brasilianische Verwaltungsrecht zum Europa- und Völkerrecht.
Gemein ist ihnen der Versuch, vorurteilsfrei und offen die Reine Rechtslehre als Zu-
gang wie als Forschungsformation, Kelsens Leben und Werk, zu durchdringen, zu ana-
lysieren und so zu neuen Ergebnissen zu gelangen.
Wir, die Herausgeber, haben uns zwecks besserer Übersicht erlaubt, eine grobe
Gliederung der Beiträge vorzunehmen, freilich ohne dass damit eine Eingrenzung und
Einschränkung der enthaltenen Ergebnisse verbunden sein soll.
1. Die Aufsätze von Gardner, Alexy, Somek, Kletzer, Heidemann und Kiener
beschäftigen sich vorrangig mit den (rechts)philosophischen bzw. wissen-
schaftstheoretischen Grundlagen der Reinen Rechtslehre.
2. Demgegenüber wird von Carpentier, Pelegrino, Hochmann, Cadore und
Pirker die Dynamik und die Beziehungen der Rechtsnormen zueinander dis-
kutiert, wie auch die den Juristen besonders wichtige hermeneutische Tätig-
keit, welche wir zusammenfassend der „Rechtstheorie“ zugerechnet haben.
3. Eine weitere Gruppe von Beiträgen  – nämlich jene von Foljanty, Menezes,
Telman, Widłak und Kühler  – beschäftigen sich mit konkreten positiven
Rechtsordnungen und deren Problemen, auf die die Argumente der Reinen
Rechtslehre angewandt werden.

Franz Steiner Verlag


14 Vorwort

4. Schließlich hat die mit Abstand größte Gruppe von Beiträgen verschiedene
Aspekte des (ideen)geschichtlichen Kontextes der Reinen Rechtslehre bzw.
von Kelsen beleuchtet, nämlich Schauer, Kosielińska-Grabowska, Losano,
Paulson, Pavčnik, Wagrandl, Lijoi sowie Paz/Wagner.

Freiburg i. Br., 9. September 2019


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


I. Die Grundlagen der Reinen Rechtslehre /
The Foundations of the Pure Theory of Law
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


Normativity (in Kelsen and otherwise)

JOHN GARDNER †*
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: This chapter deconstructs the concept of normativity by engaging with Kelsen’s
postulate of an irreducible duality between Is and Ought. Supererogatory acts are a test
case, because they demonstrate that there must be reasons beyond norms to account for
the value of actions. It critiques Kelsen’s conflation of norms, reasons for action (epistemic,
prudential and moral) and values. If the normativity of the law is escapable, because it is
morally fallible, the special problem of moral normativity arises. Hence, the question of
normativity morphs into a question about the relevance of particular inescapable reasons.

Keywords: legal philosophy, Hume’s law, Hans Kelsen, normativity, reasons for action, Is
and Ought

Schlagworte: Rechtsphilosophie, Humes Gesetz, Hans Kelsen, Normativität, Hand-


lungsgründe, Sein und Sollen

1. Hume’s law

An Ought cannot be reduced to an Is, or an Is to an Ought; and so an Is cannot be


inferred from an Ought, or an Ought from an Is. (GTN 16.I)1

* All Souls College, Oxford. This contribution is based on my talk at the IVR German Section conference in
Freiburg on 21 September 2018. [Editors’ note: John Gardner passed away on 11 July 2019, before the proofs
for his contribution were finalised. The abstract, keywords and all changes (corrections) to the paper as
delivered are the editors’.]
1 My Kelsen quotations throughout are from Michael Hartney’s translation of Kelsen’s Allgemeine Theorie
der Normen (Vienna 1979), published as Hans Kelsen, General Theory of Norms (Oxford 1991). Since most of
the chapters and sections in the book are very short, and readers may well be using different editions, I do
not give page numbers. Instead I provide locations for the quoted passages parenthetically in the text, in the

Franz Steiner Verlag


18 JOHN GARDNER †

By these words in his last book, and no less consistently in earlier books, Kelsen
endorses Hume’s Law.2 In fact he fortifies it: Hume worried only about the inference of
an ought from an is; Kelsen worries no less about the inference of an is from an ought.
That makes his claim stronger than Hume’s. Kelsen’s core argument for this claim is
summarized by him as follows:
Since something can be without being decreed to be obligatory in a norm, and something
can be decreed to be obligatory in a norm without being in reality, therefore when some-
thing is it does not follow that something ought to be, or when something ought to be, that
something is. The relation between Is and Ought is one of irreducible duality. (GTN 17)

We will come back in a moment to the specialized apparatus of norms (and decrees)
that constitute Kelsen’s interpretation of the Ought. For now, we should agree that
the argument is sound as far as it goes. But how far does it go? Suppose we replace the
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

word ‘something’ with the letter ‘P’, and see how it looks now:
Since P can be without being decreed to be obligatory in a norm, and P can be decreed to
be obligatory in a norm without being in reality, therefore when P is it does not follow that
P ought to be, or when P ought to be, that P is.

Can we really conclude from this that ‘[t]he relation between Is and Ought is one of
irreducible duality’? Surely we can conclude only that the relationship between ‘P is’
and ‘P ought to be’ is one of irreducible duality. ‘You are buying me lunch’ neither
entails nor is entailed by ‘you ought to be buying me lunch.’ True enough. But we do
not know whether the same goes for the relationship between ‘P is’ and ‘Q ought to be’.
For example, we do not know whether the same goes for the relationship between
‘you promised to buy me lunch’ and ‘you ought to be buying me lunch’, to borrow John
Searle’s famous attempt at a counterexample to Hume’s law.3 We know roughly how
Kelsen would reanalyse that claimed counterexample to explain it away. He would say
that there is a hidden ‘Ought’ in the background. He would say that there is a norm
in play sub silentio according to which you ought to do whatever you promised. You
promised to buy me lunch so, according to the sub silentio norm, you ought to buy me
lunch. That is certainly a possible reanalysis of Searle’s claimed counterexample. It is
one of the (Humean) reanalyses that Searle is attempting to resist. But offering the
reanalysis is not making an argument for it. If you think that it is the correct reanalysis,
that is probably because you already endorse Kelsen’s conclusion that ‘[t]he relation

form GTN XX.xx, where XX in Arabic numerals represents the chapter number and xx in roman numerals
represents the section number (if applicable).
2 David Hume, A Treatise of Human Nature (2nd ed, ed L. A. Selby-Bigge and P. H. Nidditch, Oxford 1975),
3.1.1.
3 John R. Searle, ‘How to Derive “Ought” from “Is”’, Philosophical Review 73 (1964), 43.

Franz Steiner Verlag


Normativity (in Kelsen and otherwise) 19

between Is and Ought is one of irreducible duality.’ You have done nothing to show
that this conclusion is correct.
Personally, I share Searle’s broadly Aristotelian resistance to Hume’s law. I mention
that mainly by way of a declaration of interest. In the following remarks I will not make
any case against Hume’s law. What I aim to do here is merely to draw attention to some
ways in which the claim that ‘[t]he relation between Is and Ought is one of irreducible
duality’ can easily be made to seem more plausible than it is. What I will suggest is that
there are several different distinctions in the neighbourhood, and that the sense of an
‘irreducible duality’ is mainly owed to their conflation. Since Kelsen is an arch-conflat-
er of these distinctions, and one of the few to conflate them openly, deliberately, and
even proudly, he will be figuring prominently in a lot of what follows. Towards the end
we will, however, turn away from our engagement with Kelsen towards some preoc-
cupations of contemporary moral and legal philosophers. I will suggest that they are
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

overblown or misdirected preoccupations.

2. Norms and Oughts

What does any of this have to do with my title, ‘Normativity’? ‘Normativity’ is an un-
countable noun confected from the adjective ‘normative’, while ‘normative’ is in turn
an adjective derived from the countable noun ‘norm’. Maybe I am too literal-minded,
but it strikes me as natural to suppose that normativity is the property that all and only
normative things have, and things are normative if and only if they are norms. This
agrees with Kelsen’s usage, although he uses the word ‘normativity’ very rarely (e. g.
GTN 45 n114).4 More interesting and important, for present purposes, is that Kelsen
regards the world of norms as co-extensive with the world of Oughts, and normativity,
accordingly, as a property of all and only Oughts. I have given this contribution the
title ‘Normativity’ partly because it helps us to focus on this Kelsenian doctrine.
Norms, for Kelsen as for most students of the subject, are of various distinct types.
In his terminology, each distinct type has a different ‘normative function’ (GTN 1.V).
The main types of norms recognized and discussed by Kelsen (as by Hart, von Wright,
Raz, and many other students of the subject5) are the obligation-imposing, power-con-
ferring, and permission-granting types.6 Yet this variety of ‘normative functions’ lands

4 He is more promiscuous with the word ‘normative’ which is used in as many as 45 of his 61 chapters.
5 H. L. A. Hart, The Concept of Law (Oxford 1961), 26–32 (on obligations v powers) and 246n (on permis-
sions); G. H. von Wright, Norm and Action (London 1963), ch V (on obligations v permissions) and ch X
(on powers); Joseph Raz, Practical Reason and Norms (London 1975), chs 2 and 3.
6 Kelsen adds the ‘derogating’ type, but this seems to reflect a muddle on his part between the typology of
norms and the sub-typology of normative powers. Derogating is exercising a power that makes an action
permissible (either in the sense of removing an obligation or in the sense of yielding a separate permissive
norm). See GTN 25.IV for evidence of the muddle.

Franz Steiner Verlag


20 JOHN GARDNER †

Kelsen with some unique terminological challenges. If, as Kelsen wants to say, the world
of norms and the world of Oughts are co-extensive, then, as he points out, ‘the word
“ought” is being used in a broader sense than is usual’ (GTN 25.II). More specifically:
According to common usage, ‘ought’ corresponds only to commanding; ‘can’ corresponds
to empowering and ‘may’ to permitting. We say ‘He “ought”’ only of the person to whom
something is commanded; we say ‘He “may”’ of the person to whom something is per-
mitted and ‘He “can”’ of the person empowered to do something. If we say that even an
empowering norm decrees an ‘ought’ and that an ‘ought’ obtains even in the case of a per-
mission  – since empowering and permitting are essentially related to an Ought  – then
the word ‘ought’ expresses the three normative functions (commanding, empowering, and
permitting). (GTN 25.II)

This is not what I find strange. What I find strange is not how broadly Kelsen uses
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

the word ‘ought’, but by contrast how narrowly he uses it. Consider this passage:
The norm which decrees a certain behaviour to be obligatory institutes a value. The
judgement that some behaviour is ‘valuable’ or ‘has value’ (and in this sense, is ‘good’)
means that this behaviour is decreed to be obligatory in a norm, is the content of an
Ought. (GTN 16.III)

We will talk more about values below, and particularly about the relationship between
obligation-norms and values that is suggested in the first sentence. But for present pur-
poses, let’s think about the second sentence. It identifies the value or goodness of an
action with the action’s obligatoriness. Thus there are no valuable or good non-ob-
ligatory actions. This seems extremely implausible. To take just one kind of valuable
action thereby annihilated, what about supererogatory actions, also known as ‘actions
beyond the call of duty’? Possibly Kelsen believes that there are no supererogatory
actions. But that hardly seems to warrant his ruling out their existence by definition,
such that anyone who thinks that there are supererogatory actions, and that it can be
valuable to perform them, must be making a mistake about the very concept of value.
Perhaps denying the intelligibility of belief in supererogation is not Kelsen’s only
option. Instead he could conceivably invoke his ‘broader sense of the word “ought”’
and argue that supererogatory actions can be fitted into his normative taxonomy in a
different place. They are actions that one is permitted not to perform. Now to fit them
into his category of valuable actions he only has to make a minor modification:
The judgement that some behaviour is ‘valuable’ or ‘has value’ (and in this sense, is ‘good’)
means that this behaviour is decreed to be obligatory or permitted to be omitted in a norm.

We may agree with the proposal that supererogatory actions are actions that one is
permitted not to perform. They are regulated, in that negative way, by a permissive

Franz Steiner Verlag


Normativity (in Kelsen and otherwise) 21

norm.7 The difficulty is, however, that this is clearly not the feature of supererogatory
actions that entails their value. Many actions are permitted to be omitted without be-
ing supererogatory and without any implications as to the value of performing them.
Under the conditions of my train fare, I am permitted to break my journey overnight or
not to do so, at my discretion. Nothing is implied about the value of doing either. True,
breaking my journey overnight becomes supererogatory if, for example, I do it at the
cost of some personal inconvenience to assist a stranger who is taken ill on the train.
The value of the action lies in the assistance I thereby provide. Relative to that value,
however, the permission not to perform the action is a surprise and a puzzle. It is even
known among philosophers as ‘the problem of supererogation’.8 Given their value, why
are supererogatory actions not obligatory? Why, in spite of that value, are they permit-
ted to be omitted? The suggested modification to Kelsen’s claim about value turns this
question upside down. It makes it the case that actions that one is permitted to omit
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

are eo ipso valuable to perform. That is a crazy position to take.


Of course Kelsen might challenge my example: he might say that my assisting the
stranger on the train is an obligation, perhaps a moral obligation of which my legal
permission to break my journey overnight happily facilitates the fulfilment. Then he is
able to accommodate the case in his original formulation. But he has to be able to give
this answer not only for this case but for all alleged cases of supererogation. And if he
does, that brings him back to his original difficulty. He is not only denying that there
are cases of supererogation. He is denying the intelligibility of the category. He is not
solving the problem of supererogation but denying that there could possibly be any
such problem.

3. Reasons beyond norms

The problem of supererogation can be restated as a problem, not about values, but
about reasons. Supererogatory actions are those that one has powerful (possibly de-
cisive) reasons to perform, but that one is permitted not to perform because of the
burden of performing them. There can also be actions that one has powerful (possibly
decisive) reasons to perform but that one has no obligation to perform, never mind
any permission not to perform them.9 Your severe symptoms give you decisive reasons
to see your doctor, but you have no obligation to see her. Your love of Hitchcock gives

7 For an influential analysis in these terms, see Joseph Raz, ‘Permissions and Supererogation’, American
Philosophical Quarterly 12 (1975), 161.
8 Roderick Chisholm and Ernest Sosa, ‘Intrinsic Preferability and the Problem of Supererogation’, Synthese
16 (1966), 321.
9 It follows that not every action that one has no duty to perform is ‘beyond the call of duty’ in the sense
that can (in suitable cases) make its performance supererogatory. Only actions that one has no duty to
perform and that one is permitted not to perform can be supererogatorily performed. Thus the kind of

Franz Steiner Verlag


22 JOHN GARDNER †

you decisive reasons to see the new print of Vertigo, but you have no obligation to see
it. Of these actions one might say: they are actions that one ought to perform even
though it is not the case that one ought to perform them. Putting it that way invites
the charge of contradiction. But there is no contradiction. It is just that there is more
than one sense in which we say of an action that it ought to be performed. Sometimes
we mean that the action is obligatory; on other occasions we mean only that there are
powerful, perhaps decisive, reasons to perform it. Kelsen does not discuss the latter
case any more than he discusses the case of supererogation. That is because these are
cases, if I may put it this way, of an ‘ought’ without any norm that corresponds to it.
The action ought to be performed, and it is a valuable action to perform, but those facts
hold irrespective of any ‘commanding, empowering, or permitting’ norm.
Just as some deny that there can be supererogation, so some deny that there can be
an action that one has decisive reason to perform but that one has no obligation to
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

perform. Some utilitarians, for example, would subscribe to the view that I have an ob-
ligation to perform whichever action I have decisive reason to perform.10 But that is a
substantive principle. It admits the conceptual distinction between having an obligation
and having a decisive reason. Does Kelsen also admit it? To put it another way, does he
admit the existence of actions that one ought to perform but not by virtue of any norm
under which one ought to perform them? Perhaps he would say that, in the examples
of the severe symptoms and of the new Hitchcock print, the way to go is to differentiate
different normative points of view. One lacks both a moral obligation and a legal obliga-
tion, perhaps, but couldn’t it be that one falls under an analogously commanding norm
from the point of view of prudence or from the point of view of aesthetic appreciation?
Perhaps he would say that.11 But the answer does not solve his problem.
That is because, in order to solve his problem, it is not enough for him to assimi-
late only decisive reasons to the world of norms. We need to know how to think about
reasons more generally, whether decisive or not. Do they or do they not belong to the
domain of Ought? So now we see the strange narrowness of Kelsen’s use of the word
‘ought’. If no norm, he says, then no ought. But what if, in spite of there being no norm,
there are reasons? If I assert that you have a reason, probably not a decisive one, to eat
a hot lunch on a winter’s day, where does my assertion stand in Kelsen’s ‘irreducible
duality of Ought and Is’? Does it stand on the ‘Ought’ side or the ‘Is’ side?

permission involved in supererogation is not just absence of duty, which Kelsen calls ‘negative permission’
(GTN 31). It is the kind of permission that exists under a distinct permissive norm.
10 E. g. Peter Singer, ‘Famine, Affluence, and Morality’, Philosophy and Public Affairs 1 (1972), 229; Shelly
Kagan, The Limits of Morality (Oxford 1991).
11 He turns to the metaphysics of ‘points of view’ for some similar purposes, e. g. in GTN 58 note 163.

Franz Steiner Verlag


Normativity (in Kelsen and otherwise) 23

4. Reasons with and without values

Ontologically, reasons are facts, and that may dispose one to place reasons on the ‘Is’
side. What reason do you have to eat a hot lunch on a winter’s day? The fact that a hot
lunch will stop you shivering is one reason; the fact that a hot lunch will help keep
winter flu at bay is another. But one may think that the ontological classification is a
red herring. As Kelsen says:
By the very nature of things, even a statement about a norm is an is-statement, a statement
about the specific existence (an Is) of a norm (an Ought). But the ‘Is’ of an ‘Ought’, the
existence of a norm, is something different from the existence of a fact; it is an ideel and not
a ‘real’ Is or existence. (GTN 41)

So there is the fact that a norm exists (e. g. the fact of my obligation). That is differ-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ent, thinks Kelsen from the fact of a ‘real’ fact’s existence. We may worry a little about
the scare-quote-marks around ‘real’. But we may yet be tempted by the thought that the
‘“Is” of an “Ought”’ is something special, and we may think that in this respect the fact
of the norm is no different from the fact that constitutes the reason. Both are facts only,
so to speak, within the normative domain.
They are something special, then, but what makes them special? It is tempting to
return here to the relation with values. If I have a reason to φ, and a fortiori if my φing is
obligatory, doesn’t that entail that there is some value in my φing? In Kelsen’s version:
When a norm decrees behaviour to be obligatory, the actual behaviour which agrees with
the norm can be judged (i. e. evaluated) as valuable or good, and the behaviour which does
not agree with the norm as ‘disvaluable’ or bad. As was indicated previously, this norm
institutes a value, if we mean by ‘value’ agreement with a norm and by ‘disvalue’ failure to
agree with a norm. When a certain behaviour is objectively – i. e. purely cognitively – as-
serted to agree or not with a norm presupposed to be valid (and in this sense is evaluated),
the norm serves as a standard of value. (GTN 30)

Notice that Kelsen is a deontologist, i. e. one who believes that the norm institutes the
value, rather than the value instituting the norm. Be that as it may, we may think that
what makes the “‘Is’ of the ‘Ought’” into something special is the place of value in it.
Where there are reasons and norms there are values, and that is what takes reasons and
norms out of ‘ordinary’ facticity. Thus:
The duality of Is and Ought coincides with that of reality and value. … In the case of an
objective value-judgment, the value cannot … be understood as a property of reality, as a
colour is a real property of a real object. That something real is objectively ‘valuable’ means
that an Is agrees with an Ought. (GTN 16.III)

Again, we may worry about the possibly tendentious use of the world ‘real’ here. But
making a value-realist attack on Kelsen is not our immediate priority. Instead, our

Franz Steiner Verlag


24 JOHN GARDNER †

question is whether the domain of Ought is coextensive with the domain of value,
once we are thinking of both norms and reasons as belonging to the domain of Ought.
The answer is that it is not.
Consider epistemic reasons, reasons to believe. That I have reasons to believe P,
even decisive reasons to believe P, does not entail or even suggest that there is any
value in my believing P. My belief that it is a warm day outside may be supported by
evidence (people are walking by without coats), by testimony (I was told by my friend
who visited earlier), and by perception (I can feel the warm air through the window).
I have decisive reasons to believe that the day is warm. But what is the value of my be-
lieving this, while I lie incapacitated in my hospital bed? Perhaps the belief only adds
to the curse of my incapacitation. Likewise, while I am bed-ridden, what is the value of
my believing that the bus to town will be leaving in seven minutes, or that today there
is a strike on the railways? There is no value in my believing either of these things since
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

I am not going anywhere and I don’t have anyone else who is going anywhere to whom
I could impart the information. And yet I have decisive reasons to believe both of these
things: for both the bus timetable and the local newspaper are right here on my bed-
side table. Epistemic reasons, to put it crudely, are reasons that do not correspond to
any value in the thing that they are reasons for.
Could we not say that the truth of one’s beliefs is a value in its own right, and that
this is the value served by epistemic reasons? Kelsen rightly rejects this option:
‘Good’ is a value, a moral or legal value, according to whether it is a moral or a legal norm
which institutes this value. If we also conceive of truth as a ‘value’, as a logical or theoretical
value (as opposed to a moral or legal value which is a practical value), we might believe
that the desired parallel or analogy between a statement and a norm can be justified by the
claim that both are related in some way to values.

But this is not possible: there is no parallel or analogy between the being-true of a state-
ment and the being-good of behaviour, between the judgment that a statement is true and
the judgment that behaviour is morally or legally good. The latter is a genuine value-judg-
ment, but not the former. (GTN 45.I)

But this leaves Kelsen with a difficulty. What are we to say about epistemic reasons,
reasons to believe? Being reasons that do not correspond to value in what they are
reasons for, do they belong to the domain of Ought or the domain of Is?
Kelsen offers no answer. He says nothing about epistemic reasons or epistemic
norms. In developing his view that truth is not a value, he thinks only of logic, not epis-
temology. He concludes that the so-called norms of logic are not truly norms:
There do not exist any norms of logic prescribing that statements are to be true, and thus
instituting truth as a logical value. Logic is not a ‘normative’ science. … Acts of thought as
such are not subject to any norms at all (GTN 45.III)

Franz Steiner Verlag


Normativity (in Kelsen and otherwise) 25

We may regard the claim that the norms of logic are not truly norms as a reductio.12 It
shows only that Kelsen’s General Theory of Norms is not an explanation of norms in
general, but only of a certain specialised type of norms. Be that as it may, the omission
of attention to epistemic norms raises independent doubts about Kelsen’s bold con-
clusion. Cannot a belief be unjustified? Cannot a believer be biased, gullible, supersti-
tious, prejudiced in her reasoning towards her beliefs? Are these not faults relative to
her cognition of truth? And if they are faults, should Kelsen not also say that they con-
stitute or entail epistemic norm-violations, notwithstanding that truth is not a value?

5. Normativity deconflated

Allow me to recap. The distinction between the world of norms and what lies outside
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

it does not correspond to the distinction between the world of reasons and what lies
outside it, for there are reasons beyond norms, for example those that tell in favour
of supererogatory action. The distinction between the world of reasons and what lies
outside it does not correspond to the distinction between the world of value and what
lies outside it, for there are epistemic reasons as well as practical ones. Finally, for com-
pleteness, the distinction between the world of norms and what lies outside it does not
correspond to the distinction between the world of values and what lies outside it, for
there are epistemic norms as well as epistemic reasons, and neither corresponds to any
kind of epistemic value. There are three quite different distinctions here. Kelsen con-
flates them. But now we know they are different: Which, we may wonder, constitutes
the ‘irreducible duality’ of Ought and Is? Now that we know that there are three differ-
ent distinctions in the neighbourhood, should we still agree with Kelsen’s claim that
The difference between Is and Ought cannot be explained any further: We are immediate-
ly aware of the difference (GTN 17)

6. The special problem of legal normativity

Many contemporary philosophers fret about normativity, and in particular about how
normativity is possible.13 But often they are not fretting about normativity in my liter-
al-minded sense, i. e. about the property or properties in virtue of which a norm is a
norm. Often their concerns extend beyond norms to reasons that are not norms. In
framing a question about reasons, irrespective of whether they are norms, one might

12 But cf Gilbert Harman, ‘Logic and Reasoning’, Synthese 60 (1984), 107.


13 For some recent high-quality fretting, see David Plunkett, Scott Shapiro, and Kevin Toh, Dimensions of
Normativity: New Essays on Metaethics and Jurisprudence (Oxford 2019).

Franz Steiner Verlag


26 JOHN GARDNER †

expect them to have used the word ‘rationality’ instead of the word ‘normativity’. They
might have asked: How is rationality possible? But that would be a rather misleading
way to express their question. For the word ‘rationality’ is widely used to designate the
capacity or propensity to respond to reasons, to follow reasons, or to use reasons in
one’s reasoning. In this sense it is a property, not of the reason, but of the reason-user.
By contrast, the word ‘normativity’ is not in everyday use so it can be given a technical
meaning for philosophical purposes. And perhaps the technical meaning is something
like this. Normativity designates the property or properties in virtue of which a reason
is a reason. Or, to put it differently, normativity is what differentiates a fact that is a
reason from a fact that is not a reason.
I said that many philosophers writing about normativity fret about how normativity
is possible. They are attempting, we may now think, to address deep doubts about the
very possibility of reasons. The doubts, however, come from two contrasting direc-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

tions, reflected in two contrasting literatures. Among philosophers of law, the doubts
are mainly about how so-called ‘legal reasons’ can possibly qualify as reasons. After all,
as Kelsen and others taught us, law is a human creation. It is found in or supervenes
upon legislative acts, judicial decisions, and customary practices. It is therefore mor-
ally fallible. At least sometimes it purports to give us reasons to act which are morally
invalid and which, apart from fidelity to the law, there is no reason to treat as reasons.
In that case fidelity to the law is surely irrational. So why do we classify legal reasons
as reasons? Why not go straight to the moral reasons, relative to which these so-called
legal reasons turn out to be nothing but extremely unreliable intermediaries? Call this
the ‘special problem of legal normativity’.
You will see right away that the special problem of legal normativity is not one that
could be encountered by Kelsen. That is because Kelsen regards morality as just anoth-
er normative order operating with a given social space:
Morality is just as much a social order as law, and an individual is just as subject to the
morality which is valid in the social group to which he belongs as he is subject to the law
which is valid in the social group to which he belongs. (GTN 1 note 4)

Like law, morality is a human creation. Law does not answer to morality any more than
morality answers to law. Nor is there any third thing to which both law and morality
answer:
The Natural Law … is [supposed to be] … a system of norms which is immanent in nature
and which is posited by the will of nature. There is, and there can be, no such law. The name
‘Natural Law’ is given to whatever appears just to a given Natural Law teacher, which he
then projects onto ‘nature’ in order to give it the necessary authority. (GTN 35 note 90)

Those who worry about the special problem of legal normativity think, on the con-
trary, that there is something, a set of reasons beyond the law, to which the law must
answer. Morality is the name generally given to this set of reasons.

Franz Steiner Verlag


Normativity (in Kelsen and otherwise) 27

One might say that, for Kelsen, both law and morality are relative, whereas for those
who worry about the special problem of legal normativity, law is relative whereas mo-
rality is real. Those who worry about the special problem of legal normativity worry
about how relative reasons are possible, given the existence of real reasons. What space
could there be in rational thought for reasons that lack the status of real reasons? There
is no such problem for Kelsen, since the relative is, so to speak, as real as anything in
the domain of Ought ever gets. Everything there is valid only relative to some system
of norms. There is no normative reality beyond that. Remember Kelsen’s rebuke:
[T]he ‘Is’ of an ‘Ought’, the existence of a norm, is something different from the existence
of a fact; it is an ideel and not a ‘real’ Is or existence. (GTN 41)

These words, however, tend to suggest a problem with talk of the ‘real’ and the ‘relative’.
It is not totally clear what is meant by either term. Even Kelsen must resort to scare-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

quotes around the word ‘real’ to convey that fact. Elsewhere, responding to that termi-
nological problem, I have suggested a different way of formulating the special problem
of legal normativity. I have suggested that we think of legal reasons as escapable and
moral reasons as inescapable.14 The idea is that there is an intelligible question of why
I should follow, or even pay attention to, legal reasons. Further reasons, not local to
the law, are needed for caring about what the law says. That is the sense in which legal
reasons are escapable. But there is no intelligible question of why I should follow, and
more generally pay attention to, moral reasons. No further reason is called for. They
are among the reasons that are mine to follow, and more generally to attend to, just in
virtue of my being a rational agent, i. e. an agent with the capacity to follow and attend
to reasons.
Formulated in these terms, the special problem of legal normativity becomes: Why
attend to these so-called escapable reasons, given that not doing so is ex hypothesi not
irrational? Why not just follow the inescapable reasons, such as the moral ones, direct-
ly? I already hinted at the answer. For escapable reasons, the question arises of why I
should follow them or more generally pay attention to them. The answer to that ‘why’
question needs to be given, ultimately, in terms of inescapable reasons. But it does not
follow that one could leap straight to the inescapable reasons and ignore the escapable
ones. Suppose I am advised to follow the law, or I promise to conform to the law, or I
debate whether I have an obligation to obey the law, or I warrant that I am acting with-
in the law, or I take on a role that requires me to support the rule of law. Naturally there

14 John Gardner, ‘Nearly Natural Law’, American Journal of Jurisprudence 52 (2007), 1, reprinted with minor
corrections in my book Law as a Leap of Faith (Oxford 2012). In his ‘Escapable Law: John Gardner on Law
and Morality’, Jerusalem Review of Legal Studies (forthcoming) Leslie Green makes some telling criticisms
of the escapable/inescapable distinction as I originally presented it. I have presented it here in a way that is
supposed to steer clear of most of his criticisms. However a couple of criticisms in section 3 of his paper do
strike me as rather relativistic and I make no attempt to steer clear of these!

Franz Steiner Verlag


28 JOHN GARDNER †

arise questions about the soundness of the advice, the bindingness of the promise, the
basis of the obligation, the truth of my warranty and the legitimacy of the role. These
questions arise at the level of inescapable reasons. But if there are such reasons they
refer to the law, which ex hypothesi a system of escapable reasons. The relevant ques-
tion is whether I have any inescapable reasons to resort to certain escapable reasons.
The escapable reasons do not become redundant just because there is a question of
whether I have inescapable reasons to resort to them. On the contrary: it is only if they
are not redundant – i. e. if resorting to them would make some difference – that I could
possibly have inescapable reasons to resort to them. There can be no reason to resort
to anything if resorting to it makes no rational difference.
Some people seem to be squeamish about calling escapable reasons, such as legal
reasons, ‘reasons’. Much ink was spilt among philosophers of law over the ages in trying
to establish that either legal reasons are inescapable or they are not reasons at all. But
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

why so? Presumably those who are squeamish about calling escapable reasons ‘rea-
sons’ would also be squeamish about calling escapable norms ‘norms’. Here they could
learn from Kelsen. Since Kelsen thought all norms escapable, his work provides a test
case for the cogency of the view that escapable norms are norms. One may think, as I
think, that he got a lot of things about norms wrong, including quite a few things about
legal norms. But one thing that he clearly did not get wrong is that legal norms are
norms in spite of being escapable (or ‘relative’) norms. Why be more squeamish about
calling escapable reasons ‘reasons’ than one would be about calling escapable norms
‘norms’? Is it is a norm of Monopoly that one must pay rent if one lands on another
player’s property. So there is a reason, so far as playing Monopoly goes, to pay such
rent. Naturally that does not allow one to avoid the further question of whether one
has any reason to play Monopoly. It is just that such a reason, which ultimately needs to
be inescapable, makes essential reference to the escapable reasons given by Monopoly
itself. For it is ex hypothesi a reason to play Monopoly, with all its crazy norms and
bizarre reasons (all fortunately escapable).

7. The special problem of moral normativity

If all this is right, then the special problem of legal normativity is not so much of a
problem. It is a storm in a teacup. So we turn our attention now to doubts that come
from precisely the opposite direction. They are doubts about what I call inescapable
reasons. How can there be such things? How can any reason, or for that matter any
norm, have such a hold over us? This might be called the ‘special problem of moral
normativity’ for it is most often raised in connection with moral reasons and moral
norms. But that already shows where one error is creeping in. For the same question
arises with various other kinds of reasons and norms. It arises with respect to pruden-
tial reasons and norms, for example. How can my own future have such a hold over me

Franz Steiner Verlag


Normativity (in Kelsen and otherwise) 29

that I ought to keep an eye on my own health and my own pension? An interesting fea-
ture of some writing that treats moral normativity as a special problem is that it treats
prudential normativity as fait accompli, and attempts (often in a contractarian spirit)
to build moral normativity out of it.15 But such a line of argument is hard to maintain
for long. What problem does morality throw up on this front that prudence does not?
Both moral and prudential reasons and norms have an inescapable hold over rational
beings, and inescapability is supposed to be the puzzling feature of reasons that is be-
ing explored under the heading of ‘normativity’.
More importantly, perhaps, the same question arises with respect to epistemic rea-
sons and epistemic norms. For they too are inescapable.16 There is no intelligible ques-
tion of why I should follow them. No further reason is called for. They are among the
reasons that are mine to follow, and more generally to attend to, just in virtue of my
being a rational agent, i. e. an agent with the capacity to follow and to attend to reasons.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Why is the intense worry about the normativity of moral reasons not paralleled, then,
by a no less intense worry about the normativity of epistemic reasons? This is a par-
ticularly important challenge if we think that what we are looking for is some mysteri-
ous ingredient X that differentiates a fact that is a reason from a fact that is not a reason.
For surely all facts are epistemic reasons. Any fact that we become aware of supports
the inference to some further belief beyond belief in that very fact. Everything is evi-
dence of something else. If that is true, and if we accept that normativity in the relevant
sense is a property of facts that are inescapable reasons, then there is no ingredient X
that must be added to a fact to turn it into a reason. Rather, the question turns into this
one: why is this particular fact a reason for this particular response? And that is not a
problem unique to inescapable reasons. Even of legal reasons and Monopoly reasons
we have the problem of why this legal or ludic fact is a reason for that response by a
competent reason-user. This is surely not a problem of normativity but of relevance. It
is not about how reasons can have their inescapable hold over us. It is about what is to
be done with them once (inescapably or otherwise) they have their hold over us. It is
about what rationalises what.
Thinking about epistemic reasons alongside moral reasons has another advantage in
defusing worries about moral normativity. It is part of the nature of a belief, as opposed

15 A no-nonsense example is David Gauthier’s Morals by Agreement (Oxford 1987): ‘Morality, we shall
argue, can be generated as a rational constraint from the non-moral [=prudential] premisses of rational
choice’ (4).
16 Brian Leiter writes: ‘Even in the theoretical domain, there is no real normativity, no norms of belief …
the agent must adhere to.’ Leiter, ‘Normativity for Naturalists’, Philosophical Issues 25 (2016), 64 at 75. This is
perhaps an attempt to deny the inescapability of epistemic reasons and norms. If so it is misleadingly put.
That epistemic reasons and norms are inescapable does not mean that any particular epistemic reason or
norm is inescapable. Probably there is room for diversity in approach or technique in arriving at defensible
beliefs. Defensibly, some rely more heavily on testimony, some on perception, etc. That is compatible with
the inescapability (or what Leiter calls ‘real normativity’) of epistemic reasons and norms taken as a set.

Franz Steiner Verlag


30 JOHN GARDNER †

to delusions, flights of fancy, etc. that it responds to reasons for believing. If one is not
believing for reasons then one is not exactly believing. Of course sometimes one may
get the reasons wrong. But even then there is the question of whether one gets the
reasons wrong for reasons. Did one miss the evidence, or misperceive what was going
on, for a reason? Responsiveness to the world – to the facts – is part of what it takes
to be a believer. Belief, in that sense, entails rationality. Is this not equally true with
action? How can one act without intending? And how can one intend if not by acting
for what one takes to be reasons? And if one mistakes what are not reasons for reasons,
again there is the question of whether one did so for reasons. As lawyers put it, if not
justified was one at least excused? Again these are questions that presuppose that we
are dealing with rational beings, beings with the capacity and the propensity for paying
attention to and making use of reasons. And there we have the answer to the supposed
problem of the normativity of moral reasons. As rational beings we cannot but engage
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

with reasons, at any rate the inescapable ones. And among inescapable reasons, mor-
al reasons are nothing special in respect of their inescapability. Their hold over us as
rational beings is no different from the hold of prudential reasons, epistemic reasons,
and various other inescapable reasons.
I say that moral reasons and epistemic reasons are alike in respect of their hold over
us. But we already know that they are unalike in another way. Where there are reasons
for action there is value in our doing as they would have us do. Where there are rea-
sons for belief, there need be no value in our believing as they would have us believe.
If there is something to discuss in ethics that does not equally arise in epistemology,
it is that. Could it be that those who believe in the existence of a special problem of
moral normativity are really just worried about the place of value in the world? If so,
their problem does not extend to epistemology, but it does still extend to prudence,
and one wonders why they regard it as a special problem for morality. But one may also
ask whether the problem is as intimately connected with our place in the world as the
talk of ‘normativity’ leads one to expect. For there is value in the world whether or
not there are extant valuers to experience it. The question is not how value comes to
be relevant to us or how we come to be locked into engaging with it. We are rational
beings, to whom (hence) reasons apply, including practical reasons, which by their
nature bring us into constant encounters with the value all around us. The problem,
rather, is in the nature of the world. I am not sure that we are well-served by thinking
of this as a problem of norms and normativity, or even of reasons and rationality. It is
a mistake to think that value is there for the sake of valuers. Valuers, on the contrary,
are there for the sake of value.17 If the special problem of moral normativity is really the
quite different problem of the existence of value, then I think we should be told.

17 Tim Macklem and I explore this theme further (and with more subtlety) in ‘Value, Interest and Well-Be-
ing’, Utilitas 18 (2006), 362.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System
of Non-Positivism

ROBERT ALEXY*
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: Hans Kelsen is a positivist. This, however, does not mean that his ideas are with-
out significance for non-positivism. The opposite is the case. Non-positivism, as a theory
of existing law, is necessarily a theory of the dual nature of law, that is, of its real or factual
dimension as well as its ideal or critical dimension. This implies that non-positivism neces-
sarily includes elements of positivism. At this point the question arises as to which kind of
positivism best fits with non-positivism. The answer to this question is Kelsen’s positivism.

Keywords: Positivism, non-positivism, Hans Kelsen’s non-naturalism, basic norm, norma-


tivity, non-positivistic basic norm

Schlagworte: Positivismus, Nichtpositivismus, Hans Kelsens Nichtnaturalismus, Grund-


norm, Normativität, nichtpositivistische Grundnorm

Some months before his death on January 7, 2018, Ralf Dreier published an article
with the title “Integratives Verstehen  – eine Bilanz” (Integrative Understanding  – a
Balance). There he takes stock of his intellectual life. We read that he has fluctuated not
only between Habermas and Luhmann but especially between Kant and Hegel, and
also between Kelsen and Radbruch.1 We learn that the Kelsen-Radbruch fluctuation
concerned him not only over a long period of time, but, indeed, right up to the day
he wrote the article.2 These fluctuations can be understood in two ways: first, as an
oscillation between all-or-nothing alternatives, that is, either to accept all the theses

* I should like to thank Stanley L. Paulson for suggestions and advice on matters of English style.
1 Ralf Dreier, Integratives Verstehen  – eine Bilanz, in: Rechtsphilosophie und Grundrechtstheorie Robert
Alexys System, ed. Martin Borowski, Stanley L. Paulson, and Jan Sieckmann, Tübingen 2017, 586.
2 Ibid., 583.

Franz Steiner Verlag


32 ROBERT ALEXY

essential for Kelsen or to accept all the theses essential for Radbruch, or, second, as an
attempt to connect elements from both camps. The concept of integrative understand-
ing in the title of Dreier’s article points in the direction of connection. Still, talk about
connections, like talk about middle ways in general, is of little value if the connections
are not constructed in a sufficiently precise way. To contribute to such a construction
is the purpose of my paper.

I. The System of Non-Positivism

Kelsen is a positivist, Radbruch a non-positivist. Positivism and non-positivism are


two theses that contradict each other. Positivism says that there exists no necessary
connection between law and morality. Non-positivism says that there does exist a
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

necessary connection between law and morality. If one could say nothing more about
positivism and non-positivism than this, there would be little reason to enquire about
connections between positivism and non-positivism. Fortunately, one can, and, in-
deed, must say more. One way to do so is to relate elements of Kelsen’s legal theory to
elements of the system of non-positivism.
All positivistic conceptions of the nature of law refer to two, and only two, elements
of the definition of law: authoritative issuance and social efficacy. These two elements
concern the real or factual dimension of law. All serious non-positivistic conceptions
add to these real or factual elements a third element. It refers to the ideal or critical
dimension of law: moral correctness, first and foremost justice.3 This is the thesis of
the dual nature of law.4 The dual-nature thesis connects a positive side of law with a
non-positive side. With this, the path leading to a construction of the connection of el-
ements of Kelsen’s theory and non-positivism is opened. But to open a path is different
from traveling on it. For this, a look at the system of non-positivism is indispensable.
The system of non-positivism consists of a center, a conceptual framework, and impli-
cations based on both.

1. The Centre: Claim to Correctness

The centre or the Archimedean point of the system of non-positivism is the argument
from correctness.5 This argument says that law necessarily raises a claim to correctness,
and that this claim refers not only to the real dimension of law, but also to its ideal

3 Robert Alexy, The Argument from Injustice A Reply to Legal Positivism (first pub. 1992), trans. Bonnie
Litschewski Paulson and Stanley L. Paulson, Oxford 2002, 3–4.
4 Robert Alexy, The Dual Nature of Law, Ratio Juris 23 (2010), 167.
5 See on this Alexy, The Argument from Injustice (n. 3 above). 35–9.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System of Non-Positivism 33

dimension,6 that is, to moral correctness, first and foremost to justice.7 The argument
from correctness is the source of the necessary connection between law and morality.
Thus, it is not surprising, that many objections have been raised against it.8 These ob-
jections cannot be discussed here. In the present context, only one point is of interest.
Kelsen apparently ascribes a certain thesis to all forms of non-positivism, namely, the
idea “that an immoral social system is not law presupposes an absolute morality, that
is to say, a morality that is valid everywhere and at all times”.9 This idea of Kelsen’s
seems to refer to a morality that is connected with three universal quantifiers, that is,
with three “alls”: all places in space, all points in time, and  – although not explicit,
this is implicit in the sentence quoted – all moral questions. I completely agree with
Kelsen that such a morality does not exist and, therefore, cannot be the result of “sci-
entific cognition”.10 But this most radical form of moral cognitivism is by no means a
presupposition of the claim to correctness thesis, and, thereby, of non-positivism. To
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

be sure, the claim to correctness thesis would lose its sense if everywhere everything
were always morally possible. In this case, nothing but decisions – in Kelsen’s theory,
decisions backed by empowerment – would rule the agenda. Countering this decision
thesis, however, there is the argumentation thesis. It says that rational argumentation
on moral questions is possible, for there exists a system of rules, principles, and forms
of rational practical discourse.11 Discourse theory does not say that there exists just one
right answer in all cases. But it says that some answers are discursively necessary, some
discursively impossible, and some merely discursively possible. Discursively necessary
is what is required by the rules of discourse either directly or indirectly, that is, with ref-
erence to the procedure of discourse. Human rights and democracy belong here.12 Dis-
cursively impossible is what is excluded by the rules of discourse. The racial principle

6 Robert Alexy, Legal Certainty and Correctness, Ratio Juris 28 (2015), 444.
7 Alexy, The Dual Nature of Law (n. 4 above), 168–72.
8 See, for instance, Eugenio Bulygin, Alexy und das Richtigkeitsargument, in: Rechtsnorm und Rechtswirkli-
chkeit Festschrift für Werner Krawietz, ed. Aulis Aarnio, Stanley L. Paulson, Ota Weinberger, Georg Henrik
von Wright, and Dieter Wyduckel, Berlin 1993, 19–24; Robert Alexy, Bulygins Kritik des Richtigkeitsar-
guments in: Normative Systems in Legal and Moral Theory Festschrift for Carlos E Alchourrón and Eugenio
Bulygin, ed. Ernesto Garzón Valdés, Werner Krawietz, Georg Henrik von Wright, and Ruth Zimmerling,
Berlin 1997, 235–50; Eugenio Bulygin, Alexy’s Thesis of the Necessary Connection between Law and Moral-
ity, Ratio Juris 13 (2000), 133–7; Robert Alexy, On the Thesis of a Necessary Connection between Law and
Morality: Bulygin’s Critique, Ratio Juris 13 (2000), 138–47; Eugenio Bulygin, Alexy Between Positivism and
Non-positivism, in: Neutrality and Theory of Law, ed. Jordi Ferrer Beltrán, José Juan Moreso, and Diego M.
Papayannis, Dordrecht 2013, 49–59; Robert Alexy, Between Positivism and Non-positivism? A Third Reply
to Eugenio Bulygin, in: Neutrality and Theory of Law (this note), 225–38.
9 Hans Kelsen, Pure Theory of Law, trans. (from the 2nd German Edition, 1960) Max Knight, Berkeley and
Los Angeles 1967, 68.
10 Kelsen, Pure Theory of Law (n. 9 above), 63.
11 Robert Alexy, A Theory of Legal Argumentation (first pub. 1978), trans. Ruth Adler and Neil MacCor-
mick, Oxford 1989, 187–206.
12 Robert Alexy, Discourse Theory and Human Rights, Ratio Juris 9 (1996), 220–33.

Franz Steiner Verlag


34 ROBERT ALEXY

(Rassenprinzip) and the leader principle (Führerprinzip), as constitutive for the Third
Reich, belong here. Discursively merely possible is what can be justified without vio-
lation of discourse rules, although the contrary can also be justified without violation
of discourse rules. This applies to the abstract level, which concerns, inter alia, human
rights as such, as well as to the concrete level, which concerns, in the case of human
rights, first and foremost, their application by means of proportionality analysis.13 The
broad area of mere discursive possibility, which comprises what Rawls calls “reasona-
ble disagreement”14, does not deprive the claim to correctness of its sense. To be sure,
when a case cannot be resolved by argument, a decision becomes necessary, in courts
and parliaments normally a majority decision. This is a solution that belongs to the
real dimension of law. But with the claim to correctness, the ideal dimension remains
alive as a regulative idea. Correctness as a regulative idea implies that the institutional
resolution of disagreement remains open for future argumentation.15
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

2. Conceptual Framework: Two Elements

Conceptual frameworks can be more or less elaborated. Here, only two elements are
of interest: the distinction between the observer’s and the participant’s perspective
and the distinction between classifying and qualifying connections between law and
morality.

a) Observer’s and Participant’s Perspective

The distinction between the observer’s and the participant’s perspective16 is of fun-
damental importance for the construction of the relation between positivism and
non-positivism. In the system of non-positivism, the perspective of the observer is,
with one small exception,17 a positivistic perspective, in contrast to the perspective of
the participant which is essentially a non-positivistic perspective. The observer asks
questions about and adduces arguments on behalf of a position that reflects how legal
questions are actually decided in a legal system, whereas the participant asks questions
about and adduces arguments on behalf of what he deems to be the correct answer to a

13 Robert Alexy, The Absolute and the Relative Dimension of Constitutional Rights, Oxford Journal of
Legal Studies 37 (2017), 46–7.
14 John Rawls, Political Liberalism, New York 1993, 55.
15 Alexy, The Absolute and the Relative Dimension of Constitutional Rights (n. 13 above), 47.
16 Alexy, The Argument from Injustice (n. 3 above), 25.
17 This exception results from the claim to correctness. A system of social rules that does not raise a claim to
correctness is not a legal system; Alexy, The Argument from Injustice (n. 3 above), 34. This conceptual truth con-
cerns, first, the ideal dimension and is, second, inevitable even for one who qualifies himself solely as an observer.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System of Non-Positivism 35

legal question in the legal system in which he finds himself. The observer’s perspective
is defined by the question “How are legal decisions actually made?” the participant’s
by the question “What is the correct legal answer?” These two questions require dif-
ferent kinds of arguments. The observer is confined to fact-based arguments, whereas
the participant, say, a judge, has to adduce, alongside fact-based arguments, normative
arguments that are not fact-based. This applies to all cases that, owing to the “open
texture”18 of law, cannot be resolved solely on the basis of fact-based arguments.19

(1) The Relation between the Two Perspectives

The relation between the two perspectives is asymmetrical, and this for two reasons.
First, all participants must also be observers. Otherwise they could not apply the posi-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

tive law. In contrast to this, it is possible to be an observer without being a participant,


that is, to be a pure observer. The second reason for asymmetry is that a legal system
can exist without pure observers, but not without participants. This implies that the
participant’s perspective is more fundamental than that of the pure observer.

(2) Naturalistic and Normativistic Positivism

There exist two kinds of pure observers: naturalistic and


normativistic pure observers, and, therefore, two kinds of NP ¬NP
positivism: naturalistic and normativistic positivism. All
normativistic positivists are non-naturalistic positivists. That
is to say, naturalistic and normativistic positivism stand in
the relation of contradictoriness. If one adds non-positivism
to this, one can formulate a triad in which non-positivism
¬P
stands both to naturalistic positivism and to non-naturalistic
or normativistic positivism in the relation of contrariness:
“P” stands here for “positivism”, “N” for “naturalistic”, and “¬” for negation. One
might call this triad the “naturalism triad”.

18 H. L. A. Hart, The Concept of Law, 3nd edn., Oxford 2012, 128.


19 The distinction between the observer’s and the participant’s perspective has been criticized from the
positivistic as well as from the non-positivistic side. See, on the one hand, Joseph Raz, The Argument from
Justice, or How Not to Reply to Legal Positivism, in: Law, Rights and Discourse The Legal Philosophy of
Robert Alexy, ed. George Pavlakos, Oxford 2007, 22–5, and, on the other hand, John Finnis, Law as Fact and
as Reason for Action: A Response to Robert Alexy on Law’s “Ideal Dimension”, The American Journal of
Jurisprudence 59 (2014), 86–90. On replies to this, see Alexy, An Answer to Joseph Raz, in: Law, Rights and
Discourse (this note), 45–8, and Robert Alexy, The Ideal Dimension of Law, in: The Cambridge Companion to
Natural Law Jurisprudence, ed. George Duke and Robert P. George, Cambridge 2017, 324–5.

Franz Steiner Verlag


36 ROBERT ALEXY

The question arises: Which of the two positivistic contraries to non-positivism fits
better into the system of non-positivism? The answer is easy: non-naturalistic positiv-
ism. Karl Olivecrona is a paradigmatic representative of naturalism. In his book “Law
as Fact”, published in 1939, Olivecrona maintains, with an eye to Kelsen, that “[t]he
rules of law are a natural cause – among others – of the actions of the judges in cases
of litigation as well as of the behaviour in general of people in relation to each other”.20
One might call this the “natural cause thesis”. In contrast to this, Kelsen, in the first
edition of his “Reine Rechtslehre”, published in 1934, defines “law as norm”,21 and norm
as “meaning”,22 and the “unique sense” of this meaning as “ought”, and “ought” as a
“category”.23 This is the language in which abstract entities are described. Kelsen insists
that norms – and thus, law – can be reduced neither to physical events nor to psychical
processes. They belong not to natural reality but to an “ideal reality”.24 Such an ideal re-
ality, which exists in addition to the physical and the psychical world, would be a “third
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

realm”25 in the sense of Frege. One might call this the “meaning content thesis”. With its
reference to Frege’s “third realm” or Kelsen’s “ideal reality”, the meaning content thesis
is a rather strong ontological thesis. Its correctness depends on the fourth concept of
Kelsen’s line of concepts used to determine what law is: norm, meaning, ought, and
category. The concept of category leads to the problem of the basic norm. This will be
our theme in the last part of this article. Here, only one point is of interest: the compe-
tition between naturalistic and non-naturalistic positivism.
This competition is decided by the claim to correctness, which stands in the centre
of non-positivism. The claim to correctness is necessarily connected with argumenta-
tion or justification. Correct is what can be justified. Justification is an inferential en-
terprise.26 A correct or rational performance of an inference has to proceed according
to the rules of logic. The rules of logic, however, cannot be directly applied to natural
facts. They can only be applied to propositions, be they propositions about what is, be
they propositions about what ought to be. I would like to call this the “propositionality
thesis”. This implies that naturalistic positivism, as represented by Olivcecrona, is as
such incompatible with non-positivism, whereas non-naturalistic positivism, as repre-
sented by Kelsen, is, in this respect, compatible. In short, the system of non-positivism
includes Kelsen’s non-naturalism.

20 Karl Olivecrona, Law as Fact, Copenhagen and London 1939, 16 (emphasis by R. A.).
21 Hans Kelsen, Introduction to the Problems of Legal Theory (first pub. 1934), trans. Bonnie Litschewski
Paulson and Stanley L. Paulson, Oxford 1992, 13.
22 Ibid., 11, 13–4.
23 Ibid., 24.
24 Ibid., 15.
25 Gottlob Frege, The Thought: A Logical Inquiry, in: Philosophical Logic, ed. P. F. Strawson, Oxford 1967,
29.
26 Robert Alexy, The Nature of Legal Philosophy, Ratio Juris 17 (2004), 162.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System of Non-Positivism 37

b) Classifying and Qualifying Connections

The second element of the conceptual framework to be considered here is the distinc-
tion between classifying and qualifying connections between law and morality. This
distinction concerns the effects of moral defects. The effect of a classifying connection
is the loss of legal validity. By contrast, the effect of a qualifying connection is legal
defectiveness or incorrectness that does not, however, undermine legal validity.27 The
decisive point here is that the effect of moral defectiveness or incorrectness is legal
defectiveness or incorrectness. That morally defective laws are morally defective is a
trivial truth. A positivist has no problem agreeing with this. But that these laws are,
over and above this, also legally defective is not only not trivial but, indeed, is of the
greatest significance for the relation between the real and the ideal dimension of law,
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

and this not only for theoretical reasons, but also for reasons practical in nature. If the
defect were merely a moral one, it would be difficult to explain why a higher court –
independently of what the positive law says – has legal power to set aside the unjust
decision of a lower court in a case in which this unjust decision is every bit as compat-
ible with positive law as a just decision would be.28

3. Implications

The distinction between the classifying and the qualifying connection implies that
three kinds of non-positivism have to be distinguished: exclusive, super-inclusive, and
inclusive non-positivism.29

a) Exclusive Non-Positivism

Exclusive non-positivism is the most radical version of non-positivism. It claims that


every injustice, every moral defect of a norm, precludes its being legally valid, its being
law. A classical version of this view is expressed by Augustine’s statement that “a law

27 Alexy, The Argument from Injustice (n. 3 above), 26.


28 Alexy, The Ideal Dimension of Law (n. 19 above), 327.
29 The logical relations between these three forms of non-positivism can be described by means of the
quantifier triad. See on this Robert Alexy, Law, Morality, and the Existence of Human Rights, Ratio Juris
25 (2012), 4–7.

Franz Steiner Verlag


38 ROBERT ALEXY

that is not just would not seem to me to be a law”.30 A recent example is Beyleveld and
Brownsword’s thesis “that immoral rules are not legally valid”.31
Kelsen attributes to non-positivism the thesis “that social norms must have a moral
content, must be just in order to be considered as law”.32 This attribution is correct
with respect to exclusive non-positivism, but incorrect with respect to super-inclusive
and inclusive non-positivism. And the argument, already considered,33 concerning the
limits of moral knowledge, which he connects with this attribution, is correct, too,
but correct only as an argument against exclusive non-positivism, the most vulnerable
form of non-positivism. It is not correct as an argument against super-inclusive and
inclusive non-positivism. Both acknowledge the limits of moral argumentation, that is,
the broad area of mere discursive possibility or reasonable disagreement. That is to say,
non-positivism is not destructed by Kelsen’s destruction of exclusive non-positivism.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

b) Super-Inclusive Non-Positivism

Super-inclusive non-positivism is the radical counterpart of exclusive non-positivism.


Super-inclusive non-positivism goes to the other extreme. It maintains that legal va-
lidity is in no way whatever affected by moral defects. At first glance this seems to be a
version of positivism, not of non-positivism. This first impression, however, is wrong
for two reasons. The first results from the distinction between classifying and quali-
fying connections which plays a central role in the system of non-positivism. Even if
legal validity is in no case affected by moral incorrectness, that is, no classifying effect
is assumed in any case of moral defectiveness, a qualifying effect can be assumed in
all cases of moral incorrectness, namely legal defectiveness. Strong tendencies in this
direction are to be found in the work of Aquinas, Kant, and Finnis.34 But this issue
shall not be pursued here. The second reason for the non-identity of super-inclusive
non-positivism with positivism is that super-inclusive non-positivism is compatible
with moral reasons for legal validity, positivism, as related only to the real or factu-
al dimension of law, not. Both reasons will play an important role in answering the
question of whether a basic norm is possible as well as necessary not only in Kelsen’s
non-naturalistic positivism but also in the system of non-positivism. This is the basic

30 Augustinus. De libero arbitrio – Der freie Wille, trans. and ed. Johannes Brachtendorf, Paderborn 2006,
86: “Nam lex mihi esse non videtur, quae iusta non fuerit” (I, 11).
31 Deryck Beylevld and Roger Brownsword, Human Dignity in Bioethics and Biolaw, Oxford 2001, 76.
32 Kelsen, Pure Theory of Law (n. 9 above), 64 (trans. alt.). See further Kelsen’s interpretation of non-posi-
tivism as the “thesis that law is moral according to its nature”, ibid., 68 (trans. alt).
33 Above, 31–2.
34 Alexy, The Dual Nature of Law (n. 4 above), 176; Alexy, Law, Morality, and the Existence of Human
Rights (n. 29 above), 6.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System of Non-Positivism 39

norm question. This question cannot be answered without reference to the third form
of non-positivism, inclusive non-positivism.

c) Inclusive Non-Positivism

Inclusive non-positivism claims neither that moral defects always undermine legal va-
lidity nor that they never do. Following the Radbruch formula,35 inclusive non-positiv-
ism maintains that moral defects undermine legal validity if and only if the threshold
of extreme injustice is transgressed. Injustice below this threshold is included in the
concept of law as defective but valid law.
All three forms of non-positivism can be reconstructed as solutions to the compe-
tition between two principles: the principle of legal certainty and the principle of jus-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

tice. Exclusive non-positivism gives unconditional precedence to justice, super-inclu-


sive non-positivism to legal certainty. Serious objections can be raised to both claims.36
The situation is different in the case of inclusive non-positivism; it establishes a condi-
tional precedence to justice in cases of extreme injustice.37 With this, the question how
the Radbruch formula is related to Kelsen’s basic norm enters the picture.

II. Basic Norm

With his basic norm, Kelsen tries to overcome two positions to which the pure theory
of law stands in fundamental opposition: naturalism on the one hand and non-posi-
tivism on the other.

1. Basic Norm and “Ought”

The main concern of Kelsen’s confrontation with naturalism is the transformation of


an “is” into an “ought”, the main concern of his confrontation with non-positivism is to
achieve this transformation without having any recourse to morality, that is, his aim is
to establish normativity without morality.

35 Robert Alexy, A Defence of Radbruch’s Formula, in: Lloyd’s Introduction to Jurisprudence, 8th edn., ed.
M.D.A. Freeman, London 2008, 426–8; Alexy, The Argument from Injustice (n. 3 above), 40–62.
36 Alexy, The Dual Nature of Law (n. 4 above),176.
37 Alexy, Legal Certainty and Correctness (n. 6 above), 445–7.

Franz Steiner Verlag


40 ROBERT ALEXY

Kelsen speaks, with respect to the formal structure of his solution of the “is”  –
“ought” problem, of a “syllogism”38 which he further qualifies as a “normative syllo-
gism”.39 This syllogism can be called “basic norm syllogism”.40 In The Argument from
Injustice, I reconstructed this syllogism as follows.41 The “is” is represented by the first
premise:
(1) Constitution C has in fact been issued and is socially efficacious.

The “ought” is expressed, inter alia, by the conclusion:


(3) It is legally prescribed that one behave in accordance with constitution C.

(3) does not follow logically from (1). For this reason, one has to introduce a further
premise which makes the transition from (1) to (3) possible:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

(2) If a constitution has in fact been issued and is socially efficacious, then it is legally pre-
scribed that one behave in accordance with this constitution.

This additional premise counts as one of many possible formulations of Kelsen’s basic
norm.
Stanley Paulson has argued that this reconstruction attributes “justified normativ-
ity”42 to Kelsen. If one understands by justified normativity morally justified norma-
tivity, I do not agree with this. To be sure, Joseph Raz attributes a justified norma-
tivity thesis to Kelsen when he says that Kelsen’s conception of the normativity of a
legal system presupposes that the system be considered as “morally just and good”.43
Nothing of the kind is contained in my reconstruction. Indeed, I use the expression
“legally prescribed” in the formulation of the basic norm. But expressions like “legally
prescribed” and “legally binding” (rechtsverbindlich)44 can be used without any moral
implications. Paulson argues that a basic norm formulation that employs such expres-
sions “appears to lend support to the justified normativity thesis”.45 Perhaps, these
expressions may seduce readers into connecting justified normativity in the sense of
morally justified normativity with these concepts. But this seduction can be coun-
tered by sound interpretative and systematic arguments. It should be added that the

38 Kelsen, Pure Theory of Law (n. 9 above), 202.


39 Ibid., 212.
40 Alexy, The Argument from Injustice (n. 3 above), 98.
41 Alexy, The Argument from Injustice (n.  3 above), 96–8; Robert Alexy. Comments and Responses, in:
Institutionalized Reason The Jurisprudence of Robert Alexy, ed. Matthias Klatt, Oxford 2012, 324–5. The num-
bering has been changed.
42 Stanley L. Paulsen, A “Justified Normativity” Thesis in Hans Kelsen’s Pure Theory of Law? Rejoinders
to Robert Alexy and Joseph Raz, in: Institutionalized Reason (n. 41 above), 68.
43 Joseph Raz, Kelsen’s Theory of the Basic Norm, in: Normativity and Norms Critical Perspectives on Kelse-
nian Themes, ed. Stanley L. Paulson and Bonnie Litschewski Paulson, Oxford 1998, 47, 58.
44 Hans Kelsen, Allgemeine Staatslehre, Berlin 1925, 99.
45 Paulson, A “Justified Normativity” Thesis in Hans Kelsen’s Pure Theory of Law? (n. 42 above), 87.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System of Non-Positivism 41

term “legally prescribed”, as used in my basic norm formulation, is just one element of
a family of concepts that can be used to express the “ought”. Others are legal validity,
objective sense, and empowerment.46 And I agree with Paulson that there is strong
textual support to the effect that in “Kelsen’s mature position”,47 already announced
in the late 1930s48 and culminating in the second edition of the Pure Theory of Law,49
empowerment is the fundamental normative modality. But in our present context
this is of no special importance. Our question is how the transition from an “is” to an
“ought” can be brought about, and the answer to this question does not depend on
the normative modality representing the “ought”. Furthermore, the textual arguments
speaking for the fundamental character of empowerment can be confronted with sys-
tematic arguments that speak against it.50 But I shall not elaborate this any further
here. For our purposes it is sufficient to state that the basic norm syllogism shows that
a basic norm is, as Kelsen puts it, “logically indispensible”51 (logisch unerläßlich) for
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

the transformation of an “is” into an “ought”.


According to Paulson all this shows nothing more than that Kelsen’s “appeal to the
normative syllogism” amounts to a “begging the question”, “a petitio principii”.52 I agree
with this on an important point, but not completely. The point on which I do not agree
concerns the analytical power of the basic norm syllogism. The basic norm premise
qua second premise of this syllogism is not merely an “assumption”.53 It is a necessary
presupposition of the transformation of an “is” into an “ought”. Kelsen himself empha-
sizes that there exists an alternative to normativism, namely naturalism, according to
which the law is nothing more than a system of “power relations”.54 This implies that
the necessity of the basic norm premise is not an unconditional necessity but only a
conditional necessity.55 This has far reaching consequences for qualifying the episte-
mological character of Kelsen’s basic norm. It is not simply transcendental, but only
“weakly transcendental”,56 and, with this, only a relative a priori category.57 But this is
more than nothing. It is a transformation from an “is” to an “ought”.

46 Ibid., 86–7.
47 Ibid., 83.
48 Hans Kelsen, Recht und Kompetenz. Kritische Bemerkungen zur Völkerrechtstheorie Georges Scelles,
in: Hans Kelsen Auseinandersetzungen zur Reinen Rechtslehre, ed. Kurt Ringhofer and Robert Walter, Vienna
1987, 75.
49 See, for instance, Kelsen, Pure Theory of Law (n. 9 above), 26.
50 Robert Alexy, Hans Kelsen’s Concept of the “Ought”, Jurisprudence 4 (2013), 240–5.
51 Kelsen, Pure Theory of Law (n. 9 above), 204.
52 Paulson, A ‘Justified Normativity’ Thesis in Hans Kelsen’s Pure Theory of Law? (n. 42 above), 70.
53 Ibid.
54 Kelsen, Pure Theory of Law (n. 9 above), 218.
55 Alexy. Comments and Responses (n. 41 above), 325.
56 Alexy, The Argument from Injustice (n. 3 above), 110.
57 Robert Alexy, Hans Kelsens Begriff des relativen Apriori, in: Neukantianismus und Rechtsphilosophie,
ed. Robert Alexy, Lukas H. Meyer, Stanley L. Paulson, and Gerhard Sprenger, Baden-Baden 2002, 193–202.

Franz Steiner Verlag


42 ROBERT ALEXY

2. The Necessity of Kelsen’s Basic Norm

Kelsen’s basic norm qua transformation rule is – with respect to its structure, not with
respect to its systematic consequences – the simplest form of a basic norm. It confines
itself to connecting issuance and efficaciousness on the side of the “is” with validity
or another normative concept on the side of the “ought”. One can call it, with respect
to its structural simplicity, the “minimal basic norm”. Kelsen’s syllogism argument is
successful insofar as it demonstrates that one has at least to presuppose this minimal
basic norm if one wants to transform an “is” into an “ought”. This is an analytical ar-
gument, and one can characterize Kelsen’s basic norm for this reason as an “analyt-
ical basis norm”.58 But to use a certain form of the basic norm in an argument that
successfully demonstrates that a basic norm is necessary for the establishment of the
“ought”, is not to demonstrate that just this form is necessary, too. To demonstrate that
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

we need a basic norm does not suffice to demonstrate that the basic norm used in this
demonstration is the basic norm. In order to show this, “alternative approaches” have
to be “ruled out”.59 It is exactly this point that Paulson addresses with his reproach of
“begging the question”.60 What is more, Paulson explicitly refers, in the context of the
alternative problem, to natural law theory, that is, to non-positivism: “In particular,
[Kelsen’s] contemptuous dismissal of natural law theory has no force as an argument,
and natural law theory remains, then, as an alternative to his own theory.”61

3. A Non-Positivistic Basic Norm

Kelsen’s thesis that a basic norm is necessary in order to perform a transformation of


an “is” into an “ought” remains, however, true. This transformation can be described
as a category transformation,62 and non-naturalistic positivism needs this transforma-
tion as well as non-positivism. Insofar as this is the case, Kelsen’s basic norm express-
es a universal truth. This universal truth establishes a necessary connection between
non-naturalistic positivism and non-positivism.
With this, however, the alternative question is not yet answered. The answer con-
sists, first, of the construction of a non-positivistic basic norm, and, second, of its jus-
tification. Here, I will focus on the construction. This construction comprises three
steps by means of which Kelsen’s basic norm is transformed into a non-positivistic
basic norm: first, the insertion of a limiting clause, second, the insertion of an opening

58 Alexy, The Argument from Injustice (n. 3 above), 96.


59 Paulson, A “Justified Normativity” Thesis in Hans Kelsen’s Pure Theory of Law? (n. 42 above), 76.
60 Ibid., 70.
61 Ibid., 76. See also ibid., 76, fn.77.
62 Alexy, The Argument from Injustice (n. 3 above), 105.

Franz Steiner Verlag


Hans Kelsen’s Legal Theory in the System of Non-Positivism 43

clause, and, third, the addition of a level of justification. With this, Kelsen’s analytical
basic norm is transformed into a normative basic norm.63

a) Limiting Clause

The Radbruch formula, which says, in its shortest form, that extreme injustice is no
law, is, as already stated, an essential element of inclusive non-positivism, which is en-
dorsed here as the result of balancing legal certainty against justice.64 Super-inclusive
non-positivism knows no limiting clause, exclusive non-positivism advocates the ex-
treme counterpart, a limiting clause that knows no limits. Here, only the basic norm
of inclusive non-positivism shall be considered. It can be formulated in different ways.
Our starting point shall be the following version of Kelsen’s basic norm:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

(1) The law includes all norms that have been authoritatively issued and are socially effi-
cacious.

The insertion of the Radbruch formula shall be expressed in the following way:
(2) The law includes all norms that, first, have been authoritatively issued and are socially
efficacious and that, second, are not extremely unjust.

b) Opening Clause

The Radbruch formula concerns the classifying connection between law and morality.
In order to make the non-positivistic basic norm complete, the qualifying connection,
too, must be included. This can be done by an opening clause which refers to the claim
to correctness:
(3) The law includes all norms that, first, have been authoritatively issued and are socially
efficacious and, second, are not extremely unjust, and, third, all principles and arguments
that are required by the claim to correctness.65

63 Ibid., 116.
64 Above, ###.
65 Alexy, The Argument from Injustice (n. 3 above), 102, 127.

Franz Steiner Verlag


44 ROBERT ALEXY

(c) Level of Justification

According to Kelsen, the basic norm is incapable of being established: “Into the basis
of the validity of the basic norm there can be no further enquiry.”66 Non-positivism
contests this.67 An example for a non-positivist who uses the term basic norm (Grund-
norm), even if in part only in quotation marks, is Gustav Radbruch.68 His justification
of the basic norm is exclusively the principle of legal certainty,69 occasionally connect-
ed with “peace, order” (Friede, Ordnung).70 This exclusive orientation on legal certainty
brings Radbruch’s position of 1932 very close to super-inclusive non-positivism, which
also could be considered as a kind of moral positivism. Moral positivism, however, is
much less acceptable than Kelsen’s detached or analytical positivism.71
All of this shows that the alternative problem, as formulated by Paulson, concerns
far more than the problem of the logical reconstruction of the transition from an “is” to
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

an “ought”. It concerns, perhaps, the deepest question of legal philosophy, the relation
of law and morality.

Robert Alexy
Christian Albrechts University, Faculty of Law, Olshausenstrasse 40, 24118 Kiel,
Germany, alexy@law.uni-kiel.de

66 Hans Kelsen, The Function of the Constitution (first pub. 1964), trans. Iain Stewart, in: Essays on Kelsen,
ed. Richard Tur and William Twining, Oxford 1986, 117 (trans. alt.); see also Kelsen, Pure Theory of Law
(n. 9 above), 195.
67 Alexy, The Argument from Injustice (n. 3 above), 113–6.
68 Gustav Radbruch, Rechtsphilosophie, 3rd edn. 1932, in: Gustav Radbruch Gesamtausgabe, ed. Arthur Kauf-
mann, vol. 2, Heidelberg 1993, 313, 324; see also 422. With and without quotation marks the term is used at
324.
69 Ibid., 314, 422.
70 Ibid., 314.
71 Alexy, The Argument from Injustice (n. 3 above), 121.

Franz Steiner Verlag


The Demystification Impasse
Legal Positivism Divided Against Itself

ALEXANDER SOMEK
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: Legal positivists emphasize the relevance of law-creating social facts. This em-
phasis bespeaks a healthy skepticism towards the idea that the law lends expression to
something natural or necessary. Hence, legal positivism seeks to demystify the law. With
regard to the task of demystification, at least two strategies can be distinguished. The first
seeks to anchor the purportedly moral foundations of the law in social conventions, while
the other reduces the alleged rational force of appeals to morality in the power to make
them stick. Interestingly, each strategy of demystification can be used to demystify the
other. The upshot is, in the end, a demystification impasse.

Keywords: Legal positivism, Inclusive legal positivism, separability thesis, social facts, de-
mystification, conventionalism

Schlagworte: Rechtspositivismus, Inklusiver Rechtspositivismus, Trennungsthese, so-


ziale Tatsachen, Demystifizierung, Konventionalismus

Introductory

What makes legal positivism special? The trivial answer has it that, as a position in legal
theory, legal positivism insists on separating law and morality. However misleading
or even wrong-headed this answer may be,1 it can be rendered least problematic by
attributing to legal positivists the view that a sharp line needs to be drawn between
propositions concerning what the law is, on the one hand, and postulates with regard

1 John Gardner, Law as a Leap of Faith: Essays on Law in General, Oxford: Oxford University Press, 2012,
19–53; Joseph Raz, The Authority of Law: Essays on Law and Morality, (2d. ed.), Oxford: Oxford University
Press, 2009, 315–316.

Franz Steiner Verlag


46 ALEXANDER SOMEK

to what it ought to be, on the other.2 Regardless of whether this so-called “separability
thesis” captures adequately, let alone fully, the core of what positivists believe there
is still something that proponents of the approach share, namely a commitment to
recognize, with almost exalted soberness, the relevance of law-creating social facts.3
This commitment implies a healthy scepticism towards the idea that the law lends
expression to something natural and necessary or is a manifestation of the order of
things or possibly even an exhibit of superhuman design. It was thus not by accident
that H. L. A. Hart attributed to one of the arch legal positivists, Jeremy Bentham, the
ambition to “demystify” the law and the legal profession.4
One should not be surprised to see the problem that demystification is supposed to
resolve arise in the legal context in the first place. The run-of-the mill operations of legal
institutions involve the exercise of force. Law supposedly renders the use of such force
legitimate, for example, by motivating the morally intransigent to do the right thing in
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

response to coercive threats. Those employing force in the name of the law are likely to
rationalize their conduct by putting the best possible face on what authorizes their con-
duct. The chief means thereto is idealization, that is, the representation of facts in light
of appealing ideals, such as justice or the rule of law, or the invocation of awe-inspiring
powers, such as the accumulated wisdom of generations.5 Social facts are then seen as
lending expression to norms or visions of the reasonable. Hence, the demystification of
legal ideas that we encounter in the works of legal positivists is, to a certain extent, also
the consequence of taking the separation of law and morality seriously. It makes an end
to viewing the substance of law as an embodiment of moral ideas.
Unsurprisingly, the world of law is replete with idealizations. Much of the more
mundane work of legal positivists, which is of relevance to doctrinal legal scholarship,
concerns debunking false idealizations and limiting their scope, if at all, to the level
indispensible to render the law as a social fact that is invested with normative signif-
icance. Not by accident, schools of legal positivism differ with regard to whether or
not they regard these idealizations as necessary in the first place;6 they diverge from
each other also inasmuch as they debunk the idealizations embraced by their peers as
superfluous and misleading.

2 H. L. A. Hart, Positivism and the Separation of Law and Morals, Harvard Law Review 71, 1958, 593–621.
3 Joseph Raz, The Authority of Law: Essays on Law and Morality, Oxford: Clarendon Press, 1979, 46–47.
4 H. L. A. Hart, Bentham and the Demystification of the Law, Modern Law Review 36, 1973, 2–17. This legacy
of legal positivism has more recently been recognized in Anglo-American circles. Joseph Raz, Ethics in the
Public Domain: Essays in the Morality of Law and Politics, (2d. ed.) Oxford: Oxford University Press, 1995,
210; Scott J. Shapiro, Legality, Cambridge, Mass.: Harvard University Press, 2011, 388–389; Brian Leiter, The
Radicalism of Legal Positivism (ssrn.com/abstract=1568333), in: Legal Realism and Legal Positivism Recon-
sidered, Ethics, 2001, 111–301.
5 The relevant Burkean pretensions were the target of Bentham’s scorn. Hart, (footnote 4), 2–3.
6 As is well known, Scandinavian legal realists do not. Torben Spaak, A Critical Appraisal of Karl Olivecro-
na’s Legal Philosophy, Heidelberg: Springer, 2014.

Franz Steiner Verlag


The Demystification Impasse 47

This is the topic of central concern here. What serves to demystify from the angle
of one version of legal positivism may appear to be guilty of re-mystification from the
perspective of another. This is, arguably, one of the major problems bedevilling the
overall approach.
Unnecessary or misleading idealizations conceal from the observer what they real-
ly are. This explains why not only legal positivists, but also other critical scholars, have
perceived the law as being surrounded with an ideological halo.7 The law comes to
this world with the aid of ideas obscuring what the law really is. Its self-presentation
is aided and abetted by social practices that endow it with an atmosphere of conse-
cration.

Demystifying the social appearance of law


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

The quest to rid the law of unnecessary idealizations can be pursued on at least three
different levels.
The first is straightforward and concerns the social aura of the law and of the legal
profession. The Levellers as well as Bentham have to be credited for incisive interven-
tions meant to expose displays of esoteric legal language and uses of the vestiges of
authority as strategies for creating an insuperable distance between those who are in
the know and the lay public that is not. Traditional apparel, the comportment of or-
gans of the courts and the ritual quality of legal procedures are means to instil into the
uninitiated a sense of awe and respect for the wisdom and dignity of those who master
a craft to which are admitted only a few. This type of demystification concerns the
broader social context in which the law operates and how an aesthetics – among which
not least the indirect communication of complexity – serves the material interests of a
privileged profession. The requisite debunking of pomp and circumstance is supposed
to pave the way for a simpler legal system and to a practice of law that is more amenable
to the horizon of notoriously bewildered and easily flabbergasted ordinary people.

Demystifying legal scholarship

The second level of demystification is slightly more difficult to pin down, yet possibly
even more important than the first. It is targeted at the manner in which legal scholars
and adjudicating bodies – courts, notoriously – make sense of (“interpret”) the legal
materials. In this context, we encounter what legal positivists usually understand by

7 D. Kairys (ed.), The Politics of Law: A Progressive Critique, New York: Pantheon Books, 1982. Alan Hunt,
The Ideology of Law: Advances and Problems in Recent Applications of the Concept of Ideology to the
Analysis of Law, Law and Society Review 19, 1985, 11–38.

Franz Steiner Verlag


48 ALEXANDER SOMEK

“ideology”, namely, a false representation of legal norms subservient to some partisan


agenda. This function is played by ideology even if those who perpetuate false ideas
may not be aware of their complicity. It is manifest in how scholars routinely reason
about law and in the vocabulary they use.8
While exposing the adverse social consequences of surrounding the law with the
aura of majesty and mystery was one great virtues of Bentham’s critique of the English
common law tradition, exposing the traces of ideology in legal reasoning was perhaps
the greatest achievement of Kelsen’s legal theory, in particular in the field of public law.
Three types of critical intervention were used by Kelsen in the pursuit of this objec-
tive: challenging dualism, identifying hypostatization, and detecting methodological
syncretism.
In a manner that anticipated what many decades later would sail under the flag of
“deconstruction”9 Kelsen attacked various binary oppositions that predominated the
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

legal thinking of his time. Just as in the case of binaries that would in the late twentieth
century become the intellectual sweethearts of deconstructivists10 one pole of a pair of
opposites was favoured over the other. Kelsen went about his deconstructive business
by showing that the meaning of the disfavoured pole accounted, at least in part, for the
significance of its preferred opposite. Kelsen attacked the duality of private and public
law and concluded that the distinction was spurious. He submitted the “dualistic” po-
sition on the relation between domestic law and public international law to a strident
critique in order to replace it with two versions of monism to which he attributed equal
plausibility. He tore down the line separating natural personhood and the personhood
of legally constituted entities, such as corporations, and upended the traditional rela-
tion between rights and laws. All of this work is neatly summarized in the first edition
of his Pure Theory of Law.11
Hypostatization is for nominalists of a Kantian persuasion12 a cardinal sin, for it mis-
takes what is merely a thought for an existing entity. For example, while the concept of

8 Otto Pfersmann‚ Kelsens Ideologiekritik, in: Hans Kelsen: Die Aktualität eines großen Rechtswissenschaft-
lers und Soziologen des 20  Jahrhunderts (ed. N. Aliprantis & T. Olechowski), Wien: Manz, 2014, 53–66.
9 Jack Balkin, Deconstructive Practice and Legal Theory, Yale Law Journal 96, 1987, 743–786.
10 For an introduction, see Jonathan Culler, On Deconstruction: Theory and Criticism after Structuralism,
London: Routledge, 1983, 86–87.
11 Hans Kelsen, Introduction to the Problems of Legal Theory (trans. Bonnie Litschewski Paulson & Stanley
L. Paulson), Oxford: Clarendon Press, 1992.
12 Immanuel Kant, Critique of Pure Reason (trans. P. Guyer & G. Wood), Cambridge: Cambridge Univer-
sity Press A 395, 1998, 439: “Thus every dispute about the nature of our thinking being and its conjunction
with the corporeal world is merely a consequence of the fact that one fills the gaps regarding what one
does not know with paralogisms of reason, making thoughts into things and hypostatizing them; from this
arises an imagined science, both in regard to affirmative and negative assertions, in that everyone either
presumes to know something about objects about which no human being has any concept, or else makes
his own representations into objects, and thus goes round and round in an eternal circle of ambiguities and
contradictions.”

Franz Steiner Verlag


The Demystification Impasse 49

the state, according to Kelsen, signifies merely the unity of the legal system (which is a
rather mystical idea, anyhow) the public scholarship of his time took the state to be a
real entity that precedes the legal order. Hypostatization concerns also the concept of
the “natural person” which is based on the belief in the reality of something that is the
mere point of imputation on which a bundle of norms converges.13 Obviously, there is
an internal link between dualisms and hypostatization, for one of the entities that in
juxtaposition with another comprises a duality is frequently only posited by thought
(such as the natural person or the state). This is part of the deconstruction strategy.
The hypostasized entity is usually the preferred pole of the binary. As it falls by the
wayside, the “dangerous supplement” is all that remains.14
Finally, uncovering methodological syncretism – the mingling of ways of reasoning
or analysis that belong to different disciplines – is the strategy that Kelsen uses early in
his work in order to emancipate the legal analysis of acts of will from its entrapment in
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

psychology.15 It is also the battle cry used against any interdisciplinary work in the field
of public law scholarship, in particular if it concerns a combination with sociological
theory.16

No fertile ground

It bears emphasis that what is perhaps the most important part of Kelsen’s theory did
not at all fall on fertile ground in the Anglo-American world even in the case where
Kelsen was blessed with the good fortune of having a most gifted student such as Hart.
Kelsen’s ideas were simply not relevant there. In the case of the United Kingdom it
stands to reason that, perhaps not only in public law, the common law tradition has
never had such a strong tradition of legal science as its civil law counterpart. A legal cul-
ture that is comfortable with reasoning from case to case while leaving the business of
legal theory to high-minded social reformers does not feel prompted to revisit its foun-
dations. By contrast, much of the controversies in German public law in the nineteenth
and early twentieth century concerned basic concepts, fundamental ideas such as the
Rechtsstaat and developing the correct understanding of relatively recent phenomena
such as rights against the state. Hence, incomparable paths of political development
and the intellectual shape of legal scholarship may explain why “Germanic” debates

13 Kelsen (footnote 11), 47–51.


14 Culler (footnote 10), 105–106.
15 Hans Kelsen, Hauptprobleme der Staatsrechtslehre entwickelt aus der Lehre vom Rechtssatz, Werke vol. 2.1
(ed. M. Jestaedt), Tübingen: Mohr Siebeck, 2008, 192–300. Stanley L. Paulson, A “Justified Normativity”
Thesis in Hans Kelsen’s Pure Theory of Law? Rejoinders to Robert Alexy and Joseph Raz, in: Institutional-
ized Reason (ed. M. Klatt), Oxford: Oxford University Press, 2011, 62–111 at 65.
16 Hans Kelsen, Der soziologische und der juristische Staatsbegriff: Kritische Untersuchung des Verhältnisses
von Staat und Recht, Tübingen: Mohr, 1922.

Franz Steiner Verlag


50 ALEXANDER SOMEK

did not find any resonance in the United Kingdom. In the United States, the situation
was even worse for Kelsen. When Kelsen arrived, the vintage critique of conceptu-
alism, as it is particularly evident in the writings of Felix Cohen,17 had already run its
course. Clearly, the deconstructive part of the program, which consisted of challenging
and overcoming “conceptual jurisprudence” had in large part been already completed
by the American legal realist. The constructive vision did not clearly emerge, but it was
clear that aside from Llewellyn‘s endorsement of “situation sense”, some social science
approach had to matter.18 Understating the point a bit, if anything did not stand out
from Kelsen’s writings then this was the idea that legal scholarship ought to be turned
into a social science. Indeed, the whole business of clearing up conceptual muddle may
have created the false impression that Kelsen was obsessed with topics of legal formal-
ism.19 And, indeed, in many respects Kelsen’s project is deeply steeped in nineteenth
century German constructivism.20
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Ironically, only the least interesting parts of Kelsen’s theory have caught the atten-
tion of most Anglo-American legal theorists, notably the Grundnorm.21 The career of
this stillborn brainchild is testament to the lack of interaction between and among the
worlds of legal scholarship of the mid twentieth century.

Demystifying legal norms

There is, finally and lest we forget, a third level of demystification that we can associate
with legal positivism. It concerns the law itself and should not be confused with the
first project of demystification that takes aim at how legal institutions generally adorn
themselves with dignity and purport to pursue high-minded aspirations. Rather, at this
level, demystificiation concerns the substance of laws.
How can this be? The critical analysis of legal institutions can become the task of le-
gal theory where the law itself, either mediated by the authority of courts or by means
of legislation, comes to incorporate an ideologically distorted understanding of itself.
Joseph Raz observes with fine irony that only bad theory can make a practical differ-

17 Felix Cohen‚ Transcendental Nonsense and the Functional Approach, Columbia Law Review 35, 1935,
808–849.
18 Laura Kalman, Legal Realism at Yale, 1927–1960, Chapel Hill: University of North Carolina Press, 1986.
19 Karl N. Llewellyn, Jurisprudence: Realism in Theory and Practice, Chicago: University of Chicago Press,
1962.
20 Stanley L. Paulson, Hans Kelsen’s Earliest Legal Theory: Critical Constructivism, Modern Law Review
59, 1996, 797–812.
21 The final word appears to have been spoken on this topic now. Stanley L. Paulson, The Great Puzzle:
Kelsen’s Basic Norm, in: Kelsen Revisited: New Essays on the Pure Theory of Law (ed. L. Duarte d’Almeida, J.
Gardner & L. Green), Oxford: Hart Publishing, 2013, 43–62.

Franz Steiner Verlag


The Demystification Impasse 51

ence.22 For example, the law may require us to sustain differences between private and
public law. While there is, according to Kelsen, no manner in which such difference can
be convincingly sustained, one cannot but muddle through with makeshift conceptu-
al arrangements simply because the law has forced the contrast upon us. Even whole
fields of law may rest upon false understandings of their function and their point. For
example, a careful analysis of the core category of anti-discrimination law, direct dis-
crimination, reveals that this field of law necessarily serves a redistributive function.
Alas, European anti-discrimination law does not recognize this fact.23 Kelsen’s critique
of the law of the United Nations was an attempt to expose the dearth of legal expertise
extant in the drafting process and the influence of phony ideas on the substance of an
international agreement.24
Ironically, thus, misapprehension of the law can be built into a very field of law.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

The real normativity of law

The remaining discussion will focus on the avowed goal of the second level of demys-
tification, namely that which by all accounts is supposed to be the enlightened appre-
hension of the law. When it comes to this, we encounter two major approaches that
have been developed by legal positivists in the twentieth century. They coincide, not
by accident, with its most salient strands, one Anglo-American, the other continental.
If one wanted to attach names to these approaches they would read “Hart” and “Kels-
en”, respectively.
Generally, any project of demystification needs to have at least a rough idea of what
the law is supposed to be like once mystifications have been removed and lost their
grip on the legal imagination. This means that such projects invariably need to commit
themselves ontologically to a certain view of what the law really is. At a most general
level, both strands of modern legal positivism agree on two things. First, the law is not
merely “patterns of behaviour” but rather manifest in certain directives that are called
either “rules” or “norms”. Second, whether or not such directives are authoritative is
to be determined with reference to social facts. Hence, both versions of legal positiv-
ism agree that claims to legal validity need to be warranted, to the greatest possible

22 The original lines read as follows: „Theory aims at understanding. By and large, only bad theory can lead
to change. If its wrong conclusions are accepted their acceptance may lead to a change in the self-under-
standing of the culture which will make the bad theory true.” Joseph Raz, Between Authority and Interpreta-
tion: On the Theory of Law and Practical Reason, Oxford: Oxford University Press, 2009, 98–99.
23 I, for one, have made an attempt to analyze Anti-Discrimination Law from such an angle. See Alexander
Somek, Engineering Equality: An Essay on European Anti-Discrimination Law, Oxford: Oxford University
Press, 2011, 19–20.
24 D. A. Jeremy Telman, Law or Politics? Hans Kelsen and the Post-War International Order, Constellations
18, 2011.

Franz Steiner Verlag


52 ALEXANDER SOMEK

extent, by propositions pointing to some law-creating factor that can be described in


value-neutral terms.25
In addition to the separability thesis, this emphasis on the factual is undoubtedly
the second defining mark of legal positivism.26 Remarkably, however, it also points us
to a type of social normativity that inheres in the law and indicates how it differs from
the persuasive force of reasons for action. Evidently, in the case of Austin, designating
the factual connotes the overpowering force of necessitation. The conception of law
as the command of the sovereign suggests strongly that the commander’s power to
have his or her way is insuperable. In the case of Kelsen, the overall effectiveness of the
coercive legal order, to which reference is made in the conditional part of the Grund-
norm, represents something akin to a force of nature. Hart’s concept of a “social rule”
insinuates that there is no way for a single person or a few forlorn dissenters to stem the
tide of convergent behaviour. The factual that gives rise to law in the guise of sources
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

originates from an alternative to the normativity of reasons and replaces the “non-co-
ercive force of better arguments” (Habermas) with the stubbornness of social power
or mindless routine. If a source lacked this nexus to might it would never merit the
attention of legal positivists. That’s right. In their eyes, anything claiming to be merely a
better argument could never pass as a source of law.27

The transformation of reasons into the having of reasons

In the eyes of Hartian Anglo-American legal positivists, real law is manifest in legal
enactments – notably rules – the validation of which is derivative of a social rule that is
accepted as the relevant (law-identifying) standard by law-applying officials.28 By con-
trast, from Kelsen’s perspective, real law is to be found in norms that originate from –
and belong to – a dynamic system of law-creation. The key role within such systems is
played by legal powers. They are the real real laws in the universe of the Pure Theory of
Law.29 As we shall see, they are so for good reasons.
In spite of different ontological commitments, it is the case that upon encounters
with idealizations both legal positivisms carry out a transformation that is followed by
a question concerning authority.

25 Raz (footnote 3), 47.


26 Raz (footnote 3), 37, where the “social thesis” is presented as saying that “what is law and what is not is
a matter of social fact”.
27 It may be doubtful whether this is true also of inclusive legal positivists.
28 H. L. A. Hart, The Concept of Law (2d. ed.), Oxford: Clarendon Press, 1994, 102–103.
29 For an alternative elaboration of the point of the pure theory that puts the reinterpretation of acts of
violence in term of sanctions at the centre, see Christoph Kletzer, The Idea of a Pure Theory of Law, Oxford:
Bloomsbury Publ., 2018.

Franz Steiner Verlag


The Demystification Impasse 53

The task of transformation consists of presenting an appeal to practical reason as


a social fact. Reason can be translated into something factual30 only if the having of a
reason is treated as a state of belief (which may possibly be caused by some other fact).
Hans: This is wrong.
We: Hans believes this to be wrong.

Legal positivism distinguishes itself from other approaches to law by recasting reasons
for action as beliefs in such reasons and, in addition, attributing these beliefs to some-
one who has them. This is not to say that this transformation already provides us with
the full picture, yet it captures adequately its demystifying ethos.
Hans: You ought to do this.
We (1): Hans believes that we ought to do this.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

We (2): Hans wants us to do this.


We (3): Hans wants us to do this because he wants it.
We (4): If Hans has authority over us we have reason to do what Hans wants us to do
because he wants it.

The transition from (1) to (2) is possible because Hans believes we have to do this. If
Hans really believes31 that we ought to do something then he wants us to do something.
Emphasizing the having of reasons for action, that is, a having of reasons regardless of
their content, accounts for the transition form (2) to (3). This is the consequence of
transforming a belief into a social fact. It marks the major step taken by legal positivists
by shifting from a content-dependent to content-independent perspective.32 What Raz
calls the “social thesis”33 turns out to be an implication of the separability thesis. If the
reason for action is not derivative of the full and unrestrained examination of pros and
cons it has to be somehow tied to a fact that is internally indifferent to reasons (for that
is what we mean by “content-independent”, aside from the fact that we have reason to
do something because somebody else has said so). Finally, (4) merely points to the
normative premise that is needed in order to invest the social fact of wanting with
authority. Transforming moral claiming into content-indifferent wanting or endorsing

30 Facts can be reasons, but reasons are not facts.


31 I believe that in an ideal world men should wear at least business attire, if not tuxedos, when they attend
an operatic performance. I believe that in an ideal world this should be so, and I am aware of the fact that it
is never going to be that way in our real world. Actually, I also believe that in the real world people shouldn’t
be forced to wear a suit and a tie if they do not want to. Nevertheless, I believe that in a counterfactual world
it should be the case that people adhere to a dress code. Such a belief does not fall into the category of the
“real” belief that is of relevance here. By contrast, I also believe that people should not kill each other. I do
not want them to kill each other. This is a real belief.
32 It is clearly the step that calls for the justification of authority. See Raz (footnote 3), 214 (on the normal
justification thesis).
33 Raz (footnote 3), 37.

Franz Steiner Verlag


54 ALEXANDER SOMEK

is the essence of the positivist project of demystificiation. Not by accident, it simulta-


neously raises the question of legitimate authority.

Anglo-American demystification

To your claiming that morality enjoins us from doing or omitting something legal pos-
itivists reply that all that we really know, if you are sincere, is that it is you who believes
in this moral demand. A high-minded moral claim is thus brought down to the realm
of social facts.
At the outset, it was mentioned that demystification aims at removing idealizing
sugar coatings from the legal materials. Here is an example of such a sugar coating,
variations of which can be encountered relatively frequently in the literature.34
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Mystification: Human dignity is at the heart of all modern constitutional systems. No-
where can a law violating this ideal pass constitutional muster.

The demystification that Anglo-American legal positivists would engage in would


roughly amount to the following:
Anglo-American demystification: The relevance of appeals to human dignity is based upon
a social rule that either accepts appeals to human dignity as appeals to valid law or permits
(or even requires) extra-legal standards to determine judicial decisions.

Insiders will immediately realize that the formulation attempts to cover (“incorporate”)
both inclusive and exclusive legal positivism.35 In both cases, an open set of substantive

34 Shapiro (footnote 4), 270; Raz (footnote 22), 193.


35 Shapiro (footnote 4), 273–274. “Inclusive” legal positivism stands for the view that moral principles can
be binding on legal officials because they are valid law. Such principles are taken to be valid law if they or any
other appeal to moral merit is recognized as a source of law within the rule of recognition. “Exclusive” legal
positivism, by contrast, insists that the morality that is thereby appealed to is not incorporated into the legal
system. A norm can count as law only if it has a social pedigree and can be identified without engaging in,
or recourse to, a moral argument. For extensive discussion of these issues, see Jules L. Coleman, The Prac-
tice of Principle: In Defence of a Pragmatist Approach to Legal Theory, Oxford: Oxford University Press,
2001, 106–109, 153. It appears to be Shapiro’s take on exclusive legal positivism that it requires judges to take
moral principles as moral principles into account. See Shapiro, note 4 at 272: “[…][T]he fact that American
judges are simply under a[] [legal, A. S.] obligation to apply nonpedigreed norms does not imply that they
are compelled to apply preexisting law; rather, they are merely under an obligation to reach outside the law
and apply norms of morality instead.” It should be noted that, at least more recently, Raz has conceived of
exclusive legal positivism differently. According to him, judges – as human beings – are always under an ob-
ligation to be guided by moral standards, for this is what it takes for human beings to act appropriately. If the
law has authority it is morality that carves out space for content-independent reasons – that is, rules that are
derivative of the authority of law – and thereby precludes a determination of conduct on the basis of a full
examination of the moral merits of the situation. Morality is that which “incorporates” the law, and not the
other way round. Therefore, it would be rather odd to believe that morality first incorporates law in order
to incorporate itself subsequently into what it has itself incorporated. It is much sounder to assume that the

Franz Steiner Verlag


The Demystification Impasse 55

moral arguments (called, euphemistically, “non-pedigreed norms”) is considered to be


legally relevant on the ground of either its “incorporation” into the rule of recognition36
or an explicit or implicit “validating purport”37 by the law with regard to this set.38
The important point is, at any rate, that Anglo-American legal positivists transform
the claims that moral standards make on us into a de facto shared and practiced belief
that standards of this type – no matter which there are39 – are relevant for the adjudi-
cation of legal disputes. Whether or not the belief in this relevance is accepted as law
depends on whether it guides the practice of legal officials and that they would be criti-
cised for committing mistakes if they failed to take these standards into account. In oth-
er words, the acceptance of moral standards is manifest in a rule-guided social practice.
It follows that the awareness of this social fact is supposed to provide us with a de-
mystified account of moral reasoning in the legal context. The aura of natural law dis-
appears. We grasp that the social fact of acceptance is that which confers relevance to
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

arguments that appear to invoke “natural law”.


Dworkin has pointed out vis-à-vis Coleman’s variety of inclusive legal positivism
that what is remarkable and puzzling about this form of demystification is that it leaves
everything as it is.40 Practice does not have to alter a bit. Indeed, it is essential for such

law, when referring to reasons of morality, makes various exemptions from the general exclusion of judging
matters on their moral merits. The “inclusion” of morality is rather a partial revocation of its self-limitation.
See Raz (footnote 22), 188–190.
36 Coleman (footnote 35), 107, distinguishes between the grounds of the criteria of legal validity and these
criteria themselves. While according to inclusive legal positivism the relevance of the criteria is grounded
in a social convention (i e , a social fact) “[…] the criteria themselves need not state social facts”. In the case
of human dignity mentioned in the example above this means that the relevance of appeals to “human dig-
nity” is conventionally established, even though there is no social rule or other set of facts that determines
what human dignity means, let alone demands, in singular cases. The meaning is relevant even though it is
up for grabs. One should not be surprised that one encounters in Coleman’s works statements such as the
following (116): “Judges may agree about what the rule is and disagree with one another over what the rule
requires.” It is puzzling to imagine an agreement over what courtesy is and disagreement over what it re-
quires in every single instance. We would call this type of agreement a merely seeming agreement. Coleman
is aware of this, of course, and adds that judges could not disagree in every case or even in most cases. But
this implies that some of the criteria of legality or some applications of these must be conventionally given.
Inclusive legal positivism cannot eliminate the social fact element from the criteria altogether and restrict it
to the formulation of rule as an empty shell.
37 I take this from Hart’s discussion of dualism in international law. See H. L. A. Hart, Kelsen’s Doctrine of
the Unity of Law, in: Normativity and Norms: Critical Perspectives on Kelsenian Themes (ed. S.L. Paulson &
B. Litschewski Paulson), Oxford: Clarendon Press, 1998, 553–581, at 560–561. The point of such validating
purport is to sustain the difference in the relation of the legal orders involved. Raz goes to some length to
point this out. See Raz (footnote 22), 193.
38 I shall leave aside, for the moment, Raz’s more sophisticated alternative to incorporationism which
treats the relevance of morality as an exclusion of exclusionary reasons.
39 This is a point that is repeatedly emphasized by Coleman. See merely Jules Coleman & Brian Leiter,
Legal Positivism, in: A Companion to Philosophy of Law and Legal Theory (ed. D. Patterson, 2d ed.), Malden:
Wiley-Blackwell, 2010, 228–248 at 238; Coleman (footnote 35), 116.
40 Ronald Dworkin, Justice in Robes, Cambridge, Mass.: Harvard University Press , 2006, 207.

Franz Steiner Verlag


56 ALEXANDER SOMEK

inclusive legal positivism not to assimilate morality to mere conventional standards.41


It is not the case, therefore, that inclusive legal positivists claim that the moral appeals
that we encounter in a legal contexts necessarily invoke conventional morality. Rather,
these appeals are intended to be direct appeals to what is morally right or wrong. What
demystification accomplishes is to highlight that their legal authority has its root in a
conventional social foundation.
They: In hard cases we have to invoke moral standards.
We: They observe a convention that requires them to invoke moral standards in
hard cases.

As Dworkin correctly observed, there is evidently something amiss here.42 The inclu-
sion of moral appeal into the sources of law does in no manner contribute to legal cer-
tainty. The import of moral standards is not at all uncontroversial. Very likely there will
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

be disagreement concerning their relevance or their proper application.43 The atten-


tion that legal positivists pay to social facts, however, would require identifying those
who are actually calling the shots when it comes to applying the convention that says
that the morally correct outcome counts. The talk of a coordinating convention – or
a shared co-operative activity44 – may in fact conceal relations of power amongst the
law-applying officials. The very idea that the identification of valid law is based on a
social rule to the application of which every official is an equal partner may reflect
a moderate determinacy of application or a “settledness” of law that does not stem
from the solidity of the conventional standard but from effective relations of subordi-
nation.45 In a field where disagreements and controversies are not unlikely to abound,
potential dissenters are hushed into acquiescence with subtle means of co-optation or
by involving them in rituals at the end of which they can hope to pass as initiates.

The predicament

But the problem cuts even deeper. Moral arguments that do not invoke conventions
(“We do not…”) aim at giving an answer that does not only state – report and reas-
sert – a belief from a particular first person plural perspective. They want to be correct
tout court. Hence, they have to base themselves on conceptions of the morally most
attractive understandings of certain terms, such as human dignity, or of moral ideas,

41 Coleman (footnote 35), 127.


42 Dworkin’s objections are usefully summarized in Coleman & Leiter (footnote 39), 237–238.
43 See also Shapiro (footnote 4), 276.
44 Coleman (footnote 35), 96–98; Shapiro (footnote 4), 204–208.
45 See the observations in Alexander Somek, The Legal Relation: Legal Theory after Legal Positivism, Cam-
bridge: Cambridge University Press, 2017, 35.

Franz Steiner Verlag


The Demystification Impasse 57

such as impartiality. Arriving, if necessary, at such a conception involves moving be-


yond that which is merely conventional. While, for example, we may regard the eating
of meat as morally impeccable it may nonetheless be true that upon reflection the prac-
tice turns out to be cruel. A distinction emerges, hence, between what is convention
and what is substantively correct. From the point of view of the latter the truth of a
moral claim does not depend on communis opinio or convergent behaviour among the
members of an in-group.
It follows that everyone who is confident enough to claim to know what is morally
right may legitimately assert what he or she believes to be the moral truth. In relation
to existing conventions this means that either everyone would have to be entitled to
say what the law is or, alternatively, nobody would have to be. If this right had to be
conceded to everyone we would end up with a situation of radical legal pluralism. It
would amount to an equivalent to Hobbes’ right to everything. Legally speaking, who-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ever claims to know the right answer would have to have the power to lay down the
right answer as law, for otherwise the inclusion of morality into the criteria for law
would not make any sense. Alternatively, everyone could be denied this right. In both
cases, however, the result would be radical legal indeterminacy. This result would be in-
consistent with what legal positivism sets out to do, for there would be no more social
fact left determining what accounts for the existence of valid law.
This is, very briefly put, the major problem of inclusive legal positivism. It seeks to
accommodate the non-conventional within the conventional. To the convention that
has the form “This is what we do, and I do what all others do” it adds “What we do is
to have each determine for him – or – herself what we do”. What one gets, then, is a
“rule” saying that what we do is to have each determine for all what we do. If each has
the power to determine for all what the law is there can be no stable social fact. The
convention that allegedly incorporates appeals to morality into a rule of recognition
does not give rise to facts that are capable of generating law. They are “fake facts”. The
convention turns out to be a self-denying ordinance.

Turning to continental demystification

Inclusive legal positivists may want to reply that the above analysis is uncharitable and
transforms their position into something outright absurd. The legal order allocates
powers to say what the law is in a hierarchical manner and confers such a power in the
final instance to the highest courts and not to lay persons.46 In the event that a conflict
arises between their respective judgments the view of the court must prevail.

46 I take it that this reply is similar to Coleman’s attempt to rebut an objection by Dworkin (footnote
35), 157–158. Dworkin pointed out against Coleman’s construction of inclusive legal positivism that if the
content of the criteria of what is valid law are not conventional the law applying officials might never agree

Franz Steiner Verlag


58 ALEXANDER SOMEK

If this is the reply given by “incorporationists” then they have already conceded that
what matters, from a legal point of view, is the allocation of powers to state authorita-
tively what morality demands and not what morality itself requires.47 They have there-
by shifted, tacitly, to another mode of demystification that no longer anchors cloudy
appeals to values or ideas in conventions for making such appeals but rather in con-
crete powers to make them stick. This is the continental way.
Continental demystification: The courts invested with the power of judicial review of leg-
islation have discretion to attribute any meaning to human dignity as they see fit and to
invalidate laws on that ground.

A different picture emerges here, and one is inclined to call it true – as opposed to
existing – exclusive legal positivism. True (or truly) exclusive legal positivism insists
on translating the determination of legally relevant moral claims into the exercise of
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

the legal power to say and to lay down what the law is.48 This, at any rate, is consistent
with a Kelsenian mindset, which appears to reflect the original critical spirit of legal
positivism. From this perspective, the alleged convention underpinning genuine ap-
peals to moral values always has be to read as a rule authorizing such appeals. Appeals
to “non-pedigreed norms” are made within the confines of powers that provide them
with a legal pedigree.49

on what these are. This is the equivalent of the observation made in the text that the quest for the morally
right answer does not commit anyone to arrive at an answer that elicits the consent of all. To this objection
Coleman replies that the ground of the inclusion of morality, the rule of recognition as a social rule, is best
understood to be a shared cooperative activity. While such an activity may give rise to a number of disa-
greements it commits law-applying officials to settle disagreements and to converge on certain views. This
reply makes it plain that morality is not relevant per se but only subject to the norms governing the shared
cooperative activity (on these norms, see ibid 96–99).
47 Coleman’s idealized account of the rule of recognition as a shared cooperative activity of course must
raise the suspicion of mystifying relations of social power.
48 The difference to Raz’s exclusive legal positivism should be duly noted. From Raz’s point of view judges
are under a moral obligation to arrive at morally defensible resolutions of legal questions. The law merely
modifies the way moral obligations affect judges. It is, however, not easily to parse from Raz’s text how this
is supposed to work. See Raz (footnote 4), 196: „References to moral considerations in constitutions are
typically not cases of the incorporation of morality but blocks to its exclusion or modification by ordinary
legislation. […] Judicial review not only makes the block to the exclusion of modification of constitution-
ally protected moral considerations by legislation enforceable; in addition in [sic] conferring on the courts
powers to enforce that block, it gives them, when adjudicating on the compatibility of legislation with
the constitutionally protected moral considerations, the power to modify the application of those moral
considerations themselves.“ It does not emerge clearly what such a modification is supposed to consist
in. Moreover, while Raz’s idea concerning the partial exclusion of the exclusion of moral consideration is
undoubtedly very elegant, it is doubtful whether it can fully account of the idea that judges are guilty of
committing legal – and not merely moral – mistakes if the do not take morality into account. Why should
the law make morality legally relevant if all that it does is to cede the ground back to morality?
49 I take it that exclusive legal positivists would agree with that view.

Franz Steiner Verlag


The Demystification Impasse 59

In 1926, in the course of a discussion at a meeting of the German public law teachers’
association, Hans Kelsen memorably put this basic demystifying thrust of legal posi-
tivism in the following famous words:50
The question that natural law theory is after is the eternal question of what lies behind pos-
itive law. Whoever tries to find an answer, I am afraid, discovers neither the absolute truth
of a metaphysics nor the absolute justice of a natural law. Whoever lifts the veil and does
not close his eye is going to be stared at by the gorgonian head of power.

Power is the fabric that legal relations are made of. The only idealisation that a non-re-
ductive form of legal positivism cannot dispense with consists of viewing the exercise
of power as authorized by power-conferring norms. These norms can be reconstructed
by interpreting an effective coercive order as composed of social practices that find
their organising core in abiding by legal norms.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Another predicament

But this is not the end of the story. The recurring question within “dynamic” legal sys-
tems that govern their own creation and application is under which conditions a legal
power is indeed successfully exercised. The question is, in other words, whether offi-
cials exercising their power do so truly or merely seemingly. If they do not exercise it
truly, their acts have to be regarded as void.
Interestingly, continental legal positivism encounters at this point a predicament
that is strikingly similar to the impasse of inclusive legal positivism. The question is
whose interpretation concerning the exercise of legal powers counts. Since the days
that Rex v. Hampden was decided,51 this has been the arch-constitutional question:
Quis iudicabit?52
It may appear that the solution to this question has to parallel the way out of the
inclusive legal positivist’s predicament. The legal order allocates powers to settle ques-
tions concerning the scope of powers. Ultimately, such a power to interpret powers
may be vested in a court. Yet, even when the whole legal order becomes construed as
a set of interlocking power-conferring rules the question remains whether those to
whom the system designates the task to serve as the ultimate interpreters of powers
have used their powers of interpretation excessively or not. It follows that, just as in

50 Hans Kelsen, in: Veröffentlichung der Vereinigung deutscher Staatsrechtslehrer 3, 1927, 54–55 (my transla-
tion).
51 Rex v. Hampden, 3 State Trails 862, 1637, partly reprinted in: The Stuart Constitution ,1608–1688, (ed., J. P.
Kenyon, 2d. ed), Cambridge: Cambridge University Press, 1986, 98–103.
52 Thomas Hobbes, Leviathan (ed. I. Shapiro), New Haven: Yale University Press, 2010, 166 (chapter 26);
Carl Schmitt, Der Begriff des Politischen: Text von 1932 mit einem Vorwort und drei Corollarien, Berlin: Dunck-
er & Humblot, 1963, 122.

Franz Steiner Verlag


60 ALEXANDER SOMEK

the case of claiming moral insight, everyone has equal standing when it comes to adju-
dicating the question whether the ultimate interpreters exceeded their powers (acted
ultra vires) In particular, there is no reason to assume that legal officials are in a bet-
ter position to say what is within an organ’s power than uninvolved and disinterested
scholars adopting the internal perspective of participants in the legal system. There-
fore, the question must arise that has been explored by Adolph Julius Merkl already a
century ago.53 The powers of government can be rendered differently from the perspec-
tive of the real insiders using the system and of outsiders assessing their behaviour by
adopting, apprehensively, the internal, critically reflective attitude. Depending on the
perspective, the legal system can be given different “faces”. The question which of the
two faces shows us the law as it really is?
Leaving aside the fierce debate that this question gave rise to among members of
the early Viennese school,54 the underlying problem can be stated as follows: The legal
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

power and the act exercising the power need to match somehow. Loosely put, the legal
power has to be comfortable with accepting an exercise as its own; the exercising act,
in turn, must be convinced that the legal power is right for what it wants to achieve. The
determination is never a one-way street. Rather, just as the power, properly interpret-
ed, needs to recognize the act that wishes to be covered by it, the act needs to recognize
the power as appropriate to its end.
Unsurprisingly, the only media available in legal systems that facilitate such mutual
recognition are the social conventions of interpretation. This point is more important
than it may seem. Unless conventions that guide the interpretation of legal rules are
practiced and provide guidance to those operating the system the normativity under-
lying the effective order could not exist. Effective legal orders are effective legal orders
only if they give effect to norms or rules. The normative quality of these norms or rules
has to be manifest in the behaviour of officials who share views as to what these rules
require, at any rate in standard cases. Without the existence of such views – shared
conventions – no legal system could ever exist and legal scholars would have nothing
to describe.
No legal system could ever do without conventions, for their existence logically
precedes and mediates the existence of law. This explains, however, why the views of
the law applying officials have priority over the views by critically reflective scholars.
Their influence accounts for the effectiveness of the overall system. Without their views
there would be nothing for scholars to describe. If a social practice merely consisted
of acts of power that do not observe any rule-like constraints it would not amount to a
legal system. While legal scholars may believe that their understandings of the limits of

53 Adolf Julius Merkl‚ Das doppelte Rechtsantlitz. Eine Betrachtung aus der Erkenntnistheorie des Rechts,
1918, 47 Juristische Blätter 425–427, 444–447, reprinted in: Die Wiener rechtstheoretische Schule (ed. H. Klecat-
sky, R. Marcic & H. Schambeck), Vienna: Europaverlag, 1968, vol. 1, 1167–1201.
54 On this see a forthcoming contribution by Rodrigo Cadore.

Franz Steiner Verlag


The Demystification Impasse 61

legal powers ultimately count these understandings are necessarily derivative of, and
secondary to, the conventions that are practiced. Practice has priority over theory. This
is the demystification that Anglo-American legal positivism has in stock for continen-
tal legal positivists that are often inclined to take for granted that the rules allocating
powers are somehow “objective”, possibly even abstract entities amenable to detached
description from some theoretical point of view. On the contrary, if the practice view
of meaning is correct then the meaning of rules resides somehow in their routine ap-
plication.55

Conclusion

Apparently, we end up, thus, with a relation of mutually assured deconstruction. We


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

are looking at a demystification impasse.


Anglo-American demystification re-mystifies. It conceals the indeterminacy that in-
heres in conventionalism. The demystification can be overcome by shifting the focus
from conventions to legal powers.
But continental demystification is also guilty of re-mystification, for it fails to recog-
nize that the limits of legal powers become originally drawn in convention-based prac-
tice. Law as a system of interlocking power-conferring rules presupposes the efficacy
of conventions.
The source of all this bewilderment may well be that legal positivism has never ex-
plained the “factness” of the social facts that are supposed to function as legal sources.
But that is a different story the answer to which leads us already beyond legal positiv-
ism.
It will be remembered what the decisive step is (see above p. 52–54). It consists of
transforming belief into the fact of belief or, alternatively put, of transforming judg-
ment into the fact of judgment. This is something that we ordinarily do in the case of
moral disagreement. We allow others to go forward with what they happen to believe.
It is from such mutual recognition of judgments that legal relations arise. That the law
is based upon sources is a reflection of this way of relating to one another.
This indicates that the perplexity surrounding legal positivism can be overcome by
altering our view of the sources. Legal relations allow someone to have a say. That’s
the whole secret of authority. Sources of law are not primarily social facts but embodi-
ments of subjectivity. They provide us with knowledge of the law. Social facts are rele-
vant to how the law is known in the forms of sources.

55 Dennis Patterson, Law and Truth, New York: Oxford University Press, 1996. For a general restatement of
this view, see Christoph Menke, Die Souveränität der Kunst Ästhetische Erfahrung nach Adorno und Derrida,
Frankfurt a. M.: Suhrkamp, 1988, 223.

Franz Steiner Verlag


62 ALEXANDER SOMEK

Legislation, for example, in virtue of resting on a decision, recognizes what the law
is with an eye to what has been decided. Knowing the law from the perspective of leg-
islation, therefore, requires adhering faithfully to the text of a statute or to its original
meaning. The pertinent maxims of interpretation mark what it takes to know the law
“legislatively”. By contrast, the law is known in a manner congenial to customary law
if claims are made that “this” or “that” is not what we are ready to condone here. If the
law is known “customarily” this knowledge is replete with appeals to common under-
standings. Finally, law that is known “scholarly” comes about in appeals to a different
form of fact, actually to argumentatively established differences in situations that are
supposed to warrant the application of the appropriate legal rule. The law emerges,
ideally, from a better argument. It is indeed only at this point that the law transcends
the domain of the factual, however, only in order to become reabsorbed by it in the
case of decisions.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

The forms of legal knowledge that are inherent in sources epitomize varieties of sub-
jectivity. They, in turn, make a legal consequence conditional upon something factual.
The sources say “you ought to” because “we see it that way”, “we have decided it that
way” or “the stronger argument requires that you follow this rule”. The law flows from
common understandings or decisions or, finally, from better views concerning what is
to be considered the applicable rule.
I can only posit here that by carrying this view of sources forward the predicaments
of legal positivism can be finally overcome.

Alexander Somek
Institute of Legal Philosophy, University of Vienna, Schenkenstraße 8–10, 1010 Vienna,
Austria, alexander.somek@univie.ac.at

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence

CHRISTOPH KLETZER
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: One important, yet still under-theorised aspect of the Pure Theory of Law is its
treatment of what Kelsen called the ‘legal technique’, ie an analysis of how exactly the law
does what it does. In this paper I try to outline both the specific intermediary theoretical
location of the legal technique and its decentralised nature by comparing the technique of
the law with another one: the technique that allows us to transfer data on the internet, the
so-called internet protocol suite. This will help us, I believe, in getting a clearer understand-
ing of the different levels or layers of analysis and keep separate what is done from how it is
done. After having determined the exact yet often misallocated level of some of the most
relevant insights of the Pure Theory I will then analyse the specific architecture of what I
call the legal stack. This is followed by a discussion of the operation of the legal protocol
and an analysis of the nature of its decentralisation. I argue that this stack analysis helps us
see more clearly that (i) the law is ultimately always a matter of the use of force; that (ii) the
law operates not by simply telling people what to do, but that it tells people what to do by
conditioning the use of force; and (iii) that whilst the use of force might be fundamental
to the legal system, it is not its most significant part: what is more significant than the use
of force is all the things the law does by means of this use of force.

Keywords: Hans Kelsen, decentralisation, networks, protocol, primitive law, Pure Theory
of Law

Schlagworte: Hans Kelsen, Dezentralisation, Netzwerke, Protokoll, primitives Recht,


Reine Rechtslehre

When thinking about the relation of the world of facts and the world of the law most
of us, jurisprudes, legal experts and laymen alike, work from a model of detached de-
pendence: we take both worlds to be inextricably entangled, yet conceptually (or at
least provisionally) separable from one another. Most contemporary jurisprudential
theories can be plotted somewhere on the spectrum marked out by the extremes of

Franz Steiner Verlag


64 CHRISTOPH KLETZER

pure detachment and pure dependence: on the extremes we have a natural law theory
based on divine revelation, stressing the complete detachedness of the law from this
world and its determination by another, transcendent world, on the one hand, and a
reductive, sociologist realism, highlighting the complete dependence of law on actual
facts and the merely illusionary or ideological character of any detachedness or inde-
pendence of the law, on the other. Legal Positivism of a Hartian tint, for instance, sits
somewhere between those extremes in that it takes the law to be both a social fact and
a normative phenomenon.
Now, whilst this dualist ontology is certainly plausible and to an extent helpful it has
the disadvantage of hiding what I believe to be a key element of a full understanding of
the law, namely, the peculiar mechanics of the relationship of the two worlds.
This mechanism is what in the context of the Pure Theory of Law has been called
the legal technique.1 The analysis of this legal technique has been one of the main, yet, I
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

believe, under-appreciated, contributions of the Pure Theory of Law: the Pure Theory
aims to enlighten us about the law not primarily by investigating what its essence or
nature is but by analysing how exactly it works.
In what follows I want to highlight the attractiveness of this approach.2 I try to do
this by discussing a feature of the legal technique which has so far been, I believe, large-
ly overlooked: the decentralised nature of the legal technique. In order to get a better
grip of both the specific intermediary theoretical location of the legal technique and
its decentralised nature, I want to start off by comparing the technique of the law with
another one: the technique that allows us to transfer data on the internet, the so-called
internet protocol suite. This will help us, I believe, in getting used to manoeuvring be-
tween different levels or layers of understanding and keep separate what a certain phe-
nomenon does from how it does it.
After that I will apply the insights to the law and analyse the architecture of what I
call the legal stack. This is followed by a discussion of the specific working of the law
and an analysis of the nature of its decentralisation.

A. Decentralisation

It is a curious fact about the internet that one cannot easily switch it off. A government
might be able to block its domestic users from accessing certain online content, but no
single actor could feasibly switch off the internet altogether. Why is that? The answer

1 See, for instance, Hans Kelsen, ‘The Law as a Specific Social Technique’ 9 University of Chicago Law Re-
view 1941: 75–97.
2 Many of the jurisprudential arguments presented in this paper have already been defended, more formal-
ly and extensively, in previous publications. See Christoph Kletzer, ‘Primitive Law’ in 4 Jurisprudence 2013:
263–72 and The Idea of a Pure Theory of Law (Oxford: Hart Publishing) 2018.

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 65
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Figure 1

is not as obvious as one might think. After all, if one considers the architecture of a
typical communication network, it would be rather easy to shut it down: when, say,
my grandfather staying in the Bristol wanted to call his friend at the Sacher Hotel, his
call had to be relayed via a couple of switchboards: the hotel-switchboard gave him an
outside line and connected him with the city-switchboard, which was connected to all
the hotels in the city. This city-switchboard then connected him to the switchboard of
the Sacher, which in turn was connected with all the rooms in the Sacher and which
connected him with his friend.
Take network A from Figure 13 to illustrate the city telephone network of Vienna in
the Thirties where each station represents a hotel or a private telephone station. The
central node is the city switch-board that connects these stations. It is clear that in
such a set-up one would only have to take out the central node to make any telephone
communication between hotels or private stations impossible and to thus shut down
the entire city network.
Now whilst the problem is easily identifiable the solution is less so. After all, this
star-shaped, hub-and-spokes network architecture is not entirely accidental: we do
need some method to keep an authoritative record of which stations are connected to
the network and how to reach them. This essential function is provided by the central
switchboard. Without it, it seems, there would be no operative network to start with. It
is thus the central switchboard that represents the unity of the network.

3 Diagram from Paul Baran, ‘On Distributed Communications Networks’ P-2626, 1962, RAND Corpora-
tion.

Franz Steiner Verlag


66 Christoph KLetzer

Yet, at the same time the centralised architecture presents an unacceptable vulnera-
bility: telephone network downtime might have been an inconvenience for my grand-
father, for military communication it would have been a disaster, especially so in a
post-nuclear-attack scenario. The fact that a star-shaped US command and control net-
work would quite certainly be rendered inoperative by a Soviet first strike also meant
that the US would likely be incapacitated to issue a timely counter-strike. Within the
logic of the cold war, which was stabilised only by the assurance of mutual complete
destruction, this presented a potentially deadly game-theoretical itch: since there was
a high possibility that the US would be rendered incapable of retaliating in time the
only option left open to the US would be not to wait for a situation where it would
need to retaliate but rather anticipate any attack by striking first. This led to a danger-
ous finger on the trigger mentality. And all of this was caused mainly by unreliable
network design. A survivable second strike capability was needed to move down from
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

a guns loaded, hair-triggered command and control doctrine.


It was within this context that Paul Baran in 1960 presented a series of papers show-
ing that with sufficient network redundancy, ie with sufficiently many extra connec-
tions between the stations, a network can become surprisingly resilient to the taking
out of individual stations.4 Reliable networks can be built out of unreliable elements by
using redundancy. Baran called such a mesh-network ‘distributed’.5
Such a network architecture was, of course, a wonderful idea. However, in the
1960ies it was but an idea. After all, what was missing was the crucial bit: finding out
how to allow the stations to communicate with each other on such a network, ie with-
out there being a central node that relays the messages from the sender to the right
recipient. As mentioned before, the hub-and-spoke design of a centralised network
did not emerge by accident. Rather, the star-shaped setup seemed necessary to make
sure a message could find its way from one node to the other. Conversely, if there were
no central node controlling communications, opening and closing lines, connecting
servers, confirming transmissions and keeping records of all nodes, it seemed impossi-
ble to make sure a message from one part of the network finds its way to another part
of the network.
What solved this problem was the development of a so-called network protocol. A
protocol is a computer programme that uniformly runs on each station and tells each
station how to package, address, relay and confirm messages sent on the mesh net-
work. It is the uniformity of this protocol that allows a decentralised or distributed
network to operate. In a sense the protocol takes over the role of the central hub and
thus represents the unity of the network. There is thus a shift in the paradigm of what

4 Paul Baran, ‘On Distributed Communications Networks’ P-2626, 1962, RAND Corporation and ‘On Dis-
tributed Communications: Introduction to Distributed Communication Networks’, Memorandum RM-
3420-PR, 1964, United States Air Force Project RAND.
5 In this paper we will not follow Baran’s nomenclature but rather call them ‘decentralised’.

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 67

represents the unity of the network. We move from a master-unity paradigm to a proto-
col-unity paradigm. In the case of the internet this network protocol is called the inter-
net protocol suite or TCP/IP.
Let us have a closer look at where exactly in the logic of the network this protocol
operates and thus also where the important paradigm shift occurred. In computer sci-
ence it is common to think of computer systems to be a stack of so-called abstraction
layers operating on top of each other. In the case of the internet this stack would look
something like this:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Figure 2

At the bottom we have the physical link layer. It represents the hardware of a variety of
different servers at different physical locations linked to each other by cables or wire-
less radio connections. This hardware alone might form a physical net but it is not yet a
network. After all, data has to be able to travel from one server to the other.
On the very top we have the application layer. In this layer we find the internet ap-
plications we use every day, like the applications that allow us to read newspapers, ac-
cess our bank account, order books, stream movies, go through the publications of
colleagues and so on.
When we think of the internet, most of us think only about these two layers: we
know there must be some kind of physical connection and we know what we want to
do with this connection. However, as we have seen in the brief historical digression,
the truly crucial bit that is missing in this picture is neither of the two but rather what
lies in between: the so called data link, transport, and network layers, which is taken care
of by the TCP/IP. This protocol makes or establishes the actual communication net-
work. It represents the unity of the internet. In a sense it is the internet. Without it, you
would only have hardware laying around. Without it, you could not watch any movies,
order any books or read any newspapers, as no station would know how to communi-
cate with any other station. The top layer tells you what you do with the internet, the

Franz Steiner Verlag


68 Christoph KLetzer

bottom layer is where you do what you do on the top layer. The intermediary layer,
however tells you how to do all of that.

B. The Legal Stack

 regulating Law Application Layer What?

 Legal protocol
transport/Network
Layer
how?
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

 Legal Facts Link Layer Who?

Figure 3

Now, in the law, too, we tend to only think about two layers in a stack. On the one hand
we see all the things that the law does: it tells people what to do, it makes demands, it
regulates affairs, sets down taxes and so on. This can be called the application layer of
regulatory law. It is what most jurisprudes in the Anglo-American tradition of jurispru-
dence focus on exclusively.
On the other hand, there are the social underpinnings of these legal affairs: people
speaking to each other, judges pronouncing things, voters voting, and so on. This is the
world of the social and psychological forces into which the law is embedded, or on top
of which the law is built. Sociologists of law and legal realists focus exclusively on this
realm. Legal reductionists claim that it is the only realm that really exists.
But in the law there is, too, the intermediary layer, a layer corresponding to the trans-
port/network layers of the internet stack. It is, I believe, this layer that Hans Kelsen was
addressing when he was writing about the legal technique and it is this layer on which
the most significant discoveries and insights of the Pure Theory of Law have to be
located. Whilst the rule of recognition, for instance, is a device of the top application
layer of regulating law, the basic norm of the Pure Theory can only be understood as
operating in this intermediary layer.
The exact manner in which I understand the Pure Theory to model the operation
of the legal protocol will be discussed in the next chapter. I want to use the remaining
paragraphs of this chapter to say a couple of words about the architecture of the legal
stack as a whole.

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 69

This architecture can be understood under a loosely emergentist paradigm.6 Under


this paradigm the world is structured in higher and lower layers of existence where the
properties of the higher layers are grounded or anchored in the realities of the lower
level and in this sense do not exist independently of them. However, at the same time
and crucially they are not reducible to them.7
The relation between the layers can be described as follows: whilst a bottom layer
is more fundamental than the others, the top layer is more significant. Whilst the cable
network is more fundamental it is certainly not the significant part of the internet,
which can be found in the applications we are using every day. The applications of the
higher abstraction layer, however, are dependent in their running on the more funda-
mental physical network. We could never understand how the application layer oper-
ates if we were completely ignorant of the other layers building its foundations. If we
knew nothing about the physical network and the internet protocol, the applications
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

of the internet would have to seem rather miraculous.


Now, importantly, the internet protocol suite inhabits a wold between what is sig-
nificant but non-fundamental, on the one hand, and what is fundamental but in itself
in-significant, on the other. This middle layer is at the same time significant and fun-
damental.
The legal stack has a similar design. On the bottom there are fundamental facts
which are fundamental but taken by themselves legally in-significant. On the top we
have significant yet non-fundamental legal phenomena of regulative law.
As mentioned, most jurisprudes in their analysis of the law focus exclusively on the
top layer and relegate everything which does not inhabit this layer to legal sociology.8
By doing so, I believe, they miss out on a good chunk of significant and fundamental
insights of the intermediary level. By neglecting the more fundamental layers many
jurisprudes face the impossible task of explaining the operation of an application layer
which without reference to the intermediary protocol layer must seem rather miracu-
lous.
In trying to provide answers on this level of legal technique, the Pure Theory is nei-
ther trying to uncover the essence of the regulating law, nor its nature. Its object might
be more fundamental than the legal applications but that does not make it more sig-
nificant. Rather the opposite. The stack analysis tells us how exactly the law does what
it does. This, however, does not mean that what the law does can be reduced to this
operation.

6 Klaus Brunner and Bert Klauninger, ‘An Integrative Image of Causality and Emergence’ in Causality,
Emergence, Self-Organisation, ed Vladimir Arshinov and Christian Fuchs, (Moscow: NIA-Parioda, 2003),
23–35, 26.
7 David Blitz, Emergent Evolution: Qualitative Novelty and the Levels of Reality (Dordrecht: Kluwer, 1992),
179.
8 See, for instance, Joseph Raz, The Authority of Law: Essays on Law and Morality (Oxford: Clarendon
Press, 1979), 104–105.

Franz Steiner Verlag


70 Christoph KLetzer

C. The Legal Protocol

Now that we have made some progress in determining the level on which many of the
explanations offered by the Pure Theory operate, let us go ahead and try to find out
what exactly this legal protocol according to the Pure Theory of Law is. On the top lev-
el the law issues demands, regulates conduct, creates norms and so on. But how exactly
does it do all of that?
If we want to answer how the law does what it does, we
cannot in our answer, by threat of circularity, make ref-
erence to the concepts and ideas that operate on the top
level. Questions of legal normativity, of the legal ought
in general and of the demands the law creates in particu-
lar operate on the top level. When answering how the
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

law can be normative and create legal demands this lan-


guage of legal normativity is as unavailable to us as the
Figure 4 techniques of the application layer of the internet are
unavailable at the protocol layer.
Rather conversely, we have to build our stack from the bottom up. Just as the internet
protocol can be understood only insofar as it makes a specific use of the underlying
physical link so our legal protocol can only be understood insofar as it makes specific
use of the facts underlying it. So to understand how the legal protocol operates, we
have to accurately model the medium in which this protocol does what it does. What
we find in this bottom layer is agents doing things, where doing things includes saying
things. At this stage of modelling it is not yet relevant what exactly these agents do.
What we are facing, rather, at any given moment is a set of agents with a certain ca-
pability to do things. We could call these capabilities the agents’ forces and represent
them as arrows pointing in different directions.
So, seen independently from the layers above, the bottom layer can be represented
as a set of agents and their forces pushing in different directions. Again, what the exact
content of these forces is is as of yet irrelevant. The forces are not necessarily aligned.
As mentioned, the agents communicate with each other and insofar as these commu-
nications affect other agents, they are part of the forces that operate between them.
Now, whatever the law does on the top layer, it certainly has to affect those forces: it
regulates behaviour, co-ordinates action and ultimately monopolises the use of force.
So in order to do anything at all, the law has to be able to influence these social forces
and to bring some kind of order to them. How exactly this order which the law brings
to the social forces looks like is not yet relevant and will vary depending on the state
of legal development and specific historic and cultural features of a given legal system.
What matters here is the more basic problem of how the law can bring any order to
these forces at all.

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 71

So how is this to be done? There are three classic answers to this problem: (i) com-
munication, (ii) self-interest, and (iii) force. I will go through each of the answers in
turn.
A first (i) answer to the question of how the law brings order to the world would
look something like this: the law influences its subjects’ actions mainly by communi-
cating standards to them. The subjects then voluntarily take an internal perspective
towards these standards and comply with them. It is roughly along these lines that
Hart wrote:
The principal functions of the law as a means of social control are not to be seen in private
litigation or prosecutions, which represent vital but still ancillary provisions for the fail-
ures of the system. It is to be seen in the diverse ways in which the law is used to control, to
guide, and to plan life out of court.9
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Under this model the law basically functions as a naked demand made on us. Such a
system of pure or un-fortified demands will likely bring about at least some level of
co-ordination of behaviour. However, there are a couple of well known difficulties with
this model: first of all, this view models the law to function in the same way as morality
does, ie by simply telling people what to do. Now whilst it might seem attractive to
have such a functional alignment of normative practices, this approach produces a host
of so far unresolved difficulties. After all, we do ascribe a different mode of origin, oper-
ation, content and effect to legal and moral rules. All these differences may be covered
up in the demand-paradigm but they re-surface at various intersections of contempo-
rary jurisprudential debates about the law, where they are pressed down only to surface
at another point. Since there is no consensus that such a Whack-A-Mole approach will
ultimately be unsuccessful (or that there can be anything else than such an approach
in philosophy in general) I do not take this to be a debunking argument. What comes
closer to such a debunking argument is the second well known criticism: it argues that
in relying entirely on the voluntary co-operation of the subjects such a pure demand
model is incapable of explaining the level of co-ordination and ultimately the monop-
olisation of the use of force you find in legal systems. Note that the claim presented in
this paper does not maintain that the law does not demand anything nor that these de-
mands do not have any effect whatsoever. The claim is only that taken in isolation the
demand model is incapable of enlightening us about how the law actually moves about
the forces in the bottom layer and how it ultimately monopolises the use of force.
A second (ii) approach do deal with the problem would be to embellish the the-
ory of pure standards or demands with a theory of self-interest. After all, it seems to
be mostly in the self-interest of the individual agents to align their acts with the legal
order. However, there is, of course, a missing link between self-interest and the actual

9 HLA Hart, The Concept of Law, 2nd edn (Oxford, Oxford University Press, 1994) 33.

Franz Steiner Verlag


72 Christoph KLetzer

alignment of action. After all, whilst it is certainly in one’s self-interest to benefit from
social order (be it in the form of peace, prosperity or security), it is not necessarily in
one’s self-interest to bear one’s share of the costs of this social order, especially given
one’s knowledge that everybody else will make similar calculations: self-interest taken
by itself advises us to appear to be aligning our actions with the legal order and we can
appear to be aligning our action in two ways, either by actually aligning our actions
or by pretending to align our action.10 The fact that in some instances it may be easier
and less costly to actually align our actions to the legal order rather than to only make
it appear that we are doing so does not weigh very heavily, since in cases where the
stakes are high, the differential payoffs of a good show will more and more outweigh
the benefits of honest play. After all, the costs of putting up a show remain relatively
constant whereas the costs of actual honest commitment vary heavily depending on
the task involved.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

The claim that there might be a strong incentive to free-ride does not, of course, by
itself refute the argument that it might nevertheless be in the self-interest of individual
agents to co-operate or to align their actions and this claim thus does not establish
that the spontaneous emergence of co-operation would be impossible. Rather, con-
versely, some kind of emergence of a propensity towards co-operation or alignment
of behaviour is likely to emerge even from evolutionarily aligned forces of self-interest
taken in isolation. Strategies which are maximally prone to free-ride or diverge from
common behaviour might actually not constitute what has been called an evolutionary
stable strategy:11 in the context of everyone free-riding or diverging it will be beneficial
to develop some kind of tendency towards co-operative or convergent behaviour since
such a tendency, if met with other such tendencies, would out-compete the lone diver-
gent free riders.
Such a spontaneous emergence of co-operation and behavioural convergence will,
however, only occur in pockets of rather limited co-operation and by the same evolu-
tionary reasons cannot grow beyond a certain limit: just as the evolution of a ‘hawk-
ish’12 strategy that never co-operates will be out-competed by gangs of co-operating
‘doves’, so a society composed of purely dovish players is easy prey for hawkish ex-

10 Some forms of social coordination do seem to require actual alignment. The traffic rule to drive on the
left hand side of the road benefits only drivers who actually drive on the left hand side. This, however, is a
special case as it is unclear how one could appear to drive on the left hand side without actually driving on
the left hand side. From most other traffic rules, however, like the rules not to park on certain spots, not to
exceed a certain speed limit, not to jaywalk, not to kill pedestrians by running them over and so on, one
can benefit without actually being in compliance with them, but rather by successfully appearing to be in
compliance.
11 John Maynard Smith defines an evolutionarily stable strategy (ESS) as “an uninvadable strategy” and
defines this quality of univadability mathematically in terms of the payoffs of certain strategies. John May-
nard Smith, Evolution and the Theory of Games (Cambridge: Cambridge University Press, 1982), 10.
12 Maynard Smith uses the terms ‘dove’ and ‘hawk’ to represent the two strategies modelled in his Hawk-
Dove game, where the hawk escalates and continues until injured or until opponent retreats, whereas the

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 73

ploitative behaviour. In a world of doves, a single hawk can make a lot of easy prey and
at a certain saturation point the benefits of turning hawkish will simply be too great to
forgo. The question thus is not whether human nature is such that it does or does not
favour behavioural alignment and co-operation but rather which behavioural strate-
gies would become too costly to resist even for the best natured being.
Whatever the level of such an emergent co-operation resulting from an equilibrium
of the hawkish and dovish forces pulling in opposite directions, I think it is safe to say
that it represents but the minimal fall-back level of locally limited pockets of co-opera-
tion and social ordering on the scale of families and tribes. Going beyond this minimal
level of spontaneous co-ordination and social ordering and moving towards some kind
of monopolisation of the use of force will demand much more of agents and will put
their willingness to forgo the benefits of free-riding under ever more strain. I think it is
safe to say that any regime that claims to have law has a level of social order that goes
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

far beyond that minimum of what people want to do ‘naturally’, ie merely out of self-in-
terests co-ordinated by evolutionary strategies. How would such level of the alignment
and ordering of social forces be achieved? How could people be made to do things that
go beyond and against what they are willing to do naturally?
The only remaining answer (iii), it seems, is force: if people do not align their be-
haviour naturally or willingly, it seems they need to be forced to do so. On the face of
it, this might appear a promising route. However, there is a principled problem with
such an approach that relates to the possible holder of such a force, to the agent that
is supposed to force the others. Let us say the amount of cumulative resistive force to
be overcome in order to bring people into alignment is Z. Now, no single individual of
the group will be strong enough so that his individual force X > Z. Thus any individual
aiming to dominate a group by force will have to make use of the forces of other mem-
bers, say a group of henchmen or myrmidons. Say, the combined force of the myrmi-
dons is quantified as Y. In order for this suggested alignment by naked force to work, Y
has to be greater than Z. At the same time, for the leader to control his helpers by force
alone it has to be the case that X > Y. But if both X > Y and Y > Z, then it follows by
necessity that X > Z: if the purported leader were strong enough to dominate by sheer
force the myrmidons which in turn were strong enough to dominate by sheer force the
wider public then the purported leader would by necessity be strong enough to domi-
nate the wider public by sheer force directly. Since this is certainly not the case – after
all we needed to go through the myrmidons only because we established that X ≯ Z – it
cannot be the case that the supposed ruler truly dominates the myrmidons by force
alone. And this makes perfect sense. It is likely that they are not truly forced by him
but follow him voluntarily, ie either by following his naked demands (see above (i))

Dove retreats at once if the opponent escalates. John Maynard Smith, Evolution and the Theory of Games
(Cambridge: Cambridge University Press, 1982), 12.

Franz Steiner Verlag


74 Christoph KLetzer

or by concluding it is in their self-interest (see above (ii)) to follow him independently


of the use of force.
But there is a last option we missed: could it not be the case that no individual agent,
but the group as a whole exercised force to keep individual agents in line? Could there
not be some some Durkheimian social forces13 at play that are in a sense external to
and coercive of individual actors? Could not ‘the community’ make sure that people
come and stay in line? This is an attractive option, however, without going into details
of Durkheim’s theory, it can be ruled out at this stage as it posits and takes as given ex-
actly what we want to explain, namely, the social force that aligns the individual forces.
The explanation of how the individual forces are aligned, how the legal force is con-
stituted, thus has to go down a different route than the ones taken by the above theo-
ries and might well involve some kind of interplay of them. It requires a shift from the
master-paradigm to the protocol paradigm.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

The legal technique as presented by the Pure Theory does just that. I think that nei-
ther communication nor self-interest nor force taken in isolation is capable of creating
the alignment of forces which would constitute the monopolisation of the use of force
we see in modern societies. So how can it work?
In discussing primitive legal systems Kelsen gives a curious answer:
In primitive law, the individual whose legally protected interests have been violated is
himself authorised by the legal order to proceed against the wrongdoer with…coercive
means. This is called self-help. Every individual takes the law into his own hands. Blood
revenge is the most characteristic form of this primitive legal technique. Neither the es-
tablishment of the delict nor the execution of the sanction is conferred upon an authority
distinct from the parties involved or interested. In both these aspects the legal order is
entirely decentralised…Nevertheless, in a primitive community the man avenging the
murder of his father upon one whom he considers to be the murderer is himself regarded
not as a murderer but as an organ of the community…The distinction between murder, as
a delict, and homicide, as a fulfilment of a duty to avenge, is of the greatest importance for
primitive society. It means that killing is only permitted if the killer acts as an organ of his
community, if his action is undertaken in execution of the legal order. The coercive act is
reserved to the community, and is, in consequence, a monopoly of this community. The
decentralisation of the application of the law does not prevent the coercive act as such
from being strictly monopolised.14

To illustrate this idea, take the following situation: say X kills A. In revenge A’s brother
B kills X. In revenge Y kills B, then C kills Y and Z kills C.

13 See, for instance, Emile Durkheim, Suicide (London: Routledge, 2002) and The Elementary Forms of
Religious Life (New York: The Free Press, 1995).
14 Hans Kelsen, General Theory of Law and State (Cambridge, MA: Harvard University Press 1949) 338–9.

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 75

Figure 5
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Figure 6 Figure 7

In terms of the pure facts at the basic level, this is but a series of mutual killings. It is
a war-like and un-ordered chaotic state of affairs. If we apply Kelsen’s logic of prim-
itive law, however, whilst the situation in terms of pure facts remains the same, the
legal assessment has fundamentally changed: now the end-of-alphabet group (X, Y, Z)
are marked as culprits or criminals and the beginning-alphabet-group (B and C) are
marked as legal organs.
In a global context we can thus see how the law picks out certain agents by reference
to their specific direction of force and marks this force as the force of the law.
Note that this force of the law is not created by the law but only co-opted by it.
The force has existed as force before the existence of law and independently of it. It
is a force which operates on the basic, ‘physical’ layer. It is not created by the law, but
it is identified by the law amongst the existing forces of society. This force is thus by
definition always already directed in the legally right direction. The law thus solves
the problem of how to create order neither by demanding that the direction of forces
be changed or by relying on the self-interest of agents to change the direction of their
force or by forcing the use of force. It does not immediately change the direction of
forces at all. Rather it confirms the direction of certain forces in contrast to others.
This may, of course, have the mediated effect of gradually changing the direction of
forces of the agents by the feeling of that being demanded, by calculation of self-inter-
est or by the actual fact that in an instance the use of force marked out as lawful might
happen to be stronger than the one marked out as unlawful. This mediated effect,
however, is but a secondary effect. The primary method, conversely, is the identifica-
tion of a social order.
The problem discussed above of how to direct social force in a certain way is here
already resolved: the law does not need to move about forces in the right direction but

Franz Steiner Verlag


76 CHRISTOPH KLETZER

it identifies the correctly directed forces as its own force. The strongest, historically
prevailing force will be considered the legal force. It is only after this establishment of
the legal force that the legal force will consolidate itself by means of centralisation and
thus by means of a successive weeding out of unlawful force.
What needs to be seen, though, is that even at this primitive stage the law, whilst
not yet being able to centralise the use of force, it is already capable of monopolising the
use of force: the law allows us to distinguish between lawful use of force (sanctions by
state organs) and unlawful use of force (crimes by criminals). Every use of force is thus
marked out as either allowed or forbidden. This is all that is needed for monopolisa-
tion, since the monopolisation can neither mean the total absence of the use of force
altogether or even the total absence of the unlawful use of force. Even in modern states,
where the use of force is uncontroversially monopolised, there is still a lot of unlawful
use of force. What monopolisation means is that all lawful force is wielded by the state
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

as identified by the law.


This analysis of the legal stack helps us to see rather clearly the importance as well
as the practical limitations of this legal protocol. Let us, again, compare the legal stack
to the internet protocol suite. TCP/IP might establish the foundation of the modern
internet, but taken in isolation, ie without a further developed application layer run-
ning on top of it, is rather useless. Similarly, the legal protocol does establish the foun-
dations of the legal system, it does create peace and order but it does so only in a very
formal sense: using the protocol we can interpret the raging war between two families
not as war but as peace-keeping. One family acts as the organs of the state, the other
as band of criminals. This is war understood as peace-keeping but, of course, not yet
peace or security in a substantive sense. But it is more than nothing. It is the germ of a
legal order around which this legal order can grow.
So when Hart argues that this primitive law is not yet law, he is both right and wrong:
he is right in pointing out that the protocol layer is not the application layer, that the
legal protocol is not yet regulatory law. He is, however, wrong, in overlooking the fact
that the protocol layer is a fundamental ingredient of all law, including our modern law.
This protocol layer describes the legal technique, it describes how the law does what it
does, it describes how the law orders and regulates society.

D. Decentralised Centralisation

Whilst the protocol layer of the internet is irreducibly decentralised, the abstraction
layer on top of it, ie the application layer, is not: whilst, say, Google is linked to millions
of different computers, the webpage of the Londonderry Bridge Club is connected
only to a couple and whilst there are millions of webpages like the Londonderry Bridge
Club, there are but a couple like Google. The internet is thus actually quite centralised.

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 77

However centralised the application layer may be, this centralisation is still created on
top of and thus by means of the irreducibly decentralised internet protocol suite.
Something similar holds true for the legal stack. Modern law is certainly radically
more centralised than the primitive law system sketched above. Still, even in modern
law the centralisation is effected by means of the fundamentally decentralised method
of the legal protocol.
How, then, does the centralisation of the law work exactly? Again, since we are work-
ing within a layered stack we cannot simply assume the availability of any techniques
or concepts of a higher stack. Rather, we have work our way from the bottom to the
top and have to explain the functioning of these higher-order techniques and concepts
by means of the lower ones. We thus have to start with the primitive legal protocol and
see how centralisation can be achieved by using only this protocol.
The protocol above can be formalised as follows:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

[1] If X, then the use of force is lawful.

According to the Pure Theory of Law, this is the fundamental form of the law in gen-
eral. It is also the so-called basic norm. What distinguishes modern law from primitive
law is not the legal form as such or the way of its operation as a legal protocol layer, but
only the specific content of X, the content of the condition of the lawfulness of the use
of force.
What we aim to show, then, is how centralisation can be implemented by changing
only the content of the condition X.
In primitive law the content of the condition X is rather straightforward: the law-
fulness of the use of force is conditioned by simple facts, eg the fact that someone has
killed someone else who himself has not killed someone else.
[2] If A has killed B who himself has not killed someone then killing A is lawful.

Now, the first step towards centralisation would be to require not only that A has killed
someone but also that a third person (call him a ‘judge’) has authorised the killing of
A. It is important to note that this requirement would not feed into the law as a bare
demand but that it would be introduced into the condition X.
[3] If A has killed B who himself has killed C but A has not been authorised to kill B by
D then killing A is lawful.

The recursive nature of this rule has the effect that now any killing has to be authorised
by a judge in order not to have the effect of rendering a killing of the last killer lawful.
Note that this centralising move is effected neither by telling A what he can do or
cannot do, nor by telling D (the “judge”) what he can do or cannot do, but by telling
everyone what they can do to A.
Take again our case illustrated in Figure 6. X kills innocent A and then B kills X.
Under rule [2], this last killing would have been lawful which means nothing else than

Franz Steiner Verlag


78 CHRISTOPH KLETZER

that a killing of B by, say, Y, would be unlawful which, in turn means nothing else than
that a killing of Y would be lawful and so on.
If we apply rule [3], however, the situation looks entirely different: in this case, if B
had not been authorised by J (the “judge”), then his killing of X remains unlawful, even
if X might have in fact previously killed A. This in turn would render a killing of B law-
ful. However, if now Y comes along and kills B without himself having been authorised
by J, his killing of B is unlawful which in turn means that killing him would be lawful,
as long as authorised by J, and so on.
Note, that rule [3] operates not by rendering more killings unlawful but by render-
ing them lawful Or, put differently: it renders certain killings unlawful by rendering
others lawful. Under rule [2] alone, killing B would not be lawful. Under rule [3] kill-
ing B is lawful. The recursive applicability of rule [3], however, makes sure that no kill-
ing unauthorised by J is lawful. The recursive nature of this rule has the effect that now
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

any killing has to be authorised by a judge in order not to have the effect of rendering
the killing of the last killer lawful.
Now, apart from the requirement that a judge authorises the killing, other require-
ments can be introduced into the condition X. These can include the rules about how
exactly the judge has to go about to authorise the killing, how the killing has to be
conducted, who can do the killing, who counts as a judge and so on. Furthermore in
referring to certain valid rules and procedures these conditions refer to the valid cre-
ation and abrogation of new rules feeding into X. Thus, for instance, the condition X
may include requirements like: ‘as has been decided by a person appointed as a judge
in accordance to the requirements as set out in the laws regulating the appointment of
judges as created by parliament in accordance with the constitution which has been
created in accordance with international law.’
It is via this route, ie as specific content of the condition X, that the entire content
of the modern, centralised legal system enters the primitive legal rule. It is thus in this
content X that the entire application layer of the regulatory law has to be located.

E. Conclusion

Understanding the law not as a flat, one-dimensional phenomenon but adding some
depth to it by understanding it as a stack of various layers helps us, I believe, get a much
clearer understanding of how it works and also what it is. It helps us, I think, see more
clearly three facts.
First, the law is ultimately always a matter of the use of force. What distinguishes
modern from primitive law is but complexity and sheer number of the conditions that
condition the lawfulness of the use of force. It is thus no wonder that we as lawyers and
jurisprudes can sometimes get lost in the legal material and start to overlook that logi-
cally it all plays the same role, namely that of conditioning the lawfulness of the use of

Franz Steiner Verlag


Decentralisation and the Limits of Jurisprudence 79

force. Of course, modern law regulates many trivial things which do not in themselves
seem to be matters of the use of force. The stack analysis presented here, however,
shows that as soon as we try to analyse how the law regulates even those trivial things
we find out that it regulates them by ultimately holding out the use of force. This use of
force, is of course, heavily conditioned, but it is always there.
Secondly, and relatedly, the law operates not by simply telling people what to do,
but it tells people what to do by conditioning the use of force. It involves a shift from
a master-paradigm of social unity in which there has to be some real or projected uni-
fied agent behind the law to a protocol-paradigm where the social unity is explained
by means of the uniform operation of a protocol or code. The ease with which the law
makes people do things might sometimes seem miraculous. It has to remain miracu-
lous as long as we do not inquire into how exactly the law can make people do things.
In its technique the law has recurse to a decentralised and in this sense ‘social’ reservoir
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

of pure force. It establishes this social force by identifying it.


Thirdly, whilst the use of force might be fundamental to the legal system, it is not its
most significant part. What is more significant than the use of force is all the things we
do by means of this use of force.

Christoph Kletzer
King’s College London, The Dickson Poon School of Law, Strand, London, WC2R 2LS,
United Kingdom, christoph.kletzer@kcl.ac.uk

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“
bei Hans Kelsen*1

CARSTEN HEIDEMANN

The ‘Fact of Legal Science’ in Hans Kelsen’s Theory of Law


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: The ‘fact of science’ is, in two different senses, at the core of Kelsen’s neo-Kantian
theory of law during the twenties. On the one hand, Kelsen defines his theory, following
Kant, Trendelenburg, and Cohen, as a critical theory of legal science, that is, of legal dog-
matics. Legal science is a fact insofar as it is established as a social practice; and it is the task
of the Pure Theory to analyse the presuppositions of the claim to objectivity, normativity
and positivity, which is raised by dogmatic statements, and to reconstruct legal science in
accordance with these presuppositions. On the other hand, Kelsen identifies the fact which
is established by legal science, the legal norm, with the hypothetical judgment of legal sci-
ence wherein antecedent and consequent are connected by the category of imputation. This
judgment is not a psychic entity or procedure, but an objective structure; its validity might
plausibly be taken to consist in the normative demand to think according to this judgment.

Keywords: Hans Kelsen, Hermann Cohen, Friedrich Adolf Trendelenburg, fact of legal
science, transcendental method, validity

Schlagworte: Hans Kelsen, Hermann Cohen, Friedrich Adolf Trendelenburg, Faktum der
Rechtswissenschaft, transzendentale Methode, Geltung

1. Einleitung

Der Bezug auf das „Faktum der Rechtswissenschaft“ ist für Kelsens neukantische
Rechtstheorie der zwanziger Jahre von zentraler Bedeutung, und der entsprechende
Ausdruck erscheint in der ersten Auflage der Reinen Rechtslehre an prominenter Stelle.
Kelsen schreibt dort, die Möglichkeit und Erforderlichkeit einer normativen Theorie

* Ich danke Monika Zalewska für hilfreiche Kommentare zum Text.

Franz Steiner Verlag


82 CARSTEN HEIDEMANN

des Rechts sei schon durch das „jahrtausendalte Faktum der Rechtswissenschaft“ er-
wiesen.1 Und in seiner einzigen ausführlicheren Auseinandersetzung mit neukanti-
schen Theorien, Rechtswissenschaft und Recht von 1922, heißt es, dass die Beziehung auf
das Faktum einer Wissenschaft das „Um und Auf der Transzendentalphilosophie“ sei.2
Im Folgenden werde ich kurz die Genese der Wendung vom Faktum der Wissen-
schaft, ihre doppelte Bedeutung und ihre Funktion innerhalb der Reinen Rechtslehre
darstellen.

2. Das Faktum der Wissenschaft bei Cohen, Kant und Trendelenburg

Das Faktum der Wissenschaft zählt, als Ansatzpunkt der „transzendentalen Methode“,
durch die in der Nachfolge Kants notwendige Bedingungen der Möglichkeit objekti-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ver Erkenntnis begründet werden sollen, zum Grundbestand neukantischer Theorien.


Dort wird das entsprechende Theorem vor allem von Hermann Cohen vertreten. 1877
schreibt Cohen:
„Nicht die Sterne am Himmel sind die Objekte, die jene Methode betrachten lehrt, son-
dern die astronomischen Rechnungen, jene Fakten wissenschaftlicher Realität sind gleich-
sam das Wirkliche, das zu erklären steht, auf welches der transzendentale Blick eingestellt
wird. […] Jene Fakten von Gesetzen sind die Objekte; nicht die Sternendinge“.3

Und in der 2. Auflage seiner Schrift Kants Theorie der Erfahrung heißt es dann 1885:
[„Den Ausweis der apriorischen Elemente] bringt die transzendentale Methode, deren
Prinzip und Norm der schlichte Gedanke ist: solche Elemente des Bewusstseins seien
Elemente des erkennenden Bewusstseins, welche hinreichend und notwendig sind, das
Faktum der Wissenschaft zu begründen und zu festigen.“4

Cohen gibt dies als Rekonstruktion des Kerns der theoretischen Philosophie Kants
aus. In der Tat lässt die Konzeption sich auf den ersten Blick auf Kants Prolegomena
zu einer jeden künftigen Metaphysik5 stützen. In dieser didaktisch vereinfachenden Zu-
sammenfassung von Teilen der Kritik der reinen Vernunft unterscheidet Kant die in der
Kritik angewandte anspruchsvolle synthetische oder progressive Begründungsweise,
die ausgehend von der skepsisresistenten transzendentalen Einheit des Selbstbewusst-
seins die apriorischen Geltungsbedingungen objektiver Erkenntnis begründet, von

1 Hans Kelsen, Reine Rechtslehre (1. Aufl.), Wien und Leipzig 1934, 37.
2 Hans Kelsen, Rechtswissenschaft und Recht. Erledigung eines Versuchs zur Überwindung der „Rechts-
dogmatik“, in: ZöR 3 (1922), 128.
3 Hermann Cohen, Kants Begründung der Ethik, Berlin 1877, 20 f.
4 Hermann Cohen, Kants Theorie der Erfahrung (2. Aufl.), Berlin 1885, 77.
5 Immanuel Kant, Prolegomena zu einer jeden künftigen Metaphysik, die als Wissenschaft wird auftreten kön-
nen, Werke (hrsg. von Wilhelm Weischedel), Bd. 5, Frankfurt/M. 1977, 134 f. (A 38 f.), 139 f. (A 46 f.).

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 83

der leichter fassbaren analytischen oder regressiven der Prolegomena, die ausgehend
von den in Mathematik und reiner Naturwissenschaft „gegebenen“ synthetischen Ur-
teilen a priori deren Ermöglichungsbedingungen nachweist.
Die synthetische Methode ist die eigentliche philosophische Methode. Sie macht
den Kern von Kants theoretischer Philosophie aus. Dieser geht es angesichts der Un-
möglichkeit, eine Erkenntnis eines Gegenstandes mit einem erkenntnisunabhängig
gedachten Gegenstand zu vergleichen, darum darzulegen, ob und auf welche Weise
Gegenstandsbezug und Objektivität innerhalb der Erkenntnis zustande kommen kön-
nen. Sie setzt nicht voraus, dass es synthetische Urteile a priori gibt (denn dies könnte
durch den Skeptiker bezweifelt werden), durch die sich die Gültigkeit der Katego-
rien sonst leicht belegen ließe; vielmehr müssen die Kategorien als notwendige Be-
dingungen der Möglichkeit von Gegenstandserkenntnis von einer unbezweifelbaren
Prämisse aus bewiesen werden. Dieser unbezweifelbare Ausgangspunkt ist für Kant
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

die transzendentale Einheit der Apperzeption, das „ich denke“, das alle unsere Vorstel-
lungen und Urteile begleiten können muss. Sie lässt sich wohl am ehesten als „Gleich-
ursprünglichkeit“ von Erkenntnissubjekt und einem kohärenten System referierender,
also sich auf einen Gegenstand beziehender, Urteile oder Propositionen begreifen.6
Ob und wie es Kant gelungen ist, von diesem formalen Ausgangspunkt aus inhaltlich
bestimmte Kategorien nachzuweisen, ist fraglich; jedenfalls dürfte es sich bei der Be-
weisführung nicht um eine einfache „Deduktion“ handeln. Möglicherweise kommt für
die Rechtfertigung der transzendentalen Einheit der Apperzeption selbst ein „elenkti-
sches“ Argument in Frage,7 während die Gültigkeit der Kategorien im Rahmen einer
„Sequenz analytischer Implikationsschritte“8 nachgewiesen werden könnte. Denkbar
ist auch, dass die synthetische Begründung stets einer Ergänzung durch einen Beweis
nach analytischer Methode bedarf.9
Die analytische Methode ist deutlich anspruchsloser. Sie ist für Kant kein philo-
sophisches Begründungsverfahren, sondern eine ausschließlich für Prolegomena zu-
lässige Darstellungsweise. Sie müsse sich auf etwas stützen,

6 Aschenberg meint, dass das Prinzip der transzendentalen Einheit der Apperzeption grundlegender
als die Kontrastrelation zwischen Subjektivem und Objektivem sein müsse (vgl. Reinhold Aschenberg,
Sprachanalyse und Transzendentalphilosophie, Stuttgart 1982, 185 f.). Das scheint mir aber nur dann der Fall
zu sein, wenn die Einheit des Urteilszusammenhangs bei Kant sich wirklich kategorial trennen lässt von der
Einheit des empirischen Gegenständlichen, was m. E. zweifelhaft ist; vgl. unten 3.3. Zur Gleichursprüng-
lichkeit der beiden Elemente vgl. auch Hansgeorg Hoppe, Die Bedeutung der Empirie für transzendentale
Deduktionen, in: Kants transzendentale Deduktion und die Möglichkeit von Transzendentalphilosophie, Frank-
furt/M. 1988, 131.
7 Reinhold Aschenberg, Sprachanalyse und Transzendentalphilosophie (Fn. 6), 383 ff.
8 Reinhold Aschenberg, Einiges über Selbstbewusstsein als Prinzip der Transzendentalphilosophie, in:
Kants transzendentale Deduktion und die Möglichkeit von Transzendentalphilosophie, Frankfurt/M. 1988, 55.
9 Vgl. etwa Christian Krijnen, Nachmetaphysischer Sinn, Würzburg 2001, 56 f. m. w. N., der von einer not-
wendigen Komplementarität der beiden Begründungsmethoden ausgeht.

Franz Steiner Verlag


84 CARSTEN HEIDEMANN

„was man schon als zuverlässig kennt, von da man mit Zutrauen ausgehen und zu den
Quellen aufsteigen kann, die man noch nicht kennt, und deren Entdeckung uns nicht al-
lein das, was man wusste, erklären, sondern zugleich einen Umfang vieler Erkenntnisse,
die insgesamt aus den nämlichen Quellen entspringen, darstellen wird.“10

Die Möglichkeit der analytischen Methode ergebe sich daraus, dass „gewisse reine
synthetische Erkenntnis a priori“ zwar nicht unbezweifelbar, aber doch „wirklich und
gegeben“ sei, nämlich die Urteile der Mathematik und der reinen Naturwissenschaft.11
Ihr Vorteil ist, dass sie nicht aus einem unbezweifelbaren, aber formalen Prinzip zu
inhaltlichen Ergebnissen gelangen muss, sondern bestimmte nichtempirische, aber
dennoch synthetische Urteile voraussetzen darf. Entsprechend heißt es bei Kant zu
den Vorzügen der analytischen Methode:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

„[Sie] erleichtert das Geschäfte sehr, in welchem die allgemeinen Betrachtungen nicht al-
lein auf Facta angewandt werden, sondern sogar von ihnen ausgehen, anstatt dass sie in
synthetischem Verfahren gänzlich in abstracto aus Begriffen abgeleitet werden müssen“.12

Doch wird für Kant die eigentliche Begründungslast von der synthetischen Methode
getragen; denn mittels der analytischen Methode kann lediglich die relative Geltung
apriorischer Elemente bezogen auf die Sätze der Mathematik und reinen Naturwissen-
schaft nachgewiesen werden.13 Wenn Cohen die „transzendentale“ Methode mit der
analytischen Methode Kants identifiziert, relativiert er den Kantschen Ansatz also er-
heblich.
Darüber hinaus findet sich bei Cohen eine zweite Relativierung des Begründungs-
ansatzes Kants, die vermutlich auf den Einfluss Friedrich Adolf Trendelenburgs zu-
rückzuführen ist. Trendelenburg fordert Mitte des 19.  Jahrhunderts angesichts des
desolaten Zustands der Philosophie nach dem Niedergang des deutschen Idealismus,
dass sie sich an die seinerzeit blühenden Naturwissenschaften anschließen möge, zum
wechselseitigen Nutzen, denn diesen würde eine kritische Methodologie fehlen. Be-
reits 1840 schreibt Trendelenburg:
„Die Tatsachen, die [die Logik, d. h. die erkenntnistheoretische Philosophie] beobachten
sollte, um sie abzuleiten, sind die Methoden der einzelnen Wissenschaften; denn diesen

10 Immanuel Kant, Prolegomena (Fn. 5), A 40.


11 A. a. O.
12 Immanuel Kant, Prolegomena (Fn. 5), A 47.
13 Vgl. zur analytischen/synthetischen Methode auch Immanuel Kant, Logik, Werke (hrsg. von Wilhelm
Weischedel), Bd.  6, Frankfurt/M. 1977, 581 (§ 117): „Die analytische Methode ist der synthetischen ent-
gegengesetzt. Jene fängt von dem Bedingten und Begründeten an und geht zu den Prinzipien fort […],
dieses hingegen geht von den Prinzipen zu den Folgen oder vom Einfachen zum Zusammengesetzten. Die
erstere könnte man auch die regressive, so wie die letztere die progressive nennen. […] Für den Zweck der
Popularität ist die analytische, für den Zweck der wissenschaftlichen und systematischen Bearbeitung des
Erkenntnisses aber ist die synthetische Methode angemessen.“

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 85

hat der erkennende Geist in den größten Abmessungen sein eigenes Wesen eingedrückt.
Die Wissenschaften versuchen glücklich ihre eigentümlichen Wege, aber zum Teil ohne
nähere Rechenschaft der Methode, da sie auf ihren Gegenstand und nicht auf das Ver-
fahren gerichtet sind. Die Logik hätte hier die Aufgabe zu beobachten und zu vergleichen,
das Unbewusste zum Bewusstsein zu erheben und das Verschiedene im gemeinsamen
Ursprunge zu begreifen. […] Ohne sorgfältigen Hinblick auf die Methode der einzelnen
Wissenschaften muss sie ihr Ziel verfehlen, weil sie dann kein bestimmtes Objekt hat, an
dem sie sich in ihren Theorien zurechtfinde.“14

Und 1870 heißt es dann in der dritten Auflage von Trendelenburgs Logischen Unter-
suchungen, die Wissenschaften stellten der Skepsis ein „Factum“ entgegen, und die Tat-
sache der Wissenschaft sei die Basis des logischen Problems.15
Damit wird Trendelenburg zu einem Begründer der Wissenschaftstheorie.16 Zwar
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

kann sein Einfluss auf Cohens Konzeption des „Faktums der Wissenschaft“ nicht
direkt nachgewiesen werden, denn Cohen hat für seine Auffassung nie explizit auf
Trendelenburg Bezug genommen. Doch hatte er als Student in Berlin in der ersten
Hälfte der 1860er Jahre nicht nur unmittelbaren Kontakt zu Trendelenburg;17 er be-
teiligte sich auch maßgeblich (und zwar im Wesentlichen auf Trendelenburgs Seite)
an dem Streit um die „Trendelenburgsche Lücke“, in dem es um die Frage ging, ob
Kant seine Einordnung von Raum und Zeit als subjektive Anschauungsformen über-
zeugend begründet habe.18 Ferner entspricht Cohens Begrifflichkeit weitgehend der-
jenigen Trendelenburgs. Darüber hinaus lässt sich seine Deutung der analytischen
oder regressiven Methode kaum ohne Trendelenburgs Einfluss erklären. Denn diesem
geht es nicht, wie Kant, um die konstituierenden Elemente apriorischer synthetischer
Urteile, sondern um den Aufweis der Methoden und Voraussetzungen der aposteri-
orischen Naturwissenschaften schlechthin. Das findet sich entsprechend bei Cohen,
der das wissenschaftliche Faktum als Ansatzpunkt der transzendentalen Methode, wie
die oben zitierte Passage belegt, nicht (nur) in den apriorischen synthetischen Sätzen
der reinen Naturwissenschaft und Mathematik, sondern generell in den Tatsachen der
wissenschaftlichen „Gesetze“ sieht.19

14 Friedrich Adolf Trendelenburg, Logische Untersuchungen (1. Aufl.), Berlin 1840, Vorrede IV.
15 Friedrich Adolf Trendelenburg, Logische Untersuchungen (3. Aufl.), Leipzig 1870, 130 f., 312.
16 Vgl. Klaus Christian Köhnke, Entstehung und Aufstieg des Neukantianismus, Frankfurt/M. 1993, 35 ff.
17 Vgl. Deutsche Biographie, Artikel „Cohen, Hermann“ (abrufbar unter https://www.deutsche-biographie.
de/sfz8558.html).
18 Vgl. dazu insbesondere Klaus Christian Köhnke, Entstehung und Aufstieg des Neukantianismus (Fn. 16),
257 ff.
19 Vgl. dazu auch Klaus Christian Köhnke, Entstehung und Aufstieg des Neukantianismus (Fn. 16), 272: „Dass
schon 1840 von Trendelenburg gesagt worden war, die Logik müsse sich auf die Wissenschaften richten,
weil sie sonst kein bestimmtes Ziel und Objekt habe, und damit erstmals eine ‚Theorie der Wissenschaft‘
ihrem Anliegen nach gefordert war, sollte den Gedanken erleichtern, der Vater des neukantianischen Scien-
tismus, Cohen, habe tatsächlich weit mehr, als bislang gesehen worden ist, von jenem gelernt.“

Franz Steiner Verlag


86 CARSTEN HEIDEMANN

Diese Kontingenz des Ausgangspunktes wirkt sich im Übrigen bereits bei Cohen,
mehr noch bei seinem Schüler Paul Natorp, in einer eigentümlichen Dynamik der Er-
kenntnisbedingungen aus: Zwar verkörpert das Faktum der Wissenschaft das erreich-
bare Höchstmaß objektiver Erkenntnis; es ist aber nichts Feststehendes, sondern ein
„Fieri“,20 und mit ihm können sich auch die notwendigen Bedingungen seiner Mög-
lichkeit verändern.
Wenn Cohen für seine Theorie dennoch den Anspruch erhebt, die theoretische
Philosophie Kants zu rekonstruieren, dann reduziert er die Transzendentalphiloso-
phie letztlich auf eine „methodenkritische Wissenschaftstheorie mit erweiterten poli-
zeilichen Vollmachten“.21 Das bedeutet allerdings nicht unbedingt eine Trivialisierung.
So wie Kant die traditionelle Ontologie nicht wirklich „zermalmt“, sondern sie in einer
Verstandesanalytik aufgehen lässt,22 so verwirft Cohen seinerseits Kants Verstandes-
kritik nicht, sondern deutet sie in eine kritische Wissenschaftstheorie um. Die klassi-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

sche Ontologie wird quasi in der neukantischen Wissenschaftstheorie „aufgehoben“.

3. Das Faktum der Rechtswissenschaft bei Cohen und Kelsen

Nach Kelsens eigener Auskunft wird er durch eine Besprechung seiner Hauptprobleme
der Staatsrechtslehre auf Parallelen zwischen seiner Theorie und der von Cohen 1904 in
der Ethik des reinen Willens entwickelten aufmerksam.23 Diese betreffen zwar vor allem
den Personen- und Willensbegriff; in Cohens Schrift heißt es jedoch auch ausdrück-
lich, dass jede Philosophie auf das „Faktum von Wissenschaften“ angewiesen sei,24 das
sei „das Ewige in Kants System“. Und die Wissenschaft, an die Cohen seine Ethik in
sehr eigenwilliger Weise andockt, ist die Rechtswissenschaft.

3.1 Die faktische Rechtswissenschaft

Dabei modifiziert er sein eben geschildertes Konzept allerdings erheblich. Cohen er-
klärt nie genauer, was er unter dem „Faktum der Rechtswissenschaft“ versteht. Einer-
seits scheint die ausgeübte Rechtswissenschaft für ihn nicht (wie für Kelsen) mit der
Rechtsdogmatik identisch zu sein. Vielmehr lokalisiert er sie auf einer abstrakteren

20 Paul Natorp, Die logischen Grundlagen der exakten Wissenschaften, Leipzig und Berlin 1910, 14.
21 So treffend Andreas Hoeschen, Das „Dostojewsky“ – Projekt Lukács‘ neukantianisches Frühwerk in seinem
ideengeschichtlichen Kontext, Tübingen 1999, 113.
22 Immanuel Kant, Kritik der reinen Vernunft, Werke (hrsg. von Wilhelm Weischedel), Bd. 3, Frankfurt/M.
1974, 275 (A 246/B 303).
23 Hans Kelsen, Hauptprobleme der Staatsrechtslehre (2. Aufl.), Tübingen 1923, Vorrede XVII; ders., Reine
Rechtslehre (Fn. 1), Vorwort IX.
24 Hermann Cohen, Ethik des reinen Willens, Berlin 1904, 62.

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 87

Ebene als die Naturwissenschaft; sie ist für ihn als Allgemeine Rechtslehre eine „Ma-
thematik“, also gewissermaßen eine Methodik, der Geisteswissenschaften schlecht-
hin.25 Diese Allgemeine Rechtslehre scheint er jedoch andererseits vor allem der insti-
tutionalisierten Rechtsdogmatik entnehmen zu wollen. Ferner eignen sich die Gesetze
der Naturwissenschaften als Ausgangspunkt der transzendentalen Methode, weil sie
das Ideal objektiver Erkenntnis weitestgehend annähern. Sie sind als allgemeingültige
Regeln experimentell belegbar und erlauben zuverlässig Vorhersagen. Rechtswissen-
schaftliche Sätze dagegen sind für Cohen nicht deswegen tauglicher Ansatzpunkt, weil
sie einem vergleichbaren Erkenntnisideal genügen, sondern weil sie Bestandteile einer
soziokulturell etablierten, systematisch vorgehenden und Objektivität beanspruchen-
den begrifflichen Praxis sind.26 Ihm scheint es ausschließlich darum zu gehen, die Be-
deutung ethischer Begriffe, die für alle Geisteswissenschaften von Interesse sind, wie
„Handlung“, „Person“, „Wille“ etc., nicht abstrakt entwickeln zu müssen, sondern sie
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

aus einer gegebenen Praxis herauspräparieren zu können.27 Unmerklich wird so das


„wissenschaftliche Faktum“ als das der transzendentalen Analyse Gegebene durch den
„faktischen Wissenschaftsbetrieb“ ersetzt; dessen Urteile erhalten ihre Dignität nicht
aufgrund ihrer Objektivität oder Allgemeingültigkeit, sondern schlicht dadurch, dass
sie in der Wirklichkeit institutionalisiert ist. Dies bedeutet eine Annäherung an den
Neukantianismus der südwestdeutschen Schule, für den die theoretische Philosophie
von vornherein deswegen an das Faktum der Wissenschaft anknüpft, weil und soweit
diese Kulturbestandteil ist.28
In seiner später erschienenen Ästhetik des reinen Gefühls radikalisiert Cohen diesen
Ansatz im Übrigen; er verzichtet dort völlig auf das Bezugsfaktum einer Wissenschaft
und lässt die transzendentale Methode direkt am „Faktum der Kunst“ ansetzen.29 Rück-
blickend bezeichnet er dort als Aufgabe der theoretischen Philosophie, zu zeigen, „wie
sich das Kulturbewusstsein, von der Mathematik geleitet, in den Gebieten der Natur-
wissenschaft darlegt“;30 soweit es um die Ethik geht, benennt er als Bezugsfaktum nicht
mehr die Rechtswissenschaft, sondern die „sittliche Kultur in Recht und Staat“.31

25 Hermann Cohen, Ethik des reinen Willens (Fn. 24), Vorrede I, 60 ff.


26 Hermann Cohen, Ethik des reinen Willens (Fn. 24), Vorrede I.
27 Vgl. Hermann Cohen, Ethik des reinen Willens (Fn. 24), Vorrede I: „[Es] wird hier der Versuch gemacht,
die Ethik auf die Rechtswissenschaft zu orientieren. Diese ist die Mathematik der Geisteswissenschaften.
Die Methodik ihrer exakten Begriffe hat die Ethik für ihre Probleme der Persönlichkeit und der Handlung
zu belauschen.“
28 Vgl. Heinrich Rickert, Vom Begriff der Philosophie, in: Logos, Internationale Zeitschrift für Philosophie
der Kultur (1910), 18. Zur Problematik vgl. auch Hans-Ludwig Ollig, Die Frage nach dem Proprium des
Neukantianismus, in: Neukantianismus und Rechtsphilosophie (hrsg. v. Robert Alexy, Lukas H. Meyer, Stan-
ley L. Paulson und Gerhard Sprenger), Baden-Baden 2002, 81.
29 Etwa Hermann Cohen, Ästhetik des reinen Gefühls, Berlin 1912, Vorrede X.
30 Hermann Cohen, Ästhetik des reinen Gefühls (Fn. 29), 4.
31 Hermann Cohen, Ästhetik des reinen Gefühls (Fn. 29), 5.

Franz Steiner Verlag


88 CARSTEN HEIDEMANN

Das ursprüngliche Anliegen des transzendentalen Beweises im Sinne Kants, aus


einem unbezweifelbaren Punkt universale Ermöglichungsbedingungen objektiver Er-
kenntnis zu deduzieren, wird damit ein drittes Mal relativiert. Trotzdem geht es auch
dieser extrem abgeschwächten Version der kritischen Methode um die Bedingungen
der Möglichkeit von (normativer) Erkenntnis, und ihr Ergebnis bietet letztlich eine
Ontologie des jeweiligen Gegenstandsbereichs; insofern ist der Bezug auf Kant gerade
noch berechtigt.
Kelsen übernimmt Cohens Ansicht, dass die transzendentale Methode nur an das
„Faktum der Wissenschaft“ anknüpfen könne, und er geht auch grundsätzlich davon
aus, dass dieses im Bereich des Rechts mit der als soziokulturelle Praxis existierenden
Rechtswissenschaft zu identifizieren sei. In zwei Punkten distanziert er sich jedoch
vom Cohenschen Ansatz: Zum einen fasst er das Faktum der Rechtswissenschaft en-
ger; er versteht darunter ausschließlich die Rechtserkenntnis der Rechtsdogmatik.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Zum anderen und vor allem teilt er – trotz seiner fundamentalen These eines Dualis-
mus zwischen Sein und Sollen  – nicht Cohens Auffassung, dass ein fundamentaler
Unterschied zwischen theoretischer und praktischer Erkenntnis besteht. Im Bereich
der theoretischen Philosophie ist die Durchführung der transzendentalen Methode
für Cohen, wie gezeigt, Aufgabe der Logik, d. h. der Erkenntnistheorie, die an die ma-
thematischen Naturwissenschaften anknüpft. Im praktischen Bereich tritt an die Stelle
der Logik die Ethik, die zwar an die Rechtswissenschaft anknüpft, aber weniger die
Bedingungen der objektiven Geltung rechtswissenschaftlicher Urteile, als vielmehr
ein System der rechtlichen Grundbegriffe herausarbeitet, um mit deren Hilfe eine
Analyse der Geisteswissenschaften schlechthin vornehmen zu können. Die Rechts-
wissenschaft spielt für Cohen eine Rolle, die der der Mathematik für die Naturwis-
senschaften analog ist. Kelsen dagegen siedelt die Rechtswissenschaft auf einer den
Naturwissenschaften analogen Ebene an; es geht ihm um die Voraussetzungen der
objektiven Geltung der Urteile einer szientistisch verstandenen Rechtsdogmatik.32
Er kombiniert gewissermaßen das Anliegen der theoretischen Philosophie Cohens,
die Voraussetzungen der Objektivität wissenschaftlicher Urteile festzustellen, das sich
daraus rechtfertigt, dass naturwissenschaftliche Urteile das höchstmögliche Maß an
Erkenntnis enthalten, mit der Relativierung der Urteile der Rechtswissenschaft auf
eine bestimmte „historische“, in der Wirklichkeit etablierte Praxis in der praktischen
Philosophie Cohens.
Für Kelsen ist diese Variante aus zwei Gründen attraktiv. Zum einen kann er die
an die „gegebene“ Rechtswissenschaft anknüpfende transzendentale Methode ohne
weiteres in sein in den Hauptproblemen entwickeltes System integrieren. Schon dort
hat er seine Theorie in erster Linie als „Methodologie“ der etablierten Rechtsdogmatik

32 Vgl. insbesondere Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 128.

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 89

aufgefasst;33 nun gelingt ihm ohne Aufwand die Einordnung in einen philosophischen
Zusammenhang. Zum anderen kommt sie der spezifischen Zielsetzung seiner Theorie
entgegen: Während Trendelenburg und der „frühe“ Cohen den Status der zweifelhaft
gewordenen Philosophie durch die Anbindung an die erfolgreichen Naturwissen-
schaften wiederherstellen wollen, möchte Kelsen umgekehrt die durch methodische
Verirrungen zweifelhaft gewordene dogmatische Rechtswissenschaft mittels des An-
schlusses an die neukantische Wissenschaftstheorie sanieren. Die Objektivität rechts-
dogmatischer Sätze ist für diesen Ansatz kein relativ skepsisresistenter Ausgangspunkt,
sondern bleibt, als bloßer Anspruch, problematisch, und die Reine Rechtslehre stellt
nur die Bedingungen fest, unter denen diesem Anspruch stattgegeben werden könnte.

3.2 Die transzendentale Methode


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Was Kelsen genau unter der transzendentalen Methode versteht, und wie diese auf
das Recht bzw. die Rechtswissenschaft angewendet werden kann, wird von ihm nie
näher erläutert. Dadurch, dass er einerseits den wissenschaftstheoretischen Ansatz der
frühen Lehre Cohens übernimmt, dessen Ansatzpunkt die Gesetze und Methoden der
exakten Naturwissenschaften sind, weil sie das Höchstmaß an Objektivität und All-
gemeingültigkeit verkörpern, andererseits seine eigene Theorie in der Folge der späten
Lehre Cohens als „transzendentale“ Analyse der soziokulturellen Praxis der Rechts-
dogmatik versteht, ergeben sich letztlich zwei verschiedene Deutungen der transzen-
dentalen Methode:
Nach der ersten setzt die transzendentale Analyse direkt am wissenschaftlichen
Urteil an, denn dieses ist
„der Grund- und Eckstein der Transzendentalphilosophie, die eben darum nur Kritik der
Wissenschaft, Erkenntniskritik, weil Analyse des synthetischen Urteils sein kann“.34

Aufgabe der transzendentalen Methode ist es nach dieser Deutung, die nichtfaktischen
Ermöglichungsbedingungen der hypothetischen Urteile der Rechtswissenschaft her-
auszuarbeiten; deren Existenz und Geltung wird vorausgesetzt. Ergebnis einer solchen
Präsuppositionsanalyse sind u. a. die Kategorie des Sollens als urteilskonstituierende
Kategorie35 und die Grundnorm als oberstes Geltungskriterium36.
Nach der zweiten Deutung setzt die Analyse an der rechtsdogmatischen Praxis
(und ihren alltagssprachlichen Derivaten) an, in der normative Sätze wie „X sei dazu
berechtigt, Y dazu verpflichtet, usw.“ verwendet werden; Ausgangspunkt sind also „all

33 Hans Kelsen, Hauptprobleme der Staatsrechtslehre (1. Aufl.), Tübingen 1911, Vorrede III.
34 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 128.
35 Hans Kelsen, Reine Rechtslehre (Fn. 1), 21 ff.
36 Hans Kelsen, Reine Rechtslehre (Fn. 1), 66 f.

Franz Steiner Verlag


90 CARSTEN HEIDEMANN

die Tausende von Aussagen, in denen das Rechtsleben sich täglich äußert“.37 Nach die-
ser Deutung ginge es allgemeiner um die Voraussetzungen der mit der Äußerung sol-
cher Sätze beanspruchten Normativität und Objektivität; der Rechtssatz als Urteil der
Rechtswissenschaft wäre weniger Ausgangspunkt der Analyse als vielmehr Ergebnis
einer kritischen Rekonstruktion faktisch geäußerter rechtswissenschaftlicher Sätze.
Für diese Deutung spricht, dass Kelsen die Rechtswissenschaft nicht nur analysieren,
sondern auch einer Methodenkritik unterziehen will.
Will man beide Deutungen miteinander vereinbaren, könnte Kelsens Konzeption
der transzendentalen Methode in drei teilweise ineinander übergehende Schritte ein-
geteilt werden:
Der erste Schritt besteht in einer interpretierenden Bestandsaufnahme. Zunächst
wird festgestellt, welche universalen formalen Eigenschaften die Sätze der Rechtsdog-
matik nach dem Anspruch der sie Äußernden in der Regel aufweisen, nämlich Ob-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

jektivität,38 Normativität39 und Positivität.40 Zusätzlich argumentiert Kelsen auf der


Grundlage seiner neukantisch-szientistischen Wissenschaftsauffassung dafür, dass
diese Eigenschaften auch notwendig beansprucht werden: Dass mit wissenschaft-
lichen Sätzen Objektivität beansprucht werden muss, ergibt sich unmittelbar aus
dem Wissenschaftsbegriff.41 Normativität muss beansprucht werden, damit eine Ab-
grenzung des Gegenstandsbereichs der Rechtswissenschaft möglich und somit ihre
Eigenständigkeit gewährleistet ist.42 Positivität muss beansprucht werden, weil wissen-
schaftliche Urteile auf empirisches „Material“ als Bedingung ihrer Geltung angewiesen
sind.43
In einem zweiten Schritt werden die notwendigen nichtfaktischen Voraussetzun-
gen der Möglichkeit solcher Urteile festgestellt. Allgemeine Voraussetzungen jeder
Wissenschaft sind zunächst die Einheit und Einheitlichkeit ihrer Urteile und die Aus-
schließlichkeit des wissenschaftlichen Urteilssystems für seinen Gegenstandsbereich.
Objektivität setzt Gesetzlichkeit voraus; diese wiederum setzt die Form des Rechts-
satzes als dem Kausalgesetz analoges hypothetisches Urteil (in einem logischen Sinn)
voraus.44 Positivität setzt die Existenz von Regeln voraus, die bestimmte empirische
Tatbestände  – insbesondere Setzungsakte  – als Normgeltungsbedingungen festle-

37 Hans Kelsen, Reine Rechtslehre (Fn. 1), 35.


38 Hans Kelsen, Reine Rechtslehre (Fn. 1), Vorwort III.
39 Hans Kelsen, Reine Rechtslehre (Fn. 1), 33 f.
40 Hans Kelsen, Reine Rechtslehre (Fn. 1), 14; ders., Eugen Hubers Lehre vom Wesen des Rechts, in: Schwei-
zerische Zeitschrift für Strafrecht (1921), 223, 233.
41 Hans Kelsen, Hauptprobleme der Staatsrechtslehre (Fn. 23), Vorrede VII.
42 Hans Kelsen, Reine Rechtslehre (Fn. 1), 35.
43 Hans Kelsen, Reine Rechtslehre (Fn. 1), 14, 64 f.
44 Hans Kelsen, Hauptprobleme der Staatsrechtslehre (Fn. 23), Vorrede VII.

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 91

gen;45 die Annahme rechtlicher Normativität schließlich setzt die Kategorie der peri-
pheren Zurechnung voraus, durch die Bedingung und Zwangsausübung innerhalb des
Rechtssatzes miteinander verbunden werden.46
Zuletzt werden in einem dritten Schritt, dem der Kritik, jene Sätze oder Konzeptio-
nen aus der Rechtswissenschaft ausgeschieden, die nicht mit diesen Präsuppositionen
zu vereinbaren sind oder sich nicht in der Gestalt des Rechtssatzes oder seiner Be-
standteile rekonstruieren lassen.47 Dadurch werden insbesondere soziologische oder
moralische Normbegründungen aus dem Recht ausgeschlossen; und die Begriffe des
subjektiven Rechts, der Pflicht und der Person werden als Implikate des Rechtssatzes,
seiner Bestandteile oder des Zusammenhangs von Rechtssätzen nachgewiesen bzw.
auf diese reduziert.48
Eine exakte Abgrenzung der drei Schritte dürfte allerdings kaum möglich sein.
Ohne eine generalisierende Interpretation rechtsdogmatischer Sätze wird man am
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Anfang der Analyse nicht auskommen können. Dass mit ihrer Äußerung ein An-
spruch auf Objektivität, Normativität und Positivität verbunden ist, ist aber nicht un-
bezweifelbar, und Kelsens Zusatzargument, dass dieser Anspruch erhoben werden
muss, wenn die Wissenschaftlichkeit der Rechtsdogmatik gewahrt bleiben soll, geht
eigentlich über die Aufgabe einer transzendentalen Analyse im Sinne Cohens hinaus
und beruht auf einer bestimmten Auffassung von Wissenschaftlichkeit. Ferner schei-
nen einige Rechtselemente, wie das subjektive Recht und die Pflicht, eine „Doppel-
rolle“ zu spielen. Zum einen gehören Aussagen über Rechte und Pflichten, wie aus
den oben zitierten Beispielen Kelsens für „naive“ Aussagen innerhalb der Rechtspraxis
hervorgeht, zum „Input“ für die transzendentale Methode. Zum anderen sind sie, in
geläuterter Form als „reflektierte“ sekundäre Rechtsbegriffe, Ergebnis einer kritischen
Rekonstruktion der dogmatischen Rechtswissenschaft.49

3.3 Das rechtswissenschaftliche Faktum

Die „faktische Rechtswissenschaft“ als Ansatzpunkt der transzendentalen Methode ist


also identisch mit der soziokulturellen Praxis der Rechtsdogmatik. Es bleibt die Frage,
was Kelsen unter der innerhalb der faktisch gegebenen dogmatischen Rechtswissen-
schaft festgestellten Tatsache versteht, dem „rechtswissenschaftlichem Faktum“ als
Äquivalent zum Naturgesetz, an das jedenfalls der „frühe“ Cohen die transzendentale

45 Hans Kelsen, Reine Rechtslehre (Fn. 1), 64 f.; ders., Der soziologische und der juristische Staatsbegriff, Tü-
bingen 1922, 94 f.
46 Hans Kelsen, Reine Rechtslehre (Fn. 1), 23.
47 Hans Kelsen, Zur Theorie der juristischen Fiktionen, in: Annalen der Philosophie (1919), 47 ff.; ders.,
Reine Rechtslehre (Fn. 1), 1.
48 Etwa Hans Kelsen, Reine Rechtslehre (Fn. 1), 47 ff.
49 Etwa Hans Kelsen, Reine Rechtslehre (Fn. 1), 47 ff.

Franz Steiner Verlag


92 CARSTEN HEIDEMANN

Methode im naturwissenschaftlichen Bereich anknüpft. Die Antwort liegt nach dem


oben Ausgeführten nahe: Das rechtswissenschaftliche Faktum ist die Rechtsnorm,
und diese ist identisch mit dem Urteil der Rechtswissenschaft selbst.
Dies erläutert Kelsen vor allem im Rahmen seiner Debatte mit Fritz Sander. Dort
heißt es zunächst zu der auch seiner Reinen Rechtslehre zugrundeliegenden neukanti-
schen Erkenntnistheorie, Erfahrung als Wissenschaft sei ein Verfahren der Erkenntnis,
in dem der reine Verstandesbegriff (die Kategorie) auf sinnliche Anschauungen bezo-
gen werde. Diese Einheitsbeziehung realisiere sich in den synthetischen Urteilen der
Naturwissenschaft, deren Zusammenhang die „Erfahrung als Wissenschaft im Sinne
Kants“, also letztlich die Natur, darstelle.50 Das Recht der Rechtswissenschaft sei ana-
log der Natur der Naturwissenschaft zu begreifen, wobei an die Stelle des Anschau-
ungsmaterials das Willensaktmaterial des Gesetzgebers trete:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

„Im erkennenden Bewusstsein, d. h. in den als Rechtssätzen auftretenden Urteilen der


Rechtswissenschaft, drückt sich das – in der Urteilssphäre objektivierte – Wollen als Sol-
len aus. Nur in dieser Sphäre kommt aber das Recht als Gegenstand der Rechtswissen-
schaft in Betracht. Als Gegenstand der Rechtswissenschaft ist dann das Recht ebenso ein
System von Urteilen über das Recht, wie die Natur als Gegenstand der Naturwissenschaft
ein System von Urteilen über die Natur.“51

Das aus Normen bestehende Recht ist daher identisch mit dem System der hypotheti-
schen normativen Urteile der Rechtswissenschaft, den Rechtssätzen. Die Rechtsnorm
selbst ist also ein hypothetisches Erkenntnisurteil, in dem die Kategorie der normati-
ven (peripheren) Zurechnung zur Anwendung kommt.52
Innerhalb der Theorie Kelsens ist diese Konzeption – die er um 1940 herum wie-
der aufgibt53 – im Rahmen seiner nichtrealistischen neukantischen Erkenntnistheorie
konsequent; sie hat auch hohes Erklärungspotential. Sie ermöglicht ihm, den Norm-
begriff ohne Rekurs auf eine Theorie erkenntnisunabhängiger abstrakter Entitäten
zu explizieren. Ferner wird die in der Reinen Rechtslehre insgesamt nicht klare und
dementsprechend in der Sekundärliteratur umstrittene Definition der Geltung als
Existenz oder Sollen der Norm54 unproblematisch; es handelt sich einfach um die Gel-

50 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 127.


51 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 181. Angesichts der eindeutigen Formulierung ist
die Auffassung Christoph Kletzers, Kelsen wolle in dieser Passage eine abbildtheoretische Erkenntnistheo-
rie vortragen, unhaltbar; sie wird von ihm auch nicht weiter begründet (vgl. Christoph Kletzer, Fritz San-
der, in: Der Kreis um Hans Kelsen, Wien 2008, 459).
52 Hans Kelsen, Hauptprobleme der Staatsrechtslehre (Fn. 23), Vorrede VI; ders., Allgemeine Staatslehre, Ber-
lin 1925, 48 ff.
53 Kenntlich insbesondere an der Unterscheidung zwischen der Rechtsnorm und dem Rechtssatz (ver-
standen als Urteil über die Rechtsnorm), die erstmals in Hans Kelsen, The Pure Theory of Law and Analy-
tical Jurisprudence, in: Harvard Law Review, Vol. 55 (1941), 44–70, durchgeführt wird.
54 Vgl. beispielhaft Joseph Raz, Legal Validity, in: ders., The Authority of Law, Oxford 1979, 146 ff.; Eugenio
Bulygin, An Antinomy in Kelsen‘s Pure Theory of Law, in; Normativity and Norms Critical Perspectives on

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 93

tung eines Urteils im logischen Sinne; diese ist dann gegeben, wenn der Urteilsakt
bzw. der das Urteil formulierende Satz wahr ist. Auch logische Beziehungen zwischen
Normen sind unkompliziert möglich, wenn diese als Erkenntnisurteile aufgefasst wer-
den.55 Zuletzt vermeidet die Konzeption die sich aus der späteren Trennung zwischen
der Rechtsnorm, als einer der Erkenntnis vorgegebenen Entität, und dem Rechtssatz,
als einer isomorphen Reproduktion der Norm auf der Erkenntnisebene, ergebenden
Probleme.56
Kelsen gibt zu, dass es einen „Schein der Paradoxie“ habe, wenn das Recht mit
den Urteilen „über“ das Recht identifiziert werde,57 doch hält er seine Auffassung für
unvermeidlich, wenn man mit Kant davon ausgeht, dass die Erkenntnis selbst ihren
Gegenstand konstituiert und Erkenntnis sich stets in logischen Urteilen äußert. Und
tatsächlich beruft Kelsen sich durchaus zu Recht auf Kant. Wenn Referenz nach Kant
als eine Angelegenheit innerhalb der Erkenntnis gedacht werden muss, dann kann sie
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

letztlich allein durch die Konstitution des Erkenntnisgegenstandes in gültigen Urtei-


len herbeigeführt werden. Denn Kant führt alle Handlungen des Verstandes auf Urtei-
le zurück;58 und um im Rahmen der metaphysischen Deduktion die Herleitung der
Kategorien aus der Tafel der Urteilsfunktionen zu rechtfertigen, schreibt er:
„Dieselbe Funktion, welche den verschiedenen Vorstellungen in einem Urteile Einheit
gibt, die gibt auch der bloßen Synthesis verschiedener Vorstellungen in einer Anschauung
Einheit, welche, allgemein ausgedrückt, der reine Verstandesbegriff heißt.“59

Begriffe werden also im Urteil durch dieselbe Funktion und dieselbe Verstandeshand-
lung zur Einheit gebracht wie Vorstellungen in einer Anschauung zur Einheit des kate-
gorial bestimmten Erkenntnisgegenstandes; Kant unterscheidet nicht klar zwischen
„propositional-logischer“ und „kategorial-erkenntnistheoretischer“ Synthesis:60 In-
dem die Kategorien Urteile konstituieren, konstituieren sie zugleich den objektiven
Gegenstand der Erkenntnis. Zwar mag es so scheinen, dass dies nur auf einer begriff-
lichen Äquivokation beruhe, weil Kant in seiner vorstellungstheoretischen Diktion
nicht den Fall der Gegenstandssynthese (Subsumtion einer Anschauung unter einen
Begriff) von dem der Urteilssynthese (Subsumtion eines Begriffs unter einen anderen

Kelsenian Themes, Oxford 1998, 297 ff.


55 Vgl. zur urteilstheoretischen Normkonzeption, ihren Vorteilen und den Gründen für ihre Aufgabe
Carsten Heidemann, Noch einmal: Stanley L. Paulson und Kelsens urteilstheoretischer Normbegriff, in:
ARSP 93, 2007, 345–362.
56 Zu diesen Problemen vgl. etwa Carlos Santiago Nino, Some Confusions surrounding Kelsen‘s Concept
of Validity, in: Normativity and Norms Critical Perspectives on Kelsenian Themes (hrsg. v. S. L. Paulson und B.
Litschewski Paulson), Oxford 1998, 253 ff.; Roberto J. Vernengo, Kelsen‘s „Rechtssätze“ as Detached State-
ments, in: Essays on Hans Kelsen (hrsg. v. R. Tur und W. Twining), Oxford 1986, 99 ff.
57 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 182.
58 Immanuel Kant, Kritik der reinen Vernunft (Fn. 22), 110 (A 69/B 94).
59 Immanuel Kant, Kritik der reinen Vernunft (Fn. 22), 117 f. (A 79/B 104 f.).
60 Vgl. dazu Reinhold Aschenberg, Sprachanalyse und Transzendentalphilosophie (Fn. 6), 106.

Franz Steiner Verlag


94 CARSTEN HEIDEMANN

Begriff) unterscheide.61 Ferner werden wichtige Teile der Kritik der Vernunft, etwas das
Schematismuskapitel, überflüssig, wenn Urteils- und Gegenstandssynthese identisch
sind. Aber wenn alle Verstandeshandlungen auf Urteilen beruhen, dann ist es kaum
denkbar, wie die Kategorien außerhalb eines Urteils angewendet werden sollten und
wie die Gegenstandssynthese anders als im Urteil erfolgen sollte. Man kann Kelsen
daher auch keine Fehlinterpretation Kants vorwerfen.
Zwar ist Kelsens Gleichsetzung von Urteil und Erkenntnisgegenstand, soweit der
letztere als „Objekt“ verstanden wird, kontraintuitiv; plausibler ist es, zwischen dem
gegenstandskonstitutiven, aber selbst nicht gegenständlichen Urteil, das durch einen
Aussagesatz ausgedrückt werden kann, und dem innerhalb eines Urteils der Meta-
ebene zum Gegenstand „gemachten“ objektiven Urteil, der Tatsache, die durch einen
dass-Satz bezeichnet werden kann, zu unterscheiden.62 Doch entspricht die urteils-
theoretische Konzeption des wissenschaftlichen Faktums durchaus dem damaligen
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Stand der akademischen Philosophie. Wenn Cohen wissenschaftliche Fakten in den


als hypothetische Urteile verstandenen „Gesetzen“ lokalisiert63 und Frege Tatsachen
als „wahre Gedanken“ definiert,64 bewegen sie sich grundsätzlich in demselben Theo-
rieraum wie Kelsen.
Charakteristisch für Kelsens Konzeption des wissenschaftlichen Urteils sind zwei
weitere Merkmale:
Zum einen ist das Urteil eine objektive Struktur;65 es darf nicht mit dem psychi-
schen Urteilsakt oder einem mentalen Gehalt identifiziert werden. Die psychologi-
schen Tatsachen seien im Verhältnis zu den logischen Geltungen nur die conditio sine
qua non, aber nicht die conditio per quam des Rechts in seinem Geltungsaspekt,66
Dieser „Logizismus“ lässt sich bereits auf Kant zurückführen, der in den Prolegomena
schreibt, dass
„hier nicht vom Entstehen der Erfahrung die Rede sei, sondern von dem, was in ihr liegt.
Das erstere gehört zur empirischen Psychologie, und würde selbst auch da ohne das zwei-
te, welches zur Kritik der Erkenntnis und insbesondere des Verstandes gehört, niemals
gehörig entwickelt werden können.“67

61 So Reinhold Aschenberg, Sprachanalyse und Transzendentalphilosophie (Fn. 6), 105 f.


62 Vgl. dazu Carsten Heidemann, Die Norm als Tatsache, Baden-Baden 1997, 297 ff.
63 Etwa Hermann Cohen, Kants Begründung der Ethik (Fn. 3), 20 f. Vgl. auch Hermann Cohen, Logik der
reinen Erkenntniss, Berlin 1902, 299: „Die Kategorie des Gesetzes erweist sich als solche auch in den Geistes-
wissenschaften: in der Ethik, als ihrer Logik; sowie in der Jurisprudenz, als ihrer Mathematik […]. Alle
haben in dem hypothetischen Urteil und in der Bedingung ihren Typus.“
64 Gottlob Frege, Der Gedanke, in: ders., Logische Untersuchungen (hrsg. v. G. Patzig, 3. Aufl.), Göttingen
1986, 74.
65 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 205 ff.
66 Etwa Hans Kelsen, Das Problem der Souveränität und die Theorie des Völkerrechts, Tübingen 1920, Vorrede
VI.
67 Immanuel Kant, Prolegomena (Fn. 5), 304 (A 87).

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 95

Die Abgrenzung logischer bzw. erkenntnistheoretischer Argumente von psycho-


logischen Argumenten wurde später von Frege und Husserl aufgenommen und ge-
hört seither zum Grundbestand des Neukantianismus Marburger68 und südwestdeut-
scher69 Prägung.
Zum anderen ist das objektive, mit dem Erkenntnisgegenstand identische Urteil
stets ein „geltendes“ Urteil;70 Geltung ist gewissermaßen die logische Daseinsweise
des Urteils. Dabei tendiert Kelsen, ohne dass dies eindeutig wird, zu einem normati-
ven Geltungsbegriff. Das entspricht einem zentralen Theorem der südwestdeutschen
Schule; und so nimmt Kelsen nur zweimal kurz und etwas unklar Bezug auf Windel-
bands Gleichsetzung von Geltung und Sollen der logischen Gesetze71 und geht sonst
ohne weitere Erläuterung davon aus, dass die Geltung eines objektiven Urteils, sein
„Tatsachesein“, nicht der Existenz eines Dinges zu vergleichen ist und sich auch nicht
seiner Übereinstimmung mit einer von ihm unabhängigen Wirklichkeit verdankt.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Vielmehr sei darunter zu verstehen, dass es den Erkenntnisregeln genügt72 – „Tatsa-


che ist, was ich denken soll“, wie Rickert dies formuliert.73 Diese normative Geltungs-
konzeption geht auf Hermann Lotze zurück, für den Geltung die nichtgegenständ-
liche normative Daseinsform von „Wahrheiten“ ist.74 Sie macht Kelsens Sollensbegriff
ambivalent:75 Denn das Sollen der geltenden Norm liegt zum einen darin, dass eine
bestimmte Handlung „gesollt“ ist, zum anderen in der normativen Anforderung, ent-
sprechend zu urteilen. Folglich gilt die Norm „man soll x tun“, als normatives Urteil,
genau dann, wenn man denken oder urteilen soll, dass man x tun soll.76
Auf diese Weise können der Norm- und der übergeordnete Tatsachenbegriff ohne
eine „starke“ Ontologie abstrakter Entitäten erklärt werden; zugleich liefert diese
Konzeption ein Argument gegen jede Form eines Psychologismus. Allerdings darf
das „Denkensollen“ des Urteils nicht dem robusten inhaltlichen Sollen der Norm-als-
Faktum verglichen werden – es ähnelt eher dem impliziten praxisimmanenten Sollen
einer Wittgensteinschen Regel, die keine gegenständliche Existenz hat. Eine vergleich-
bare zeitgenössische Version dieser Konzeption findet sich etwa in Hilary Putnams
internem Realismus; dort heißt es – wenn auch tentativ – :

68 Etwa Hermann Cohen, Logik der reinen Erkenntniss (Fn. 63), 52 ff.


69 Etwa Heinrich Rickert, Der Gegenstand der Erkenntnis (2. Aufl.), Tübingen 1904, Vorrede VI.
70 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 205.
71 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 207 ff.; ders., Der soziologische und der juristische
Staatsbegriff, Tübingen 1922, 81 f.
72 Hans Kelsen, Die Grundlagen der Naturrechtslehre und des Rechtspositivismus, Charlottenburg 1928, 62.
73 Heinrich Rickert, Zwei Wege der Erkenntnistheorie, in: Kant-Studien 14, 1909, 185. Diese These macht
Rickert sich in Auseinandersetzung mit der Theorie Theodor Lipps zu eigen.
74 Hermann Lotze, Logik, Leipzig 1874, 501 ff.
75 Hans Kelsen, Rechtswissenschaft und Recht (Fn. 2), 208.
76 Vgl. dazu näher Carsten Heidemann, Geltung und Sollen: Einige (neu-)kantische Elemente der Reinen
Rechtslehre Hans Kelsens, in: Neukantianismus und Rechtsphilosophie (hrsg. v. Robert Alexy, Lukas H. Mey-
er, Stanley L. Paulson und Gerhard Sprenger), Baden-Baden 2002, 203–222.

Franz Steiner Verlag


96 CARSTEN HEIDEMANN

„Würde ich es wagen, ein Metaphysiker zu sein, dann würde ich wohl ein System schaffen,
in dem es nichts als Verpflichtungen gibt. […] In meinem Phantasiegebilde von mir als
dem metaphysischen Superhelden würden sich alle Tatsachen zu Werten auflösen. Dass
in meinem Zimmer ein Stuhl steht, würde als eine Menge von Verpflichtungen analysiert
[…]: etwa die Verpflichtung, zu denken, dass in diesem Zimmer ein Stuhl steht, falls die
epistemischen Bedingungen ‚gut‘ genug sind.“77

Putnams Theorie des internen Realismus lässt sich durchaus der neukantischen Ideen-
welt zuordnen. Im Übrigen lassen sich auch bei Kant selbst die Grundzüge einer nor-
mativistischen Erkenntnis- und Geltungstheorie nachweisen – so werden Begriffe von
ihm (auch) als Regeln aufgefasst, und die nach der Kritik der reinen Vernunft grund-
legende Eigenschaft des Verstandes, Spontaneität, hängt letztlich von seiner Fähigkeit
zum Regelfolgen ab.78
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

4. Resümee und Ausblick

Ich fasse zusammen: Das „Faktum der Wissenschaft“ macht in doppelter Bedeutung
den Kern des Kelsenschen Neukantianismus in den zwanziger Jahren aus.
Zum einen versteht Kelsen seine Lehre in der Folge von Kant, Trendelenburg und
Cohen als methodenkritische Theorie der als soziale Praxis etablierten und insofern
als Faktum gegebenen dogmatischen Rechtswissenschaft; ihre wesentliche Aufga-
be ist es, die Präsuppositionen der mit der Äußerung rechtswissenschaftlicher Sätze
beanspruchten Objektivität, Normativität und Positivität herauszuarbeiten und die
Rechtsdogmatik so zu rekonstruieren, dass sie mit den dabei gewonnenen Resultaten
im Einklang steht. Für diesen Theoriestatus beruft Kelsen sich zwar zu Recht auf die
Kantsche Philosophie; er ist aber das Resultat einer dreifachen Depotenzierung der
transzendentalen Analyse Kants: Kants Ausgangspunkt ist die unbezweifelbare trans-
zendentale Einheit der Apperzeption, aus der „progressiv“ die Kategorien als notwen-
dige Ermöglichungsbedingungen jeglicher Gegenstandserkenntnis gewonnen werden
können. Schon bei Kant wird dieser Ansatz ein erstes Mal relativiert, indem er die
transzendentale Analyse für den Zweck einer leichter erfassbaren Darstellung an den
synthetischen apriorischen Urteilen, die in Mathematik und reiner Naturwissenschaft
„gegeben“ sind, ansetzen lässt. Eine zweite Relativierung findet sich in der Theorie

77 Hilary Putnam, Wozu die Philosophen?, in: ders., Von einem realistischen Standpunkt, Reinbek 1993,
203–220, 215.
78 Vgl. dazu Markus Willaschek, Die „Spontaneität des Erkenntnisses“. Über die Abhängigkeit der „Trans-
zendentalen Analytik“ von der Auflösung der dritten Antinomie, in: Metaphysik und Kritik Interpretationen
zur „Transzendentalen Dialektik“ (hrsg. v. Jiří Chotaš, Jindřich Karásek und Jürgen Stolzenberg),Würzburg
2010, 165–184.

Franz Steiner Verlag


Das „Faktum der Rechtswissenschaft“ bei Hans Kelsen 97

Trendelenburgs und den frühen Schriften Cohens, deren Ansatzpunkt die (aposte-
riorischen) Methoden und Ergebnisse der systematisch vorgehenden und „erfolgrei-
chen“ Naturwissenschaften sind. Zu einer dritten Relativierung kommt es in der Ethik
Cohens, die an der Rechtswissenschaft als einer „gegebenen“ soziokulturellen Praxis
ansetzt. Kelsens neukantische Erkenntnistheorie verbindet den szientistischen Ansatz
der Relativierung der transzendentalen Methode auf die objektiven, exakten Natur-
wissenschaften mit dem Ansatz der Relativierung dieser Objektivität auf die Praxis
der Rechtsdogmatik.
Zum anderen setzt Kelsen das rechtswissenschaftliche Faktum, die Rechtsnorm,
mit dem hypothetischen Zurechnungsurteil der Rechtswissenschaft gleich, das nicht-
psychologistisch als objektive Denkstruktur zu verstehen ist und dessen Daseinsweise
oder Geltung in der Folge Lotzes und Windelbands nicht analog der Existenz eines
Dinges aufgefasst werden darf, sondern am ehesten als normative Anforderung zu ver-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

stehen ist, dieses Urteil so zu denken.


Beide Konzeptionen sind von zentraler Bedeutung innerhalb der Reinen Rechts-
lehre. Legt man Kelsens Deutung der transzendentalen Methode zugrunde, so ist es
zwar einerseits Aufgabe seiner Theorie des Rechts eine „Rechtsontologie“ auszuarbei-
ten; insofern liefert sie durchaus eine Metaphysik des Rechts. Andererseits ergibt sich
diese Rechtsontologie aber (nur) als Ergebnis einer Präsuppositionsanalyse der in der
existierenden soziokulturellen Praxis der Rechtsdogmatik erhobenen Ansprüche. Das
ist eine metaphysisch bescheidene Konzeption, die auch unabhängig vom Theorie-
rahmen der Reinen Rechtslehre plausibel ist. Innerhalb der Reinen Rechtslehre hat
sie hohen Erklärungswert  – beispielsweise dürfte der Status der Grundnorm ohne
sie kaum verständlich sein79 – und bedingt zugleich ihre kritische Funktion: Die zur
Rechtsontologie gehörenden notwendigen Elemente des Rechts umfassen (nur) das
rechtswissenschaftliche Zurechnungsurteil, seine Bestandteile, seine Implikate und
seine Voraussetzungen.80 Das bedeutet, dass alle Rechtsbegriffe entweder auf diese
Elemente zurückgeführt oder aus der Rechtserkenntnis ausgeschieden werden müs-
sen bzw. als Norminhalt nur kontingenter Bestandteil des Rechts sein können. Dies

79 Eine Präsupposition eines Urteils ist eine notwendige Voraussetzung dafür, dass es überhaupt geltungs-
bzw. wahrheitsdifferent ist. Die Grundnorm ist unter Zugrundelegung der Darstellung Kelsens z. B. inso-
fern eine Präsupposition jeder Rechtsnorm / jedes rechtswissenschaftlichen Urteils, als sie überhaupt erst
die Möglichkeit schafft, geltende von nichtgeltenden rechtswissenschaftlichen Urteilen zu unterscheiden.
80 Dabei kommt es in den zwanziger Jahren zu einer merkwürdigen „Ungleichzeitigkeit“ der Konzeptio-
nen Kelsens. Die Auffassung der Rechtsnorm als dem Kausalgesetz analog strukturiertes Zurechnungs-
urteil passt zu der eindimensionalen statischen Normauffassung, die Kelsen in den Hauptproblemen der
Staatsrechtslehre entwickelt hat. Mit der Übernahme der Merklschen Stufenbautheorie sind Formen der
Rechtsnorm erforderlich geworden (Ermächtigungsnorm, individuelle Norm), die nicht auf das hypothe-
tische Zurechnungsurteil reduziert werden können; dies macht letztlich Kelsens Unterscheidung zwischen
einer eindimensionalen „statischen“ Rechtsbetrachtung (die mit der Form des hypothetischen Urteils aus-
kommt) und einer „dynamischen“ Rechtsbetrachtung erforderlich, die nur bedingt ineinander übersetzbar
sind.

Franz Steiner Verlag


98 CARSTEN HEIDEMANN

betrifft u. a. die Begriffe des Staates und der Person, die jeweils als Mengen von Rechts-
sätzen aufgefasst werden, sowie der Pflicht und des subjektiven Rechts, die als Impli-
kate des Rechtssatzes bzw. als auf diesen reduzierbar aufgefasst werden.81
Noch positiver muss die Bewertung des urteilstheoretischen Normbegriffs ausfal-
len. Mit ihm werden die Aporien von Kelsens späterer korrespondenztheoretischen
Erklärung der Normerkenntnis ebenso wie die Probleme einer Willenstheorie der
Norm vermieden,82 und die schwierige Aufgabe, Objektivität, Normativität und Posi-
tivität des Rechts miteinander zu vereinbaren, lässt sich ohne großen Aufwand bewerk-
stelligen: Normativität ist eine Frage der spezifischen Urteilsmodalität; Objektivität
wird gewährleistet durch die Urteilsform; Positivität ist eine Frage der empirischen
Kriterien für die Geltung der normativen Urteile.83 Insgesamt handelt es sich damit
um eine Konzeption, mit deren Hilfe der komplexe Status der Rechtsnorm zwischen
Sein und Sollen auch außerhalb des Kontextes der Reinen Rechtslehre ohne großen
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

metaphysischen Aufwand plausibel erklärt werden kann.

Carsten Heidemann
Bordesholm, heidemann.carsten@gmail.com

81 Vgl. Hans Kelsen, Reine Rechtslehre (Fn.  1), 46 (zur Pflicht), 40 (zum subjektiven Recht), 52 ff. (zur
Person), 115 ff. (zum Staat).
82 Vgl. dazu Carsten Heidemann, Noch einmal: Stanley L. Paulson und Kelsens urteilstheoretischer
Normbegriff (Fn. 55), 352, 354.
83 Zu einem Versuch einer Rekonstruktion des urteilstheoretischen Normbegriffs mittels der Methodik
der sprachanalytischen Philosophie vgl. Carsten Heidemann, Die Norm als Tatsache (Fn. 62), 324 ff.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law

MAXIMILIAN KIENER
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: This essay raises a new challenge for Hans Kelsen’s The Pure Theory of Law, i. e.
what I call the Cognitivist Challenge This challenge concerns the question as to how to
retain Kelsen’s view about relative legal norms, i. e. that one can objectively cognise these
norms, while not rejecting his view about absolute moral norms, i. e. that one cannot objec-
tively cognise those norms. After explaining the Cognitivist Challenge in detail, I will pres-
ent Positive Legal Fictionalism as a solution to it. I will claim that Positive Legal Fictionalism
is rooted in Kelsen’s own statements and provides the specific epistemic advantages that
Kelsen’s project of cognising the law requires. Positive Legal Fictionalism thereby deepens
the understanding of Kelsen’s project as a ‘theory’ and as ‘pure.’

Keywords: Hans Kelsen, Legal Positivism, Fictionalism, Cognition, Validity, Basic Norm

Schlagworte: Hans Kelsen, Rechtspositivismus, Fiktionalismus, Erkenntnis, Geltung,


Grundnorm

Introduction

Hans Kelsen (1881–1973) was one of the leading so-called ‘legal positivists’ in the 20th
century. He was a legal positivist because he exclusively focused on “questions of what
the law is and how the law is made [i. e. the law as it is posited], not [on] (…) questions
of what the law ought to be.”1 Kelsen separated his position from natural law theory by
denying that law, in order to be law, “must have some concern for justice, be it a matter

1 Kelsen, Hans. Introduction to the Problems of Legal Theory A Translation of the First Edition of the Reine
Rechtslehre or Pure Theory of Law (Henceforth abbreviated as PTL1). Translated by B. L. Paulson and S.
Paulson. 2004, Oxford Clarendon, 7. Emphasis added. In addition to referring to the English translation, I
will also provide references to the German original. See Kelsen, Hans. Reine Rechtslehre Studienausgabe der
1  Auflage 1934 Edited by M. Jestaedt. (Henceforth abbreviated as RR1). 2008, Tübingen Mohr Siebeck, 15.

Franz Steiner Verlag


100 MAXIMILIAN KIENER

of assuring an ethical minimum, be it a matter of attempting, however inadequately, to


be ‘right’ law, that is, simply, to be just.”2
However, unlike many other positivists, Kelsen denies that the validity of legal
norms (or in other words their normative bindingness) can ever be derived from the
fact that people by and large comply with them. Kelsen adheres to a strict distinction
of ‘is’ versus ‘ought’ and claims that a norm (i. e. an ‘ought’) can never receive its valid-
ity from a fact (i. e. an ‘is’) but only from another norm. Accordingly, a norm is valid if
and only if it can be traced back to a higher norm which confers validity upon it. As the
chain of validity-conferring norms cannot go back infinitely, Kelsen claims that the va-
lidity of legal norms ultimately depends on a first norm, the validity of which no longer
depends on another norm. Kelsen calls such a norm the basic norm and presents it as
a non-positive norm (i. e. a norm that is not posited and therefore not part of a given
legal system) and as a norm that is merely presupposed in thought. The basic norm
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

authorises the highest positive norm, e. g. the constitution, and can be schematically
formulated as follows:
“Coercive acts ought to be performed under the conditions and in the manner in which
the historically first constitution, and the norms created according to it, prescribe. (In
short: one ought to behave as the constitution prescribes.)”3

As the basic norm is necessary for the validity of norms, it will also be crucial for the
cognition of the law as objectively valid, which Kelsen describes as his main objective:
“my aim from the very beginning was to raise it [i. e. jurisprudence] to the level of a genu-
ine science, a human science. The idea was to develop those tendencies of jurisprudence
that focus solely on cognition of the law rather than on the shaping of it, and to bring the
results of this cognition as close as possible to the highest values of all science: objectivity
and exactitude.”4

In this essay, I aim to make two contributions to the debate on Kelsen. Firstly (1), I
want to draw attention to a serious challenge to Kelsen’s project of the cognition of the
law, pertinent to both editions of The Pure Theory of Law, which I call the Cognitivist
Challenge and which has not received attention in the literature. Secondly (2), I want
to show how trying to solve the Cognitivist Challenge leads to a deeper understanding
of Kelsen’s project as a ‘theory’, i. e. a project concerned with cognition, and how it is
‘pure’, i. e. separated from morality as well as separated from sciences concerned with

2 Kelsen, PTL1, 22. (RR1, 32.)


3 Kelsen, Hans. Pure Theory of Law Translated from the Second (Revised and Enlarged) German Edition by
Max Knight. (Henceforth abbreviated as PTL2). 2009, Clark The Lawbook Exchange, Ltd., 201. Also with
regards to the second edition, I will provide references to the German original: Kelsen, Hans. Reine Re-
chtslehre Studienausgabe der 2  Auflage 1960 Edited by M. Jestaedt. (Henceforth abbreviated as RR2). 2017,
Tübingen Mohr Siebeck, 359. See also Kelsen, RR1, 76.
4 Kelsen, PTL1, 1. (Footnote 1).

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 101

matters of fact as opposed to norms. In particular, I will aim to solve the Cognitivist
Challenge by presenting an account I call Positive Legal Fictionalism. From an exeget-
ical perspective (2.1), I will claim that Positive Legal Fictionalism can be firmly rooted
in Kelsen’s own statements, albeit it sometimes needs to develop Kelsen’s statements
further rather than taking them for granted. Therefore, Positive Legal Fictionalism will
be a Kelsenian but not quite Kelsen’s own account. From a systematic perspective
(2.2), I will then explain the specific epistemic advantages that Positive Legal Fictional-
ism provides for Kelsen’s project of cognising the law, thereby solving the Cognitivist
Dilemma, and emphasise the value of fictionalism for legal positivism. Finally (3), I
summarise my results and draw conclusions.
This essay is very ambitious. To address and solve the Cognitivist Challenge, I will
need to take into account a very wide scope of Kelsen’s primary texts: Kelsen’s early
writings, the two editions (from 1934 and 1960) of Kelsen’s The Pure Theory of Law,
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

work from Kelsen’s late period, as well as Kelsen’s correspondence with other scholars.
In addition, I will need to link the debate on Kelsen to claims from the philosophical
and metaethical debate on so-called fictionalism, i. e. a link which has not been ex-
plored in the debate on Kelsen so far. And finally, I will also always need to balance
exegetical and systematic aspects in the course of developing my arguments.
The combination of such ambition and the usual space constraints will prevent me
from discussing the wide body of secondary literature on Kelsen’s work in detail. On
the one hand, I will thereby have to leave the reader with room for further inquiry, i. e.
an inquiry into how my approach exactly aligns or conflicts with the majority view on
some of Kelsen’s claims. On the other hand, it is precisely such an approach that will
enable me to focus on very important exegetical and systematic aspects in sufficient
depth, thereby providing a fresh take on many of Kelsen’s claims. Hence, I hope that
the way in which I will proceed in this essay, including my scare treatment of secondary
literature, is not only excused with reference to space constraints but also appreciated
as intrinsically linked to the object to my inquiry and the ambition I pursue.

1. The Cognitivist Challenge

Hans Kelsen presents The Pure Theory of Law explicitly as a “theory.” By “theory”, he
means that The Pure Theory of Law “aims solely at cognition of its subject-matter”5 and
is therefore a genuinely scientific project. Kelsen then identifies as the subject-matter
the legal norm, which he claims can be stated in the form of a conditional: it (legally)
ought to be the case that if some natural event X occurs then the legal consequence Y
is imposed. Kelsen continuously emphasizes that one can reach objective cognition

5 Kelsen, PTL1, 7. (Footnote 1).

Franz Steiner Verlag


102 MAXIMILIAN KIENER

of the legal norm, and at its core the ‘legal ought’, thereby grounding a genuine “Re-
chtswissenschaft”6 (legal science). Objective cognition consists in the cognition of a
legal norm’s objective validity. A legal norm is objectively valid if and only if it is genu-
inely normatively binding, which means that the legal norm’s normative force cannot
be reduced to either subjective wishes or preferences or to any other psychological or
sociological facts.
On the other hand, Kelsen endorsed a diametrically opposed view about moral
norms. Most notably in the first edition of The Pure Theory of Law, Kelsen contends
that a moral norm is a “logosfremdes Objekt”7 (alien to objective inquiry8) and “can-
not be accounted for by way of rational cognition.”9 That something morally ought
to be the case does not qualify as a suitable object of a “Wissenschaft.” Moral norms
cannot be objectively valid because, if they exist at all, they always depend on psycho-
logical or sociological facts. And if they cannot be objectively valid, there cannot be
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

objective cognition of them in the sense outlined.


Hence, taken at face value, Kelsen draws a stark contrast between legal and moral
norms in the first edition of The Pure Theory of Law. However, a closer look reveals that
there is another hidden distinction. When Kelsen talks about moral norms in the first
edition, he always talks about absolute moral norms;10 and when he talks about legal
norms, he always talks about relative legal norms. A norm is absolute if its validity does
not depend on any other norm, and relative if its validity depends on another norm.
Therefore, Kelsen does not contrast moral norms with legal norms after all. He con-
trasts absolute moral norms with relative legal norms.
Unfortunately, employing these two different distinctions at once (i. e. moral ver-
sus legal, relative versus absolute) obfuscates what Kelsen aims to say. This is because
the distinction between moral and legal norms does not map on to the distinction
between absolute and relative norms: there may be relative moral norms (as ethical
relativism claims) and absolute legal norms (as natural law theory claims). And Kelsen
explicitly recognises this in the second edition of The Pure Theory of Law. But if there
can be relative moral norms and absolute legal norms, this leaves us with the following
question: do absolute moral norms fail as objects of cognition because they are moral
norms or because they are absolute norms? What exactly is Kelsen’s position? Unfortu-
nately, both alternatives lead to problems.
Let us first assume that norms cannot be an object of cognition if they are mor-
al norms and thereby follow what Kelsen seems to claim in the first edition. On this

6 Kelsen, RR1, 15. See also the foreword. “Rechtswissenschaft”, unlike “legal science”, does not suggest af-
finity to natural sciences. (Footnote 1).
7 Kelsen, RR1, 27. (Footnote 1).
8 The translation in Kelsen PTL1 as “alien to logic” seems too narrow. See Kelsen, PTL1, 17. (Footnote 1).
9 Kelsen, PTL1, 17. (Footnote 1).
10 In PTL1, Kelsen always implies that moral norms are absolute. See Kelsen, PTL1, 16–18, 22–23, 28, 34.
(RR1, 26–29, 33, 40, 47). (Footnote 1).

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 103

reading, the combination of Kelsen’s view about legal and moral norms becomes very
puzzling. This is because Kelsen himself stressed that the ‘legal ought’ and the ‘moral
ought’ converge on the category of ‘ought’ as their genus proximum.11 Therefore, one
might think if the ‘legal ought’, unlike the ‘moral ought’, can be an object of cognition,
this must be because of what is distinct about the ‘legal ought.’ However, Kelsen’s an-
swer is that the “distinction between law and morality consists in the fact that the law is
a coercive order”12 whereas morality is not such a coercive order, viz. morality does not
impose a “coercive act of the state.”13 But this does not explain why one ‘ought’ (legal)
is an object of cognition while the other ‘ought’ (moral) is not. Hence, this first inter-
pretation of Kelsen’s contrast of different norms seems to lead to a dead end.
But how about the interpretation, according to which Kelsen claims that norms
cannot be an object of cognition if they are absolute? This seems to be the position
that Kelsen later endorsed in the second edition of The Pure Theory of Law. Kelsen
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

claims that “an absolute value in general and an absolute moral value in particular is
rejected from the point of view of scientific cognition.”14 On this basis, both moral and
legal norms can be objectively valid and an object of cognition, as long as they are
relative. Consequently, Kelsen concedes in the second edition, opposing his claims in
the first edition, that ethics is the “discipline directed toward their [i. e. moral norms’]
cognition”15 and refers to ethics as “the science of morals.”16 However, even on this read-
ing, it remains puzzling why only relative but not absolute norms can be an object of
cognition? How can relativity facilitate objective validity and objective cognition? And
moreover, after giving up the distinction between the moral and the legal, how can
Kelsen preserve the distinction between morality and the law, which he is so keen to
stress as part of legal positivism? Hence, the second interpretation of Kelsen’s contrast
of different norms also leads to difficult questions.
In the following, I will use the term Cognitivist Challenge to describe the challenge of
how to retain Kelsen’s view about the relative legal norms, i. e. that one can objectively
cognise these norms, while not rejecting his view about the absolute moral norms, i. e.
that one cannot objectively cognise these norms A solution to the Cognitivist Chal-
lenge will be of vital importance for understanding Kelsen’s project as a ‘theory’, viz. a

11 See Kelsen, PTL1, 26. (RR1, 37.) (Footnote 1).


12 Kelsen, Hans. The Function of a Constitution, in: Essays on Kelsen. Edited. by R. Tur and W. Twining.
1986, Oxford Clarendon, 112. See also Kelsen, RR1, 37. (Footnote 1).
13 Kelsen, PTL1, 26. (Footnote 1).
14 See PTL2, 63. (RR2, 128). See also Kelsen, PTL2, 18, 67, 68–69. (RR2, 50, 134, 137). Kelsen, Problem der
Gerechtigkeit, in Kelsen, RR2, 689 In my translation: Kelsen rejects “from the point of view of scientific
knowledge (…) the existence of an absolute in general and absolute values in particular (…) and only
[recognises] the validity of relative values.” (Footnote 3).
15 Kelsen, PTL 2, 59. (RR2, 119). (Footnote 3).
16 Kelsen, PTL2, 86. (RR2, 169). See also Kelsen, PTL2, 8, 19, 22, 64. (RR2, 33, 51–52, 56, 126). (Footnote 3).

Franz Steiner Verlag


104 MAXIMILIAN KIENER

project of objective cognition and as a project that is ‘pure’, viz. a project separated not
only from empirical sciences but also from morality.

2. Positive Legal Fictionalism as a Solution to the Cognitivist Challenge

In this section, I will develop what I call Positive Legal Fictionalism as a solution to the
Cognitvist Challenge. Support for Positive Legal Fictionalism comes from an exegetical
perspective, showing that it can be firmly rooted in Kelsen’s own statements, and from
a systematic perspective, demonstrating how it effectively assists Kelsen’s project of
cognising the law and thereby solves the Cognitivst Challenge. However, before I elab-
orate on these two perspectives, I want to outline my proposal. Positive Legal Fictional-
ism consists of the following two claims:
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Error Theory: There are no objectively valid norms. Hence, there is no objective cog-
nition of norms17 and any statement asserting the existence of objectively valid norms
is false.
Legal Fiction: Employing a fiction facilitates objective cognition of relative legal
norms, which would otherwise not be possible, and provides further epistemic advan-
tages for the cognition of the law.

In a nutshell, the solution that Positive Legal Fictionalism proposes to the Cognitivist
Challenge is the following: it is correct that no norm, legal or moral, relative or absolute,
can be an object of cognition – actually speaking. However, resorting to a fiction allows
us nevertheless to separate the class of relative legal norms from all other norms and
promote legal norms in a way that facilitates objective cognition of them. One caveat
is required here: there can be objective cognition of these norms qua norms insofar
these norms’ validity does no longer depend on any facts or psychological preferences
but only on another norm, i. e. the fiction, which I will shortly explain with reference
to the basic norm; and therefore the legal norms’ validity as well as their cognition
can be objective, notwithstanding the fact that such validity will be fictional just like
the fiction itself. Positive Legal Fictionalism understands ‘fiction’ as a false assumption
and identifies as the fiction the assumption of an existing basic norm. The basic norm
qua fiction is supposed to ensure that Kelsen’s project is a theory, i. e. concerned with
objective cognition, as well as that Kelsen’s project is pure, i. e. separated from morality
and empirical sciences. The fiction can do this by providing the basis for objective va-

17 Note, it is about objective cognition of norms qua norms This is still consistent with saying that there
can be objective cognitions of norms understood as sociological or psychological facts.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 105

lidity, i. e. validity which cannot be reduced to either subjective wishes or preferences


or to any other psychological or sociological facts, but which depends on norms only.
This solution to the Cognitivist Challenge sidesteps the dead end from the first inter-
pretation of Kelsen’s contrast mentioned earlier: Positive Legal Fictionalism does not
consider the law as a coercive order as being the reason why legal norms, but not moral
norms, can be an object of cognition. Moreover, Positive Legal Fictionalism also side-
steps the difficult questions for the second interpretation: it does not imply that it is
just the relativity of norms that makes them suitable objects of cognition. It is the fic-
tion applied to the positive law that renders relative legal norms suitable for objective
cognition.
As the fiction is applied to legal norms only, Positive Legal Fictionalism is consistent
with Kelsen’s claim in the first edition of The Pure Theory of Law that only legal norms,
but not moral norms, can be objects of cognition. Moreover, Positive Legal Fictionalism
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

is also in accordance with Kelsen’s claim in the second edition of The Pure Theory of
Law that only relative norms can be objects of cognition: the fiction is an absolute
norm (i. e. the basic norm) which is not itself an object of cognition but only facilitates
the cognition of other relative norms, the validity of which depend on the fiction in
the relevant way.
One caveat is required at this point: Error Theory must not be confused with Adolf
Julius Merkl’s “Fehlerkalkül”, which is an entirely different idea. The term Error Theory,
as I use it, is not concerned with the question as to whether or to what extent those
legal norms which contradict other higher legal norms are still legally valid, but rather
focuses on the systematic falsity of certain statements. So understood, Error Theory is
common terminology in philosophical metaethics and, in particular, stems from John
Mackie’s seminal work in his book Ethics: Inventing Right and Wrong.
To further explain and vindicate Positive Legal Fictionalism as a solution to the Cog-
nitivist Challenge, I will firstly highlight the exegetical basis of Positive Legal Fictionalism
in Kelsen’s work (2.1) and then, on this basis, elaborate on how, from a systematic per-
spective (2.2), Positive Legal Fictionalism solves the Cognivist Challenge.

2.1 The Exegetical Perspective: The Basis of Positive Legal Fictionalism


in Kelsen’s Work

Error Theory generalises Kelsen’s claims about absolute moral norms to all norms. Not
only moral and absolute norms but also legal and relative norms are barred from being
an object of cognition.
Error Theory is supported by Kelsen’s own statements. Emphasizing a clear dis-
tinction between ‘is’ and ‘ought’, Kelsen held that any norm (legal or moral), i. e. any
‘ought’, could derive its validity (i. e. objective validity) only from another norm, i. e.

Franz Steiner Verlag


106 MAXIMILIAN KIENER

an ‘ought’, but never from facts, i. e. an ‘is’.18 However, the chain of norms conferring
validity on other norms cannot extend infinitely. Therefore, if a particular norm is to be
objectively valid, there has to be one first norm the validity of which no longer depends
on another norm, i. e. there has to be one absolute norm. However, Kelsen rejects the
existence of absolute norms, and in the absence of absolute norms no norm can be ob-
jectively valid. Hence, Error Theory Note that this is consistent with anything Kelsen
says about the basic norm. Kelsen does not claim that the basic norm actually exists but
only presents it as a condition necessary for the validity of norms.
While Error Theory corroborates Kelsen’s statements about norms not being suita-
ble objects of cognition, Legal Fiction aims to explain how Kelsen can still claim that
there is objective cognition of relative legal norms. Legal Fiction claims that a fiction
can separate the class of relative legal norms from all other norms and promote these
relative legal norms in such a way as to make them objects of cognition. Legal Fiction
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

proposes to recast Kelsen’s account of ‘presupposing the basic norm’ as ‘employing a


fiction’ and thereby explain the role of the fiction.
As already mentioned, Kelsen introduced the basic norm to explain the validity of
relative legal norms. The basic norm is the first and absolute norm that stops an infinite
regress of the chain of validity conferring norms. Kelsen explains the basic norm as
merely ‘presupposed’ (“voraus-gesetzt”) and as a purely non-positive norm. Kelsen
calls the basic norm a ‘hypothesis’ and a transcendental condition of the very possi-
bility for positive law to be objectively valid.19 Kelsen claimed that all norms in a legal
system have to converge on one single fundamental norm, the ‘basic norm’, if they are
to be considered objectively valid. Converging on the basic norm makes these norms
objectively valid because their validity cannot be reduced to either subjective wishes
or preferences or to any other psychological or sociological facts. Their validity ulti-
mately depends on a norm, i. e. the basic norm.
Recasting Kelsen’s account of ‘presupposing the basic norm’ as ‘employing a fic-
tion’ then leads to the following explanation of Legal Fiction: the fiction is the basic
norm and the fiction is employed by presupposing the basic norm in thought. The
basic norm qua fiction is only applied to positive legal norms within a legal system
and thereby separates the class of legal norms from other norms, e. g. moral norms.
The basic norm qua fiction makes such legal norms objects of cognition by providing
the necessary basis for their objective validity: a first and absolute norm, i. e. the basic
norm which invests all the positive legal norm with objective validity. The basic norm

18 See Kelsen, PTL 1, 193. (RR 2, 346). (Footnote 1 and 3).


19 See Kelsen, PTL1, 58, 64, 70. (RR1, 77, 85, 93) Kelsen, PTL2, 101–107. (RR2, 197–207). (Footnote 1 and
3). Kelsen, Hans. General Theory of Law and State. Translated by A. Wedberg. 1945, Cambridge, Harvard
University Press, 110, 134, 401, 406, 437. Kelsen abandons the notion of “hypothesis” in favour of a fiction in:
Kelsen, Hans. General Theory of Norms. Translated by Michael Hartney. 1991, Oxford Clarendon, 256. I will
later take up this point once more.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 107

renders all these legal norms into relative legal norms, which ultimately depend on the
basic norm. So understood, Legal Fiction is an interpretation as well as a further devel-
opment of Kelsen’s doctrine of the basic norm. To see where it is interpretation and
where it is further development, I need show in greater detail how Legal Fiction accords
with Kelsen’s own statement on fictions (2.1.1) and his alleged method for cognising
the law: Kantian transcendentalism (2.2.2).

2.1.1 Kelsen on Fictions

Kelsen thought about fictions at several points in his career. Two such points deserve
special attention. Firstly, as early as in 1919, Kelsen writes on Vaihinger’s account of
legal fictions. Agreeing with Vaihinger, Kelsen claims that there are “true, i. e. episte-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

mological fictions in legal science (…) of the attempt to know the law, dictions of the
intellectual mastery of the legal order.”20 It is indeed most noteworthy that, from a very
early stage, Kelsen claims that there are epistemological fictions, i. e. fictions that are
supposed to assist the “intellectual mastery of the legal order.”21 However, Kelsen does
not elaborate much on this point. He only mentions the ‘person’ as a legal fiction and
argues that using such a fiction may provide the “advantages of illustration and simpli-
fication ”22 The ‘person’ provides these advantages by offering a vivid picture instead of
the mere juridical and technical description of a bundle of obligations and rights.
Secondly, from 1960 onwards, Kelsen draws his attention again to legal fictions. Two
significant differences occurred at this point, however. Firstly, Kelsen now considers
his own account from a fictionalist standpoint and not only other accounts like Vai-
hinger’s. He thinks of his basic norm as a fiction.23 Secondly (and related to the first
point), whereas Kelsen’s 1919 essay was only concerned with fictions within the law
(the ‘person’) Kelsen’s consideration of the basic norm as a fiction now focuses on
fiction about the law as a whole.
Unfortunately, Kelsen’s position from 1960 onwards is inconsistent with his posi-
tion in 1919. This inconsistency is most obvious when Kelsen, after 1960, claimed that
the basic norm is a fiction insofar as it “contradict[s] reality”24 and explicitly revokes
earlier statements about norms (‘Ought’) not being able to contradict reality (‘Is’),

20 Kelsen, Hans. On the Theory of Juridic Fictions. With Special Consideration of Vaihinger’s Philosophy
of the As-If, in: Legal Fictions in Theory and Practice. Edited by M. Del Mar and W. Twining. 2015, Switzer-
land Springer, 5.
21 Ibid., 5.
22 Ibid, 7. Emphasis in original.
23 See for instance Kelsen, Hans. On The Pure Theory of Law. Israel Law Review, 1966, 1–7.
24 Kelsen, Hans. The Function of a Constitution, in: Essays on Kelsen. Edited. by R. Tur and W. Twining.
1986, Oxford OUP, 117.

Franz Steiner Verlag


108 MAXIMILIAN KIENER

which formed the basis of his view in 1919. As a result, any fictionalist interpretation of
Kelsen needs to take sides.
Following Kelsen completely in his view in 1960 and later would partly sidestep
the Cognitivist Challenge This is because his 1960 view less emphatically stresses the
distinctness of the realm of ‘ought’ and norms, i. e. what Kelsen aims to reach objective
cognition of. My aim in this essay is therefore to retain Kelsen’s view that norms cannot
contradict reality. I will consider the basic norm as a fiction, not in the sense of fictions
contradicting reality (i. e. in the realm of ‘Is’), but in the sense of stating something that
does not exist in the realm of ‘Ought’.
Hence, Positive Legal Fictionalism embraces Kelsen’s claims concerning fictions
about the law (i. e. the fiction is the basic norm), the claim about fictions providing
genuinely epistemic advantages, and claims about the distinctness of ‘is’ and ‘ought’.
Positive Legal Fictionalism is thereby maximally close to Kelsen’s own statements. It
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

only departs from Kelsen where Kelsen’s contradicting statements made a departure
inescapable and the choice between contradicting statements was made with an eye to
upholding rather than sidestepping the Cognitivist Challenge.
However, it must also be stressed that Positive Legal Fictionalism goes beyond Kels-
en’s owns statements. It goes beyond Kelsen’s claims by recommending that the view
on fictions, developed from Kelsen’s own claims, be applied to all phases of Kelsen’s
work, including the first and second edition of The Pure Theory of Law, and not only to
the later sceptical phase in Kelsen’s life that is usually associated with Kelsen being a
fictionalist if he is considered a fictionalist at all. This is a first reason why Positive Legal
Fictionalism is a Kelsenian but not quite Kelsen’s own account.

2.1.2 Kelsen on Kantian Transcendentalism

By expanding the fictionalist reading to all phases of Kelsen’s work, Positive Legal Fic-
tionalism competes with the most widely accepted interpretation of Kelsen’s method
in cognising the law: Kantian transcendentalism. In this subsection, I will claim that
fictionalism is a better exegetical basis for Kelsen’s method in cognising the law than a
qualified version of Kantian transcendentalism.
In The Pure Theory of Law, Kelsen focuses on the conditions under which the cog-
nition of positive law as objectively valid becomes possible. Kelsen claims that this ap-
proach resembles Kant in his inquiry into the conditions of the possibility of cognition
in the natural sciences.25 Therefore, the basic norm, i. e. what Kelsen identifies as the

25 See Kelsen, PTL2, 202. (RR2, 361). (Footnote 3).

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 109

condition for cognising the positive law as objectively valid, is to be considered “cogni-
tively and theoretically transcendental in terms of the Kantian philosophy.”26
Despite such alleged common ground, many commentators correctly pointed out
that Kelsen’s link to Kantian transcendentalism is rather loose.27 To begin with, where-
as Kant’s inquiry (and also the inquiry of Neo-Kantians like Cohen whom Kelsen re-
fers to) was about the very conditions of cognition as such, Kelsen’s inquiry is only
about one way of looking at one particular area of the world, i. e. the positive law. Kelsen
indeed claims that:
“The Pure Theory of Law is well aware that the specifically normative meaning of certain
material facts, the meaning characterized as ‘law’, is the result not of a necessary interpreta-
tion but of a possible interpretation, possible only given a certain basic presupposition.”28

As a result, Kelsen admits that one could simply refuse to accept his conditions of
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

cognising the law as objectively valid and instead consider legal norms as mere acts of
power and domination. However, in Kant’s account, in contrast, it was not possible
simply to reject the transcendental conditions of cognitions and choose a different way
of looking at things.
Secondly (and as a result of the first difference), Kant’s and Kelsen’s uses of the term
‘hypothesis’ diverge considerably. When Kant identifies the conditions of cognition as
such, he states these conditions as ‘hypotheses’. By ‘hypotheses’, he means claims that
he cannot directly prove but only postulate through the transcendental method. Nev-
ertheless, for Kant, a hypothesis is something that can be true and is indeed assumed
to be a good candidate for a true claim. This is different in Kelsen. Although the basic
norm is also called a ‘hypothesis’, Kelsen does not mean something that can be true.
The basic norm, being a norm, cannot be true, according to Kelsen. And Kelsen explic-
itly states that the basic norm does not exist.
Referring to these differences, commentators suggested that Kelsen lacked suf-
ficient acquaintance with Kantian transcendentalism.29 However, this is a rather un-

26 Kelsen, PTL1, 25. (RR1, 36). (Footnote 1) Kelsen expresses his allegiance to Kant in other places too. See
Letter to Treves in Kelsen papers, 171. Similar statements also appear in earlier and later work. For earlier
work, see Kelsen, Hans. Hauptprobleme der Staatsrechtslehre Entwickelt aus der Lehre vom Rechtssatze (from
1923). 1960, Aalen Scientia, Vorrede, XVII.
27 See Wilson, Alida. Joseph Raz on Kelsen’s Basic Norm. The American Journal of Jurisprudence 27 (1982),
49. Dreier, Horst. Rechtslehre, Staatssoziologie und Demokratietheorie bei Hans Kelsen, 1986 Nomos Baden-
Baden, 44, 81–89. Winkler, Günther. Rechtstheorie und Erkenntnislehre Kritische Anmerkungen zum Dilemma
von Sein und Sollen in der Reinen Rechtslehre aus geistesgeschichtlicher und erkenntnistheoretischer Sicht 1990,
Wien New York Springer, 107–108.
28 Kelsen, PTL1, 34. Emphasis added. (RR1, 47). He repeats this point in Kelsen, PTL2, 218 (RR2, 393).
(Footnote 1 and 3) See also Kelsen, Hans. Value Judgments in the Science of Law, in: What is Justice? Justice,
Law, and Politics in the Mirror of Science Collected Essays by Hans Kelsen 2013. Clark The Lawbook Exchange,
Ltd., 226–227.
29 See for instance Wilson, 1982, 48. (Footnote 27).

Franz Steiner Verlag


110 MAXIMILIAN KIENER

charitable interpretation and in this subsection I want to claim that Kelsen’s departure
from Kant was to a great extent deliberate and should therefore be appreciated rather
than criticised. Framing Kelsen’s method in fictionalist terms rather than transcenden-
tal terms can charitably capture Kelsen’s departure from Kant and should therefore be
the preferred interpretation of Kelsen’s method in cognising the law.
Support for a fictionalist interpretation comes from what commentators sometimes
overlook, namely that Kelsen himself stressed that he employs the Kantian philoso-
phy only “by analogy”30: it is just the focus on conditions of the possibility of cognition
that links Kelsen to Kantian Philosophy in some loose sense. There are other aspects
of (Neo)-Kantian transcendentalism that Kelsen explicitly rejects. In a little-known
letter to Renato Treves from 1933, Kelsen contrasts the conditions of the possibility of
cognition developed by Kant and Cohen with his own conditions of the possibility
of cognizing the positive law. Kelsen claims that Kant and Cohen adhered to a “con-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

tentually constituted, materially determined a priori”31, by which he meant that Kant


and Cohen adhered to substantial value judgements and potentially a form of value
realism, i. e. the assumption of absolute values. Kelsen, in contrast, rejects such values
and is looking only for “purely formal categories of a priori validity.”32 The basic norm
is supposed to be completely formal in the sense of not including any substantial or
absolute values, or any restriction on what legal norms may proscribe. The basic norm
merely supports whatever the highest positive legal norm, i. e. the constitution, states.
Moreover, in the same letter, Kelsen claims that moral relativism is “something that
Cohen – exactly like Kant on this point – was not prepared to accept, if only because of
his religious convictions.”33 However, moral relativism is key to Kelsen’s own account.
Kelsen fiercely rejects absolute norms and denies they could be accounted for by way
of rational cognition.
In order to keep the basic norm purely formal as well as defend ethical relativism,
Kelsen had to claim that the basic norm is indeed just one contingent way of looking
at the positive law as well as deny that the basic norm can be true in the sense Kan-
tian hypotheses can be true. If the basic norm were a materially-determined, true, and
absolute norm, Kelsen would undermine his relativism about norms and the purely
formal nature of the conditions of cognising the positive law.
Therefore, Kelsen’s letter to Treves shows that his departure from presenting the
very conditions of cognition as such and the adherence to ‘hypotheses’ as truth-apt
was not simply due to a lack of acquaintance with Kant. It can be seen as deliberate and

30 Kelsen, PTL2, 202. (RR2, 361). (Footnote 3).


31 Kelsen, Hans. The Pure Theory of Law, ‘Labandism’, and Neo-Kantianism. A Letter to Renato Treves,
in: Normativity and Norms: Critical Perspectives on Kelsenian Themes, ed. by Stanley L. Paulson and Bonnie
Litschewski. 1998, Oxford Clarendon, 173. Original emphasis.
32 Ibid.
33 Ibid.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 111

necessary to uphold central features of his account, even if Kelsen did not explicitly say
so in the Pure Theory of Law. Therefore, it is no surprise that Kelsen, even though he
still declares allegiance to Kant and Cohen in focusing on an inquiry into the conditions
of the possibility of cognition, rejects Kant’s and Cohen’s overall framework as versions
of “natural law theory.”34
These exegetical data require a choice: one could either continue interpreting Kels-
en as employing Kantian transcendentalism but with all these necessary qualifications
that (as commentators correctly claim) take away what is genuine about Kantian tran-
scendentalism; or one could develop a different interpretation that aligns more closely
with all of Kelsen’s claims. In the rest of this subsection, I will opt for the second route
and claim that a fictionalist framework, as given by Legal Fiction, can account for all of
Kelsen’s claims and should therefore be adopted.
To begin with, Legal Fiction recognises Kelsen’s attempt to identify the conditions
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

under which cognising the law as objectively valid is possible. Legal Fiction recognises
the basic norm qua fiction as the necessary condition for the possibility of objection
cognition of relative legal norms. As such, Legal Fiction can be understood per analo-
giam to Kantian transcendentalism in the restricted sense Kelsen refers to. But operat-
ing with a fiction leaves no doubt that employing the fiction is only one way of looking
at the positive law. My fictionalist proposal does not imply that the basic norm is the
only way, or even an especially privileged way, of looking at the positive law. And it is
precisely the contrast between a fiction and a Kantian hypothesis that maps on to the
contrast between Kelsen’s basic norm and a Kantian hypothesis. A fiction is something
we do not assume to be true but actually know is false or non-existing. This is exactly
what Kelsen says about the basic norm but what would be misleading to say about a
‘hypothesis’. And in fact, Kelsen later abandoned the notion of ‘hypothesis’ in favour
of the notion of a “fiction”.35 As a result, the fiction is purely formal and not, as Kelsen
thought about Cohen’s and Kant’s approach, contentually-constituted and material-
ly-determined. The fiction is the basic norm and as such comprises only what Kelsen
attributes to the basic norm. Finally, since the basic norm is a fiction, applied to the law,
it does not imply any claims about morality. Therefore, it is compatible with ethical
relativism as Kelsen defends it.
Hence, Legal Fiction can account for all of Kelsen’s claims. On this basis, I propose
reconstructing Kelsen’s method of cognising within a fictionalist framework rather
than within a framework of Kantian transcendentalism. However, it must be noted
that preferring fictionalism over Kantian transcendentalism does not imply abandon-
ing the characterisation of Kelsen’s account as Kantian or Neo-Kantian altogether.
Other aspects of Kelsen’s account that are usually considered (Neo)-Kantian can still

34 Ibid.
35 Kelsen, Hans. General Theory of Norms, 1991, Oxford Clarendon, 256. (See Footnote 18).

Franz Steiner Verlag


112 MAXIMILIAN KIENER

remain in place. These aspects are: (i) the assumption of ‘is’ and ‘ought’ as separate and
irreducible categories; (ii) the assumption of the validity of norms in terms of binding
force; (iii) the emphasis of genuine normativity of legal science, where legal science is
sharply contrasted with more empirical approaches like legal sociology.36

2.1.3 Summary

This section (2.1) outlined the exegetical basis of Positive Legal Fictionalism in Kelsen’s
own claims. However, Positive Legal Fictionalism did not merely adapt Kelsen’s claims
but went beyond them in two ways: firstly, as there was no consistent account of a legal
fiction in Kelsen, Positive Legal Fictionalism had to reconstruct such an account from
Kelsen’s contradictory claims. Positive Legal Fictionalism then applied this account to
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

all of the different phases in Kelsen’s work and not only to the latter sceptical phase.
Secondly, Positive-Legal-Fictionalism suggested reinterpreting Kelsen’s method in cog-
nising the law as a fictionalist approach rather than as a Kantian transcendental ap-
proach. In the next section (2.2), I will elaborate on how Positive Legal Fictionalism
contributes to a project of cognition required to solve the Cognivist Challenge.

2.2 The Systematic Perspective: The Contribution of Positive Legal Fictionalism to


a Project of Cognition

It may seem paradoxical that a fictionalist account, i. e. an account relying on false


claims, could ever contribute to a project of cognition. However, if Positive Legal Fic-
tionalism is to be a solution to the Cognitivist Challenge, I need to explain how it con-
tributes to objective cognition. To do this, I will firstly elaborate on fictionalism as a
scientific method more generally (2.2.1), then specify the particular epistemic advan-
tages that come with the assumption of the basic norm qua fiction (as made by Positive
Legal Fictionalism) (2.2.2), and finally explain how Positive Legal Fictionalism not only
provides epistemic advantages but also remains loyal to legal positivism (2.2.3).

36 See Bulygin, Eugenio. An Antinomy in Kelsen’s Pure Theory of Law, in: Normativity and Norms: Critical
Perspectives on Kelsenian Themes, ed. by Stanley L. Paulson and Bonnie Litschewski. 1998, Oxford Claren-
don, 297–315.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 113

2.2.1 Fictionalism as a Scientific Method

Recall that Kelsen claimed that there are “epistemological fictions in legal science (…)
of the attempt to know the law, dictions of the intellectual mastery of the legal order.”37
In order to illustrate how a fiction could assist such intellectual mastery, Kelsen gives
the following example:
“And we have to speak of a fiction as soon as cognition (and especially juridic cognition)
takes a detour in knowing its object (and in juridic knowledge this object is the law, the
legal order, the legal ought), a detour in which it consciously sets itself in contradiction to
this object; and be it only in order to better grasp it: just like a rock-climber, in order to
avoid an obstacle and reach his goal more easily, is sometimes forced to temporarily climb
downwards, i. e. in a direction directly opposed to his goal, the peak.”38
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

So, according to Kelsen, a fiction can be a device that helps to overcome an obstacle
to cognition. Such an obstacle can be overcome, Kelsen argues, if we allow the fiction
to take us on a “detour” and thereby identify an alternative path to cognition. Unfor-
tunately, Kelsen does not elaborate on how such a “detour” can be mapped out in de-
tail. However, the metaphor of rock-climbing is instructive. And most interestingly, it
resembles a method that Vaihinger, whom Kelsen knew, once attributed to Fermat’s
solution of a mathematical problem. Therefore, Vaihinger’s interpretation of Fermat
may help to specify Kelsen’s take on how fictions assist cognition, even if Kelsen does
not explicitly refer to it.
Vaihinger presented his reader with the following mathematical task: suppose you
would like to divide a line into two parts, x and a-x, so that the term x2(a-x) is maxi-
mally large. Vaihinger claims that it was impossible to provide a solution until Fermat
employed a particular trick.
Fermat substituted the x in the term x2(a-x) with (x+e), leading to the new term:
(x+e)2(a-x-e). He then equated the original term with the new one, reaching the fol-
lowing equation: x2(a-x) = (x+e)2(a-x-e)
By equating both terms, Fermat consciously made a false assumption, Vaihinger
claims. Fermat knowingly equated two unequal terms. After multiplying all the terms,
Fermat then reached the following equation: 2ax + ae = 3x2 + 3xe + e2
In order to reverse his initial false assumption (i. e. arbitrarily adding e to the equa-
tion), Fermat cancelled all the terms containing an e. However, arbitrarily cancelling
all the terms containing an e is in itself another false assumption, Vaihinger claims: one
must not simply leave terms out of an equation.

37 Kelsen, Footnote 19, 5.


38 Ibid.

Franz Steiner Verlag


114 MAXIMILIAN KIENER

After cancelling all terms containing an e and shifting everything but x to the left
side, Fermat reaches 2a/3 = x and this is indeed the solution to the problem: if the line
x is divided into parts, x and a-x, so that 2a/3 = x, the term x2(a-x) is maximally large.
On Vaihinger’s interpretation, Fermat used two false assumptions as tools in his
arithmetic operation. He made a first false assumption (i. e. arbitrarily adding e to the
equation) to facilitate an easier arithmetic calculation and made a second false as-
sumption (i. e. dropping all terms containing an e) in order to reverse the initial false
assumption. Vaihinger calls this the “method of opposite errors”, i. e. errors that neu-
tralise each other.
Fermat’s solution, at least in Vaihinger’s description, seems to illustrate what Kels-
en says about the metaphor of the rock-climber: “just like a rock-climber, in order to
avoid an obstacle and reach his goal more easily, is sometimes forced to temporarily
climb downwards, i. e. in a direction directly opposed to his goal, the peak”39, so is
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Fermat, in order to avoid a mathematical obstacle, forced to temporarily depart from


the truth, i. e. think in a direction directly opposed to his goal, the truth. Fermat does
this by means of his first false assumption. And just as the rock climber’s adjustment of
his position allows him to “reach his goal more easily” 40 later, so can Fermat reach the
true solution to the problem more easily after relying on a false equation that he later
neutralised by means of a second well-chosen false assumption.41
Vaihinger’s interpretation of Fermat’s solution illustrates how false assumptions, i. e.
fictions, may help to achieve cognition. Moreover, Vaihinger’s interpretation is strik-
ingly close to how Kelsen himself describes how fictions can help legal science, and
Kelsen in fact knew Vaihinger’s book containing this interpretation.42 Therefore, ap-
plying Vaihinger’s interpretation to Kelsen’s Pure Theory of the Basic Law may be a
plausible way of explaining the epistemic significance of the fiction in Kelsenian terms.
So, let us transfer this take on fictions to Kelse’s Pure Theory of Law:
Just as Fermat faced an arithmetic predicament, Kelsen can be seen to face a pre-
dicament concerning the validity of norms. According to Error Theory, there are no
objectively valid norms. However, legal science in Kelsen’s sense, if it is to be possible
at all, is only possible if there are objectively valid norms.
Just as Fermat employed a false assumption to overcome an arithmetic predicament
and facilitate an easier arithmetic calculation, Kelsen can be seen as employing the

39 Ibid.
40 Ibid.
41 Vaihinger, Hans. Die Philosophie des Als Ob System der theoretischen, praktischen und religiösen Fiktionen
der Menschheit auf Grund eines idealistischen Positivismus Mit einem Anhang über Kant und Nietzsche 1918,
Leipzig Meiner, 200 ff.
42 Admittedly, it is not certain that Kelsen knew this passage when writing about fictions, but it is nev-
ertheless likely. Vaihinger’s book was published in 1918 and Kelsen published a review of it in 1919, the
same year in which his other paper appeared from which the rock-climbing metaphor is taken. See Kelsen,
Footnote 20.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 115

basic norm qua fiction to overcome the legal predicament of there being no objective
validity. Once the basic norm, understood as a fiction, is presupposed, all relative legal
norms within a particular legal system are invested with objective validity.
However, just as Fermat reversed the initial false assumption in order to reach the
correct result, Kelsen also has to prevent the fiction from distorting the results reached.
After the basic norm qua fiction is presupposed, Kelsen’s can be interpreted as adding
something like a tag to any legal norm saying that legal norms are valid only on the ba-
sis of the fiction, i. e. on the basis of this mere product of thought, but not valid in any
more robust sense. Kelsen may have done this by introducing the Rechtssatz (rule of
law) as opposed to the Rechtsnorm (legal norm). Kelsen explains both terms as follows:
“Rules of law [i. e. Rechtssätze] (…) are hypothetical judgments stating that according
to a national or international legal order, under the conditions determined by this order,
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

certain consequences determined by the order ought to take place. Legal norms [i. e. Re-
chtsnormen] are not judgments, that is, they are not statements about an object of cogni-
tion. According to their meaning they are commands, they may also be permissions and
authorizations.”43

Hence, Rechtsnormen are the legal norms that I presented as the subject-matter of The
Pure Theory of Law. Rechtssätze are statements about those norms and usually take the
following form: according to the legal system X, such and such ought to be the case.
The prefix ‘according to the legal system X’ can be read as the required tag. On a fiction-
alist interpretation, this prefix can be read as “according to the fiction in legal system
X”. This tag makes clear that objective validity still depends on the fiction and prevents
the results from being distorted, that is from stating objective validity independent of
a mere product of thought. Hence, similarly to how Fermat’s dropping all terms con-
taining an e neutralised his initial false assumption, Kelsen’s prefix in his Rechtssätzen
prevents the potential false assumption of objective validity that exists independently
of the fiction and makes clear that objective validity is still dependent on arbitrary
presupposition.
However, there seems to be a difference between Fermat and Kelsen. Whereas Fer-
mat’s second false assumption neutralised his first false assumption, Kelsen – on this
interpretation – only highlights but does not neutralise the use of a fiction. The falsity
of stating norms of objective validity does not disappear. Error Theory is still in place
and, as I will show in section 2.2.3, still plays an important role. At this point, I only
want to highlight this difference from Fermat. Kelsen prevents distorted results by a
different strategy from that which Vaihinger attributes to Fermat.

43 Kelsen, PTL2, 71. (RR2, 141). (Footnote 3).

Franz Steiner Verlag


116 MAXIMILIAN KIENER

2.2.2 The Epistemic Advantages of the Legal Fiction

In the last subsection (2.2.1) I tried to explain the seemingly paradoxical claim that fic-
tions, i. e. deliberate departures from truth, can sometimes be used to reach the truth.
I claimed that both Fermat and Kelsen can use an initial fiction to overcome a pre-
dicament and later prevent the fiction from leading to distorted results. This method
strikingly resembles Kelsen’s own description of the use of fictions in his metaphor of
the rock-climber, and it is a plausible assumption that Kelsen knew about Vaihinger’s
interpretation of Fermat and was inspired by it.
In this subsection (2.2.2), I want to be more specific and describe the particular
epistemic advantages that come with the basic norm interpreted as a fiction. These
epistemic advantages fall into two groups: firstly, there are all the epistemic advantages
that Kelsen attributes to the basic norm independent of considering it a fiction. Since I
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

proposed to recast Kelsen’s account of presupposing the basic norm as employing the
fiction, the epistemic advantages of the basic norm can be re-interpreted as the epis-
temic advantages of the fiction. Secondly, however, there are also epistemic advantages
that specifically stem from the basic norm being a fiction. Such epistemic advantages
are not already among those which Kelsen highlights but still contribute to his overall
project of cognising the law.
Let me start with the first set of advantages, those already associated with the basic
norm independent of considering it a fiction. Consider the following passage from
Kelsen:
“That an assembly of people is a parliament, and that the result of their activity is a statute
(in other words, that these events have this ‘meaning’), says simply that the material facts
as a whole correspond to certain provisions of the constitution. That is, the content of an
actual event corresponds to the content of a given norm.”44

Kelsen here claims that a legal norm renders “actual events” into something with le-
gal significance: there is a parliamentary law (i. e. legal event) and not only a group of
people writing down some rules they wish to be complied with (i. e. natural event);
there is an instance of murder (i. e. legal event) and not just an instance of killing (i. e.
natural event), etc. Without a legal norm, or more precisely without objectively valid
legal norms, there would only be natural but not legal events. Kelsen seems to think
of something like a parallel or augmented reality. 45 However, all this is only possible

44 Kelsen, PTL1, 10. (RR1, 19). (Footnote 1).


45 See Kelsen, Hans. Value Judgments in the Science of Law, in: Kelsen, H. What is Justice? Justice, Law, and
Politics in the Mirror of Science Collected Essays by Hans Kelsen 2013. Clark, The Lawbook Exchange, Ltd.,
227: “If the system of legal norms is an ideology, it is an ideology parallel to a definite reality.” See also 2008,
49, where Kelsen qualifies the pure theory as an ideology behind a reality which is supposed to conform
to this ideology. See also “Welt des Sollens” and “Welt des Seins” in Kelsen, 1960, e. g. 5 ff. (Footnote 26).

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 117

after presupposing the basic norm because only the presupposition of the basic norm
ensures objective validity. Therefore, legal practice arises and is intelligible as a legal
practice only by presupposing the basic norm.
Hence, Kelsen’s own statements allow the extracting of a specific reason for em-
ploying the basic norm as a fiction: presupposing the basic norm makes the legal prac-
tice – insofar as it is concerned with legal cognition – intelligible. More particularly,
these passages allow us to extract two ways in which presupposition of the basic norm
contributes to “cognizing positive law.”46 Firstly, it facilitates legal propositional knowl-
edge, viz. knowing that there are certain legal things (parliamentary law, murder, etc.),
which would not legally exist without the presupposition of the basic norm. Secondly,
it facilitates legal understanding, viz. understanding why actual events bring with them
certain legal characteristics. It is because they correspond to norms which are invested
with validity by the presupposition of the basic norm. On this basis, the legal scientist
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

can know that there are certain norms, certain legal events, also that a group of norms
leading back to the same basic norm belong to the same legal system, and understand
why all this is true: namely because of the link to the basic norm.
However, over and above those epistemic advantages associated with the basic
norm independent of its interpretation as a fiction, there are also epistemic advantages
specifically connected with the basic norm qua fiction
Firstly, fictionalist accounts can avoid the disadvantages of realist accounts or natural
law theories but still have the same advantages as these views. A disadvantage of realist
accounts of norms or natural law theory is often seen in their commitment to compli-
cated ontology. However, the advantage of these theories is their simple semantics.
Realist accounts and natural law theory make genuine propositional assertions about
norms. As a result, such statements can be robustly true or false and they are subject to
principles of logic and inference. Such simple propositional semantics, truth-aptness,
and the applicability of logical principles seem to be conducive to a scientific project,
other things being equal.
Fictionalism can avoid the disadvantage of complicated ontology but still provide
the advantages of realist accounts. Fictionalism can avoid complicated ontology by ad-
hering to an error theory, e. g. rejecting any absolute norms, which natural law theory
claims exist and which Kelsen continuously highlighted as ontologically incompre-
hensible. At the same time, however, fictionalism can still claim that there are genuine
assertions about the subject matter. It may be true that there are no norms. However,
one can still state that ‘according to the legal fiction, such and such a norm exists’ (i. e.
Kelsen’s Rechtssätze). Such statements are genuine assertions with propositional con-
tent. As such, these statements fall under simple realist semantics and can therefore be
robustly true or false as well as subject to principles of logic and inference. Hence, a first

46 Kelsen, PTL1, 58. Emphasis added. (RR1, 77). (Footnote 1).

Franz Steiner Verlag


118 MAXIMILIAN KIENER

epistemic advantage of fictionalist accounts like Positive Legal Fictionalism is that it can
avoid complicated ontology while providing the same advantages as natural law theory.
Secondly, another advantage of Positive Legal Fictionalism comes from a distinction
available to all fictionalist accounts, namely the distinction between accepting and be-
lieving claims. Accepting a claim means endorsing it for the time being without believing
it to be true. Believing a claim in contrast means taking the claim to be true. Fictionalist
accounts can make this distinction by interpreting certain claims as systematically false
(e. g. due to a Error Theory) and thereby oppose believing in those claims. However,
fictionalist accounts can still argue that one should accept these claims for reasons oth-
er than their literal truth (e. g. the specific advantages connected to the fiction, which
may be epistemic, pragmatic, political, or aesthetic.)
L. J. Cohen used this distinction in a discussion of methods in science more gener-
ally. He argued that a scientist who only accepts his hypotheses, claims, and theories is
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

superior to a scientist who believes them. Concerning the later, he says:


“possession of a belief that p might make him less ready to change his mind about accept-
ing that p if new evidence crops up or a better theory becomes available. It might even
make him less ready to look for new evidence or a better theory, when otherwise he would
have done so.”47

Hence, ‘accepting’ instead of ‘believing’ claims leads to a greater open-mindedness and


willingness to revise one’s theories, which may be conducive to a project aiming at
cognition.
Positive Legal Fictionalism, being a fictionalist account, can resort to the distinction
between acceptance versus belief and thereby provide an epistemic advantage for a
Kelsenian project of cognising the law. Due to Error Theory, a legal scientist embracing
Positive Legal Fictionalism would never believe in legal norms, that is take them to be
objectively valid, full stop, or even take them to be true. A legal scientist would only
accept the legal norms provisionally, that is accept them as part of a contingent legal
system and on the contingent fictionalist assumption of a presupposed basic norm. As
a result, a legal scientist embracing Positive Legal Fictionalism does not cling on to legal
norms in any way but is indifferent to changes in them. As a result, the statements about
those norms, the Rechtssätze, are also subject to radical contingency. They are radical-
ly contingent because they would need to change whenever Rechtsnormen change.
And this is the case even though Rechtssätze – unlike Rechtsnormen – can be robustly
true or false. So, the fictionalist legal scientist only accepts Rechtsnormen and, since
Rechtsnormen are the basis for Rechtssätze, he is also maximally open-minded for
changing Rechtssätze whenever “new evidence crops up or a better theory becomes
available”, to use Cohen’s phrase again.

47 Cohen, Laurence Jonathan. An Essay on Belief and Acceptance, 1992, Oxford OUP, 88.

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 119

Hence, insofar as open-mindedness can be considered an epistemic virtue in line


with Cohen’s description, Positive Legal Fictionalism – being a fictionalist account – can
claim to provide this virtue for a project of cognising the law.

2.2.3 Fictionalism and Legal Positivism

In the last two subsections (2.2.1 and 2.2.2), I tried to explain the use of a fiction as a
scientific method more generally as well as specify the particular epistemic advantages
of the basic norm qua fiction. In this subsection (2.2.3), I wish to highlight how my
fictionalist proposal, amidst all cognitivist aspirations, still remains loyal to legal posi-
tivism, thereby also returning to a claim I made at the end of section 2.2.1.
It is indeed the specifically fictional take that aligns with legal positivism: the take
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

on there being no objectively valid norms (Error Theory) but only a fiction that ensures
fictional objective validity (Legal Fiction). In order to see how these claims provide
advantages, let us consider the opposite account of a richer cognitivist project that
abandons Error Theory and the assumption of a fiction. Such a project does not seem
to be terribly attractive vis-à-vis Kelsen’s legal positivism. The last thing that Kelsen
should arrive at is a cognitivist account that tends towards recommending certain
norms are really true and ought to be complied with.’ Kelsen indeed stresses that the
Pure Theory is not supposed to legitimise the law in any way but remain agnostic on the
moral worth of the positive law.48 And Dreier is correct when he says that it was never
Kelsen’s aim to hedge the arbitrariness of positive law.49 Error Theory is in accordance
with these claims and emphasizes that full normative necessity in law is still compati-
ble with the utmost moral depravity of the legal system. As such, keeping Error Theory
in place facilitates an interpretation that is carefully nuanced between cognitivist and
positivist aspirations. This justifies labelling my proposal Positive Legal Fictionalism.

2.2.4 Summary

This section (2.2) aimed to explain how Kelsen’s Pure Theory of Law can be a cogni-
tive project after the adoption of Error Theory. Referring to Positive Legal Fictionalism,
I made three claims: firstly (2.2.1), the epistemic function of a fiction can be explained
according to Kelsen’s metaphor of the rock-climber. Just as Fermat employed a false
assumption to solve a mathematical challenge, Kelsen can use a fiction to solve a legal
challenge, i. e. how to explain objective validity. Secondly (2.2.2), presupposing the ba-

48 Most notably Kelsen, PTL1, 7, 18. (RR1, 15, 29). Kelsen, PTL2, 1, 68–69. (RR2, 21, 136–138). (Footnote
1 and 3).
49 See Dreier, 1986, 59 (Footnote 27).

Franz Steiner Verlag


120 MAXIMILIAN KIENER

sic norm qua fiction then specifically facilitates the following epistemic advantages: it
provides legal propositional knowledge and it provides legal understanding. Further-
more, it avoids complicated ontology while facilitating simple propositional semantics
and leads to a particular open-mindedness in legal theorizing. Finally (2.2.3), over and
above these epistemic advantages, a fictionalist account is also particularly suited to
legal positivism, thereby balancing cognitive and positivist aspirations.

3. Conclusion

At the beginning of this essay, I identified two objectives: drawing attention to the
Cognitivist Challenge as a serious challenge to Kelsen’s account and proposing Positive
Legal Fictionalism as a solution to the Cognitivist Challenge.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

The Cognitivist Challenge was a challenge of how to retain Kelsen’s view about the
relative legal norms, i. e. that one can objectively cognise these norms, while not reject-
ing his view about the absolute moral norms, i. e. that one cannot objectively cognise
these norms I claimed that a solution to the Cognitivist Challenge is of vital importance
for understanding Kelsen’s project as a ‘theory’, viz. a project of objective cognition,
and as a project that is ‘pure’, i. e. separated not only from empirical sciences but also
from morality.
The solution that I proposed, Positive Legal Fictionalism, proceeded in two steps:
firstly, relying on Error Theory, Positive Legal Fictionalism claimed that no norms are
objectively valid – actually speaking. It thereby accounted for all of Kelsen’s statements
about certain norms, e. g. absolute moral norms, not being objects of cognition. Sec-
ondly, relying on Legal Fiction, Positive Legal Fictionalism presented the use of a fiction
as the tool to facilitate objective cognition of relative legal norms. Presupposing the
basic norm qua fiction invests all relative legal norms with objective validity and facil-
itates several epistemic advantages: legal propositional knowledge, legal understand-
ing, realist semantics without complicated ontology, open-mindedness. These advan-
tages form the basis of Kelsen’s project of cognising the law.
Hence, Kelsen’s project should be understood as a ‘theory’ along these fictionalist
lines and considered ‘pure’ to the extent that the legal fiction applies only to positive
legal norms and not moral norms, as well as to the extent that the fiction facilitates the
genuine normative validity, which cannot be reduced to facts.
I hope to have shown that Positive Legal Fictionalism is also firmly rooted in Kelsen’s
account from an exegetical basis. However, Positive Legal Fictionalism is only a Kelseni-
an but not quite Kelsen’s own account. This is because Positive Legal Fictionalism went
beyond Kelsen’s own claims by (i) applying a unified account of fictions about the law
to all the different phases in his work, (ii) explaining Kelsen’s claims about the method
of cognising the law in terms of a fictionalist rather than a qualified Kantian frame-
work, (iii) explicitly linking Kelsen’s take on the epistemic use of fictions to Vaihinger,

Franz Steiner Verlag


Fictionalising Kelsen’s Pure Theory of Law 121

and (iv) developing from the previous points specific epistemic advantages which
Kelsen himself did not highlight: i. e. simple semantics without complicated ontolo-
gy, open-mindedness on the basis of the distinction between acceptance and belief.
Therefore, Positive Legal Fictionalism may be seen as a proposal for how to develop
Kelsen’s Pure Theory of Law further, rather than as account that merely summarises
its present status.

Acknowledgment

I want to thank Leslie Green, John Hyman, Jörg Kammerhofer, and Matthias Jestaedt
for their detailed and very helpful comments on an earlier draft of this essay. I also ben-
efitted considerably from presenting an earlier version of this essay at the 2018 Confer-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ence of the International Association for the Philosophy of Law and Social Philosophy
(IVR): Kelsen’s Pure Theory of Law: Conceptions and Misconceptions, at the Univer-
sity of Freiburg (Germany).

Maximilian Kiener
University of Oxford, St. Peter’s College, OX1 2DL Oxford, UK,
maximilian.kiener@philosophy.ox.ac.uk

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

II.

Franz Steiner Verlag


Rechtstheorie / The Mechanics of Law
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts
An Essay in Critical Reconstruction

MATHIEU CARPENTIER
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Abstract: This article focuses on Hans Kelsen’s late period, in which, breaking from views
that he had previously held dear, Kelsen espoused a novel theory of normative conflicts.
He claimed that since the principle of non-contradiction does not apply to norms, norma-
tive conflicts are not logical impossibilities. Two equally valid norms may conflict; wheth-
er the law provides the tools to solve such a conflict is not a logical matter, but is dependent
on contingent positive legal norms, also known as “metarules” such as Lex posterior. How-
ever, Kelsen did not want to cut the link between normative conflicts and invalidity (or
non-validity). He claimed that conflicts are solved through derogation. This is the claim
the present paper intends to analyse and ultimately refute. I then go on to introduce a new
conception of metarules conceived as norms of applicability, with no special bearing on
the conflicting norm’s validity.

Keywords: Hans Kelsen, derogation, normative conflicts, normative systems, legal valid-
ity, applicability

Schlagworte: Hans Kelsen, Derogation, normative Konflikte, normative Systeme, rechtli-


che Geltung, Anwendbarkeit

In1 1962, Hans Kelsen published an article called “Derogation”2, in which he shook up
some of his own well-entrenched beliefs. Even if the shortcomings and inconsistencies
of his earlier views did not totally disappear, that article made a crucial step in (what

1 Previous versions of this paper were presented at the Legal Philosophy Seminar of the University Pom-
peu Fabra in Barcelona and at the Conference of the German Section of the IVR in Freiburg. I would like to
thank Jose Juan Moreso, Chiara Valentini, Stanley L. Paulson, Jörg Kammerhofer and Thomas Hochmann
for invaluable comments.
2 Hans Kelsen, Derogation in: Essays in Jurisprudence in Honour of Roscoe Pound, ed. Ralph Newman, 1962.
This essay was reprinted in Hans Kelsen, Essays in Legal and Moral Philosophy, 1973. I cite and quote it here
from the 1973 edition.

Franz Steiner Verlag


126 MATHIEU CARPENTIER

I believe is) a good direction. It opened the way to Kelsen’s so-called late, sceptical3
period and was re-used – sometimes verbatim – in his last opus magnum, the General
Theory of Norms.
In this paper I shall try to re-elaborate and defend some of the basic tenets devel-
oped by Kelsen in the “Derogation” article, while trying to solve some of its main in-
consistencies. Here are some of the things which, controversial as they may be, I think
Kelsen basically got right:
(1) Repeal – or abrogation, or derogation4 – is a specific, self-standing normative
function. It does not prescribe, nor prohibits nor authorizes any behaviour. It
suppresses the validity of another norm from the normative system they both
belong to.
(2) Normative conflicts are not a matter of logical contradiction. Two conflicting
norms may be equally valid in a legal system; therefore, two conflicting nor-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

mative propositions (Sollsätze, descriptive propositions about norms) may be


true in the same time5. This is a major move by Kelsen away from what he

3 “Sceptical” should not be understood here as referring to so-called “rule-scepticism”, i. e. the thesis ac-
cording to which rules can never guide behaviour. Although Kelsen has sometimes been understood as
somehow paving the way for an interpretation-centred rule-scepticism, it is not what is at stake here – see,
on such an understanding, Michel Troper, La Théorie du droit, le droit, l’Etat, 2001, 67 ff.; Riccardo Guastini,
Rule-Scepticism Restated, in: Oxford Studies in Philosophy of Law Volume 1, ed. Leslie Green / Brian Leiter,
2011, 138 ff.; for a critique, see Christoph Kletzer, Kelsen’s Development of the Fehlerkalkül-Theory, Ratio
Juris 18 (2005), 53. The later Kelsen is sometimes dubbed sceptical because of his newfound stance that
there are no logical relations between norms (Hans Kelsen, General Theory of Norms (trans. M. Hartney),
1991, 189–193), or at least that norms are subject neither to the principle of non-contradiction nor to rules
of inference; but as Bruno Celano notes, Kelsen has to grant that at least some logical relations between
norms may exist (see Bruno Celano, Norm-Conflicts: Kelsen’s View in the Late Period and a Rejoinder, in:
Normativity and Norms, ed. Stanley Paulson / Bonnie Litschewski-Paulson, 1998, 345 ff.).
4 In what follows, I will use all those terms as broadly synonymous. Kelsen himself argues that the classical
distinction between abrogation (as total repeal) and derogation (a partial repeal) is mistaken, due to a
failure to appreciate the importance of (1). See Hans Kelsen, General Theory of Norms, 111–112. Therefore, I
will use the three words interchangeably. I must warn at the outset that in my own native language, French,
“derogation” does not mean the suppression of a norm, or of the validity thereof, from a normative system;
rather it means the creation of an exception in an individual case (as in “accorder une dérogation”, which
means roughly “to grant an exception”). When one uses the word “dérogation” in French, it is obvious to any
speaker that what is at stake is not a validity problem, but rather a momentary non-application of a norm.
Since I will argue that what is at stake in most normative conflicts is a matter of applicability and not of
validity, I shall make it clear that I will only use “derogation” in the technical meaning defined above, lest I
am accused of begging the question by using a French-infested notion of “derogation”.
5 Indeed, the downside of this conception is that two conflicting normative propositions (i. e. descriptive
propositions about norms) may be true in the same time. Depending on whether normative propositions
describe the norm-content or the norm’s validity, such a claim may entail very unfortunate consequences.
Kelsen himself does not seem to choose between the two functions of normative propositions, although
he only mentions propositions about validity (See Hans Kelsen, General Theory of Norms, 164). As Kelsen
shows (Hans Kelsen, General Theory of Norms, 223) statements about the validity / existence of the conflict-
ing norms are not contradictories: if p = there is a valid norm to the effect that that it is obligatory that P and
q = there is a valid norm to the effect that it is forbidden that P, since q ≠ ~ p, p and q are not contradictories.
Whether such statements of validity amount to statements of fact will not be discussed here (see generally

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 127

wrote just two years before in the second edition of the Pure Theory of Law: “A
conflict of norms is just as meaningless as a logical contradiction”6. In “Dero-
gation”, Kelsen claims that when two descriptive propositions contradict each
other, one of them is false from the outset, whereas in the case of a conflict of
norms, both norms are valid until one is repealed – or both are.
(3) Metarules (such as, per Kelsen, Lex Posterior7) are not logical principles but
positive legal norms8.

And here are the main flaws and shortcomings of Kelsen’s new theory:
(4) Kelsen’s classification of normative conflicts is clumsy at best. It confuses two
distinct problems: the conflict’s scope (regarding the classes of cases covered
by the norm’s antecedent) and the conflict’s nature (regarding the deontic fea-
tures of the norm’s consequent).
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

(5) More importantly, due to a confused notion of legal validity, Kelsen could not
take himself to apply (2) above to conflicts arising between norms of different
ranks within the normative hierarchy. Just a few paragraphs after writing that a
conflict between two norms does not entail anything about either norm’s va-
lidity, Kelsen writes that in the case of “unconstitutional statutes” “no conflict
exists”. This is due to a confusion made by Kelsen (and many others) between
validity as membership (of a norm within a legal system) and validity as con-
formity with higher ranking norms, including those prescribing the procedure
by which lower-ranking norms are to be created. More on this later.
(6) Although Kelsen’s formulation on this matter is quite ambiguous9, he seems
to believe that the only solution to a normative conflict is through repeal and

Carlos E. Alchourrón / Eugenio Bulygin, The Expressive Conception of Norms, in: Normativity and Norms,
ed. Stanley Paulson / Bonnie Litschewski-Paulson, 1998, 383 ff.). On the other hand, if normative proposi-
tions describe the norm-content, the description of a normative conflict necessarily yields two contradicto-
ry descriptive statements: if p = “it is forbidden to phi” and q = “it is not forbidden to phi”, then p and q are
contradictory. The same goes for r = “it is obligatory to phi” since r → q according to the axiom D of standard
deontic logic. Kelsen’s new tenet that the principle of non-contradiction does not hold for norms entails that
it does not hold any more for descriptive propositions about the content of norms.
6 Hans Kelsen, Pure Theory of Law (2nd ed., trans. Max Knight), 1967, 206; Hans Kelsen, Reine Rechtslehre,
1960, 210.
7 For reasons soon to be obvious, Kelsen refers only to Lex Posterior, and sets aside Lex Superior and Lex
Specialis.
8 For an overview of Kelsen’s shifting positions throughout his lifetime, see Stanley L. Paulson, On the
status of the Lex Posterior Derogating Rule, Liverpool Law Review 5 (1983), 5–18.
9 Kelsen writes: “The conflict can, but need not be, solved by derogation” (Hans Kelsen, Derogation, 271).
There are two ways to understand this sentence: on the first reading, Kelsen means that conflicts can be left
unsolved, but that when they are solved, the solution is always derogation; on the second reading, Kelsen
means that conflicts can be solved by derogation, but may also be solved through other means. Context in-
dicates that the first interpretation is correct as an exegetical matter – although I will argue that, as a matter
of fact, Kelsen should have meant the second one.

Franz Steiner Verlag


128 MATHIEU CARPENTIER

that metarules are “norms which regulate derogation” (in the sense of repeal).
But this is clearly not the case as a matter of empirical fact. Judges often solve
normative conflicts without altering the validity of either norm.
(7) Even if we grant that metarules are “norms which are about derogation”  –
which I am going to argue is not the case – there still remains an ambiguity
regarding what they actually do. On one interpretation, they are (pro futuro)
derogating norms. For instance, Lex Posterior means: “whenever a conflict oc-
curs, the prior norm ceases to be valid whenever the posterior norm becomes
valid”. On another interpretation, they are norms regulating derogation, that is,
power-conferring norms giving some officials (e. g. judges) the power to repeal
one of the two conflicting norms (or both).

In this paper, I will try to vindicate propositions (1)–(3) while showing that most of
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

the shortcomings of (4)–(7) can be avoided. Throughout, my claims will not be exe-
getical in nature; my aim is to offer a conceptual and critical reconstruction of Kelsen’s
later views, by trying to build a coherent picture of normative conflicts on what I be-
lieve are correct intuitions on Kelsen’s part.
My main claim is that normative conflicts are a matter of applicability, not validity,
and that metarules are not norms of derogation but norms of (in)applicability, even
if some legal systems also empower certain law-applying organs to repeal one of the
conflicting norms (or both). The paper is comprised of four parts. In the first part, I
will propose a tentative and incomplete classification of normative conflicts; in the
second part I will try to adumbrate the distinction between validity and applicability;
in the third part, I will show that metarules are norms of applicability, beginning with
Lex posterior. This claim vindicates Kelsen’s intuition that Lex Posterior is a positive
law norm and not a logical principal and in the same time rebuts his tenet that the only
way to deal with normative conflicts is through derogation. In the last part I will try to
show that the same holds for Lex Superior, against Kelsen’s clear and trenchant claims
to the contrary.

I. Types of Normative Conflicts

Let us assume that legal norms (and perhaps other norms as well) are conditional
norms. A conditional norm has the following structure: a factual antecedent specify-
ing the norm’s scope, that is, the properties of the generic or individual case regulated
by the norm; a normative consequent (or as Alchourrón and Bulygin would put it,
a normative solution), prescribing the behaviour to be adopted when the antecedent
is instantiated; and between the two a logical link (a conditional) which I will leave

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 129

non-specified10. In that perspective, categorical norms are conditional norms whose


antecedent is T, i. e. whatever tautology. This presentation of legal norms as condition-
al norms11 leaves unanswered the question of the individuation of norms. I shall not try
to defend any theory of norm-individuation here.
A crucial distinction has to be made between, on the one hand, the nature of a nor-
mative conflict, i. e. the ways two norms may conflict with each other, and, on the oth-
er hand, the scope of the normative conflict, i. e., the classes or subclasses or cases in
which the conflict actually occurs.

A. Contradiction, contrariety, incompatibility

The conflict occurs when the normative consequents of both conflicting norms cannot
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

be applied in the same time. Contrary to a persistent misconception, the criterion of


a normative conflict is not that both norms cannot be obeyed at the same time12. As a
matter of fact, what I will call normative contradiction is about two conflicting norms
which can both be obeyed. However, they cannot both be applied in the same timeby
law-applying organs: a law-applying organ cannot use both norms to justify its deci-
sion. If we were to say that norms conflict only when they fail to motivate behaviour,
we would reduce norms to motivational reasons. Here I will primarily take norms as
providing justificatory reasons for action. If you disagree with that, it does not matter
terribly, since it is rather a side issue for the present purposes.
There are three kinds of normative conflicts13.
The first kind of normative conflict can be called normative contrariety. In such a
case, the conflict involves two normative solutions with the same deontic operator and
opposite internal negations: for instance, OA (it is obligatory to do A) and O~A (it is

10 Here I will not elaborate on the problems that representing that link as the material implication (wheth-
er internal or external) creates (see on this Mathieu Carpentier, Norme et exception Essai sur la défaisabilité
en droit, 2014, 92–106). Neither will I ask here whether normative conflicts – and, more specifically, the kind
of normative conflict which is called in the literature “defeasibility” – call for a specific, defeasible condi-
tional. On this, see for instance, Mathieu Carpentier, Norme et exception, 268–281. For the present purposes,
I shall assume that all these problems have already been settled, one way or another.
11 Which is endorsed by Kelsen himself, see Hans Kelsen, Pure Theory of Law, 100.
12 A pathbreaking analysis in this respect is H. Hamner Hill, A functional taxonomy of normative conflict,
Law and Philosophy 6 (1987), 227 ff.
13 This typology borrows a lot from H. Hamner Hill (footnote 12), and also Stephen Munzer, Validity and
Legal Conflicts, Yale Law Journal 82 (1973), 1041–1043 and Risto Hilpinen, Normative conflicts and Legal
Reasoning, in: Man, Law and Modern Forms of Life, ed. Eugenio Bulygin / Jean-Louis Gardies / Ikka Niinu-
luoto, 1985, 195). However, neither Munzer nor Hilpinen consider what I call normative incompatibility,
and they both write about normative subcontrariety (PA and P~A) which does not strike me as involving
a genuine normative conflict (which Hilpinen candidly acknowledges). The very distinction between nor-
mative contrariety and contradiction, which I share with Hilpinen and Munzer, can already be found in
Jeremy Bentham, Of Laws in General, 1970, 110–111.

Franz Steiner Verlag


130 MATHIEU CARPENTIER

obligatory not to do A), which is the same as ~PA. Those solutions are deontic contra-
ries. In such a case, both norms cannot be obeyed in the same time14.
The second kind of normative conflict can be called normative contradiction: here
again we have two normative solutions with the same deontic operator but opposite
external negations: for instance, OA (it is obligatory to do A) and ~OA (it is not oblig-
atory to do A), which is the same as P~A. Obviously such a conflict does not involve
a motivational conflict. One can obey both norms, simply by doing A. You cannot be
said to have disobeyed a norm which permits you not to help your neighbour if you
do in fact help your neighbour, unless, of course there is also a norm forbidding you to
help your neighbour, which ex hypothesi is not the case here (if it were we would have
a case of normative contrariety, not contradiction). So, both norms can be obeyed.
However, both norms cannot simultaneously justify one’s behaviour. If a norm forbids
you to steal and another one authorizes you to steal, a judge who has to apply both is
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

faced with a genuine dilemma.


The third kind of normative conflict can be called normative incompatibility15 It in-
volves two actions A and B which can only be described by two incompatible propo-
sitions. Two propositions p and q are incompatible if and only if they cannot both be
true, but they can both be false: p and q are incompatible iff p ^ q is always false and it is
false that ~(~p^~q). It means that although p →~q and q →~p, it isn’t the case the p↔~q
nor that q↔~p, because it is not the case that ~p→q nor that ~q→p. Two actions can be
called incompatible if they can be described by two incompatible propositions. For
instance: it is obligatory that all houses are entirely paint in white and it is obligatory that
all houses are entirely paint in blue. Such norms are incompatible. Normative incompat-
ibilities break into two distinct subclasses, according to the kind of deontic operator at
stake. For any two incompatible actions A and B, there is a normative contradiction-in-
compatibility when OA and PB (or ~O~B) and there is a normative contrariety-in-
compatibility when OA and OB.
What is the point of distinguishing normative incompatibility (and its two sub-
classes) from normative contradiction and contrariety? When two actions A and B
are incompatible, it is not the case that ~A→B (and vice versa). Therefore, even though

14 According to Lars Lindahl, Conflicts in Systems of Legal Norms: A logical point of View, in: Coherence
and Conflict in Law, ed. Bob W. Brouwer / Ton Hol / Arend Soeteman / Willem van der Velden / Arie de
Wild, 1992, 39, only normative contrarieties are genuine normative conflicts since only them cannot be
simultaneously obeyed. But as I have hinted, I will consider normative conflicts as involving not only con-
flicting motivations, but also, and mainly, conflicting normative justifications. I would add that what some
normative incompatibilities (see infra) also involve conflicting motivations.
15 Such cases are not to be confused with what Uta Bindreiter calls “normative incompatibility”, which
is what happens when obeying simultaneously both norms is materially possible but is morally – or oth-
erwise normatively – subject to blame. (See Uta Bindreiter, Why Grundnorm?, 2000, 139–141). Bindreiter
also refers to what he calls empirical incompatibility, when there is an empirical impossibility to perform
both actions: for instance: you ought to work from noon to midnight; you ought to work from midnight to noon.
Neither cases are normative conflicts for the purpose of this paper.

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 131

A→~B, O~A^O~B is not a normative conflict, whereas O~A^O~~A is a normative


conflict: it is a normative contrariety. If it is forbidden to paint one’s house in white
and not to paint one’s house in white, we have a normative contrariety; if it is forbidden
to paint one’s house in white and to paint one’s house in blue, there is no normative
conflict whatsoever. Just paint your house in pink or in purple.
This triple (or even quadruple) distinction between contrariety, contradiction and
incompatibility is indeed schematic. It does not account for a certain type of norma-
tive conflicts which are not predicated on the classical deontic modalities (permission,
prohibition, obligation). For instance, it does not prove useful in case of a conflict be-
tween two power-conferring norms16.

B. The Scope of Normative Conflicts


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Normative conflicts do not always occur in all cases. They occur modulo the facts spec-
ified in the norm antecedent. Therefore, conflicts differ in their scope.
Here I will use (and marginally modify) Alf Ross’ famous classification of norma-
tive conflicts in On Law and Justice17. He distinguishes between three kinds of conflicts
in what concerns their scope.
The first kind is what Ross calls total (or, more precisely, total-total) conflicts, when
the sets of cases X and Y regulated by both norms are identical. For instance: for every
x, if x is a vehicle, it is forbidden for x to enter the park; for every x, if x is a vehicle, it is per-
mitted for x to enter the park.
The second kind is what Ross calls total-partial conflicts, when the set of cases X reg-
ulated by one norm is a subset of the set of cases regulated Y by the other. The conflict
occurs only when an instantiation of X obtains, and not when other cases belonging to
Y but not to X obtain. For instance, for every x, if x is a vehicle, it is forbidden for x to enter
the park; for every x, if x is an on-duty ambulance, it is permitted for x to enter the park.
The third kind is what Ross calls partial (or, more precisely, partial-partial) conflicts
when the set of cases X regulated by one norm intersects with the set of cases Y reg-
ulated by the other norm (a non-normative example of such a conflict is the famous
Nixon Diamond used in most systems of non-monotonic logic). The conflict occurs
only when an instantiation of X∩Y occurs; it does not in other instantiations or X nor

16 See for instance what Lindahl and Reidhav call “capacitative conflicts” (Lars Lindahl and David Reidhav,
Conflict of Legal Norms: Definition and Varieties, in: Logic in the Theory and Practice of Lawmaking, ed.
Michał Araszkiewicz / Krzysztof Płeszka, 2015, 49 ff.).
17 Alf Ross, On Law and Justice, 1958, 129. For a recent formalisation of Ross’s taxonomy, see Abdullatif
Elhag / Joost Breuker / Bob Brouwer, On the Formal Analysis of Normative Conflicts, in: Legal Knowledge
Based Systems (JURIX ’99), ed. Jaap van den Herik, 1999, 37–38.

Franz Steiner Verlag


132 MATHIEU CARPENTIER

Y. To borrow an example from Jorge Luis Rodriguez18: if the light is red, one should stop
one’s vehicle; if one is driving through a military zone, one should not stop one’s vehicle.
This classification seems correct. I will add a caveat however. In cases of partial-par-
tial conflicts, no specificity-based criterion of conflict-resolution (such as Lex Specialis)
can be used, because none of the conflicting norms is “more special” than the other.
Or so Ross claimed19. It is however a bit more complicated than that. In most legal sys-
tems, what seems to be a prima facie partial-partial conflict will often be transformed in
total-partial conflict due to certain (conscious or unconscious) interpretative assump-
tions. Consider the two following norms. N1: For every x, if x wilfully causes the death of
another human being, x shall be punished with thirty years in prison N2: For every x, if x is
suffering from a psychological disorder and commits a crime, x shall not be punished Now
it seems that what we have here is a partial-partial conflict, which cannot be solved
using Lex Specialis. However, most lawyers share an interpretative assumption of spec-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ificity which operates in favour of N2. Either they presuppose an additional norm N3:
whoever commits a crime shall be punished, in which case there is a total-partial conflict
between N2 and N3; or they conceive N2 as being not a single norm, but the sum of a
great number of specific norms, such as N2’: For every x, if x is suffering from a psycho-
logical disorder and commits murder, N2’’: For every x, if x is suffering from a psychological
disorder and commits theft, etc , in which case there is a total-partial conflict between
N2’ and N1.
This is why Lex Specialis applies even in cases where the conflict at hand looks like a
partial-partial one, due to some specificity assumption.

C. What About Kelsen’s Own Classification?

In “Derogation”20 and in General Theory of Norms21, Kelsen worked out a theory of nor-
mative conflicts which rests on a specific typology. Conflicts are distinguished accord-
ing to whether they are “necessary” or “possible”, “bilateral” or “unilateral”, and “total”
or “partial”. I have criticized this classification elsewhere, and I will not rehearse my
arguments here22. Suffice it to say that it is riddled with confusions, insofar as it fails to
distinguish between the nature of a conflict and the scope thereof.
For instance, the notion of a necessary conflict rests on considerations about the
conflict’s scope: a conflict is necessary when it occurs in every case, whereas it is

18 Jorge Luis Rodríguez, Contradicciones Normativas. Jaque a la concepción deductivista de los sistemas
jurídicos, Doxa 17/18 (1995) 372.
19 Alf Ross, On Law and Justice, 131.
20 Hans Kelsen, Derogation, 269
21 Hans Kelsen, General Theory of Norms, 123 ff.
22 See Mathieu Carpentier, Norme et exception, 228–230.

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 133

merely possible when it occurs only on some cases. The notion of a bilateral conflicts
deals with the presence of a conflict “on the side” of the norm: in what Ross calls a
total-partial conflict, the conflict is necessary on the side of the special norm, and
merely possible on the side of the general norm. The problem with such a classifica-
tion is that what Kelsen calls an “unilateral” conflict is not a conflict at all. His exam-
ple of such a “conflict” is the following: “Norm 1: Murder is to be punished by death,
if the murderer is over 20 years of age. Norm 2: Murder is to be punished by death, if
the murderer is over 18 years of age”. There is no conflict on the side of Norm 1, but
according to Kelsen there is a possible conflict on the side of Norm 2 (since Norm 2
may entail the punishment of a 19-year-old murderer for instance). Whatever prob-
lems are raised by the formalization of conditional norms, it is common ground that
the antecedent of a norm is a sufficient condition of the realization of the consequent.
But Kelsen’s notion of a unilateral conflict rests on a more robust conception of con-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ditional norms, which is predicated on the antecedent being a necessary condition as


well. There is a conflict between Norm 1 and Norm 2 only if Norm 1 is reinterpreted
as “Norm 1a: Murder is to be punished by death, if and only if the murderer is over 20
years of age”. Norm 1a implies Norm 1b: “If it is not the case that the murderer is over
20 years of age, then murder is not to be punished by death” which then (and only
then) conflicts with Norm 2. It is what Ross calls a partial-partial conflict, since the
class of cases about murderers over 18 intersects with the class of cases about murder-
ers under 20. But if Norm 1 expresses only a sufficient condition, there is no conflict
since the same normative solutions is applied to two classes of cases one of which is
included in the other. There is no conflict between Norm 1 and Norm 2, not any more
than between Norm 3: “If one owns farm animals, one ought to pay a tax” and Norm
4: “If one owns a cow, one ought to pay a tax”.
To sum up: Kelsen’s classification of conflicts is flawed; hence I will use the classifi-
cation elaborated above instead of Kelsen’s, however partial and incomplete it may be.

II. A Few Words on the Validity / Applicability Distinction

It is well known that Kelsen defines validity as the specific mode of existence of norms23.
This definition rests on an assimilation of the norm’s existence with its bindingness.
There exists a norm to the effect that one ought to φ if and only if it is the case that one

23 On this, see (among an extensive literature), Dick W. P. Ruiter, Legal validity qua specific mode of ex-
istence, Law and Philosophy 16 (1997), 479–505; Carlos S. Nino, Confusions surrounding Kelsen’s concept
of validity, in: Normativity and Norms, ed. Stanley Paulson / Bonnie Litschewski-Paulson, 1998, 255 ff.; An-
dras Jakab, Problems of the Stufenbaulehre. Kelsen’s Failure to Derive the Validity of a Norm from Another
Norm, Canadian Journal of Law and Jurisprudence 10 (2007), 35–68. A very clear introduction can be found
in Clemens Jabloner, Der Rechtsbegriff bei Hans Kelsen, in: Rechtstheorie: Rechtsbegriff – Dynamik – Ausle-
gung, ed. Stefan Griller / Heinz Peter Rill, 2011, 24 ff.

Franz Steiner Verlag


134 MATHIEU CARPENTIER

ought to φ. A norm’s existence coincides with the duty to obey it. Such a conception
of validity may be labelled a strong conception of validity. It can be contrasted with a
weak notion of validity which treats validity as a matter of membership of a norm within
a normative system. According to the weak conception, a norm exists only if it belongs
to a normative system. A legal norm exists if and only if it belongs to a legal system. The
notion of membership is useful insofar as it has no stakes in the ontological question
of what makes a norm exist qua norm. In that sense, legal validity, understood as mem-
bership within a legal system, is what endows a norm with its legal character – i. e. what
makes it exist qua law –, regardless of what makes it exist qua norm in the absolute. It
answers the question: why is it the law that one ought to φ? And it sets aside the quite
trickier question: what makes it the case that one ought to φ? On the weak conception,
a legal system uses a set of criteria of membership which will be picked out by a specific
rule or set of rules. This is roughly the Hartian notion of a rule of recognition, and it is
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

quite remote from the Kelsenian conception of validity as bindingness (although Kels-
en occasionally comes close to using “validity” in the weak sense24).
On the weak conception, derogation is merely about the suppression of a norm’s
membership, and it does not bear on other dimensions of what is usually called “va-
lidity”. Indeed, a weak notion of validity allows us to distinguish membership from
applicability.
The notion of applicability in contemporary jurisprudential literature (flowing
mainly from Eugenio Bulygin’s seminal work25) is a reinterpretation of the notion of
bindingness, severed from the membership dimension of normative validity. The ap-
plicability of a legal norm is to be understood as the duty (or power) that a law-apply-
ing organ has to apply it. As Bulygin has shown, a norm’s applicability is contingent on
the legal system containing norms of applicability, i. e. norms which regulate the way
other norms ought to be applied. Such norms of applicability regulate not only norms
of the legal system but norms belonging to other normative (not necessarily legal) sys-
tems as well. Norms of applicability come in various types. They may be general norms
meant to be applied to a vast range of norms: for instance, the nullum crimen principle
and the principle of retroactivity in mitius are such general norms of applicability. They
also may be particular norms regulating the applicability of a specific norm or set of

24 For instance, Kelsen writes: “Why does a certain norm belong to a certain order? And this question
is closely tied to the question: Why is a norm valid, what is the reason for its validity” (Hans Kelsen, Pure
Theory of Law, 193). In the German original text: “warum gehört eine bestimmte Norm zu einer bestimmten
Ordnung? Und diese Frage steht in einem engen Zusammenhang mit der Frage: Warum gilt eine Norm,
was ist ihr Geltungsgrund?” (Hans Kelsen, Reine Rechtslehre, 196). However, Kelsen is clear that “closely
tied” is not the same as “equivalent”. And in the following paragraph, Kelsen does equate validity with
bindingness (Verbindlichkeit).
25 See mainly Eugenio Bulygin, Time and Validity, in: Essays in Legal Philosophy, ed. Carlos Bernal et al.,
2015, 171–187.

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 135

norms: for instance, the section of a statute which deals with the statute’s own “entry
into force”.
Applicability thus understood is a kind of bindingness. In order to grasp the dif-
ference between applicability and membership, we can go back to the Ross/Kelsen
debate on validity. As Alf Ross showed in a famous article26, a distinction must be made
between a legal norm’s bindingness and a legal obligation. For instance, legally you
may not kill or steal. Legally a judge may not refuse to adjudicate a matter brought
before her. Those are legal statements, referring to legal obligations (or rather prohibi-
tions). Bindingness, on the other hand, must be understood as a duty to obey. As Ross
astutely noticed, there is no legal obligation to obey the law. There only is a legal obli-
gation or permission to φ. This is why Ross dismissed Kelsen’s conception of validity
as bindingness as hopelessly infected with some natural law disease. An obligation to
obey the law cannot be legal; it must be a moral obligation.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

As Bulygin observed27, the nexus of Ross’s objection to Kelsen’s notion of validity


is that it ultimately rests on an absolute conception of bindingness: absolute binding-
ness means that a norm is binding on its own (moral) merits. However, the notion
of applicability reveals a thoroughly relative conception of bindingness, which allows
for a specifically legal kind of bindingness: courts are vested with a duty (or at least a
power) to apply legal norms, which is grounded in other legal norms, which Bulygin
calls norms of applicability. When we say that laws are binding on judges we do not
mean that they have a duty to obey, but rather a duty to apply. Legal norms are legally
binding when a certain set of officials is under the duty to apply them. This kind of
bindingness is strictly legal, since it is predicated on the existence of a specific set of
legal norms.
The distinction between membership and bindingness understood as applicability28
is not necessarily incompatible with Kelsen’s general outlook since it does not say an-
ything about the question of absolute bindingness. And it is able to account for legal
complexities that the mere identification of validity with bindingness cannot allow.
For instance, it allows us to account for situations where applicability and member-
ship are somehow disjoined. In cases of vacatio legis, the norm belongs to the legal
system (which is proven by the fact that normative operations such as amendment
or derogation are possible in the interim) but it is not yet applicable: though it is not

26 Alf Ross, Validity and the Conflict between Positivism and Natural Law, in: Normativity and Norms, ed.
Stanley Paulson / Bonnie Litschewski-Paulson, 1998, 153.
27 Eugenio Bulygin, The Problem of Legal Validity in Kelsen’s Pure Theory of Law, in: Essays in Legal Phi-
losophy, ed. Carlos Bernal et al., 321.
28 I have raised doubts about this distinction in Mathieu Carpentier, Validity versus Applicability: a
(Small) Dose of Scepticism, Diritto & Questioni Pubbliche, 18 (2018), 107–132. However, my doubts do not
as much bear on the correctness of the conceptual distinction itself as they are concerned with the poor
heuristic value of the notion of “mere membership”.

Franz Steiner Verlag


136 MATHIEU CARPENTIER

quite normatively inert29 in the meantime, judges and other officials have no duty to
apply them. This shows that membership is not a sufficient condition of applicability.
Conversely, it is not a necessary condition either: a judge is sometime bound to apply
a foreign legal norm when a choice-of-law rule directs her to do so. And in cases of
normative post-activity, a judge is sometimes bound to apply a norm to a set of cases
which occurred after the removal of the norm from the legal system.

III. Metarules as Norms of Applicability

My claim in the present paper is that metarules are norms of applicability. I will first
show it using the very metarule Kelsen focuses on: Lex posterior. Following the dis-
tinction sketched out in (7) above, I will try to show that Lex posterior is neither a pro
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

futuro derogating norm, nor a power-conferring norm regulating derogation.

A. Meta-rules are not pro futuro derogating norms

Derogation is a central feature of legal systems. A legal authority could not properly
discharge her job or fulfil her function if she were not able to cancel past pronounce-
ments made by her predecessors or by herself. It would be difficult to conceive a legal
system close to the ones we know, and work with, that would not have such a feature.
Derogation is a specific normative function, as Kelsen pointed out. A derogating norm
has the sole effect of removing the norm from the legal system; that is, because the leg-
islature enacted the relevant norm, a previous norm disappears from the legal system.
Derogation is what allows a legal system to change; it is a function of what Hart called
the “rules of change” of a given legal system30, which empower lawmakers to repeal
earlier las and to make new ones.
It is tempting to think that this is the kind of process which is captured by the max-
im Lex posterior derogate a priori31. However, I will argue that Lex Posterior captures
another kind of situation. Indeed, it should be stressed out that there is no conflict be-
tween a derogating norm and the norm it derogates. If Lex Posterior is to play the role
of a metarule allowing officials to solve a normative conflict that, it makes no sense to
argue that it should be used when a norm derogates another. On the contrary, Lex Pos-
terior, qua metarule, deals with situations where a norm is neither a derogating norm,

29 Mathieu Carpentier, Validity versus Applicability: a (Small) Dose of Scepticism, 122.


30 On which see infra.
31 See e. g. Jörg Kammerhofer, Uncertainty in International Law, 2011, 157–158: “it seems obvious how the
problem of change is a problem of norm-conflict, what is being argued is the reverse, namely that change in
law presupposes the lex posterior maxim”.

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 137

nor a compound of norms one of which being a derogating norm, yet it conflicts with
a previous norm of the legal system32.
From here, one of the easiest moves is to claim that Lex posterior itself is a derogat-
ing norm: whenever two norms conflicts, Lex posterior derogates the prior norm. This
is the view Kelsen seems to espouse, both in “Derogation” and in the General Theory of
Norms: derogation is “the function of (…) a positive legal norm; this norm [Lex Poste-
rior] is not one of the two conflicting norms but a third norm which specifies that one
of the two conflicting norms loses its validity, or that both norms lose their validity”33.
This view leads to a theory of implied repeal. Due to norm A (Lex posterior), whenever
norm B conflicts with posterior norm C, then B is not valid, and is thereby repealed
whenever C becomes valid34. When we say that C impliedly repealed B, what we mean
is that B is repealed by Lex posterior itself.
Kelsen maintains that this a matter of positive law, not logical truth. He is right in
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

that respect. However, this picture is not without problems in other respects.
First, it happens that metarules conflict one with another. Whenever the posterior
norm is more general than the earlier norm, there is a conflict between Lex posterior
and Lex specialis35. For instance, let us suppose that, due to norm N1, vehicles have
permission to enter the park between 6PM and 6AM, and they are forbidden to do so
between 6AM and 6 PM. Then, let us say that a norm N2 is enacted to the effect that
ambulances on duty are allowed in the park at all times and that a few months later a
norm N3 is enacted to the effect that N1 is derogated and vehicles are banned from
the park at all times. Does Lex posterior or Lex specialis apply? Which of these two
metarules ought to prevail? It is plain that N3 derogates N1 insofar as N1 permits cars

32 This I take to be Kelsenian orthodoxy. See Hans Kelsen, General Theory of Norms, 108. See also on the
same idea Giovanni Battista Ratti, Negation in Legislation, in: Logic in the Theory and Practice of Lawmaking,
ed. Michał Araszkiewicz / Krzysztof Płeszka, 2015, 147, who accepts this picture but admits that “it is inter-
esting to observe that, strictly speaking, no act of derogation is carried out here”.
33 Hans Kelsen, General Theory of Norms, 125 (see also Hans Kelsen, Derogation, 273).
34 Kelsen would accept, I think, that this picture is oversimplified. Norm B may remain applicable however
repealed by C if C is due for entry into force at a later date. But it does not affect the basic point made by
Kelsen here.
35 In cases of total-partial conflicts Lex specialis seems particularly well suited to resolve the conflict, for
instance by creating an exception. It may seem strange to claim that creating an exception derogates, that is
repeals, the rule from which it is excepted; in other words, it seems strange to conceive Lex specialis as a
derogating norm. However, it is plain that if Lex Specialis is a derogating norm, it is only insofar as it der-
ogates the general norm pro tanto. What is derogated from the legal system is not the defeated norm, but
som” logical consequence of that norm (on the idea of legal systems are deductively closed, insofar as they
comprise not only valid norms but also the logical consequences thereof, see Carlos E. Alchourrón / Euge-
nio Bulygin, Normative Systems, 1971, 185; Carlos E. Alchourrón / Eugenio Bulygin, Análisis lógico y Derecho,
1991, 129, 163, 219…). In the context of a discussion of defeasibility and implicit exceptions, J. Ferrer Beltran
and G. B. Ratti have recently made a case for this notion of exception as pro tanto derogation (see Jordi
Ferrer Beltran, Giovanni Battista Ratti, Validity and Defeasibility in the Legal Domain, Law and Philosophy
29 (2010), 601–626). I will not have the time to study this theory in detail; suffice it to say that I intend to
defend a contrarian view in the present paper.

Franz Steiner Verlag


138 MATHIEU CARPENTIER

to enter the park between 6AM and 6PM. But does N3 “implicitly” (or “impliedly”)
derogates N2 as well? (to reformulate: does Lex Posterior derogates N2 as well?) Or is
it the case that N2 is still valid, and that N3 is “impliedly” derogated by Lex specialis
insofar as it prohibits ambulances on duty? If both metarules are derogating norms36,
this conflict of metarules entails that N2 and N3-insofar-as-it-covers-ambulances are
both valid and invalid. This seems rather unfortunate. On the other side, the idea of a
conflict of norms of applicability is much more appealing since it is common-sensical
that a judge can be subject to two conflicting duties.
Another defect with the view of metarules as derogating norms is that in such a con-
ception metarules (that is, per Kelsen, mainly Lex Posterior) play the same role in the
positive law conception of normative conflicts than in the logical conception, which
Kelsen once espoused before rejecting it at a later stage, as proposition (2) above has
shown. Under the logical conception, when two norms conflict, it is impossible that
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

both be valid in the same time, so it must be the case that one of them ceases to be
valid. So if Kelsen’s new stance is that although it is not a logical necessity, it is always
the case that a normative conflict produces derogation as all legal system have a Lex
posterior metarules as part of their positive law, one cannot see clearly what is gained
in Kelsen’s new view and why it should be deemed ground-breaking. Once we dis-
tinguish between membership and applicability, we can understand how a norm may
cease to operate due to some normative conflicts without being forced to draw the
conclusion that it stops belonging to the normative system.
A third defect is that such a conception of metarules in general, and Lex Posterior
in particular, does not fit with some of our basic intuitions about what is commonly
called – in UK law for instance – “implied repeal”, which, it turns out, is not about re-
peal at all. The very fact that judges use Lex Posterior shows that there is a conflict in
the first place, and that the conflict has not been pre-emptively solved by an implicit re-
peal. If there is a conflict to be solved, then there is a conflict in the first place. Moreover,
if there is a conflict between N1 and N2 (N2 having been enacted at a later time than
N1) and if N2 is itself repealed by a subsequent repealing norm N3, then the earlier law
N1 becomes applicable again, without any further act from the lawmaker37 (whereas in
cases of derogation, it is commonly admitted that the derogation of a derogating norm
does not reinstate the norm that was derogated in the first place). Interpreting that
repealing norm, to wit N3, as not only abrogating N2, but also as somewhat implicitly
“re-enacting” N1 seems quite a heavy price to pay. Therefore, Lex Posterior is best un-

36 I set aside the hypothesis that Lex specialis and Lex posterior do not discharge the same function. I will
say more about this at the end of this paper.
37 As Jeffrey Goldsworthy (who writes about the UK doctrine of implied repeal) puts it : “The inconsistent
provisions of the earlier statute are not, as it were, expunged from the statute book: if the later statute were
to be formally repealed, the earlier one should be fully revived” ( Jeffrey Goldsworthy, Parliamentary Sover-
eignty: Contemporary Debates, 2010, 289).

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 139

derstood as a norm of applicability directed at legal officials, and mainly law-applying


organs. It directs officials not to apply the earlier of two conflicting norms.
One last point. As Kelsen observed38, in some legal systems Lex Posterior coexists
with Lex Prior, according to which the earlier of the two conflicting norms is repealed
in case of a conflict. For instance, in the UK, the doctrine of “constitutional statutes”
asserts that when a statute is of constitutional significance, the doctrine of implied re-
peal cannot be used to solve conflicts with later statutory norms. Instead, the courts
ought to disapply posterior statutory norms which conflict with constitutional statutes
insofar as they clash with them39. But it would make little sense to claim that the later
norm is repealed in the process, unless we are ready to be committed to the idea of a
norm’s nullity (rather than its annullability): the later norm would then never exist in
the legal system, even though it was duly passed and enacted. This is not a conceptual
impossibility, but it does sound counter-intuitive. Indeed, if the idea of Lex posterior
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

as a derogating rule is attractive, it is because of the intuition I mentioned in the begin-


ning of the present section: a future legislator must be able to change past legislation.
As there is “explicit repeal”, there should be “implied” repeal when none of the con-
flicting norms are derogating norms. But as soon as we notice that Lex prior cannot be
reasonably described as operating this way, we can have doubts about the derogating
function of Lex posterior itself.

B. Meta-rules are not power-conferring norms regulating derogation

As we saw in (6) there is another option on the table: we can analyse metarules as rules
regulating derogation. On such a conception, meta-rules are norms empowering a cer-
tain subset of officials to repeal whichever of the two conflicting norms is defeated. Let
us examine this view in further detail.
First, a basic distinction between law-creating and law-applying organs should be
assumed40. Such an assumption will not be disputed here, however debatable it may be.
There is no doubt that law-creating organs are empowered to derogate earlier norms.
That’s part of the very notion of normative empowerment. But again, Lex Posterior,
and metarules in general, do not apply to derogating norms enacted by the lawmaker

38 Hans Kelsen, General Theory of Norms, 126.


39 See the opinion by Lord Justice Laws in Thoburn v Sunderland City Council, [2002] EWHC 195 (Admin).
This doctrine has recently been espoused by the Supreme Court: see H v The Lord Advocate, [2012] UKSC
24; R (HS2 Action Alliance Ltd) v Secretary of State for Transport, [2014] UKSC 3. See also, for an audacious
use of the notion of “constitutional statutes” (although it does not concern the implied repeal doctrine) R
(on the application of Miller and another) v Secretary of State for Exiting the European Union, [2017] UKSC 5.
40 See on this distinction Paolo Sandro’s important new book, The Creation and Application of Law: A
Neglected Distinction, 2020 (forthcoming).

Franz Steiner Verlag


140 MATHIEU CARPENTIER

itself: it is agreed that law-creating organs are empowered to repeal norms from the
legal system, because it is what lawmakers do.
In the view discussed here, law-applying organs, which have to deal with normative
conflicts, are empowered to repeal the norm which is defeated due to the application of
a meta-rule, sometimes with retroactive effect. Thus understood, meta-rules empower
judges and other officials to derogate, that is suppress the norm from the legal system.
I am not sure such an interpretation is consistent with what Kelsen has to say about
Lex posterior. However, as we shall see in the next section, the idea that one of the
conflicting norms is resolved through derogation by a certain organ other than the law-
maker is essential in the way Kelsen deals with conflicts (or, according to Kelsen, the
absence thereof) arising between lower- and higher-ranking norms.
For now, suffice it to say that the claim that normative conflicts are necessarily
solved through derogation of either conflicting norms by the law-applying organ does
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

not square well with the common conception of law-application. Removing a norm
from a legal system is not normally part of the job a law-applying organ. Such is rather
the lawgiver’s task. The distinction between membership and applicability allows us to
understand what happens when a judge resolves a normative conflict: meta-rules do
not as much direct her to remove the norm from the legal system (which is a matter of
membership) as they empower her to disapply the defeated norm.
Of course, further power-conferring norms, which are not identical to metarules
themselves, may empower the law-applying organ to resolve the conflict through der-
ogation. But it is by no means a conceptual necessity. It is contingent on the exist-
ence of further norms within particular legal systems. The default option available
to judges and other law-appliers is the use of metarules which are merely norms of
applicability. Of course, as Kelsen claimed throughout his later writings, the existence
of meta-rules themselves is a contingent matter, since it is thoroughly dependent on
positive law. However, from a conceptual point of view, it makes much more sense
to claim that metarules are norms of applicability rather than norms regulating dero-
gation, given that they are to be used by law-appliers rather than by law-creators. Be-
sides, although it is true that no metarule expresses a logical law, it is necessarily41 the
case that a legal system should at least comprise one metarule, whatever it is. Judges
and other officials are always bound to be faced with normative conflicts, and neces-
sarily at least one metarule will emerge, if only by way of custom. Sometimes they will
also be empowered to derogate norms, but that is quite distinct from applying Lex
posterior or Lex specialis.

41 This may not be a logical or conceptual kind of necessity. A weaker conception of necessity (e. g. mate-
rial, or natural, necessity) will suffice for the present purposes.

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 141

IV. And the Same Goes for Lex Superior

Until now, I have talked of metarules in the abstract, although my main example was
Lex posterior. A special case seems to be Lex superior. On Kelsen’s view, Lex superior
is not a meta-rule proper, it is rather a structural feature of every legal system. In other
words, the reason why Kelsen never pays attention to Lex Superior is that he is still
committed to the logical conception of it: Lex superior as a logical principle is pre-
cisely what Kelsen means by legal dynamics. In any legal system a norm is valid if and
only if it has been created according to the procedure set out in a higher-ranking norm.
In this fourth and last section I intend to claim that Kelsen should have bitten the
positive-law bullet; he should have acknowledged that Lex superior is no more a log-
ical principle of legal systems than Lex posterior ever was. In other terms, he should
have realized that Lex superior is a meta-rule like any other, and that it has no bearing
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

on the conflicting norms’ membership, but only on their applicability.

A. The Problem of Unlawful Law

The problem of unlawful law is as old as legal philosophy itself42. Hobbes famously
pronounced that no law may be deemed unlawful since all law derives from the sover-
eign. Modern constitutionalism grappled with this problem, never to completely solve
it. If law is to be the expression of the sovereign (even more so in case of a popular
sovereign), it cannot be circumscribed by legal limits: the sovereign cannot bind itself.
However, modern legal doctrine during after the French and American Revolutions
evolved a series of conceptual tools43 which allowed to solve this paradox – or so it was
claimed.
The notion of unconstitutional statutes, and, more generally, of unlawful law, creat-
ed a different kind of problem for Kelsen, and for other members of the Vienna School
as well44. The paradox is well known: if a norm is valid, then its creator must ex hypoth-

42 See generally Paulo Sandro, Unlocking Legal Validity: Some Remarks on the Artificial Ontology of Law,
in: Legal Validity and Soft Law, ed. Pauline Westerman / Jaap Hage / Stephan Kirste / Anne-Ruth Mackor,
2018, 112 ff.
43 Such as the pouvoir constituant vs constitué distinction by Sieyès (on which the best analysis remains Carl
Schmitt, Verfassungslehre, 1928, 98 ff.).
44 See Christoph Kletzer, Kelsen’s Development of the Fehlerkalkül-Theory, 46 ff.; Jörg Kammerhofer, Un-
certainty in International Law, 189 ff.; Johannes Buchheim, Fehlerkalkül als Ermächtigung? Kelsens Theo-
rie des Rechts letztverbindlicher Entscheidungen vor dem Hintergrund von H. L. A. Harts Rechtstheorie,
Rechtstheorie 14 (2014), 59–78; Thomas Hochmann, Les théories de la “prise en compte des défauts” et de
l’“habilitation alternative”, in: Un classique méconnu: Hans Kelsen, ed. Thomas Hochmann / Xavier Mag-
non / Regis Ponsard, 2019.

Franz Steiner Verlag


142 MATHIEU CARPENTIER

esi have been empowered to take such a norm45. Therefore, if a norm is valid it cannot –
as a matter of logical principle – conflict with higher-ranking norms, since its very con-
formity with higher-ranking norms grounds its legal validity. The logical conclusion
is that whenever a conflict occurs between two norms whose rank in the normative
hierarchy is different, the conflict is only apparent. This conclusion is paradoxical in-
sofar as unconstitutional statutes do not look like they are logical incompatibilities. If
they were, one could hardly understand why Kelsen would (rightly) be celebrated as
the creator of European-style constitutional courts46. If unconstitutional norms were a
logical impossibility, that is if unconstitutional norms were null ab initio, there would
be no need for a special organ empowered to remove them from the legal system. Only
if unconstitutional statutes are valid in the first place is such an organ necessary. And,
of course, Kelsen rightly endorsed the idea that legal norms are not null, insofar as they
are only annullable47.
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

It is well known that Kelsen solved this paradox by developing a theory of alter-
native authorisations. The basic idea is that a Constitution does not only comprises
norms empowering the legislature to legislate in such-and-such manner. The Consti-
tution also contains an “alternative” clause which empowers the legislature to act as it
sees fit. Whenever the legislature (for instance) passes a statute that conflicts with a
constitutional norm, this statute must be deemed consistent with the alternative au-
thorization, which is why it is valid. It is valid insofar as, though conflicting with a con-
stitutional norm, it conforms with an alternative constitutional norm which empowers
the legislature to pass this statute. This is a strange idea, after all: a constitution is meant
to be a frugal menu, not an all-you-can-eat buffet. If the Constitution creates a special
procedure for legislation, then it should implicitly preclude any other procedure from
being lawfully followed by the legislature48. Expressio unius, exclusio alterius.
The crucial point is that, according to Kelsen, when a constitutional court finds a
statute to be unconstitutional and thereby repeals it, it does not entail that the statute

45 I gloss over some serious problems which I will not be able to address at length: for instance, there may
be an asymmetry between on the one hand procedural constitutional norms which regulate the procedure
the legislature must follow if it is to legislate and on the other hand substantive constitutional norms which
regulate the content of legislation. For brevity’s sake, I will lump those (arguably) very different dimensions
together and focus only on the scope of the constitutional empowerment.
46 A very interesting contribution on Kelsen’s role in the creation and early life of the Austrian Consti-
tutional Court is Christian Neschwara, Kelsen als Verfassungsrichter, Hans Kelsen: Staatsrechtslehrer und
Rechtstheoretiker des 20  Jahrhunderts, ed. Stanley L. Paulson / Michael Stolleis, 2005, 353 ff.
47 Hans Kelsen, Pure Theory of Law, 276 ff. It bears noticing that in his seminal 1929 article on constitution-
al adjudication, Kelsen was not very clear on that problem, since he claimed that nullity and annulability
were equivalent, and equally plausible, ways of organizing a normative system (see Hans Kelsen, Wesen
und Entwicklung der Staatsgerichtsbarkeit. 2. Mitbericht von Professor Dr. Hans Kelsen in Wien, Veröffen-
tlichungen der Vereinigung der Deutschen Staatsrechtslehrer 5 (1929), 44–49). He later repudiated this view.
48 Kelsen sometimes seems to think that the only thing a statute must have to be valid is to be published in
the Official Gazette (see Hans Kelsen, Wesen und Entwicklung der Staatsgerichtsbarkeit, 46).

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 143

was void ab initio, even though the court’s decision may be retroactive. What the court
does is repealing a perfectly valid statute, just as a legislature would do, hence the fa-
mous theory of the “negative legislator”49. However, the court can only do so on the
grounds of the statute’s incompatibility with the explicit constitutional norm, whereas
the legislature may repeal an earlier statute for whatever reason it likes. But when no
organ (other than the lawgiver) is empowered to “annul” the defective norm, then the
norm remains fully valid (it fully belongs to the legal system).
However, Kelsen, adhered until the end of his life to the notion of alternative author-
ization, and he did so well after his “sceptical turn”50. In the General Theory of Norms, he
wrote: “[In the case] of an unconstitutional statute, it should be noted that according
to positive law a so-called ‘unconstitutional’ statute can be valid, but its validity can be
repealed by a special procedure provided for in the constitution, e. g. by the decision of
a special court. In such a case, there is no conflict of norms. For if the norm in question
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

is valid, it is also constitutional, that is, the constitution empowers the legislator to en-
act the statute in question but provides that it can be repealed by a special procedure”51.

B. The Possibility of Genuine Normative Conflicts Between Lower-


and Higher-Ranking Norms

There are many problems with Kelsen’s view. The main one is that Kelsen refuses to
allow that there are genuine normative conflicts between lower- and higher-ranking
norms. I aim to show that such genuine conflict is possible, although the conflicting
norms are equally valid. Such a hierarchical normative conflict is to be solved by me-
tarules. Lex superior is one such metarule, and like any other, it is a norm of applica-
bility. On such a view there is no need to have recourse to the baroque notion of an
alternative authorization. The downside is that the very notion of legal dynamics is put
into question; law is then seen as a system of non-hierarchical sources of validity.
How can such a bizarre (but sound) theory be achieved? There are actually many
ways to that, and although I have advocated for one specific theory somewhere else52, I

49 Hans Kelsen, Wesen und Entwicklung der Staatsgerichtsbarkeit, 54.


50 For a very interesting reconstruction of Kelsen’s later views on the unlawful law problem, see Jörg Kam-
merhofer, Uncertainty in International Law, 192–193. Kamerhofer argues that in his sceptical phase, Kelsen
would have distinguished between a norm’s existence (which results from any kind of act of will) and its
membership (which is a specific property of certain norms). However, such a distinction does not explain
how un unlawful law may still belong to the legal system, which is the very problem addressed here.
51 Hans Kelsen, General Theory of Norms, 125 (taken verbatim from Hans Kelsen, Derogation, 272).
52 Mathieu Carpentier, Sources and validity, in: Legal Validity and Soft Law, ed. Pauline Westerman / Jaap
Hage / Stephan Kirste / Anne-Ruth Mackor, 2018, 84 ff. In the same vein, see Michael Giudice, Understand-
ing the Nature of Law, 2015, 113 ff., where Giudice argues for what he calls a “contingent relation between
invalidity and unconstitutionality”.

Franz Steiner Verlag


144 MATHIEU CARPENTIER

will not rehearse my arguments here. Suffice it to say that such a theory must somehow
assume that a norm’s membership within a legal system is to be divorced from its con-
formity with higher-ranking norms53. Under such an assumption, a norm may belong
to a legal system although it conflicts with higher-ranking norms. In this outlook, high-
er-ranking norms are power-conferring norms – what Hart calls “rules of change”54.
They regulate the way legal norms ought to be posited – or derogated – and to some
extent their content. Whenever a norm conflicts with a higher-ranking norm, it does
not mean that it does not belong to the legal system; it just means that the law-maker
acted ultra vires. So, we do not need to abandon the idea of a hierarchical structure of
the legal system; what we need to do is to divorce this legal structure from the notion
of a dynamic production of legal validity – insofar as validity is understood as mere
membership within a given legal system55.
One way to achieve such a theory is to use the notion of a rule of recognition, con-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

ceived as a social fact. The rule of recognition picks out the legal system’s criteria of va-
lidity. In a positivist outlook, sources are such criteria. A norm is valid, i. e. it belongs to
the legal system, if and only if it can be traced to a source of law, that is to a complex set
of social facts. Whether it conflicts with higher-ranking norms is prima facie irrelevant
as long as the basic facts for there to be a legal norm are present. Of course, such facts
overlap to some extent with the “facts” that make up the procedure through which the
norm ought to be produced according to higher ranking norms. When the facts in the
overlap are missing, not only is the norm contrary to a higher-ranking norm, but there
may be doubts whether the “norm” is a legal norm at all. For instance, if the French
President enacts a “statute” which has not been passed in either House of Parliament,
I am not sure such a statute would be at all treated as a legal norm. However, this ex-
ample shows that two different questions are at stake. First, there is a demarcation
problem: is this “thing” a law, i. e. a legal norm, or is it a non-law? If I sign a piece of
paper where I “enact” a “law”, no one will dispute that this is not a law: the reason why
it is not a law is not that I have violated some constitutional requirement or any other
power-conferring norm; it is because the basic facts for there to be a law in the first
place are missing. We could argue that the same goes in our example, even though the
President is a legal authority. The second question is whether the president has acted
ultra vires, i. e. whether he has been violating a higher-ranking norm56. These are two
distinct questions.

53 A careful (albeit different from mine) analysis of the distinction between conformity and validity can
be found in: Régis Ponsard, Validité et conformité juridiques, in: Un classique méconnu: Hans Kelsen, ed.
Thomas Hochmann / Xavier Magnon / Regis Ponsard, 2019.
54 HLA Hart, The Concept of Law (3rd ed.), 2012, 95 ff.
55 Of course this is debatable, and it is clear that Kelsen’s is a much more robust notion of validity. But as I
made clear in Section II above, I am working with a weak notion of validity.
56 One objection could be that in our example, there is not a conflict in the sense elaborated in Section I
above, since the conflict is not about the content of the norms at stake. This is correct; however I have hint-

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 145

There are many other ways through which such a result may be achieved without
divorcing membership from conformity. You can claim that validity is a compound of
membership and conformity – that is you can save to some extent Kelsen’s dynamic
theory of law – and, in the same time, accept that norms may be valid even though
they conflict with higher-ranking norms. In other words, you do not have to agree with
the previous paragraph in order to accept that genuine conflicts between equally valid,
different ranking norms exist. For instance, some authors distinguish existence from
validity57, in which existence is understood as some kind of presumptive or prima facie
validity.
The upshot is that any theory that allows for genuine normative conflicts between
lower-and higher-ranking norms serves better the Kelsenian motto that “laws are not
null but only annullable” than Kelsen’s own theory of alternative authorizations. Kels-
en is right to argue that before it is repealed by a constitutional court, an unconstitu-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

tional statute is perfectly valid within the legal system. As we framed it in section III,
Kelsen is right to claim that Lex superior is not a pro futuro derogating norm (even if he
erroneously thought such was the case with Lex posterior) However, as we shall now
see, he is wrong to argue that the conflict must be solved through derogation.

C. Lex Superior as A Norm of Applicability

We saw in Section III that metarules are not power-conferring norms regulating dero-
gation. The same goes for Lex superior. Indeed, the reason why Kelsen was so reluctant
to mention at all Lex superior as a metarule is not only that Kelsen refused to treat con-
flicts across the normative hierarchy as genuine conflicts; it is also that Kelsen thought
that metarules (such as Lex posterior) function as derogating norms. And since there
is no nullity, but only annullability, Lex superior cannot be such a derogating norm.
However, Kelsen’s theory of alternative authorizations and of the role of constitutional
courts may be reframed as a specific theory of Lex superior, understood as a pow-
er-conferring norm directing law-applying organs (or a subset thereof) to derogate
an inferior norm whenever it conflicts with a higher-ranking norm. Kelsen seems to
presuppose that the only way to solve the “conflict” (which according to him is only

ed that the classification sketched out in Section I was incomplete (it excludes what Lindahl and Reidhav
call capacitative conflicts). However, I needed such an example because the overlapping facts I focused on
are generally facts about procedure and not facts about content. The fact that a norm has such-and-such
content has nothing to do with the source of that norm’s legal character (i. e. with its membership). How-
ever, the fact that a norm was passed in Parliament and enacted by the head of state is generally part of the
reason why it is a legal, rather than a non-legal, norm.
57 In the recent literature, see Paolo Sandro, Unlocking Legal Validity: Some Remarks on the Artificial
Ontology of Law, 114 ff.; see also Matthew Grellette, Legal Positivism and the Separation of Existence and
Validity, Ratio Juris 23 (2010), 22 ff.

Franz Steiner Verlag


146 MATHIEU CARPENTIER

apparent) is through derogation, i. e. repeal. But, as Kelsen was well aware58, it need not
be so. Kelsen was right to insist that before the court rules, the statute is still valid. But
there are many cases in which the statute remains valid after the court’s decision. Many
courts (be they constitutional courts or supreme courts) are not empowered to repeal
unconstitutional statutes. They are only empowered to disapply them or to enjoin low-
er courts from applying them59. This mundane observation points to another solution,
which is predicated on the general tenets outlined in Section IV.B above. Once we
understand that Lex superior is not about membership at all and that the reason why
norms belong to a legal system is not because they are created within the scope of the
powers conferred upon their creators, it makes it easier to treat Lex superior as a mere
norm of applicability. In that respect, Lex superior is a metarule like any other.
Such is actually the crux of Chief Justice John Marshall’s reasoning in Marbury v
Madison (and Alexander Hamilton’s reasoning in the Federalist 78)60 Interestingly,
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

when discussing unconstitutional statutes, Marshall does not use the common phrase
“null and void”; he only says that such statutes are “void”. Moreover, Marshall insists
that judicial review is part of the ordinary province of the judiciary, because an un-
constitutional statute does not differ from ordinary normative conflicts. “If two laws
conflict with each other, the Courts must decide on the operation of each. So, if a law
be in opposition to the Constitution, if both the law and the Constitution apply to a
particular case, so that the Court must either decide that case conformably to the law,
disregarding the Constitution, or conformably to the Constitution, disregarding the
law, the Court must determine which of these conflicting rules governs the case. This
is of the very essence of judicial duty”61. Judges have always been faced with normative
conflicts and they solve them by disapplying the defeating norm. The same goes for
unconstitutional statutes, and for unlawful laws in general.
Of course, in many legal systems, judges are specifically barred from applying Lex
superior, at least in what regards unconstitutional statutes. Some legal systems have
not adopted judicial review, be it American- or European-style. This shows that Lex
superior as a norm of applicability is not part of their positive law, or that, in what
regards legislation, Lex posterior defeats Lex superior. In other legal systems, further
norms empower law-applying organs to repeal (i. e. derogate) the inferior norm, but,
as we saw at the end of Section III, it is a purely contingent matter. Sometimes law-ap-
plying organs (or a subset of them, e. g. Constitutional courts) are further empowered
to repeal unconstitutional statutes: such is the case in France, where the Constitutional

58 See Hans Kelsen, Judicial Review of Legislation: A Comparative Study of the Austrian and the Ameri-
can Constitution, The Journal of Politics 4 (1942), 184 ff.
59 Mathieu Carpentier, Validity versus Applicability: a (Small) Dose of Scepticism, 115 ff.
60 Marbury v Madison, 5 U. S. 137 (1803). I am aware of the political overtones of Marshall’s opinion, which
must be taken with a grain of salt. I should be clear that I do not use Marbury to prove my point, but rather
to illustrate it by way of a very famous example.
61 Marbury v Madison, 5 U. S. 137, 177–178.

Franz Steiner Verlag


Kelsen on Derogation and Normative Conflicts 147

council is empowered to repeal an unconstitutional statute. But remember that Euro-


pean-style constitutional courts are specialised organs. They are law-applying organs
insofar as they apply the Constitution, but they are not part (not even at the top) of the
ordinary judiciary, and they are never called upon to apply statutes62. Hence, they are
never really faced with genuine normative conflicts, insofar as they are not bound to
apply one of the two conflicting norms in the first place. Only when a court must apply
both the constitution and the unconstitutional statute is there a genuine normative
conflict. When it is the case, the court will apply Lex superior and disapply the statute.
Whether the court can also repeal the statute is a different question entirely.

Conclusion
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

In this paper I have tried to vindicate some intuitions which are central to Kelsen’s
last, sceptical move. In the same time, I have aimed to identify and solve some flaws
in Kelsen’s account of normative conflicts and derogation by showing that it was not
a radical enough move. Kelsen is right to claim that metarules (such as Lex posterior)
are not logical necessities. They are part of a legal system’s positive law: even though
they are not necessarily formally enacted, some metarules at least exist under a cus-
tomary form.
Kelsen is also right to argue that derogation is a specific normative function. But he
is wrong to assume that derogation is a function of metarules. The link between nor-
mative conflict resolution and derogation is weaker than Kelsen is willing to accept.
This is shown by the fact that metarules are most and foremost norms of applicability.
This goes for Lex posterior, which is the only metarule actually discussed by Kelsen.
But as I have shown, the same goes for Lex Superior. Deflating it as a logical principle
gives room for a much more flexible view of hierarchical normative conflicts.
It could be objected that the unified conception of metarules which has been ad-
vocated here is mistaken. Metarules, the objection goes, serve each a different func-
tion. Such is the view espoused by Riccardo Guastini63 and Giovanni Battista Ratti64.
They claim that Lex superior is a criterion of invalidity ex tunc, whereas Lex posterior
is a criterion of derogation ex nunc, and Lex specialis is a criterion of priority, in the
sense that it derogates some logical consequences of the more general norm. It bears
stressing out that this picture still rests on a strong link between normative conflict

62 This is over-simplistic. I am mainly talking here about constitutional review; constitutional courts gen-
erally have other functions (such as adjudicating electoral disputes), where they will have to apply statutes
as well as other sources of law.
63 Riccardo Guastini, Interpretare e argomentare, 2011, 113 ff.
64 Giovanni Battista Ratti, Normative Inconsistency and Logical Theories: A First Critique of Defeasibi-
lism, in: Coherence: Insights from Philosophy, Jurisprudence and Artificial Intelligence, ed. Michał Araszkiew-
icz / Jaromír Šavelka, 2013, 133; Giovanni Battista Ratti, Negation in Legislation, 147

Franz Steiner Verlag


148 MATHIEU CARPENTIER

resolution and derogation. As such, there are many problems with it. For instance, the
idea that Lex superior is a criterion of invalidity ex tunc can be understood two ways.
On a “strong” reading, it means that in a hierarchical conflict, the defeated norm has
never been valid in the first place: I refer the reader to Section IV.A and B above for my
arguments against this idea. On a weak reading, it means that when courts invalidate
an unconstitutional statute, such an invalidation has retroactive effect. But this claim is
wrong as an empirical matter of fact. Moreover, I could argue that a unified conception
of metarules as rules of applicability fits better our intuitions about how legal systems
function and law-applying organs solve normative conflicts. It rests on a simple dis-
tinction between membership and applicability and does not need any further distinc-
tion or superfluous elaboration. As such it serves better than its rivals the requirement
of Denkökonomie65 which is central to Kelsen’s jurisprudential method.
One last thought. The reader may ask: what is the philosophical relevance of all this?
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Why analyse normative conflicts in the first place? Pierluigi Chiassoni once quipped
that there are two traditions in modern jurisprudence: one the one side, there are the
watch makers “who deal with a clumsy conceptual machinery laid down by tradition
and embodied in lawyers’ common sense”; on the other side, there are the philoso-
phers who address “the real, big, theoretical (and practical) issues at stake”66. The pres-
ent paper is definitely an essay in watch-making. However normative conflicts have
philosophical relevance as well. They are ubiquitous in practical reasoning, and moral
philosophy has been debating for centuries the question whether there are genuine
moral normative conflicts. Law and legal reasoning are the product of human reason;
the way they deal with normative conflicts is instrumental for any study of law’s inner
(ir)rationality.

Mathieu Carpentier
Université Toulouse 1 Capitole, 2, rue du Doyen-Gabriel-Marty, 31042 Toulouse cedex 9,
FRANCE, mathieu.carpentier@ut-capitole.fr

65 Kelsen borrowed this methodological principle from Mach via Pitamic. Hans Kelsen, Hauptprobleme
der Staatsrechtslehre entwickelt aus der Lehre vom Rechtssatze, 1922, xv
66 Pierluigi Chiassoni, A Tale from Two Traditions: Civil Law, Common Law, and Legal Gaps in Analisi e
diritto 2006, ed. Paolo Comanducci and Riccardo Guastini, 2007, 49. Chiassoni uses the distinction between
the watchmaker and the philosopher in the course of an analysis of the literature on legal gaps, but it can
be generalized to cover the topic of normative conflicts as well. However, Chiassoni identifies these two
traditions with the civil law / common law divide. This seems to me to be over-simplistic. Joseph Raz for
instance is certainly both a watchmaker (see for instance his Concept of a Legal System) and a philosopher
in that respect; so is Robert Alexy.

Franz Steiner Verlag


Suspension als eine Art von Derogation
oder als weitere Funktion der Rechtsnorm?

MATHEUS PELEGRINO DA SILVA


Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Suspension as a Kind of Derogation or as Another Function of the Legal Norm?

Abstract: According to Robert Walter the suspension should be understood as a kind of


derogation and not as a different normative function. In the article it is highlighted that
the consequences of the suspension of a norm are different from the consequences of the
derogation of a norm, since a derogated norm must be considered as completely invalid,
while the suspended norm may eventually continue to be valid. In order to support the
conception of the suspension as a normative function and not as a kind of derogation one
of Merkl’s ideas is employed, according to which the derogation of a norm requires an
empowerment for such a legal act, so that it can be analogically argued that the possibility
of suspending a norm presupposes an empowerment to suspend norms.

Keywords: Derogation, Hans Kelsen, Invalidation, Suspension, Theory of the hierarchical


structure, Robert Walter

Schlagworte: Derogation, Hans Kelsen, Invalidation, Suspension, Stufenbaulehre, Robert


Walter

I. Einleitung

Nach Adolf Julius Merkl sind Rechtsnormen in der Regel unveränderlich, und ihre
Veränderlichkeit muss positiv bestimmt werden, d. h. jede Änderung ist von einer Er-
mächtigung zur Derogation abhängig.1 Das Ziel dieses Aufsatzes ist, Robert Walters

1 „Weyr sieht den Verfassungssatz, der die Unabänderlichkeit postuliert, für abänderbar an – sein logisches
Prinzip kann ihm selbst durch einen Rechtssatz für den bestimmten Fall nicht entkräftet werden –, ich
postuliere dagegen, um zur Erkenntnis der Abänderlichkeit zu gelangen, angesichts der über ihre Abän-
derbarkeit schweigenden Verfassung einen Rechtssatz, der Abänderbarkeit statuiert; denn die Unabänder-

Franz Steiner Verlag


150 MATHEUS PELEGRINO DA SILVA

Auffassung von der Suspension als einer besonderen Art von Derogation anhand von
drei Fragen darzustellen und zu zeigen, dass anders als von Walter behauptet die Sus-
pension als eine weitere Funktion der Rechtsnormen zu identifizieren ist. Um Klarheit
über die Natur der Suspension zu gewinnen, wird auch erläutert, aus welchen Grün-
den die Invalidation, im Gegensatz zu Suspension und Derogation, keine Normfunk-
tion ist. Schließlich wird die Relevanz der Unterscheidung zwischen Derogation und
Suspension für die Diskussion über die Möglichkeit eines Stufenbaues nach der de-
rogatorischen Kraft aufgezeigt.

II. Die Suspension als eine eigene Normfunktion

In „Der Aufbau der Rechtsordnung“ hat Robert Walter die Suspension auf folgende
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Weise beschrieben:
Eine gewisse […] derogatorische Kraft käme den Anordnungen in einer bestimmten
Rechtsform zu, wenn sie nach den Bestimmungen des positiven Rechts zwar die Geltung
einer Anordnung beeinträchtigen, diese jedoch nicht vollständig vernichten können, da
sie wegen dieser eintretenden Derogationswirkung fehlerhaft und deshalb zu beseitigen
sind, wenn durch die Beseitigung die in ihrer Geltung beschränkte Anordnung wieder un-
eingeschränkte Rechtswirkung erlangt.2

Walter präsentiert die Suspension als eine Wirkung der Erzeugung einer Norm, die
angesichts der direkten Ermächtigung nicht hätte erfolgen dürfen. Das heißt: Wäre
der Erzeuger der Norm ermächtigt, diese Norm zu erzeugen, hätte er eine derogatori-
sche Norm erzeugt. Weil er dazu nicht ermächtigt ist, es ihm wegen einer alternativen
Ermächtigung aber rechtlich möglich ist, diese Norm zu erzeugen, wird die von ihm
erzeugte Norm eine andere Norm suspendieren.
Im Folgenden werden nun einige Aspekte von Walters Auffassung von der Suspen-
sion mittels dreier Fragen behandelt, um deren Unvollkommenheit aufzuzeigen.

1. Frage: Ist die Suspension eine notwendige Wirkung der gültigen Erzeugung einer
Norm, die das Ziel hat, eine andere Norm zu derogieren, aber nach den positivrecht-
lichen Bestimmungen keine derogatorische Kraft in Bezug auf diese andere Norm hat,

lichkeit ist mir normlogisches Prinzip. Ist mir aber die Abänderung einer über ihre Abänderungsmöglich-
keit schweigenden Verfassung eine Denkunmöglichkeit, so wird mir bei verfassungsmäßiger Statuierung
der Abänderbarkeit (der Verfassung und gleich ihr anderer Gesetze) die Unabänderbarkeit im einzelnen
Fall zur Normwidrigkeit, welche, um diesen Charakter zu verlieren, die Formen der Verfassungsänderung
annehmen muß.“ Adolf Julius Merkl, Die Unveränderlichkeit von Gesetzen – ein normlogisches Prinzip,
Gesammelte Schriften (hg. von Dorothea Mayer-Maly / Herbert Schambeck / Wolf-Dietrich Grussmann),
Bd. I/1, 1993, 165.
2 Robert Walter, Der Aufbau der Rechtsordnung, 1964, 58.

Franz Steiner Verlag


Suspension als eine Art von Derogation oder als weitere Funktion der Rechtsnorm? 151

sodass die Autorität, die diese Norm erzeugt, keine Ermächtigung erhalten hat, diese
Norm zu derogieren?
Zwei der Probleme bezüglich Walters Auffassung der Suspension sind einmal die
Idee, dass die suspendierende Norm irgendwann notwendigerweise als rechtswidrig an-
erkannt werden wird,3 und zum anderen die Ansicht, dass die suspendierende Norm
die Geltung der schon gültigen Norm notwendigerweise beeinflusst. In einem konkre-
ten Fall kann man von „Suspension“ sprechen, wenn zwei Bedingungen erfüllt sind:
1. Eine mit derogierender Absicht erzeugte Norm4 wird definitiv beseitigt; in dieser
Situation kann man behaupten, dass die vermeintlich derogierte Norm als gültig be-
trachtet werden kann. 2. Die Geltung dieser zweiten Norm wird für den Zeitraum zwi-
schen der Erzeugung der mit derogierender Absicht erzeugten Norm und der Erklä-
rung, dass diese Norm ungültig ist, als provisorisch aufgehoben betrachtet. Diesfalls
kann auch eine generelle positivrechtliche Bestimmung mittels über die Alternativer-
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

mächtigung erzeugte Normen existieren,5 nach der die in dieser Situation zu derogie-
rende Norm als provisorisch ungültig betrachtet werden soll.6
Die Suspension als Normfunktion muss zuerst (mittels genereller oder individu-
eller Normen) positiviert werden, bevor es möglich ist, sie in einer Entscheidung als
konkretisiert betrachten zu können. Wenn das nicht passiert, dann ist die obige Frage
zu verneinen, und beide Normen sind gültig und stehen (möglicherweise)7 miteinan-
der in Konflikt.8 Dies entspricht Merkls Auffassung, dass die Derogation einer Norm

3 Anders ist der Fall bei Lippold, der die Möglichkeit der Nicht-Anerkennung der rechtswidrigen Norm
problematisiert. Vgl. Rainer Lippold, Recht und Ordnung Statik und Dynamik der Rechtsordnung, 2000,
417.
4 Diese Absicht kann auf zwei unterschiedliche Arten und Weisen erfolgen: Es kann sein, dass die zweite
Norm nur das Ziel hat, eine andere Norm zu derogieren, oder dass die zweite Norm denselben Tatbestand
der anderen Norm mit einer von dieser Norm abweichenden Folge verbindet, sodass sie mittelbar die erste
Norm derogiert.
5 Bezüglich der österreichischen Rechtsordnung ist das der Fall. Vgl. Robert Walter, Der Stufenbau nach
der derogatorischen Kraft im österreichischen Recht, Österreichische Juristenzeitung (1965), 171.
6 Auch wenn keine generelle positivrechtliche Suspensionsbestimmung existiert, kann die in einer Ent-
scheidung angewendete Suspension trotzdem dank Alternativermächtigung als gültig betrachtet werden.
Zur Bedeutung der alternativen Ermächtigung, siehe Hans Kelsen, Reine Rechtslehre, 2. Aufl., 2017, 472–
487.
7 Sie sind im Konflikt miteinander, wenn eine Norm etwas vorschreibt und die andere Norm etwas ande-
res vorschreibt, aber nicht wenn die zweite Norm das einzige Ziel hat, die erste Norm zu derogieren. Vgl.
Hans Kelsen, Allgemeine Theorie der Normen, 1979, 86.
8 Hier ist zu berücksichtigen, dass Walter möglicherweise Folgendes angenommen hat: Weil es in einigen
Fällen der Normerzeugung mittels Alternativermächtigung passieren kann, dass aufgrund der Erzeugung
dieser Norm ein Normenkonflikt entsteht, und solange vorausgesetzt war (sowie Walter früher vertreten
hat), dass kein Normenkonflikt entstehen kann, dann ist anzuerkennen, dass es für das Vermeiden von Nor-
menkonflikten notwendig ist, auf irgendeine Weise – beispielsweise durch Postulierung eines suspensiven
Effektes – diese Normenkonflikte zu vermeiden. Es ist nicht möglich zu bestimmen, aus welchem Grund
Walter trotz des (durch Merkls Auffassung inspirierten) Einwands, dass die Ausübung einer Funktion von
der Ermächtigung für die Ausübung dieser Funktion abhängig ist, die Suspension und den suspensiven Ef-
fekt als notwendige Eigenschaft einer Rechtsordnung identifiziert hat. Eine mögliche Erklärung ergibt sich

Franz Steiner Verlag


152 MATHEUS PELEGRINO DA SILVA

von einer Ermächtigung zur Derogation abhängt. Anhand einer analogen Betrachtung
kann gesagt werden, dass eine Norm als suspendiert betrachtet werden kann, wenn
eine Autorität in der einen oder anderen Form die Ermächtigung bekommen hat, die-
se Norm zu suspendieren.
Einerseits könnte die Suspension in einer Rechtsordnung als Normfunktion vor-
bestimmt werden, sodass in solchen Fällen suspensive Normen existieren würden, die
nicht wegen einer „fehlerhaften“ Normerzeugung zustande gekommen sind, sondern
dank einer normgemäßen Normerzeugung – d. h. die Rechtsordnung würde bestim-
men, dass durch Erfüllung bestimmter Bedingungen eine Autorität ermächtigt (direk-
te Ermächtigung) war, suspensive Normen zu erzeugen. Ist das nicht der Fall, dann
kann die Suspension in individuellen Rechtsnormen konkretisiert werden und das
auch, wenn die Autoritäten, die diese Normen erzeugen, keine generelle Ermächti-
gung bekommen haben, um Normen als suspendiert betrachten zu können. Existiert
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

keine generelle positivrechtliche Bestimmung über die Suspension (keine Ermächti-


gung zur Normsuspension), dann kann (muss aber nicht) die in einer Entscheidung
angewendete Suspension dank einer Alternativermächtigung als gültig betrachtet wer-
den.

2. Frage: Kann die Suspension als eine Art Derogation betrachtet werden, oder ist sie
(neben den von Hans Kelsen identifizierten Normfunktionen)9 eine weitere Funktion
der Rechtsnorm?
Walter präsentiert uns die Suspension als eine Art von Derogation, die eine un-
vollständige „Vernichtung“10 einer Norm bewirkt.11 Aber abgesehen von einem gleich
zu behandelnden Aspekt, kann die Suspension nicht als eine Art von Derogation be-
trachtet werden, da ein wesentliches Merkmal der Derogation die Tatsache ist, dass
die derogierte Norm nicht mehr gültig werden kann, da sie (vollkommen) vernichtet

aus Walters Auffassung hinsichtlich der Existenz von Normkonflikten, denn wie Kammerhofer bemerkt,
habe Walter erst ab 1966 anerkannt, dass es Normenkonflikte geben könne: „Entgegen Walters Aussagen im
1955 erschienenen ‚Widerspruch‘ [gemeint ist Robert Walters Schrift „Über den Widerspruch von Rechts-
vorschriften“], wo die von Kelsen definierte Situation wohl in die Kategorie der faktischen Unvereinbarkeit
der Rechtsfolgen gefallen wäre, liegt der Normkonflikt jetzt [in Walters Artikel „Recht und Logik“ aus dem
Jahr 1966] in der ‚Situation, wonach a (eine Rechtsfolge) nach der einen Norm sein soll, wogegen nach der
anderen Norm diese Rechtsfolge nicht – somit: non a – sein soll.‘“. Jörg Kammerhofer, Robert Walter, die
Normkonflikte und der zweite Stufenbau des Rechts, in: Gedenkschrift Robert Walter, hg. von Clemens
Jabloner u. a., 2013, 243.
9 Diese sind: Ermächtigen, Erlauben, Gebieten / Verbieten sowie Derogieren. Vgl. Kelsen (Fn. 7), 3.
10 Walter (Fn. 2), 57.
11 Früher hatte Walter die Ansicht vertreten, dass beide Normen als gültig zu betrachten wären, aber auch
danach hat er die Situation nicht als Konflikt betrachtet, da für ihn eine der Normen zurückgedrängt war.
Da in diesem spezifischen Fall Walters Schrift die Problematik in Bezug auf die österreichische Rechts-
ordnung betrachtet, sind seine Bemerkungen im Hinblick auf diese Rechtsordnung begründet. Vgl. Robert
Walter, Können Verordnungen Gesetzen derogieren?, Österreichische Juristenzeitung (1961), 3 und 7.

Franz Steiner Verlag


Suspension als eine Art von Derogation oder als weitere Funktion der Rechtsnorm? 153

ist.12 Es kann von Teilvernichtung (und Teilderogation) gesprochen werden, da eine


derogatorische Norm ein Gesetz (als Sammlung von Normen)13 zum Teil derogieren
kann, die Idee einer provisorischen Vernichtung aber ist als contradictio in adiecto zu
identifizieren.14
Der Wesensunterschied zwischen Derogation und Suspension zeigt sich in folgen-
der Überlegung: Ist die Beeinträchtigung der Geltung einer Norm das Kriterium für
die Klassifizierung, dann können Derogation und Suspension nebeneinandergestellt
werden. Auch unter diesem Kriterium kann die Suspension nicht als eine Art Dero-
gation betrachtet werden. Ist die Suspension in der einen oder anderen Weise in der
Rechtsordnung positiviert, dann hebt die suspensive Norm die Geltung von Rechts-
normen provisorisch auf. Diese Funktion kann die Derogation nicht ausüben, denn es
existiert keine provisorische Derogation einer Norm. Deshalb ist die Suspension als
eine weitere Funktion der Rechtsnormen und nicht als eine Art von Derogation zu
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

identifizieren.
Die Suspension wie die Derogation können aber als zwei Arten von „Beseitigung“15
oder als zwei Arten von Geltungsvernichtung von Normen einer Rechtsordnung be-
trachtet werden. Wegen der Erzeugung einer suspensiven Norm ist die Geltung einer
anderen Norm in einer änderbaren Weise beeinträchtigt, sodass, solange die suspensi-
ve Norm gültig bleibt, die suspendierte Norm als eine ungültige Norm zu betrachten
ist. Sie ist also mit Vorbehalt aus dem Bereich der gültigen Normen ausgeschlossen.
Wegen der Erzeugung einer derogatorischen Norm ist die Geltung einer anderen
Norm in einer unveränderbaren Weise beeinträchtigt mit der Folge, dass die derogier-
te Norm immer und ohne Vorbehalt aus dem Bereich der gültigen Normen ausge-
schlossen ist.
Die Suspension unterscheidet sich von der Derogation, denn jene ermöglicht
die Beschreibung einer Eigenschaft der gültigen Normen einer Rechtsordnung, die
durch die Anwendung des Begriffes „Derogation“ nicht erfolgen kann. Entweder sind
Normen derogiert und deswegen für immer als ungültig zu betrachten, oder sie sind

12 „The attempt to repeal the validity of a norm which has derogated the validity of another norm in re-
gard to this norm, by means of a derogating norm, would be without effect. The norm whose validity was
terminated by the first derogating norm would not regain its validity by the second derogating norm.“ Hans
Kelsen, Derogation, in: Die Wiener rechtstheoretische Schule, hg. von Hans R. Klecatsky / René Mar-
cic / Herbert Schambeck, Bd. 2, 2010, 1170.
13 Walter benennt die nach seiner Auffassung existierenden zwei Arten von Derogation folgendermaßen:
“Derogation als vollständige Vernichtung“ und „Derogation als beschränkte Beseitigung“. Walter (Fn. 2),
57 f.
14 In einer früheren Schrift hat Walter behauptet: „Anders als im Bereich der Natur kann die Vernichtung
wieder vernichtet, das Gesetz rückwirkend wieder zum Leben erweckt werden.” Walter (Fn. 11), 3, Fn. 28.
Walters Argumentation kann nicht gefolgt werden, da keine überzeugenden Gründe angeführt werden,
warum man eine neue Bedeutung des Wortes „vernichten“ schaffen sollte, nur um die Suspension als eine
Art Derogation klassifizieren zu können.
15 Walter (Fn. 2), 58.

Franz Steiner Verlag


154 MATHEUS PELEGRINO DA SILVA

nicht derogiert und deswegen als gültig zu betrachten. Die Eigenschaft einer gültigen
Norm, dass sie unter bestimmten Bedingungen als ungültig betrachtet werden soll,
diese schwebende Natur der Geltung dieser Norm, ist eine Besonderheit, die durch
die Anwendung des Begriffes „Derogation“ nicht ausgedrückt werden kann. Aus die-
sem Grund ist die Suspension eine Funktion der Normen, die nicht als eine Art von
Derogation betrachtet werden kann, denn der Begriff „Suspension“ drückt eine Situ-
ation aus, die durch die Anwendung des Begriffes „Derogation“ nicht ausdrückbar ist.

3.  Frage: Sind angesichts der Beschreibung selbständiger Normen (für die Rechts-
ordnung) die Wirkungen der Erzeugung einer suspensiven Norm dieselben wie die
Wirkungen der Erzeugung einer derogatorischen Norm?
Ausgehend von Kelsens Definition einer unselbständigen Norm ist es möglich, die
selbständige Norm zu definieren. Nach Kelsen ist eine Norm unselbständig, denn „sie
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

bestimmt nur […] die Bedingung, an die die zweite [Norm] die Sanktion knüpft“.16
Die Norm ist als selbständige Norm zu verstehen, die die Bedingungen einer Sanktion
oder eines Zwangsaktes und die bedingte Sanktion oder den bedingten Zwangsakt
miteinander verknüpft.
Ist die Suspension in einer Rechtsordnung für den Fall der Rechtserzeugung dank
der Alternativermächtigung tatsächlich positiviert, kann gefragt werden, ob eine sus-
pensive Norm mit einem bestimmten Inhalt genau wie eine derogatorische Norm
mit demselben Inhalt eine selbständige Norm ändern wird. Sind in diesem Zusam-
menhang die Wirkungen der Erzeugung einer suspensiven Norm dieselben wie die
Wirkungen der Erzeugung einer derogatorischen Norm, dann kann die Suspension in
dieser Hinsicht dennoch als eine Art von Derogation betrachtet werden, d. h. als eine
Art von Derogation, die eine provisorische Wirkung hat.
Diese Situation involviert zwei mögliche Antworten, von denen eine Walters Auf-
fassung der Suspension als eine Art von Derogation in einem relativen und qualifizier-
ten Sinn bestätigen wird. Ist beispielsweise wegen der Erzeugung einer Erlaubnisnorm
eine Gebotsnorm suspendiert, dann wird eine selbständige Norm, die früher diese
Gebotsnorm umfasste, jetzt und in einem ganz bestimmten Sinne einen anderen In-
halt haben. Dank der Erzeugung dieser suspensiven Norm wird die Gebotsnorm jetzt,
in diesem Moment, nicht gültig sein. In diesem Sinne und nach einer Perspektive, die
nur einen bestimmten Zeitpunkt bei der Beschreibung der selbständigen Normen be-
rücksichtigt, wird eine suspensive Norm eine selbständige Norm so ändern, wie eine
derogatorische Norm mit demselben Inhalt diese selbständige Norm ändern würde.
Unter diesem Aspekt ist es in der Tat so, dass die Suspension als eine Art von Deroga-
tion betrachtet werden kann, sodass angesichts eines (vordefinierten) Zeitpunkts über
die „Beseitigung“ einer Norm gesprochen werden kann.

16 Kelsen (Fn. 6), 112.

Franz Steiner Verlag


Suspension als eine Art von Derogation oder als weitere Funktion der Rechtsnorm? 155

Es gibt aber auch einen anderen Aspekt, der für den Fall der analysierten selbstän-
digen Norm zu beachten ist. Hat beispielsweise eine Erlaubnisnorm eine Gebotsnorm
derogiert, dann wird die selbständige Norm diesen neuen Inhalt ohne Vorbehalt aus-
drücken, sodass von nun an ein bestimmter Tatbestand erlaubt ist. Hat anderseits
eine Erlaubnisnorm eine Gebotsnorm suspendiert, dann wird die selbständige Norm
diesen neuen Inhalt mit einem Vorbehalt ausdrücken, sodass von nun an und solange
die suspensive Norm gültig ist, ein bestimmter Tatbestand erlaubt ist. Somit kann die
Suspension nicht als eine Art von Derogation betrachtet werden, denn sofern die Be-
schreibung der selbständigen Norm nicht nur die aktuellen Effekte der Erzeugung
einer Norm beachtet und ausdrückt, sondern auch mögliche zukünftige Wirkungen,
wird die Suspension einer Norm etwas anderes bedeuten als die Derogation dieser
Norm. Weil die Derogation demnach mit dem zeitlichen Element „nie mehr“ ver-
bunden ist, während die Suspension ein anderes zeitliches Element ausdrückt, das
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

Element der Änderbarkeit, der Vorläufigkeit, müssen Derogation und Suspension als
unterschiedliche Normfunktionen betrachtet werden.

III. Gründe für den Ausschluss der Invalidation als weitere Normfunktion

Die Behandlung der Derogation und der Suspension als zwei Arten von Exklusion
von Normen aus dem Bereich der gültigen Normen einer Rechtsordnung macht auch
deutlich, warum die Invalidation nicht als Normfunktion zu verstehen ist.
Walter präsentiert als eine Situation von Invalidation den Fall, in dem „eine (neue-
re) Verfassungsregelung nur zur Erzeugung von Normen bestimmter Art ermächtigt,
jedoch – auf Grund bisheriger Verfassung – Gesetzesnormen bestehen, die nach der
neueren Verfassungsregelung nicht erzeugt werden dürfen“.17 Bei der Invalidation einer
Norm geht es um die Situation, in der eine schon gültige Norm angesichts einer späte-
ren Änderung der Rechtsordnung nicht mehr so erzeugt werden kann, wie sie vorher
erzeugt wurde. Wäre diese schon erzeugte Norm nach der Änderung der Rechtsord-
nung erzeugt worden, könnte sie nicht als eine rechtmäßige Norm betrachtet werden.
Weil diese Norm aber vor dieser Änderung erzeugt wurde, entsteht die Frage, wie die-
se Norm nun betrachtet werden muss, ob sie gültig bleibt oder nicht.
Ohne Bezug auf eine konkrete Rechtsordnung ist es nicht möglich, eine Lösung für
dieses Problem anzubieten, sondern es können nur drei (grundsätzliche) Alternativen
präsentiert werden. Aufgrund der Änderung der Bedingungen für die Erzeugung einer
Norm einer bestimmten Art (mit einer bestimmten Rechtserzeugungsform) kann ent-
weder behauptet werden, dass die frühere Norm (1. Alternative) als eine invalidierte

17 Robert Walter, Derogation oder Invalidation, in: Hundert Jahre Verfassungsgerichtsbarkeit, fünfzig Jah-
re Verfassungsgerichtshof in Österreich, hg. von Felix Ermacora / Hans R. Klecatsky / René Marcic, 1968,
218.

Franz Steiner Verlag


156 MATHEUS PELEGRINO DA SILVA

Norm zu verstehen ist, d. h. als eine gültige, aber normwidrige Norm interpretiert wer-
den muss; oder (2. Alternative) dass die Frage über die normgemäße oder normwid-
rige Natur der früheren (und vielleicht zu invalidierenden) Norm offen bleibt; oder
(3. Alternative) dass die frühere Norm als gültig und normgemäß verstanden wird mit
dem Argument, dass der Normerzeuger, obwohl er von dieser Norm wusste (oder wis-
sen konnte), sie hätte derogieren können, dies aber nicht getan hat.
Unabhängig von der Alternative, die gewählt wird, muss die (vielleicht) invalidier-
te Norm als eine gültige Norm verstanden werden, denn es gibt keinen rechtlichen
Grund, diese Norm als ungültig zu betrachten. Weil die Invalidation einer Norm keine
Änderung in Bezug auf die Beschreibung der gültigen Normen einer Rechtsordnung
verursacht und diese Art von Änderung von derogatorischen und suspensiven Nor-
men in der einen oder anderen Weise provoziert wird (d. h. durch die Behauptung der
provisorischen oder definitiven Beseitigung einer Norm), ist die Invalidation, anders
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

als die Derogation und die Suspension, keine Art von Normfunktion.

IV. Der Unterschied zwischen Derogation und Suspension und die


Möglichkeit eines Stufenbaues nach der derogatorischen Kraft

Walter hat in einer sehr deutlichen Weise dargestellt,18 dass nach Merkls Auffassung
die Verhältnisse zwischen den Normen einer Rechtsordnung nicht unbedingt als eine
Stufenordnung, als ein hierarchischer Stufenbau (der Stufenbau nach der rechtlichen
Bedingtheit) repräsentiert werden sollten, sondern als zwei unterschiedlich konfigu-
rierte Stufenbauten (der Stufenbau nach der rechtlichen Bedingtheit und der Stufen-
bau nach der derogatorischen Kraft).19 Die Behauptung der Existenz eines Stufenbaues
nach der derogatorischen Kraft beinhaltet die Idee, dass wegen positivrechtlicher Be-
stimmungen Normen mit einer bestimmten Rechtserzeugungsform20 Normen einer

18 Walter (Fn. 2), 53–66.


19 Wie Merkl hervorgehoben hat, sind hierarchische Verhältnisse zwischen Rechtserzeugungsformen
(Arten von Normen) nicht eine notwendige Eigenschaft aller Rechtsordnungen, sondern von konkreten
positivrechtlichen Bestimmungen abhängig. „Der rechtliche Stufenbau ist nicht etwas Rechtsimmanen-
tes, das die Rechtssetzung hinnehmen müßte, sondern ist selbst ein willkürliches, verwandlungsfähiges
Produkt der Rechtsordnung.“ Adolf Julius Merkl, Prolegomena einer Theorie des rechtlichen Stufenbaues,
Gesammelte Schriften (hg. von Dorothea Mayer-Maly / Herbert Schambeck / Wolf-Dietrich Grussmann),
Bd. I/1, 1993, 482.
20 Kelsen hat den Begriff „Rechtserzeugungsformen“ verwendet, während bei Merkl der Begriff „Rechts-
satzformen“ zu finden ist. Siehe Kelsen (Fn. 6), 499; Merkl (Fn. 19), 437. Hinter diesen Begriffen steht die
Idee, dass mehrere Arten von Normen erzeugt werden können, die wegen der Bedingungen, die von der
Rechtsordnung gesetzt sind, voneinander hierarchisch unterschieden werden können. Merkl präsentiert
diese Idee mit den folgenden Worten: „Den Hauptanteil an der Differenzierung der Rechtssatzformen hat
indes die Komplikation der Rechtserzeugung durch das je nach den verschiedenen Staats- und Regierungs-
formen wechselnde Zusammenwirken verschiedener Organe, z. B. verschiedener Faktoren der Gesetzge-

Franz Steiner Verlag


Suspension als eine Art von Derogation oder als weitere Funktion der Rechtsnorm? 157

anderen Rechtserzeugungsform derogieren können oder nicht. Sind in einer Rechts-


ordnung Bestimmungen dieser Natur zu finden, dann kann behauptet werden, dass in
dieser Rechtsordnung die Normen von unterschiedlichen Rechtserzeugungsformen
in hierarchischen Verhältnissen miteinander stehen.21
Im Rahmen der Darstellung der Existenz eines Stufenbaues nach derogatorischer
Kraft kann der Fall analysiert werden, in der eine Norm mittels Alternativermächti-
gung erzeugt wurde, die im Konflikt mit einer anderen Norm steht. Es sind grundsätz-
lich zwei Möglichkeiten denkbar.
Erstens: Enthält die Rechtsordnung eine generelle Normierung für diese Fälle von
Normkonflikten und gleichzeitig eine oder mehrere Bestimmungen über die suspen-
sive Kraft von Normen mit unterschiedlichen Rechtserzeugungsformen, dann kann
behauptet werden, dass eine vollständige Darstellung der Rechtsordnung auch die An-
wendung einer Stufung nach der suspensiven Kraft erfordert, mit der Konsequenz,
Lizenziert für Universitätsbibliothek Kiel am 03.04.2020 um 09:54 Uhr

dass es auch einen Stufenbau nach der suspensiven Kraft geben kann. Darüber hinaus
ist zu beachten, dass der Stufenbau nach der suspensiven Kraft eventuell anders darge-
stellt werden muss als der Stufenbau nach der derogatorischen Kraft, sodass beispiels-
weise Verordnungen Gesetze suspendieren können, aber keine derogatorische Kraft
über Gesetze haben.
Zweitens: Enthält die Rechtsordnung keine generelle Normierung für die Fälle von
Normenkonflikten, die das Ergebnis der Ausübung einer alternativen Ermächtigung
konstituieren, und ist auch nicht der Fall, dass durch Gewohnheitsrecht22 eine allge-

bung an der Staatswillensbildung. Je komplizierter die Staatswillensbildung, desto mannigfaltiger die in der
Rechtserzeugung benutzten Rechtssatzformen.“ Merkl (Fn. 19), 438 f.
21 Dazu schreibt Merkl: „Ein Rechtssatz, der gegenüber einem anderen Rechtssatz derogierende Kraft
hat, während dieser andere Rechtssatz ihm gegenüber keine derogierende Kraft hat, ist aus diesem Grunde
von höherem Rang und der derogierbare Rechtssatz im Vergleich mit dem derogierenden Rechtssatz von
niedrigerem Rang. Wenn dagegen – zum Unterschied von den besprochenen Fällen bloß einseitiger De-
rogierbarkeit – zwei Rechtssätze gegenseitig derogierbar sind, so ist dies Erkenntnisgrund ihres gleichen
Ranges. Z. B. ist das Verfassungsgesetz von höherem Rang als das einfache Gesetz, das Grundsatzgesetz von
höherem Rang als das Ausführungsgesetz, das einfache Gesetz wiederum und im besonderen das Ausfüh-
rungsgesetz von höherem Rang als die Vollzugsverordnung, das Gesetz und die Vollzugsverordnung von
höherem Rang als das Gerichtsurteil und der individuelle Verwaltungsakt, wenn und weil der jeweils zuerst
genannte dem jeweils an zweiter Stelle genannten Akte derogieren kann, jedoch nicht umgekehrt. Und ein
Bundes- und Landesgesetz oder ein einfaches Gesetz und eine sogenannte selbständige Verordnung sind
darum als rangsgleich zu erkennen, weil sie einander gegenseitig derogieren können.“ Merkl (Fn. 19), 468 f.
22 Kelsens Auffassung bezüglich des Gewohnheitsrechts ist ambivalent. An einer Stelle der Reine[n] Rechts-
lehre betont er, dass das Gewohnheitsrecht nur als Rechtsquelle anerkannt werden kann, wenn durch eine
oder mehrere positivrechtlichen Normierungen das Gewohnheitsrecht als Rechtsquelle gesetzt wird. „Als
objektiv gültige Norm kann aber der subjektive Sinn der die Gewohnheit konstituierenden Akte nur ge-
deutet werden, wenn die Gewohnheit durch eine höhere Norm als normerzeugender Tatbestand eingesetzt
wird.“ Kelsen (Fn. 6), 34 f. Andererseits ist an einer anderen Stelle der Reine[n] Rechtslehre eine Aussage zu
finden, nach der das Gewohnheitsrecht eventuell auch ohne positivrechtliche Bestimmung als Rechtsquel-
le betrachtet werden darf: „Wenn Gerichte als ermächtigt angesehen werden, auch Gewohnheitsrecht anzu-
wenden, so müssen sie hiezu – ganz so wie zur Anwendung der Gesetze – durch die Verfassung ermächtigt
sein; das heißt: die Verfassung muß die Gewohnheit, die durch das habituelle Verhalten der der staatlichen

Franz Steiner Verlag


158 MATHEUS PELEGRINO DA SILVA

meine Lösung identifiziert werden kann, dann müssen beide in Konflikt stehenden
Normen als gültig betrachtet werden. Doch im Gegensatz zu Lippolds Behauptung23
wird diese Situation keine Folge in Bezug auf die Vertretung der Existenz einer Stufung
nach der derogatorischen Kraft verursachen, denn allein die Existenz von Normen, die
in Konflikt mit anderen Normen stehen und mit dem Ziel erzeugt sind, diese anderen
Normen zu derogieren, reicht noch nicht, um die Stufung nach der derogatorischen
Kraft in Frage zu stellen. Diese Stufung könnte ihre Nützlichkeit nur verlieren, wenn
die mittels Alternativermächtigung erzeugten Normen in der Tat in der Lage wären,
die mit ihnen in Konfl