Proportionality Balancing and Rights Robert

Alexy s Theory of Constitutional Rights 1st

Edition Jan R Sieckmann Ed install download

https://ebookmeta.com/product/proportionality-balancing-and-
rights-robert-alexy-s-theory-of-constitutional-rights-1st-
edition-jan-r-sieckmann-ed/

Download more ebook from https://ebookmeta.com

We believe these products will be a great fit for you. Click
the link to download now, or visit ebookmeta.com
to discover even more!

The Internet And Constitutional Law The Protection Of

Fundamental Rights And Constitutional Adjudication In
Europe 1st Edition Oreste Pollicino

https://ebookmeta.com/product/the-internet-and-constitutional-
law-the-protection-of-fundamental-rights-and-constitutional-
adjudication-in-europe-1st-edition-oreste-pollicino/

Reasons Rights and Values 1st Edition Robert Audi

https://ebookmeta.com/product/reasons-rights-and-values-1st-
edition-robert-audi/

Constitutional Law and Politics, Volume Two: Civil

Rights and Civil Liberties Twelfth Edition O'Brien

https://ebookmeta.com/product/constitutional-law-and-politics-
volume-two-civil-rights-and-civil-liberties-twelfth-edition-
obrien/

Billionaire Brother's Nanny 1st Edition Rebel Bloom

https://ebookmeta.com/product/billionaire-brothers-nanny-1st-
edition-rebel-bloom/
Blindsided Love is Blind Book 1 1st Edition Matilda
Martel

https://ebookmeta.com/product/blindsided-love-is-blind-
book-1-1st-edition-matilda-martel/

Public Images Celebrity Photojournalism and the Making

of the Tabloid Press 1st Edition Ryan Linkof

https://ebookmeta.com/product/public-images-celebrity-
photojournalism-and-the-making-of-the-tabloid-press-1st-edition-
ryan-linkof-2/

Semiotics of religion Signs of the sacred in history

1st Edition Robert A. Yelle

https://ebookmeta.com/product/semiotics-of-religion-signs-of-the-
sacred-in-history-1st-edition-robert-a-yelle/

The Romanesque Abbey of St Peter at Gloucester 1st

Edition Carolyn Heighway Richard Bryant

https://ebookmeta.com/product/the-romanesque-abbey-of-st-peter-
at-gloucester-1st-edition-carolyn-heighway-richard-bryant/

Proceedings of International Conference on Scientific

and Natural Computing Proceedings of SNC 2021 Dipti
Singh Editor Amit K Awasthi Editor Ivan Zelinka Editor
Kusum Deep Editor
https://ebookmeta.com/product/proceedings-of-international-
conference-on-scientific-and-natural-computing-proceedings-of-
snc-2021-dipti-singh-editor-amit-k-awasthi-editor-ivan-zelinka-
The Other Russia Local Experience and Societal Change
1st Edition Leo Granberg Ann Mari Satre

https://ebookmeta.com/product/the-other-russia-local-experience-
and-societal-change-1st-edition-leo-granberg-ann-mari-satre/
Law and Philosophy Library 136

Jan-R. Sieckmann Editor

Proportionality,
Balancing, and
Rights
Robert Alexy's Theory of Constitutional
Rights
Law and Philosophy Library

Volume 136

Series Editors
Francisco J. Laporta, Autonomous University of Madrid, Madrid, Spain
Frederick Schauer, University of Virginia, Charlottesville, VA, USA
Torben Spaak, Stockholm University, Stockholm, Sweden
Editorial Board Members
Aulis Aarnio, Secretary General of the Tampere Club, Tampere, Finland
Humberto Ávila, University of São Paulo, São Paulo, Brazil
Zenon Bankowski, University of Edinburgh, Edinburgh, UK
Paolo Comanducci, University of Genoa, Genova, Italy
Hugh Corder, University of Cape Town, Cape Town, South Africa
David Dyzenhaus, University of Toronto, Toronto, Canada
Ernesto Garzón Valdés, Johannes Gutenberg University, Mainz, Germany
Riccaro Guastini, University of Genoa, Genova, Italy
Ho Hock Lai, National University of Singapore, Singapore, Singapore
John Kleinig, City University of New York, New York City, NY, USA
Claudio Michelon, University of Edinburgh, Edinburgh, UK
Patricia Mindus, Uppsala University, Uppsala, Sweden
Yasutomo Morigiwa, Meiji University, Tokyo, Japan
Giovanni Battista Ratti, University of Genoa, Genova, Italy
Wojchiech Sadurski, University of Sydney, Sydney, Australia
Horacio Spector, University of San Diego, San Diego, USA
Michel Troper, Paris Nanterre University, Nanterre, France
Carl Wellman, Washington University, St. Louis, USA
The Law and Philosophy Library, which has been in existence since 1985, aims to
publish cutting edge works in the philosophy of law, and has a special history of
publishing books that focus on legal reasoning and argumentation, including those
that may involve somewhat formal methodologies. The series has published numer-
ous important books on law and logic, law and artificial intelligence, law and
language, and law and rhetoric. While continuing to stress these areas, the series
has more recently expanded to include books on the intersection between law and the
Continental philosophical tradition, consistent with the traditional openness of the
series to books in the Continental jurisprudential tradition. The series is proud of the
geographic diversity of its authors, and many have come from Latin America, Spain,
Italy, the Netherlands, Germany, and Eastern Europe, as well, more obviously for an
English-language series, from the United Kingdom, the United States, Australia and
Canada.

More information about this series at http://www.springer.com/series/6210

Jan-R. Sieckmann
Editor

Proportionality, Balancing,
and Rights
Robert Alexy’s Theory of Constitutional
Rights
Editor
Jan-R. Sieckmann
Faculty of Law
FAU University Erlangen-Nuremberg
Erlangen, Germany

ISSN 1572-4395 ISSN 2215-0315 (electronic)

Law and Philosophy Library
ISBN 978-3-030-77320-5 ISBN 978-3-030-77321-2 (eBook)
https://doi.org/10.1007/978-3-030-77321-2

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland
AG 2021
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by
similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Switzerland AG.
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

Robert Alexy’s theory of constitutional rights has become a paradigm for the
analysis of rights and their application in constitutional law. On November 13–14,
2019, a conference was organised at the Friedrich-Alexander-University (FAU) in
Erlangen, which one might call a meeting of the European Alexy School with the
Lisbon Legal Theory Group. The conference coincided with the visit of David
Duarte and Jorge Sampaio (both from the University of Lisbon) to Erlangen, funded
by the Visiting Professorship Programme of the FAU.
The aim of the conference was a critical and constructive discussion of some
themes of Alexy’s principles theory of constitutional rights. With few exceptions,
the papers presented at this conference are published in this volume.
The first paper is a contribution of Robert Alexy, which outlines the necessary
link between the theory of constitutional rights and the theory of rational argumen-
tation. Alexy explains this link as provided by the principle of proportionality and, in
particular, his “weight formula”.
The subsequent contributions centre on five topics of Alexy’s theory of consti-
tutional rights: the method and criteria of balancing, the adequacy of the principles
theory in the sphere of fundamental rights, problems of judicial review regarding
balancing, the application of principles theory in the field of social rights, and the
extension of the method of balancing to the decision-making of private agents.
David Duarte explains how proportionality limits balancing discretion and claims
that Alexy’s theory of balancing, his “weight formula” in particular, does not
represent these limits correctly. As an alternative, he presents his own “positivist
weight formula”, which claims to reconcile the method of balancing with legal
positivism.
Jorge Silva Sampaio explores balancing as part of a discourse of reasons and
suggests that a crucial step in this discourse is a second-order balancing regarding the
intensities of interference and certainties attributed to the reasons in conflict. In the
end, he suggests a model of balancing designed as a form of “local particularism”.

v
vi Preface

Carsten Bäcker refers to the debate on an absolute or relative reading of the

guarantee of human dignity, which excludes or admits the balancing of the demands
of human dignity, respectively. He points out some ambivalences in the jurisdiction
of the German Federal Constitutional Court and suggests a theory that combines
absolute and relative elements of the guarantee of human dignity, thus embedding
this guarantee in the theory of principles.
In my contribution, I discuss the deficiency of a simple model of balancing as
optimisation regarding the character of fundamental rights and present a conception
of fundamental rights as rights that are in principle immune to balancing. In this
sense, fundamental rights include prohibitions of balancing, which, however, only
hold in principle and can be overridden.
Martin Borowski analyses problems of judicial review based on balancing. After
reconstructing jurisprudence of German and European courts, he discusses problems
of judicial review with an eye on Alexy’s “weight formula” and applies the notion of
formal principles in the balancing.
Matthias Klatt analyses the alleged problem of justiciability of positive rights and
rejects the democratic objection against adjudication in the field of positive rights.1
He suggests that judicial review and judicial deference can be displayed in different
degrees and develops a model of the balancing of judicial competences.
Federico de Fazio presents a rational reconstruction of balancing of social rights.
He analyses the structure of the proportionality test regarding omissions, that is,
applying positive rights, compared to interferences with (negative) rights, and out-
lines the rules and forms of the proportionality test when applied to omission.
Laura Clérico and Martin Aldao explore how proportionality analysis can display
structural inequality patterns. Analysing the opinions delivered in a case of the Inter-
American Court of Human Rights, they point out that a broad use of the propor-
tionality test disguises structural inequality, whilst a thorough test of proportionality
lays open discrimination.
Last but not least, Daniel Oliver Lalana analyses a rather new development, in
which the legislature delegates balancing to the recipients of laws, who are expected
to establish a general rule by means of balancing, under which they subsume their
own case. He explores the implications of this “proportionality-promoting legisla-
tion” (PPL) and the bearing of PPL on both legisprudence and legal theory.
The book ends with an index of key notions and main theses of the contributions,
however, referring only to the site where they are introduced, not to each page where
they occur.
In sum, this volume presents contributions to crucial issues of the theory of
principles, which highlight some topics of current research, although they cannot
address all open aspects of this theory. Many issues are still in dispute.

1
This article replaces the contribution of the Erlangen conference, which is published as Klatt,
Matthias (2020): Balancing Rights and Interests. Reconstructing the Asymmetry Thesis. In: Oxford
Journal of Legal Studies 2020, gqaa051. https://doi.org/10.1093/ojls/gqaa051.
Preface vii

Finally, I should like to thank Rosaline Bates Anoma and Mongui Ojong
Eyumeneh for suggestions regarding the English language for those contributions
where revision by a native speaker is not provided by the author.

Erlangen, Germany Jan-R. Sieckmann

November 2020
Contents

Constitutional Rights, Proportionality, and Argumentation . . . . . . . . . . 1

Robert Alexy
From Constitutional Discretion to the Positivist Weight Formula . . . . . . 11
David Duarte
Brute Balancing, Proportionality and Meta-Weighing of Reasons . . . . . 49
Jorge Silva Sampaio
Limited Balancing: The Principle of Human Dignity and Its
Inviolability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Carsten Bäcker
To Balance or Not to Balance: The Quest for the Essence of Rights . . . . 113
Jan-R. Sieckmann
Limited Review in Balancing Constitutional Rights . . . . . . . . . . . . . . . . 135
Martin Borowski
Positive Rights: Who Decides? Judicial Review in Balance . . . . . . . . . . . 163
Matthias Klatt
Proportionality Test and Constitutional Social Rights . . . . . . . . . . . . . . 197
Federico De Fazio
An Argument for the Test of Proportionality in Concreto: Silenced
Voices from the Margins to the Center . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Laura Clérico and Martín Aldao
‘Thou Shalt Balance’: Making Sense of the Delegation of
Proportionality Testing to the End Users of Laws . . . . . . . . . . . . . . . . . 231
A. Daniel Oliver-Lalana

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

ix
Editor and Contributors

About the Editor

Jan-R. Sieckmann Dr. jur., Professor of Legal Theory and Philosophy of Law at
the University of Erlangen-Nürnberg. Author of Regelmodelle und
Prinzipienmodelle des Rechtssystems, Nomos Verlag, 1990, Recht als normatives
System, Nomos Verlag, 2009, The Logic of Autonomy, Hart Publishing, 2012,
Rechtsphilosophie, Mohr Siebeck, 2018, and Logik juristischer Argumentation,
Nomos Verlag, 2020. Main fields of research: legal philosophy, argumentation
theory, and human and constitutional rights.

Contributors

Martín Aldao Faculty of Law, University of Buenos Aires (UBA), Buenos Aires,
Argentina
Robert Alexy Faculty of Law, Christian-Albrechts-Universität Kiel, Kiel,
Germany
Carsten Bäcker Faculty of Law, University of Bayreuth, Bayreuth, Germany
Martin Borowski Faculty of Law, University Heidelberg, Heidelberg, Germany
Laura Clérico Faculty of Law, University of Buenos Aires (UBA), Buenos Aires,
Argentina
FAU University of Erlangen-Nuremberg, Erlangen, Germany
Federico De Fazio Faculty of Law, University of Buenos Aires, Buenos Aires,
Argentina
David Duarte Faculty of Law, University of Lisbon, Lisbon, Portugal

xi
xii Editor and Contributors

Matthias Klatt Faculty of Law, University of Graz, Graz, Austria

A. Daniel Oliver-Lalana Faculty of Law, University of Zaragoza, Zaragoza, Spain
Jorge Silva Sampaio Faculty of Law, University of Lisbon, Lisbon, Portugal
Jan-R. Sieckmann Faculty of Law, FAU University of Erlangen-Nuremberg,
Erlangen, Germany
Constitutional Rights, Proportionality,
and Argumentation

Robert Alexy

Abstract In this article I argue that the rational application of constitutional rights
necessarily presupposes proportionality analysis as defined by principles theory and
that proportionality analysis can be rational only if it is embedded in a theory of legal
argumentation. By this means, argumentation theory is linked to the theory of
constitutional rights in a single system.

1 Thesis

My thesis is, first, that the rational application of constitutional rights necessarily
presupposes proportionality analysis and that, second, proportionality analysis can
be rational only if it is embedded in a theory of legal argumentation. I will begin with
an explication of proportionality analysis by means of principles theory. In a second
step the relation between balancing and argumentation will be analysed.

2 Proportionality and Principles Theory

2.1 Rules and Principles

Principles theory begins with the norm-theoretic distinction between rules and
principles.1 Rules are norms that require something definitively. They are definitive
commands. Their form of application is subsumption. If a rule is valid and if its

I should like to thank Stanley L. Paulson for suggestions and advice on matters of English style.
1
See Alexy (2002a), pp. 47–49.

R. Alexy (*)
Faculty of Law, Christian Albrechts University Kiel, Kiel, Germany
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 1

J.-R. Sieckmann (ed.), Proportionality, Balancing, and Rights, Law and Philosophy
Library 136, https://doi.org/10.1007/978-3-030-77321-2_1
2 R. Alexy

conditions of application are fulfilled, it is definitively required that exactly what it

demands be done. If this is done, the rule is complied with; if this is not done, the rule
is not complied with. By contrast, principles are optimization requirements. As such,
they demand that something be realized ‘to the greatest extent possible given the
legal and factual possibilities’.2 Rules aside, the legal possibilities are determined
essentially by opposing principles. For this reason, principles, each taken alone,
always comprise a merely prima facie requirement. The determination of the
appropriate degree of satisfaction of one principle relative to the requirements of
other principles is brought about by balancing. Thus, balancing is the specific form
of application of principles.

2.2 Proportionality

The nature of principles as optimization requirements leads straightaway to a

necessary connection between principles and proportionality. The principle of pro-
portionality (Verhältnismäßigkeitsgrundsatz), consists of three sub-principles: the
principles of suitability, of necessity, and of proportionality in the narrower sense.
All three sub-principles express the idea of optimization. Principles qua optimization
requirements require optimization relative both to what is factually possible and to
what is legally possible. The principles of suitability and necessity refer to optimi-
zation relative to the factual possibilities. The principle of proportionality in the
narrower sense concerns optimization relative to the legal possibilities.

2.2.1 Suitability

The first sub-principle, the principle of suitability, precludes the adoption of means
that obstruct the realization of at least one principle without promoting any principle
or goal for which it has been adopted. If a means M, adopted in order to promote the
principle P1, is not suitable for this purpose, but obstructs the realization of P2, then
there are no costs either to P1 or P2 if M is omitted, but there are costs to P2 if M is
adopted. Thus, P1 and P2 taken together may be realized to a higher degree relative
to what is factually possible if M is abandoned. P1 and P2, when taken together—that
is, as elements of a single system—proscribe the use of M. This shows that the
principle of suitability is nothing other than an expression of the idea of Pareto-
optimality. One position can be improved without detriment to the other.
An example of the violation of the principle of suitability is found in a decision of
the German Federal Constitutional Court. The law in question required that not only
persons who apply for a general hunting licence have to pass a weapons examination
but also persons who apply exclusively for a falconry licence. The Court argued that

2
Ibid. 47.
Constitutional Rights, Proportionality, and Argumentation 3

the weapons examination for falconers is not suitable for promoting the “proper
exercise of these activities as intended by the legislator”.3 Therefore, “no substan-
tially clear reason”4 existed for the infringement of the general freedom of action on
the part of the falconer, as guaranteed by Article 2 (1) of the Basic Law. For that
reason, the regulation was declared disproportional5 and, consequently,
unconstitutional.

2.2.2 Necessity

Cases in which laws are declared unconstitutional for reasons of unsuitability are
rare. Normally the measure applied by the legislator will at least promote his aims to
a certain degree. This suffices for suitability. For this reason, the practical relevance
of the sub-principle of suitability is rather low. The situation is completely different
where the second sub-principle of the principle of proportionality, the principle of
necessity, is concerned. This principle requires that of two means promoting P1 that
are, broadly speaking, equally suitable, the one that interferes less intensively with
P2 has to be chosen. If there exists a less intensively interfering and equally suitable
means, one position can be improved at no cost to the other.6 Under this condition,
P1 and P2, taken together, require that the less intensively interfering means be
applied. This is, again, a case of Pareto-optimality.
An example is the decision of the Federal Constitutional Court on sweets,
especially in the form of Easter rabbits and Santa Claus figures that consist of puffed
rice. In order to protect consumers from mistakenly taking those puffed rice sweets
to be chocolate products, a ban on puffed rice sweets was issued. The Court argued
that consumer protection could be achieved “in an equally effective but less incisive
way by a duty of labelling”.7 For this reason, the ban was declared to be a violation
of the principle of necessity and, therefore, disproportional.

3
Decisions of the Federal Constitutional Court (Entscheidungen des Bundesverfassungsgerichts;
hereafter: BVerfGE) 55, 159 (166).
4
BVerfGE 55, 159 (167).
5
BVerfGE 55, 159 (166).
6
Where the question of whether the less or the more intensively interfering means is chosen arises,
the principle of necessity presupposes that all other principles or goals are indifferent with respect to
this choice. If, however, there exists a third principle or goal, P3, that is affected negatively by the
adoption of the means interfering less intensively with P2, than the case cannot be decided by
considerations concerning Pareto-optimality. When costs are unavoidable, balancing becomes
necessary.
7
BVerfGE 53, 135 (146).
4 R. Alexy

2.2.3 Proportionality in the Narrower Sense

Just as with the principle of suitability, the principle of necessity concerns optimi-
zation relative to the factual possibilities. Optimization relative to the factual possi-
bilities consists in avoiding avoidable costs. Costs, however, are unavoidable when
principles collide. Then balancing becomes necessary. Balancing is the subject of the
third sub-principle of the principle of proportionality, the principle of proportionality
in the narrower sense. This principle tells us what optimization relative to the legal
possibilities is. In the Theory of Constitutional Rights I expressed the requirement of
optimization relative to the legal possibilities by means of a rule that I called the
“Law of Balancing”.8 This rule states:
The greater the degree of non-satisfaction of, or detriment to, one principle, the greater must
be the importance of satisfying the other.

The Law of Balancing excludes, inter alia, an intensive interference with princi-
ple P1 that is justified only by a low importance assigned to the satisfaction of the
colliding principle P2. Such a solution would not be an optimization of P1 together
with P2.

2.2.4 Weight Formula

The formulation of the Law of Balancing has been developed with an eye to a
standard situation that appears with great frequency.9 For a deeper, more precise, and
complete analysis, the Law of Balancing has to be elaborated further. This further
elaboration amounts to the Weight Formula.10 The Weight Formula defines the
weight of a principle Pi in a concrete case, that is, the concrete weight of Pi relative
to a colliding principle Pj (Wi, j), as the quotient, first, of the product of the intensity
of the interference with Pi (Ii) and the abstract weight of Pi (Wi) and the degree of
reliability of the assumptions concerning what the measure in question means for the
non-realization of Pi (Ri), and, second, the product of the corresponding values with
respect to Pj, now related to the realization of Pj. It runs as follows:

I i  W i  Ri
W i,j ¼
I j  W j  Rj

A formula like the Weight Formula, which expresses a quotient of two products,
is sensible only if all of the factors can be represented by numbers. This is the
problem of graduation. Elsewhere,11 I have proposed a discrete, that is, a

8
Alexy (2002a), p. 102.
9
See Alexy (2007), pp. 15–17.
10
See Alexy (2003), pp. 443–448.
11
Alexy (2007), pp. 20–26.
Constitutional Rights, Proportionality, and Argumentation 5

non-continuous triadic scale, in which geometric sequences are implemented. This

scale assigns the values “light” (l), “moderate” (m), and “serious” (s) to the intensity
of interference and to the abstract weights. These values are expressed by the
numbers 20, 21, and 22, that is, by 1, 2, and 4. Where the epistemic side is concerned,
that is, Ri and Rj, one can work with the stages “reliable” or “certain” (r), “plausible”
( p), and “not evidently false” (e), to which the numbers 20, 2–1, and 2–2, that is, 1, ½,
and ¼, are to be assigned.12 By means of these triads, most of the decisions of
constitutional courts can be grasped. Where they do not suffice, that is, where one
has to introduce still more attenuated graduations, they can be extended to double-
triadic scales.13

3 Balancing and Argumentation

3.1 The Formal and the Substantive Dimension of Rationality

Heretofore I have considered only the formal rationality of balancing, expressed

especially by the Weight Formula. A series of objections has been raised against the
formal character of the Weight Formula. Matthias Jestaedt, for example, has
objected to the Weight Formula on the ground that it “holds out the promise of a
degree of certainty and precision in application which it is quite unable to keep”.14
“[I]ts claimed, or at least implied, precision is revealed to be a mere illusion, a
methodological chimera”.15 Stavros Tsakyrakis speaks of “the myth of mathematical
precision”, which must be rejected where “judicial reasoning” is concerned.16
Alexander Somek objects that the balancing model of principles theory does not
go beyond “an analytically careful formalization of moral intuitionism”.17 And, just
to mention a fourth voice, Kai Möller, in his book The Global Model of Constitu-
tional Rights, begins with a juxtaposition of his theory as a “substantive moral
approach . . . with a formal theory such as Robert Alexy’s . . . theory of rights as
principles or optimization requirements”.18
If principles theory were only a formal theory, these objections would indeed
apply. But principles theory is a theory that intrinsically connects the formal and the
substantive dimension of constitutional rights, and also connects both dimensions
with respect to law in general. Ralf Poscher says, albeit with a critical undertone, that
principles theory comprises “a theory about the concept of law, another about the

12
Ibid. 25.
13
Ibid. 22–23.
14
Jestaedt (2012), p. 163.
15
Ibid. 165.
16
Tsakyrakis (2009), p. 472.
17
Somek (2006), p. 135.
18
Möller (2012), pp. 1–2 (emphasis in the original).
6 R. Alexy

structure of norms, a theory of adjudication, one of argumentation, and finally,

different doctrinal theories”.19
A formal theory that is in need of being supplemented by a substantive theory is
as such, one might well think, of no value at all. This, however, is not the case. A
purely formal theory would, indeed, be empty. A purely substantive theory has,
however, another defect. It would be blind in all questions concerning the structure
of balancing, and this defect would be a serious defect, for the structure of balancing
is complex. Even the best substantive theory would always remain vulnerable to
error for reasons of lack of conceptual clarity, lack of system, and lack of argumen-
tative conclusiveness. Such a state of affairs would amount to a serious restriction of
rationality.
Gottlob Frege compares in his Conceptual Notation (“Begriffsschrift”) the rela-
tion of his “Formula Language”20 to “ordinary language”21 with “the relation of the
microscope to the eye. The latter, because of the range of its applicability and
because of the ease with which it can adapt itself to the most varied circumstances,
has a great superiority over the microscope. Of course, viewed as an optical
instrument it reveals many imperfections”.22 This directly applies to the formal
analysis of proportionality. The ordinary language, the eye, must remain the lan-
guage of the courts in cases of balancing as well as in cases of subsumption. In the
scientific analysis of subsumption as well as of balancing, however, a formula
language, the microscope, is indispensable, in order to achieve as precise, as deep,
and as easily comprehensible insight into the structure of these two forms of the
application of law.23
The question remains: how can the formal dimension be connected with the
substantive dimension? The answer consists in embedding the Weight Formula in
a theory of rational legal argumentation.

3.2 Numbers, Classification Propositions, and their

Justification

At first glance, such an embedding seems to be excluded by the nature of the Weight
Formula on the one hand, and by the nature of legal argumentation or discourse on
the other. The basic elements of the Weight Formula are numbers, the basic elements
of legal discourse are arguments. Numbers can well be connected with measuring
instruments, say, with thermometers, but how can they be linked with arguments?
Alongside the thermometer, there exists no argumeter. The answer is this. The

19
Poscher (2012), p. 221.
20
Frege (1972), p. 101.
21
Ibid. 104.
22
Ibid. 105.
23
On a third form see Alexy (2010), pp. 9–18.
Constitutional Rights, Proportionality, and Argumentation 7

numbers that have to be substituted in the variables of the Weight Formula represent
classifications, in case of the variables Ii and Wi, as well as of their counterparts, the
classifications as “light”, “moderate”, and “serious”. An example is an interference
with the freedom of speech (Pi), which serves to protect the personality right (Pj). If
interference with the freedom of speech is classified as serious, that is, if Ii receives
the value 4, then the number 4 stands for nothing other than the proposition: “This
interference with the freedom of speech is serious”. The same applies if the intensity
of the interference with the personality right through non-protection is classified as
light. The number 1, then, stands for the proposition “This interference with the
personality right through non-protection would be light”. The pivotal point, now, is
that such propositions – one might call them “classification propositions”—are as
propositions or assertions in general in need of justification. The question is whether
they are capable of justification.
Authors who contest the rationality of balancing must deny the justifiability of
classification propositions. Examples of such denials are Jürgen Habermas’s argu-
ment from arbitrariness,24 Bernhard Schlink’s argument from decisionism,25 and
Poscher’s argument from intuitionism.26 Against all of these positions I set out the
argumentation thesis. The argumentation thesis says that propositions about inten-
sity of interference and degrees of importance lend themselves to rational justifica-
tion. An example is presented by a decision of the German Federal Constitutional
Court from 1997 on the duty of manufacturers of tobacco products to place health
warnings respecting the dangers of smoking on their products.27 This is a minor or
light interference with the freedom to pursue one’s profession. By contrast, a total
ban on all tobacco products would count as serious interference. Between such
minor and serious cases, others of moderate intensity of interference can be found.
An example would be a ban on cigarette machines along with the introduction of
provisions restricting the sale of tobacco to selected shops. If it were true, as Poscher
maintains, that balancing depends on nothing other than “our intuition”,28 then every
interpreter of a constitution would be free to say, appealing to his intuitions, that the
duty to set down health warnings is a serious inference and the total tobacco ban a
light interference with the tobacco producers’ freedom of profession. But it would
not be easy to take such an approach seriously. In any case, it would be easy to give
cogent reasons as to why these classifications are mistaken. If, in this way, the
intensity of interference, as in the tobacco case, is established as minor and the
importance of protecting the population from health risks as high,29 the result is
easily recognizable. The reason that weighs heavily on behalf of interference justifies
the minor interference.

24
Habermas (1996), p. 259.
25
Schlink (1984), p. 462.
26
Poscher (2012), p. 241.
27
BVerfGE 95, 173.
28
Poscher (2012), p. 241.
29
BVerfGE 95, 173 (183–187).
8 R. Alexy

3.3 Disagreement, Discourse, and Rationality

A balancing skeptic might concede that it is possible to justify classifications of the

intensity of interference in the case considered but nevertheless insist that there are
cases in which different classifications can be justified equally well, which, so the
objection continues, is to say that none of them is justifiable. As an example, the
skeptic might refer to the decision of the German Federal Constitutional Court in
2006 on electronic data-screening. A student with Moroccan citizenship brought a
constitutional complaint against an order of a court, upheld in the higher courts,
concerning electronic data-screening. This order required residents’ registration
offices, the central aliens’ registry, and universities to transmit data to the police
concerning male persons between 18 and 40 regarding, inter alia, their faith, native
country, and subject of study. These data were submitted to automatic data
processing with an eye to identifying potential Islamist terrorists. The majority of
the First Panel of the Federal Constitutional Court classified the measure as an
interference of “considerable weight”,30 that is, as a serious interference with the
complainant’s right to data self-determination. Such a serious interference would be
justified in cases of concrete danger but not in cases of abstract danger, characterized
by a “generally threatening situation” as it has existed “at the latest since September
11, 2001”.31 It was deemed, however, that no concrete danger was at hand. There-
fore, the constitutional complaint was considered to be justified. Completely differ-
ent is the classification stemming from the dissenting opinion of Judge Haas.
According to her assessment, the interference is “of low weight” and has to be
accepted as being in the “public interest”.32 From this point of view, the constitu-
tional complaint would not be justified.
Here two points are of interest. The first is that both sides presented numerous
arguments for their classifications. The arguments of the majority on the assessment
of the interference as serious ran over eleven pages,33 those of Justice Haas for her
assessment as light ran over four pages.34 This shows that balancing is not a matter
of naked classification. It is, essentially, a matter of argumentation. Thus, if rational
legal argumentation is possible, rational balancing is possible, too.
The second point is that disagreement does not count as an argument for irratio-
nality if “reasonable disagreement”35 can be distinguished from unreasonable dis-
agreement. Here only normative or practical argumentation is of interest. Now the
decisive point is that there exist criteria of the rationality of practical arguments and
practical argumentation or discourse. The conditions of discursive rationality can be
made explicit by means of a system of principles, rules, and forms of general

30
BVerfGE 115, 320 (348).
31
BVerfGE 115, 320 (364).
32
BVerfGE 115, 320 (379).
33
BVerfGE 115, 320 (347–357).
34
BVerfGE 115, 320 (371–374).
35
See Rawls (1993), p. 55.
Constitutional Rights, Proportionality, and Argumentation 9

practical discourse.36 This system comprises criteria that demand non-contradiction,

clarity of language, reliability of empirical premises, and sincerity, as well as rules
and forms that speak to the consequences, to universalizability, and to the genesis of
normative convictions.
A sceptic might concede that there are, indeed, some criteria of the rationality of
practical arguments but insist that this does not suffice to establish rationality in
cases like the electronic data-screening case. The issue has not been resolved
discursively by the power of the better argument but institutionally by a majority
decision. The point is correct, but it does not count as a complete analysis of the
situation. Both sides raise a claim that is not a claim to majority but a claim to
correctness, necessarily connected with law.37 This claim remains alive even after
the authoritative majority decision has been taken. This implies the possibility of
changes by way of future argumentation. In this respect, rationality has an ideal
dimension.

References

Alexy R (1989) A theory of legal argumentation (1978) (trans: Adler R and MacCormick N).
Clarendon Press, Oxford
Alexy R (2002a) A theory of constitutional rights (1985), (trans Rivers J). Oxford University Press,
Oxford
Alexy R (2002b) The argument from injustice. A reply to positivism (1992), (trans: Paulson SL,
Litschewski Paulson B). Clarendon Press, Oxford
Alexy R (2003) On balancing and subsumption. Ratio Juris 16:433–449
Alexy R (2007) The weight formula. In: Stelmach J, Brożek B, Załuski W (eds) Frontiers of the
economic analysis of law. Jagiellonian University Press, Krakow, pp 9–27
Alexy R (2010) Two or three? Archiv Philosophy Law Soc Philosophy, Supplement 119:9–18
Frege G (1972) Conceptual notation. A formula language of pure thought modelled upon the
formula language of arithmetic (1879). In: Frege, Conceptual Notation and Related Articles,
(trans: Terrell Ward Bynum). Clarendon Press, Oxford, pp 101–203
Habermas J (1996) Between facts and norms. Contributions to a discourse theory of law and
democracy (1992) (trans: William Rehg). Polity Press, Cambridge
Jestaedt M (2012) The Doctrine of balancing – its strengths and weaknesses. In: Klatt M
(ed) Institutionalized reason. The Jurisprudence of Robert Alexy. Oxford University Press,
Oxford, pp 152–172
Möller K (2012) The global model of constitutional rights. Oxford University Press, Oxford
Poscher R (2012) The principles theory. How many theories and what is their merit? In: Klatt M
(ed) Institutionalized reason. The Jurisprudence of Robert Alexy. Oxford University Press,
Oxford, pp 218–247

36
Alexy (1989), pp. 188–206.
37
Alexy (2002b), pp. 35–39.
10 R. Alexy

Rawls J (1993) Political liberalism. Columbia University Press, New York

Schlink B (1984) Freiheit durch Eingriffsabwehr – Rekonstruktion der klassischen
Grundrechtsfunktion. Europäische Grundrechte-Zeitschrift 11:457–468
Somek A (2006) Rechtliches Wissen. Suhrkamp, Frankfurt
Tsakyrakis S (2009) Proportionality: an assault on human rights? I•CON 7:468–493

Robert Alexy is Emeritus Professor of Public Law and Legal Philosophy at the Christian
Albrechts University in Kiel. He is the author of A Theory of Legal Argumentation, Clarendon
Press 1989 (first published in German 1978), A Theory of Constitutional Rights, Oxford University
Press 2002 (first published in German 1985), and of The Argument from Injustice, Clarendon Press
2002 (first published in German 1992). In his later work, he has attempted to refine and to transform
the ideas first developed in these three books into a system.
From Constitutional Discretion
to the Positivist Weight Formula

David Duarte

Abstract A correct understanding of balancing discretion shows how proportion-

ality works as an internal limit of discretion and how suitability, necessity and
proportionality in the narrow sense constrain the choices to be made by a legislative
normative authority. On the other hand, an accurate contraposition between the laws
of balancing (the content of proportionality in the narrow sense) and Alexy’s Weight
Formula leads to the conclusion that this equation does not express (as it should) the
role those laws have in constraining the range of choice presented by “prima facie
discretion”. Coming from the problems therein detected, a new formula is presented;
a formula that fully matches the substantive and the epistemic laws and that, to
satisfy its function of internal justification of a balancing decision, only entails the
variables those laws support.

1 A Brief Scheme of Constitutional Discretion

Discretion may be described as the normative scenario where, for the enaction of a
norm (or set of norms), more than one alternatives are legally admissible. These
alternatives may be related to different aspects of that norm, such as time or
procedural fact finding actions.1 It is known, however, that discretion gets its

I am deeply thankful to Margarida Duarte for her support with the mathematical features of
proportionality. For their critical approach, I am also grateful to Miodrag Jovanović, Bojan
Spaić, Wojciech Załuski and Katarzyna Eliasz (Belgrade Seminar), Robert Alexy, Jan
Sieckmann, Laura Clerico, Martin Borowski, Matthias Klatt, Carsten Bäcker and Daniel Oliver-
Lalana (Erlangen Seminar), and Pedro Moniz Lopes, Jorge Silva Sampaio and Ana Escher from
our Lisbon Legal Theory Group.
1
For instance, to enact a norm in time1 or time2 or to ask for expertise1 or expertise2. For procedural
aspects of legislative discretion, Borowski (2010), p. 135; Walker et al. (2015), p. 358.

D. Duarte (*)
University of Lisbon, Lisbon Legal Theory Group (LxLTG), Lisbon, Portugal
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 11

J.-R. Sieckmann (ed.), Proportionality, Balancing, and Rights, Law and Philosophy
Library 136, https://doi.org/10.1007/978-3-030-77321-2_2
12 D. Duarte

major relevancy when it concerns the content of the norm: after all, alternatives
regarding content are no other than different deontic statuses for an action, or
different courses of action under the same conditions, and that relevancy rests
exactly on this “legal possibility” of formulating distinct deontic consequences for
the same reality.2 Evidently, it is only this content related discretion that
matters here.
A normative scenario of discretion can be found at different levels of a legal order
and seen from the distinct perspectives of various types of norm production.
Nonetheless, and for the present purposes, discretion matters only when it is given
by a constitution (despite the possibility of applying its general theory to all levels of
norm production). Within this context, discretion is also seen here from the stand-
point of the decision-maker: that is, from the perspective of a legislative authority
enacting norms under a constitution. It is worth adding, though, that the present
approach is immune to the contingencies of a specific constitution: it only deals with
the invariants of constitutional discretion.3
Constitutional discretion may be classified under the most diverse criteria. Nev-
ertheless, the one that seems to be the most informative is its source, meaning where
in law (here, in the constitution) does discretion come from. It is a criterion that,
consequently, points out to where we can find the “openness” that confers a plurality
of deontic alternatives.4 With this criterion, discretion can be divided into:
(i) linguistic; and (ii) normative. Within the latter, a distinction has to be made
between: (iia) uni-normative; and (iib) pluri-normative.
Linguistic discretion is the one given by constitutional norm sentences whenever
they entail linguistic uncertainties. As it happens with the common usage of a natural
language by any normative authority, norm sentences in constitutions are also
vulnerable to those uncertainties, regardless of whether they come from the way
words are organized within sentences (syntactical) or from the range of meanings a
word might have (lexical).5 Bearing in mind that diverse meanings in a norm
sentence imply different possibilities of norms expressed, it follows that linguistic

2
On content related discretion, Hart (2013), p. 656; materially, Sartor (2013), p. 1450.
3
Provided that the legal order at hand entails a principle of proportionality, a contingent state of
affairs (as with all norms enacted by normative authorities or produced by customary sources). As a
matter of fact, it is not impossible (although absurd) to insert into a constitution a principle of
disproportionality. On the contingency of the content of legal orders, Bulygin (2010), p. 285;
Mendonca (2000), p. 63. Differently, claiming the universality of proportionality, Sieckmann
(2018), p. 4; Klatt and Meister (2012), p. 1.
4
This criterion is particularly operative here since it gives the best view to distinguish balancing
discretion from other kinds of constitutional discretion. For other classifications, Alexy (2002),
p. 394; Rivers (2007), p. 170.
5
Nino (2003), p. 256; Duarte (2018), p. 136.
From Constitutional Discretion to the Positivist Weight Formula 13

discretion is exactly on the point that a choice on a specific meaning has to be carried
out. After all, only one norm can be assigned to a text.6
This kind of discretion can be exemplified with the linguistic uncertainty pro-
vided by vagueness. If a constitutional norm sentence entails a vague word, then
there is no certainty on the limits of the sentence and, consequently, no certainty on
the precise scope of the constitutional norm expressed.7 In bordeline cases, where it
is doubtful if the word covers the world, a linguistic choice with obvious deontic
consequences emerges: if the word is instantiated, the norm is applicable to the case;
if it is not, the case will be forwarded to another norm. It is worth noting, though, that
this is just the outcome of vagueness (and referring to the antecedent); other
linguistic uncertainties create other discretion schemes.8
(i) “The constitution of fascist associations is forbidden”, an exception to “freedom of associa-
tion”, can be confronted with borderline cases due to the vagueness of the word “fascist”.
(ii) In such normative scenario, two deontic statuses compete: with “q ¼ to constitute an
association”, we will have a permission (P q) and a prohibition (F q).
(iii) Linguistic discretion is, then, on the exclusive disjunction of deontic consequences provided
by the alternativity given by the word: “borderline cases of fascist associations ) F q ⊻ P q”.

Normative discretion is the one given by constitutional norms regardless of the

language used to express them.9 Within this kind, we have uni-normative discretion
whenever it comes from a single norm. Obviously, nothing prevents a constitutional
norm from presenting different deontic statuses for the same action or to allow for
different courses of action under the same conditions. It is the case, for instance, of a
bilateral permissive norm addressed to the normative authority or the one imposing a
duty with different degrees of realization.10 Nonetheless, the same point is always
present: an area of discretion entirely assignable to a specific norm follows whenever
this norm presents alternatives of deontic action.
(iv) With a constitutional norm such as “p ) O q”, where “q ¼ to promote the environment”, a
clear duty with variable degrees ascertains (q ¼ q1 _ q2 _ . . . _ qn ); thus, the duty holder has
discretion (and totally assignable to “p ) O q”) regarding the choice among alternatives (q1 _
q2 _ . . . _ qn).

6
If for “norm sentence” we understand a closed text strictly expressing the totality of elements that
match a certain criterion of norm individuation (NS ! N). Obviously, as an utterance, a norm
sentence (understood just as a written text) can entail an unlimited quantity of norms (NS ! N1 ^
N2), something quite common in long constitutional articles (for instance, article 26/1 of the
Portuguese Constitution has, at least, seven norms).
7
On vagueness, Moreso (2017), p. 49; Escher (2019), p. 163.
8
For instance, the syntactical uncertainty provided by the disjunctive or conjunctive use of the word
“and”: in a norm sentence expressing “O q _ r “and” s”, we can read the following possible norms
“O q ^ s”, “O r ^ s”, or “O q _ r _ s”. The discretion present is, however, and regardless of its own
proper scheme, totally due to a linguistic uncertainty.
9
A kind of discretion that, consequently, presupposes that the linguistic moment of interpreting the
norm sentence is already behind. Navarro and Rodríguez (2014), p. 180; Sardo (2018), p. 58.
10
By bilateral permission it is meant the permission of an action (φ) and its contradictory (~φ); a
faculty. Hansson (2013), p. 199; Ratti (2018), p. 89.
14 D. Duarte

(v) With a constitutional norm such as “p ) P r”, where “r ¼ award grants” (under a bilateral
permission), an alternativity between “P r” and “P ~r” is given; if the permission is directed to
a normative authority, this authority has a choice to carry out solely assignable to “p ) P r”.

Normative discretion is pluri-normative whenever it arises from two or more

norms entering into a conflict. However, there is no discretion when the conflict of
norms is solvable by a norm of conflicts: given that this latter norm solves the
contradiction, the original scenario of two deontic statutes for the same action
disappears.11 Thus, and just for the reason of the conflict, no discretion is given
here. Pluri-normative discretion exists, effectively, whenever the conflict of norms is
not solvable by any third norm, which means that this kind of discretion is entirely
filled by “balancing discretion”: that is, the one coming from two applicable norms
with contradictory deontic statuses without a solution given by a third norm.
Balancing discretion is quite common at the constitutional level. Since constitu-
tions are full of norms that are principles and considering that principles always
originate conflicts of the “partial-partial” type (unable to be solved by norms of
conflicts), it follows that balancing discretion takes a relevant slice of the whole of
constitutional discretion.12 In brief, it may be described as the discretion related to
the choice within the deontic contradiction placed by the conflict: if two applicable
principles give irreconcilable deontic outcomes, then an assessment of preference
has to be made. Accordingly, selecting one or the other (or enacting a norm
satisfying one of them) is, exactly, what balancing discretion is all about.
(vi) With “freedom of expression” as Pi and “protection of honor” as Pj, both norms enter into a
conflict whenever their intersection is instantiated: an insulting speech act, for instance.
(vii) If Pi is a permission to speak (P s) and Pj a prohibition to affect the honor (F h), their
intersection (converting F h into ~P s) will show the area of discretion: “insult ) P s ⊻ F h
(~P s)”.

It is worth mentioning that with a classification of discretion such as the one based
on its sources a complete and exhaustive list of kinds has to follow. Accordingly, the
dual typology of linguistic and normative discretions, with the mentioned subdivi-
sion in the latter, denotes all possible kinds of constitutional discretion. It is also
worth noting that, independently of the kind of discretion at stake, a common (and
unifying) property is present: a specific range of alternatives of action is given to a
normative authority.13 This range might be designated as “prima facie discretion”

11
Wróblewski (1992), p. 289; Moreso and Vilajosana (2004), p. 106. It is relevant to note, however,
that any kind of uncertainty coming from the norm of conflicts (its sentence) is necessarily due to
linguistic indeterminacy, which means, evidently, that in this case we are again within the first
category of constitutional discretion.
12
Moreso (2002), p. 21; Chiassoni (2019), p. 189. One note: it is only based on statistical
preponderance that principles are being used as examples of constitutional norms in a conflict
conferring balancing discretion (Pino [2014], p. 612). Since rules are as defeasible as principles (and
in the same way), nothing prevents similar constitutional conflicts with rules.
13
For instance, Hart (2013), p. 661; Klatt (2007), p. 507.
From Constitutional Discretion to the Positivist Weight Formula 15

since it entails all the alternatives immediately given by the source of discretion at
hand and regardless of any “all norms considered” assessment.
It follows, then, that “prima facie discretion” has to be distinguished from “all
norms considered discretion”, being the latter the discretion still present after the
interposition of other, eventually applicable, constitutional norms (internal limits of
discretion): those that, depending on the properties of the case, might reduce (or even
eliminate) the original range. Although contingent, multiple norms can play this role
of internal limits of discretion: equality, proportionality or public interest, for
instance. If they do not eliminate discretion, after their interposition there is (still)
an all norms considered set of alternatives, which also is, and for the sake of the
separation of powers, a margin of inadmissible review.14

2 Balancing Discretion

As mentioned above, balancing discretion is the discretion given by a conflict of

constitutional norms unsolvable by norms of conflicts. Therefore, and using princi-
ples at the constitutional level, it happens when two are applicable to the same action
and provide contradictory deontic statuses. In a normative scenario such as this one,
an area of choice is given: from a decision that fully satisfies one principle to the
decision fully satisfying the other, a large range of alternatives appear. This can be
exemplified with a case: the (imaginary) vaccination case. Specifically, it reads as
follows: a legislative normative authority has to deal with the spread of an infectious
disease; it is known, nonetheless, that the spread of the disease can be prevented
through the general administration of a vaccine.
Enacting a norm imposing general vaccination will be framed by two competing
principles: P1, the prohibition to injure physical integrity; and P2, the duty addressed
to the State to promote public health. Naturally, P1 points out to a decision of
non-vaccination, since any will imply the perforation of the body and the insertion
of a liquid; while P2 points out to a decision of vaccination for the reason that it is
(or might be) a suitable course of action in order to eradicate the illness. Though, if
there are more than one active ingredients suitable to fight the disease, with distinct
efficiencies, various alternatives (rules ¼ R) can be placed within the range that starts
in P1 and finishes in P2. The figure below shows this range.
(viii) When enacting a R on vaccination, various Rs are conceivable:

14
Internal limits of discretion can provide, then, a total “reduction of discretion to zero” or,
differently, a maintenance of alternatives (same or less than initially): here, without legal parame-
ters, judicial review is constitutionally forbidden. Framing the problem as one of competence,
Alexy (2016), p. 69; Bernal Pulido (2009), p. 116.
16 D. Duarte

P1 --------------------------------------------------------------- P2
R1 R2 R3 R4

(ix) These Rs are: R1, no vaccination; R2, vaccination with active ingredient x; R3, vaccination with
active ingredient y; R4, killing those affected by the disease.

As it is visible, R1, R2, R3 are all prima facie permitted. In a way or another, and
on the condition of R2 and R3 being suitable to promote public health, they are all
sustained by the two constitutional Ps at hand: (i) R1 fully promotes P1; and (ii) R2
and R3 both promote P2, although in different degrees, depending on their
therapeutical efficiency. Therefore, the range between being sustained by P1 and
being sustained by P2 is a margin of admissible alternatives immediately given by
those principles: it is, then, the “prima facie discretion” present in this balancing
discretion. As it is also visible, R4 is out of that range: it is not sustained by P1
(actually, it is even contrary to P1) and does not promote P2 in any way.
It is inside the range given by P1 and P2, and regarding the properties of the case,
that internal limits of discretion might interfere: the exercise of “prima facie discre-
tion” always depends on the compliance with those other norms. This can be
exemplified with a refinement of our case: imagine that R3 is only addressed to
people with black hair (R3bh). If this is so, then it seems clear that the factor of
comparison (colour of hair) is not suitable for the purposes of the underlying analogy
(in selecting addressees).15 Thus, people in analogous conditions (those that might
be affected by the disease, the suitable factor) are being treated differently: accord-
ingly, R3bh, though within the range of “prima facie discretion”, is all norms
considered unconstitutional for the reason that it breaches the principle of equality.
(x) Being inside the range from P1 to P2, R3bh is prima facie permitted: regardless of other
considerations, it is a way of satisfying P2; however, R3bh contains an unsuitable assessment
of analogy: for the purposes at stake, there are people without black hair also in risk of getting
the disease.
(xi) If equality is “analogue ) O equal treatment”, then it follows that R3bh violates the conse-
quence of equality: those excluded due to the chosen comparison factor ought to be equally
treated; equality works, then, as an internal limit of discretion, removing R3bh from definitive
discretion.

In balancing discretion, proportionality is omnipresent. Given that within the

conflict one principle will be necessarily defeated, it follows that a means-end
scheme always occurs: the sacrifice of the defeated principle is the means to the
end of satisfying the defeater one. This means-end scheme is, exactly, the antecedent
of proportionality.16 Accordingly, any R adopted within “prima facie discretion”
(any R expresses the prevalence of a principle over another) has to be suitable,
necessary and proportional in the narrow sense. If these consequences are respected,
then the R is also within definitive discretion, being an eligible solution (at least

15
For the underlying scheme of analogy, Golding (2018), p. 133; Duarte (2015), p. 144.
16
On the topic, for instance, Clérico (2009), pp. 39, 101; Bernal Pulido (2007), p. 620.
From Constitutional Discretion to the Positivist Weight Formula 17

regarding proportionality). If not, this R is disproportional and, thus,

unconstitutional.
(xii) Within the conflict between P1 and P2, R2 is enacted inside the “prima facie discretion” given
by the conflict; if R2 expresses that P2 has defeated P1, then R2 expresses the sacrifice of P1 in
order to satisfy P2; R2 is, consequently, a means to satisfy the end P2 is.
(xiii) If proportionality is “means-end ) O for the means to be suitable, necessary and proportional
in the narrow sense”, then R2 has to comply with all these consequences; if it does not (if R2 is
not necessary, for instance), then it is out of “all norms considered discretion” (being
disproportional).

3 What About “Epistemic Discretion”?

Beyond the previous framework of discretion, it is common to see references to a

kind of discretion designated as “epistemic discretion”: in the way as it is usually
identified, it is related to the knowledge underlying the choices given by a set of
lawful alternatives.17 Epistemic discretion is regularly divided in two types:
(i) empirical; and (ii) normative. The former regards the factual premises that are
presupposed (or implied) within those alternatives. It is a discretion that has reality as
its object.18 The latter is also knowledge related, but specifically regarding the law:
normative epistemic discretion would be, therefore, the discretion given by the
uncertainties about what is permitted, forbidden or mandatory.19
(xiv) Empirical knowledge related to balancing discretion might be exemplified with the premise
“active ingredient x foreseen in R2 is 90% efficient”.
(xv) Normative knowledge related to balancing discretion might be exemplified with the premise
“R2 is both permitted and forbidden”.

Being related to knowledge, epistemic discretion would be accessed on the basis

of a reliability variation. Regardless of the type of knowledge at hand, empirical or
normative, the “value” of what we know about “something” is said to be more or less
reliable.20 However, it seems that reliability is not the best predication for the quality
of our knowledge: if it covers the consistency of the methods, it seems to disconsider
the truth value of a premise.21 Differently, “certainty” seems to offer a much better
approach. While predicating our knowledge about something as certain, and in order
to dissipate the known «activity-product» ambiguity, we are simultaneously stating

17
For example, Alexy (2014), p. 519; Klatt and Meister (2012), p. 80.
18
Borowski (2013), p. 1418; Klatt and Schmidt (2012), p. 75.
19
Alexy (2002), p. 415; Rivers (2007), p. 177.
20
Alexy (2003), p. 447; Klatt and Schmidt (2012), p. 76.
21
On the discussion between reliability and truth, Olsson (2007), p. 343; Bird (1998), p. 147.
18 D. Duarte

both its trueness and its reliability.22 Therefore, the variation mentioned above has to
be seen as regarding the «certainty» of a premise.
(xvi) Reliable falsity is still falsity; if underlying R2 there is the factual premise “active ingredient
x is 90% efficient”, this would be false if, after serious pharmaceutical research, it is stated as
such; even though reliable, the premise lacks the value of trueness or alike.
(xvii) Unreliable truth is still truth; if underlying R2 there is the factual premise “active ingredient x
is 90% efficient”, this would be reliable if stated after serious pharmaceutical research; but it
would be unreliable, but still truth, if it was just said by John, a football player.

A relevant precision is, however, required. Assuming that the value of our
knowledge is measured with a certainty variation and that certainty entails both
reliability and truth (or alike) does not mean that certainty is being solely opposed to
unreliability or falsity. Since certainty consists in a variation regarding the value of
knowledge, it opposes “justified true belief” to its absence, thereby also describing,
on the other side, the epistemic situation where one does not know whether a premise
is reliable and true or not.23 Therefore, certainty stands for the opposite of ignorance,
and the variation from certainty into uncertainty may be seen as the expression of its
growth: inexistent in the former, dominant in the latter.
(xviii) The premise “active ingredient x is 90% efficient” is certain (c) whenever it is reliable (r) and
true (t); however, the premise is uncertain (u) if we know that it is false or that it is
unreliable; thus, one may say that: “t ^ r ! c”; and “~t _ ~r ! u”.
(xix) The premise “active ingredient x is 90% efficient” is also uncertain, though, whenever we do
not know whether it is reliable and true or not: the more ignorant we are, the more remote we
are from certainty; thus, one may also say that: “~([t _ ~t] ^ [r _ ~r]) ! u”.

With certainty as measure, it can be said that a premise might be certain, uncertain
or something in between in a sort of Soritian variation. With a scale for defining
magnitudes, we may subsequently say that our knowledge of something has some
degree of certainty within the items of such scale. But if this is the case, then there is
a problem: what might be similar or even related to discretion if the point is just to
assess which degree of certainty underlies our knowledge ? At first sight, nothing of
this sort can be envisioned here: no alternatives are visible and no choice ascertains
(merely an object to be measured). Well, it seems that a second look is needed,
specifically separating empirical and normative knowledges.

3.1 Empirical Epistemic Discretion

In order to understand the domain of the factual in balancing discretion, it is relevant

to remember that, when enacting a R, some losses in one principle and some gains in

22
Obviously, truth entails all its “parallels”, such as truthlikeness (verisimilitude), probability
(objective chance), or inference to the best explanation (best competing explanation). See, respec-
tively, Niiniluoto (2004), p. 538; Bird (1998), p. 123; Lipton (1991), p. 56.
23
On justified true belief, Shope (2004), p. 285; Bird (1998), p. 156.
From Constitutional Discretion to the Positivist Weight Formula 19

the contradictory one will be obtained. Those losses and gains following this R are,
exactly, the result of its consequence. If the R consists of an obligation of general
vaccination, then it is its consequence () O vaccinate) that defines which are the
losses and gains. Therefore, in balancing discretion, underlying empirical premises
are all the factual premises that sustain the losses and gains the consequence of a R
imply into the principles at hand. The antecedent of a R [p in the following
illustration (xx)] is, for this matter, irrelevant.
(xx)

P1 ------------------------------------------------------------- P2
R2 (p O vaccinate with active ingredient x)

(xxi) the scheme in (xx) shows that it is the “how” the consequence of R2 implies gains in P2 and
the “how” it implies losses in P1 that establishes the universe of what might be called as
“underlying empirical premises” of this balancing discretion.

Simultaneously, it is visible that R2 has effects both on P1 and on P2. Regarding

the latter, this principle will be satisfied by R2 depending on some factual premises:
for instance, the degree of efficiency of active ingredient x or the possible secondary
health effects provoked by that ingredient. The same is valid for the losses in the
defeated principle: for instance, the amount of pain implied by the vaccine admin-
istration or even the eventual skin complications related with the type of needle used.
Thus, there are factual premises specifically related (and solely based on R2’s
consequence) with the gains in P2 and with the losses in P1. Any other considerations
out of this factual spectrum are irrelevant.
(xxii) Examples of underlying empirical premises in connection with P2 might be: a) “the grade of
efficiency of active ingredient x is 90%”; b) “there are no secondary health effects”.
(xxiii) In connection with P1: c) “the amount of pain with vaccine administration is minimal”; and
d) “a common needle is used without any kind of differentiated complications”.

We can think about these four premises {a, b, c, d} as the underlying premises
intrinsically related with R2 (obviously, competing alternatives, such as R1 or R3,
have other underlying factual premises). Each one of these four premises represents
a specific statement about the world, consisting of a description about a given state of
affairs. Regarding each one, we may have different degrees of certainty, in the way
mentioned above. Irrespective of everything else, each premise brings with it a
specific degree of certainty, expressing how much certain one can be about
it. Together, we can have an overall degree of certainty on the premises related to
a specific P: the degree of certainty of {a, b} to P2, for instance.24

24
The (negative powers) double triadic scale will be used in the next example. Later on, a more
precise explanation for this scale will be given. On the other hand, and just for the sake of simplicity,
it is assumed that each premise contributes with the same weight to the overall degree of certainty
given to the premises related to a specific P. However, and obviously, this is just a simplification
(it is possible but not necessary).
20 D. Duarte

(xxiv) We can have 2-1 ( 12 ) of certainty on premise a), 2-3 ( 18 ) on premise b), 2-2 ( 14 ) on premise c),
and 2-4 ( 16
1
) on premise d); therefore, this means certainty 2-2 ( 14 ) on the premises
underlying the losses in P1 and certainty 2-3 ( 18 ) on the premises underlying the gains in P2.

As it seems visible, there is nothing of discretion here, mainly when one bears in
mind that this all comes from balancing discretion. The alternativity between R2 and
the other Rs is already behind, no alternativity exists between these premises and
others, and each one of them stands for a knowledge with a specific (amount of)
value. In straight words, knowledge of the world is not selectable. As far as one can
see, the sole possibility of a “discretion resemblance” here appears whenever there is
uncertainty about the degree of certainty: for instance, if it is uncertain how certain
premise a) is. But, again, being this the case, then empirical epistemic discretion
would be just a choice on competing measurements of knowledge.
(xxv) In (xxiv) it was conceived that we can have 2-1 ( 12 ) of certainty about premise a); if we are
uncertain about that 2-1 ( 12 ), we probably consider the possibility of a variation between 2-1 (
1 -3 1
2 ) and 2 ( 8 ); we now have alternatives, giving a setting similar to the one used to describe
discretion.
(xxvi) However, and considering that these alternatives are just different possibilities of how
certain we are [from 2-1 ( 12 ) to 2-3 ( 18 )], this would show empirical epistemic discretion
as no more than a «meta-certainty» choice (thus, not at all clear on how it would influence
the choice of R2).

3.1.1 Empirical Epistemic Discretion in Klatt’s Approach

A very precise explanation about the content of empirical epistemic discretion is

given by Klatt; an explanation that, moreover, goes exactly on the line of connecting
such discretion with measuring.25 Klatt’s main point is that the real epistemic
discretion is discretion of classification, understanding this, and roughly speaking,
as the mental action by which one assigns a value to «something». Obviously, Klatt
is considering the Weight Formula and, now just for the empirical domain, presents
discretion of classification within the problem of classifying an intensity of interfer-
ence while facing empirical uncertainties.26 Using our case, it regards the setting of
empirical uncertainty when we have to define losses or gains in P1 and P2.
(xxvii) in order to quantify the losses in P1 and the gains in P2 implied by R2 we depend on the
certainty of the factual premises: when uncertain, ignorance prevents us to assign a specific
value: if we do not know how effective active ingredient x is, we do not how much health is
being promoted (P2).
(xxviii) the uncertainty about these losses and gains (intensities of interference) based on factual
uncertainty gives rise to Klatt’s (empirical) discretion of classification: the choice between
different possible amounts of losses and gains in P1 and P2 coming from uncertain
empirical premises.

25
On the present topic, Klatt writes with Meister (2012) and Schmidt (2012). However, and for the
sake of simplicity, only his name is mentioned in text (differently from footnotes, of course).
26
Klatt and Meister (2012), p. 110; Klatt and Schmidt (2012), p. 75.
From Constitutional Discretion to the Positivist Weight Formula 21

Regardless of everything else, Klatt points out an unavoidable problem: how can
we measure losses and gains in P1 and in P2 (which is decisive for balancing) if we
are uncertain about the (empirical) consequences of R2 in those principles? Clearly
not discussing competing measurements of empirical knowledge, Klatt presents the
problem as the consequence of empirical uncertainty into the classification of losses
and gains in the principles at hand.27 It is a problem no one has addressed yet and a
crucial one for balancing discretion: the prevalence of P2 over P1 depends on their
corresponding losses and gains and these, on the other hand, are totally dependent on
a solution for empirical uncertainty.
(xxix) Applying the double triadic scale both to the degree of certainty about an empirical premise
(negative powers) and to the amount of losses and gains in P1 and P2 (positive powers), Klatt
is not discussing if the degree of certainty we have about premise a) is 2-1 (12) or 2-2 (14).
(xxx) Differently, Klatt is pointing out into the problem of how can we assign 27 (128) of gains in
P2, for instance, if premise a) has been assessed with some degree of uncertainty; for
instance, how can we assign 27 (128) of gains in P2 when premise a) has 2-5 ( 32 1
) of
empirical certainty.

In order to solve the problem, Klatt created his “classification balancing”, a

scheme of the balancing type where two different possible degrees of losses or
gains in a specific P are weighted (upon an empirical uncertainty scenario, obvi-
ously). For the purpose, Klatt assumed that the most certain intensity has to be
balanced against the most pessimistic (which seems to be totally correct): by doing
so, we will obtain as quotient a specific degree of interference (loss or gain) that
follows from the Law of Classification.28 Klatt explains the equation as a tool with a
heuristic function regarding the Weight Formula: the quotient of classification
balancing, with an amount of losses or gains in a P, is then inserted in that equation.
(xxxi) If we do not know, regarding premise c), how much pain is caused by vaccination, we do
not how much physical integrity is being harmed (P1); however, we can conceive that the
most pessimistic interference is 4 (with 14 of certainty) and that the most certain is 1 (with 12 of
certainty).29
(xxxii) With these amounts, the classification of P1 to be inserted in the Weight Formula is 4: the
most pessimistic intensity (denominator) prevails over the most certain (numerator):

1 12 0, 5
CP1 ¼ ¼ ¼ 0, 5
4 14 1

Klatt’s classification balancing is an indispensable and extremely sophisticated

tool: without it we cannot use the Weight Formula in cases of empirical uncertainty;

27
Klatt and Meister (2012), p. 113; Klatt and Schmidt (2012), p. 77.
28
That reads as follows: “the more reliable a more intensive classification of intensity of interference
is, the more reliable must be a less intensive classification of intensity of interference”. Klatt and
Meister (2012), p. 118; Klatt and Schmidt (2012), p. 81.
29
Using now the single triadic scale as Klatt does. Klatt and Meister (2012), p. 120; Klatt and
Schmidt (2012), p. 82.
22 D. Duarte

moreover, the practical problem of assigning an interference to a P in an empirical

uncertainty situation seems to have no other solution. Nevertheless, there are good
reasons to think that, within the engineering of classification balancing, and in order
to make it totally operative, two refinements must be carried out. First, regarding the
possibility of certainty on the intensity of interference with empirical uncertainty.
Second, regarding the underlying idea that the single triadic scale (and without any
constraints) is suitable to classification balancing.
First, Klatt considers two situations of uncertainty: (i) uncertainty on empirical
premises with certainty on the intensity of interference; and (ii) uncertainty on
empirical premises with uncertainty on the intensity of interference.30 However,
the first seems inexistent: if we do not know how the world is regarding the effective
consequences of a R, how can we know how much those consequences will affect a
specific P? Probably, this misstep is due to the fact that Klatt goes beyond the
spectrum of the factual domain relevant for balancing: as it is visible with his own
example, he uses an irrelevant premise.31 But if one stays within the spectrum of
relevant premises, uncertainty on the intensity will always be the case.
(xxxiii) Klatt uses as an example a rule taken from the Aviation Act, a R solving a conflict (and for
simplicity sake) between different holders of the right to life given by a P (Pi):

Pi holder y --------------------------------------------------------- Pi holder x

Rz
(Rz: if the aircraft is a danger for civilian lives O to shoot the
aircraft).

(xxxiv) Klatt analyzes whether the antecedent of Rz is or not fulfilled on a case with Rz; the point is
that this uncertainty does not belong to the factual spectrum of the consequences of Rz in
Pi; it comes from the vagueness of the word “danger” and it has nothing to do with
balancing (is on Rz applicability).
(xxxv) Bearing in mind that “underlying empirical premises” of balancing are just those coming
from the consequences of a R on a P (here Rz on Pi), it follows that classification balancing
has to limit itself to the set of empirical premises effectively relevant for balancing.

Second, Klatt uses the single triadic scale with the following correspondences:
(i) 20 ¼ 1 ¼ certainty; (ii) 21 ¼ ½ ¼ average certainty; and (iii) 22 ¼ ¼ ¼
uncertainty. This scale, even though not being a review scale, which is a good move
from Klatt, is unusable.32 On the one hand, because classification balancing depends
on uncertainty of interference caused by uncertainty about the underlying premises:

30
Klatt and Meister (2012), p. 114; Klatt and Schmidt (2012), p. 77.
31
Klatt and Meister (2012), p. 113; Klatt and Schmidt (2012), p. 75.
32
For the scale, Klatt and Meister (2012), p. 112; Klatt and Schmidt (2012), p. 76. Not being a
review scale means that it is not dependent on (or strictly usable by) a reviewer body.
From Constitutional Discretion to the Positivist Weight Formula 23

if this is so, then no premise can be classified as certain (20 ¼ 1).33 On the other hand,
because if there is (total) uncertainty (22 ¼ ¼), then it is almost guaranteed that the
R is unsuitable in the first place: the R at hand most probably had already failed in the
first (eliminatory) test of proportionality.
(xxxvi) We can conceive, regarding how much active ingredient x is effective [premise a)], that
the most pessimist interference of R2 in P2 (not promoting it) is 1 (with 14 of certainty) and
that the most certain interference of R2 in P2 is 2 (with 1 of certainty).
(xxxvii) Then: (i) interference 1 in P2 based on 14 of certainty expresses a degree of ignorance that
points out into the very unsuitability of R2; (ii) interference 2 in P2 based on 1 of certainty
expresses inexistent uncertainty on the factual premise; classification balancing is dis-
pensable in both cases.

Both difficulties emerging from the used scale imply that Klatt is condemning
classification balancing to just one measure: average certainty (21 ¼ ½). As it
seems clear, following the reasons given, the single triadic scale is not appropriate
for classification balancing: a tool designed for uncertainty cannot contain a value for
(total) certainty, in the same way that, given the wider context of proportionality
(where it still stands), it cannot include a value for (total) uncertainty. For these
reasons, classification balancing has to adopt the double triadic scale and, more than
that, has to eliminate the higher and the lower degrees. Only with such scale
classification balancing becomes coherent and operative.
(xxxviii) The following correspondences may be given to the double triadic scale, first for certainty
in classification balancing and second for gains and losses in interferences34:

– «high maximum certainty» (1 ¼ 20 ¼ 1) – high maximum gain or loss (256 ¼ 28)

– high moderate certainty ( 12 ¼ 2-1 ¼ 0,5) – high moderate gain or loss (128 ¼ 27)
– high light certainty ( 14 ¼ 2-2 ¼ 0,25) – high light gain or loss (64 ¼ 26);
– medium high certainty ( 18 ¼ 2-3 ¼ 0, 125) – medium high gain or loss (32 ¼ 25);
– medium moderate certainty ( 16
1
¼ 2-4 ¼ 0,0625) – medium moderate gain or loss (16 ¼ 24)
– medium light certainty ( 1
32 ¼ 2 ¼ 0, 03125)
-5
– medium light gain or loss (8 ¼ 23)
– low high certainty ( 64
1
¼ 2-6 ¼ 0, 015625) – low high gain or loss (4 ¼ 22)
– low moderate certainty (128
1
¼ 2-7 ¼ 0, 0078125) – low moderate gain or loss (2 ¼ 21)
– «low light certainty» ( 1
256 ¼ 2 ¼ 0, 00390625)
-8
– low light gain or loss (1 ¼ 20)

(xxxix) «High maximum certainty» and «low light certainty» have to be blocked; they represent
total certainty and (almost) total uncertainty, values that do not trigger classification
balancing.
(xl) With the double triadic scale, the classification balancing carried out above in (xxxii) will
be as follows (mathematically equivalent):

33
In other words, certainty means that classification balancing is dispensable.
34
It must be said that the values with negative powers are just to classification balancing (and
Alexy’s Weight Formula). To the Positivist Weight Formula, as it will be seen later on, only the
values with positive powers are needed (and regarding the measurement of “certainty” one only has
to substitute “gain or loss” by “certainty”).
24 D. Duarte

4 18 0, 5
CP1 ¼ 1
¼ ¼ 0, 5
16 16 1

The utility of classification balancing for balancing discretion is undeniable. As

said, it is indispensable for assigning amounts of losses and gains in situations of
empirical epistemic uncertainty. However, Klatt also uses classification balancing to
frame his account of empirical epistemic discretion: to him, this discretion (named as
“discretion of classification”) is present whenever we obtain a stalemate as the
quotient of the classification formula.35 With a quotient such as this one, it is
irrelevant which intensity will be inserted in the Weight Formula: both possibilities
(the most pessimistic and the most certain) lead to the same result and, for this
reason, assigning one or the other is just a matter of choice.36
(xli) A stalemate can be exemplified with a most pessimistic intensity of 16 with medium moderate
1
certainty ( 16 ) and a most certain intensity of 4 with high light certainty ( 14 ):

4 14 1
CP1 ¼ 1
¼ ¼1
16 16 1

(xlii) From this stalemate it follows that inserting 4 or 16 in the Weight Formula is just a matter of
choice: under Klatt’s approach, it expresses the sole situation where our empirical uncer-
tainties give raise to two different, but equally admissible, intensities of interference.

It is not convincing, though, that classification balancing provides a good account

of what empirical epistemical discretion might be. Observing carefully what the
stalemate outcome stands for, it becomes visible that it states no more than the
equality between two hypotheses of intensity with their respective amounts of
uncertainty. It gives, accordingly, a choice among intensities of interference, saying
nothing on “how much” certain we are about the “certainty” of our empirical
premises.37 Furthermore, while exclusively presenting that choice, the stalemate
also presupposes exact amounts of uncertainty: for this reason, the sole variation
that could give sense to a «proper» empirical discretion also disappears.38
(xliii) As seen above, a stalemate occurs with an intensity of 16 with medium moderate certainty
1
( 16 ) and an intensity of 4 with high light certainty ( 14 ); the choice given is, therefore, between

35
Klatt and Meister (2012), p. 119; Klatt and Schmidt (2012), p. 82.
36
Klatt and Meister (2012), p. 120; Klatt and Schmidt (2012), p. 83.
37
It has to be stressed that, with the stalemate, what is given is a choice on intensities of interference
(bringing with them specific uncertainties). Thus, classification balancing does not deal directly
with empirical uncertainty.
38
Given that there are two alternative uncertainties (because they come with the most pessimistic
interference and with the most certain one), with the stalemate we will insert in variable Re of the
Weight Formula one of those degrees of uncertainty (the one corresponding to the intensity chosen);
then, the Weight Formula will always receive a «certain» amount of uncertainty independently from
how uncertain we are about the degree of uncertainty.
From Constitutional Discretion to the Positivist Weight Formula 25

inserting 16 or 4 into the Weight Formula as the degree of intensity of interference of R2

in P1.
(xliv) In a scenario such as this one there is nothing to be chosen within the realm of the factual; if it
1
is exactly the variation between (16 ) and ( 14 ) that could give some sense to «empirical
discretion», its combination with intensities makes that variation irrelevant.

Stalemates in classification balancing do not seem to be, thus, a satisfactory

explanation for the so called empirical epistemic discretion. In balancing discretion,
uncertainty about empirical premises is a matter of degree when one has to measure
the quality of knowledge, weakening the value of our assessments of losses or gains
while progressing into higher degrees of ignorance. Nevertheless, a degree or
another is not a matter of choice: it is what it is within a given scale. Accordingly,
something resembling discretion will be present here only if an amount of empirical
certainty is legally required and if there is uncertainty about it. Only this would make
“empirical discretion” a meaningful token of discretion.

3.2 Normative Epistemic Discretion

Normative epistemic discretion is usually described as the discretion regarding the

certainty of the normative premises underlying balancing discretion. For instance,
Alexy says that normative epistemic discretion ascertains when it is unknown which
is the best weight or once there is uncertainty about the result of balancing.39
Differently, both Klatt and Rivers assign to the concept the uncertainties regarding
the classification of intensities of interference and abstract weight of principles.40 For
them, normative epistemic discretion arises in any setting providing uncertainty on
the value to be inserted in variables I and W of the Weight Formula: here, a choice
between alternative values has to be carried out.41
(xlv) For Alexy, normative epistemic discretion is the uncertainty about the outcome of balancing;
in other words, and with the Weight Formula, when we have ignorance about which P
prevails (the superscript «v» shows the locus of discretion [variation]):

I W Re Rn
Wv ¼
I W Re Rn

(xlvi) For Klatt and Rivers, normative epistemic discretion is the uncertainty about intensities of
interference and abstract weights; with the Weight Formula, the choice within variables I and
W (the superscript «v» shows the locus of discretion [variation]):

39
Alexy (2002), pp. 415, 421.
40
Klatt and Schmidt (2012), p. 87; Rivers (2007), p. 178.
41
Klatt and Schmidt (2012), p. 88; Rivers (2007), p. 178.
26 D. Duarte

Iv Wv Re Rn
W¼
Iv Wv Re Rn

Alexy’s proposal for the content of normative epistemic discretion is not really
convincing: being an arithmetic operation with a necessary precise outcome, it is
difficult to conceive what is doubtful about the result of the Weight Formula.42 On
the other hand, Klatt’s and River’s solutions face the same problems previously
assigned to empirical epistemic discretion: alternative values on the quality of our
knowledge, now regarding how much a R will interfere with a P, are relevant if there
is some legal requirement to measure. Therefore, confining this discretion to alter-
native values of I and W seems insufficient: it says nothing about why we should
have a precise amount of losses and gains in the first place.
(xlvii) Alexy’s answer seems unsatisfactory because the outcome of balancing is necessarily
certain; once values are inserted into the Weight Formula, there is always one mathematical
outcome:

2 2 1 1
W¼ ¼2
1 2 1 1

(xlviii) Klatt’s and Rivers’ answers are (partially) unsatisfactory: even if recognizing a variation
similar to discretion, it is not known which is the legal basis for that variation; actually,
values on intensities of interference and abstract weights are supposed to have some legal
support.

3.3 Understanding Epistemic Discretion Through the Laws

of Balancing

All the skepticism here expressed regarding epistemic discretion, whether empirical
or normative, is mainly due to the fact that its common scientific explanation
confuses two different things: (i) balancing discretion in itself (and all information
associated, the principles at hand and their underlying premises); and (ii) the obli-
gation to respect proportionality and, inside its frame, to comply with the two laws of
proportionality in the narrow sense. It is, then, a confusion between the exercise of
balancing discretion and the (posterior) definition of one of its own internal limits.
Therefore, when they are separated, it follows that, before applying proportionality,
law does not give any relevance to the value of knowledge.
(xlix) As seen before, at the level of “prima facie discretion”, nothing more exists than the
contradictory principles and the choice to be carried out among the alternatives present:

42
This criticism already in Klatt and Schmidt (2012), p. 87.
From Constitutional Discretion to the Positivist Weight Formula 27

P1 --------------------------------------------------------------------- P2
R1 R2 R3

At an initial moment, the exercise of balancing discretion is independent from the

certainty of underlying empirical and normative premises. When we choose between
R2 and R3 we have «prima facie discretion» whether we are more or less certain
about those premises: at this level, we may choose R2 regardless of total ignorance
about the efficiency of its active ingredient or about the amount of losses in P1.
Everything becomes different when proportionality jumps into the stage: when the
principle is triggered, both laws of balancing remove options with “some” degree of
ignorance: they state that beyond a specific limit of uncertainty the choice of R2 (for
instance, if it is the case) may be effectively inadmissible.
(l) Accordingly, R2 passes from a (prima facie) selectable option into a (all norms considered)
non-selectable one when, for reasons related to proportionality (as an internal limit of
discretion), it is known that its underlying premises have an inadmissible degree of ignorance.

This becomes understandable once the laws of balancing are analyzed.

The substantive law imposes that “the more the losses in a P, the more the gains in
the opposite P”. In turn, the epistemic law imposes that “the more the losses in a P,
the more the certainty about the underlying premises”. As it is visible, both laws
have a similar structure: (i) the first connects a consequence of equal and higher gains
in a P regarding the losses in another; and (ii) the second connects a consequence of
equal and higher certainty about the underlaying premises regarding the losses in a
P.43 Taken together, both laws presuppose for their own application accurate
amounts of “losses”, “gains”, and “certainties”.
(li) The substantive law is “the more the losses in a P, the more the gains in the opposite P”; thus,
it is a normative sequence such as: “if losses in a P ) O gains  in the other P”.
(lii) The epistemic law is “the more the losses in a P, the more the certainty about the underlying
premises”; it is a normative sequence such as: “if losses in a P ) O certainties  losses”.
(liii) Both laws can be expressed in a single sentence: «if there are losses and gains and uncer-
tainties ) O gains  losses ^ certainties  losses».
(liv) Both laws require (and presuppose) the measurement of losses, gains and certainties; the
underlying ratio (of losses and gains, for instance) depends on those measures.

All this explains why assigning values for losses, gains, and certainties is no more
than making the assessments both laws presuppose for their application and,

43
There has been some divergence on the wording of both laws of balancing, specifically regarding
the epistemic one. For instance, Julian Rivers formulates the second law as follows: “the more
serious a violation of a right is, the greater must be the reliability of the legislature’s assessment that
a competing interest will be realized to a sufficiently great extent” (2007, p. 183). Matthias Klatt and
Johannes Schmidt put it as follows: “the more serious an interference with a principle Pi is, the more
certain must be those premises that justify the classification of intensity of interference Ii” (2012,
p. 85). No doubt about the importance of the wording. However, as a part of proportionality, the
main point is that its core has to be intersubjectively defended from specific theories or perspectives.
Thus, it is with the purpose of preserving that core that such an aseptic wording (as the one above)
was conceived.
28 D. Duarte

subsequently, for the application of their consequences: without a given amount of

losses, gains and certainties it is impossible to apply proportionality in the narrow
sense. Hence, it becomes visible: (i) that the value of knowledge solely becomes
legally relevant when proportionality is applied; and (ii) that the so called epistemic
discretion is no other than the discretion coming from the comparative quantifier “the
more” related to “losses”, “gains” and “certainties”.44 It is, therefore, a discretion
given by the linguistic expression of both laws of balancing.
(lv) Looking into the wording of their common structure “the more X, the more Y” (that expresses
“if there are losses and gains and uncertainties ) O gains  losses ^ certainties  losses”), we
see that it implies the comparative “quantification” of “losses”, “gains” and “certainties”.
(lvi) Therefore, to know whether there is a loss and into what extent in a specific P (caused by the
specific consequence of a R) stands for the instantiation of the comparative quantifier «the
more»; the same is valid, obviously, for gains and certainties.

What has been said can also refine the connection between balancing discretion
and the alleged epistemic one, effectively a “discretion of quantification”.45 Hence,
while balancing discretion is the area of choice given by a conflict of norms not
solved by a third norm, discretion of quantification is a kind of discretion of
linguistic nature given by an internal limit of balancing discretion: proportionality
in the narrow sense. As it is typical to discretion coming from vague quantification,
this discretion only emerges when there is uncertainty about the measure: if amounts
of losses or gains are evident, then there is no discretion at all; the only discretion
present is, thus, the one in borderline cases of measurement.46
(lvii) As seen, balancing discretion can be illustrated with the line of opposite principles among
which there is a choice to be carried out (R1 or R2 or R3;); it comes from the conflict between
P1 and P2:

P1 --------------------------------------------------------------------- P2
R1 R2 R3

(lviii) Differently, discretion of quantification comes from the quantification that is required for
amounts of “losses”, “gains” and “certainties” by the expression «the more»: it is given by
(and it belongs to) an internal limit of balancing discretion, proportionality in the narrow
sense.

44
On “the more” as a quantifier, Lassiter (2009), p. 40. Also, Sauerland and Stateva (2011), p. 125.
45
Because it turns out that such a thing as «epistemic discretion», understood as the variation
regarding what we know about one possible world at some specific time, does not exist: in a space-
time point, we know what we know. What exists, instead, is the discretion related to the measure-
ment of how much we do know (this is, the quantification of losses, gains and certainties).
46
Being this a coherent outcome with what has been previously said: if our knowledge is what it is,
the sole possibility of discretion in the epistemic realm comes from the uncertainties about the
“quantification” of that knowledge.
From Constitutional Discretion to the Positivist Weight Formula 29

(lix) However, if R2 implies (under the substantive law) 27 (128) of gains in P2 and 23 (8) of losses
in P1, and these amounts are evident, then a clear quantification of “losses” and “gains”
ascertains (not a borderline case): here, the first law does not give any discretion of
quantification.

4 From Necessity (in Proportionality) to Alexy’s Weight

Formula

As claimed above, proportionality is a contingent norm of a legal order: indepen-

dently from its source, it either exists or not in a specific set of norms.47 Whenever a
legal order entails a norm such as proportionality, it follows that a precise content is
recognized: it imposes, for means-end normative connections, that the means ought
to be suitable, necessary and proportional in the narrow sense. From these three
consequences, three proportionality tests follow: (i) if R2 is suitable to satisfy P2;
(ii) if R2 is necessary (for instance, when compared with R3); and (iii) if R2 complies
with both laws of balancing. Following this order, these tests are also eliminatory:
failing one implies immediate disproportionality.48
(lx) The first test of proportionality is suitability; thus, we have to see if R2 is appropriate to satisfy
P2; accordingly, if the active ingredient foreseen in R2 is 0% effective, then R2 is not suitable;
this means, evidently, that R2 is immediately disproportional and, thus, unconstitutional.
(lxi) This eliminatory effect, now with the suitability test, also means that the following test (here,
necessity) becomes irrelevant (or, perhaps better said, useless): no matter what might be the
outcome of the second test, the unconstitutionality of R2 is already irreversible.

Necessity follows suitability. In necessity, an evaluation is required in order to

assess whether a R, in comparison with other possible Rs on the matter, satisfies the
condition of being “necessary”. Here, the initial problem is that the necessity of a R
might have, at least, three different meanings: (i) the R with more gains than the
others, with equal losses; (ii) the R with less losses than the others, with equal gains;
and (iii) the R with the better ratio of gains and losses.49 One may call these three
meanings of necessity, and respectively, “gains necessity”, “losses necessity”, and
“ratio necessity”.50 As it is perceptible, the third one is more demanding (strong
necessity) than the first and the second (weak necessity).

47
It is important to note that a legal order entails proportionality not only when explicitly stated in a
constitution, but also when it has been produced as a customary norm (the cases of Germany or
Brasil). On the topic, Barak (2012), p. 211; Bernal Pulido (2007), p. 505.
48
Klatt and Meister (2012), p. 71; Bernal Pulido (2007), p. 693. For a different perspective, Clerico
(2018), p. 32.
49
On different notions of necessity, Sartor (2010), p. 198. Generally, Barak (2012), p. 326.
50
Of course, there are also ratios in “gains necessity” and in “losses necessity”. However, the
purpose here is just to isolate the cases where there are no equal gains or equal losses (and name
them as “ratio necessity”).
30 D. Duarte

(lxii) In “gains necessity”, the necessary R is the one with more gains: if R2 has 25 (32) of gains in
P2 and 24 (16) of losses in P1 and R3 has 24 (16) of gains in P2 and 24 (16) of losses in P1, then
R2 is “gains necessity” necessary; R3 is, consequently, unnecessary.
(lxiii) In “losses necessity”, the necessary R is the one with less losses: if R2 has 26 (64) of gains in
P2 and 24 (16) of losses in P1 and R3 has 26 (64) of gains in P2 and 25 (32) of losses in P1, then
R2 is “losses necessity” necessary; R3 is, consequently, unnecessary.
(lxiv) In “ratio necessity”, the necessary R is the one with the better ratio: if R2 has 26 (64) of gains
in P2 and 23 (8) of losses in P1 and R3 has 25 (32) of gains in P2 and 24 (16) of losses in P1,
then R2 (23 ratio) is “ratio necessity” necessary; R3 (21 ratio) is, consequently, unnecessary.
(lxv) “Gains necessity” and “losses necessity” express a “weak necessity”: they just impose as
necessary a R with more gains and equal losses or less losses with equal gains; “ratio
necessity” expresses a “strong necessity”: it imposes the more demanding limit of the better
ratio.

It is well known in legal science (and in the legal practice as well) that these
meanings have been mixed: generally, necessity is understood whether as “gains
necessity” or as “ratio necessity”.51 However, since we are dealing with an internal
limit of discretion (establishing a limit beyond which a R is disproportional), an
intersubjectively shared unique level of limitation would be seriously desirable.
Despite its attractiveness, it seems that “ratio necessity” is too demanding as a
limit to the decision-maker’s options: why should it be unconstitutional to choose
the R with less losses even if it has a worse ratio ? Whenever preventing losses is
somehow crucial, nothing justifies such R to be disproportional.52
(lxvi) If R2 has 27 (128) of gains in P2 and 24 (16) of losses in P1 and R3 has 25 (32) of gains in P2
and 23 (8) of losses in P1, then R2 has a better ratio: 23 > 22; however, R3 has less losses: 8 <
16; if necessity is «ratio necessity», then choosing R3 leads to an unconstitutionality.
(lxvii) However, necessity is applied to each and every conflict of principles, which means that a R
similar to R3 would always be unconstitutional (in any scenario); forbidding the decision-
maker to promote less losses (no matter the field at stake) makes proportionality a legislative
blocker.

Taking all this into account, it seems that necessity has to be seen as solely
requiring a weak limit (weak necessity): this is, to qualify as necessary the R with
more comparative gains or the R with less comparative losses.53 Accordingly,
whenever R2 and R3 have the same gains, the R with lesser losses is necessary,
being necessary the R with more gains whenever R2 and R3 have the same losses. It
stays unsolved, however, the ratio case: to know which R is necessary when both
have distinct amounts of losses and gains with unequal ratios.54 In situations such as
these, it seems, though, that necessity should not reproach any of the alternative Rs:
without any evident «superiority», none of them is clearly unnecessary.

51
Sartor (2018), p. 628; Pinker (2013), p. 130; Klatt and Meister (2012), p. 79.
52
Similarly, Sartor (2018), p. 629.
53
But it seems mistaken to sustain, nonetheless, that necessity regards the R that is less intrusive in a
right (Rivers 2007, p. 171). The simple situation of a collision between two rights shows this
formula to be empty. See, for instance, Sieckmann (2012), p. 220; Pinker (2013), p. 30.
54
Because with equal ratios, the solution is undisputed: necessity never implies disproportionality.
From Constitutional Discretion to the Positivist Weight Formula 31

(lxviii) Therefore, necessity (under the present understanding) only eliminates Rs with equal
amounts of gains or losses whenever one R has, respectively, less losses and more gains;
“gains necessity” and “losses necessity” can be conjointly applied in the second test of
proportionality.
(lxix) Regarding Rs with different losses and gains but the same ratio (R2 has 26 (64) of gains in P2
and 25 (32) of losses in P1 and R3 has 25 (32) of gains in P2 and 24 (16) of losses in P1 [21 ¼
21]), necessity does not interfere: the decision-maker may choose one or the other.
(lxx) Regarding Rs with unequal ratios (R2 has 27 (128) of gains in P2 and 24 (16) of losses in P1
and R3 has 25 (32) of gains in P2 and 23 (8) of losses in P1 [23 > 22]), necessity should not
interfere as well: the intrinsic value of less losses should also allow the decision-maker to
choose.

It sounds quite a strong statement, then, to say that necessity asks for Pareto
optimality (when a R makes one P better off without making the conflicting P worse
off).55 Rigorously, it depends on what is understood by necessity: for instance, “ratio
necessity” has no necessary connection with Pareto optimality. Actually, when the
Rs in comparison have different gains, losses and ratios, Pareto optimality is
occasionally not even possible: whenever one of the Rs does not have more gains
and less losses.56 However, things change when it is assumed that necessity stays off
ratio cases (as previously suggested): only with such premise it becomes true that
necessity implies the disproportionality of non-optimal Rs.57
(lxxi) With “ratio necessity” (and unequal ratios), Pareto optimality might be impossible; if R2 has
27 (128) of gains in P2 and 24 (16) of losses in P1 and R3 has 25 (32) of gains in P2 and 23
(8) of losses in P1 (23 > 22), then R2 makes P2 better off but P1 worse off and R3 makes P2
worse off but P1 better off.
(lxxii) However, in all remaining cases where necessity interferes, the necessary R expresses a
Pareto improvement: R3 is Pareto inferior when has 26 (64) of gains in P2 and 25 (32) of
losses in P1 and R2 has 26 (64) of gains in P2 and 24 (16) of losses in P1; R2 is, therefore,
Pareto optimal.

It is also worth mentioning (within necessity) that consideration for the propor-
tionality test is limited to Rs that are possible. For instance, it has no place in
necessity the imaginary (but impossible) R that fully satisfies P2 without any kind
of loss in P1: given their inverse satisfiability (gains in one P imply losses in the other
P), it is impossible to satisfy both into their maximum extent.58 Since possibility
evidently means empirical feasibility, this becomes particularly relevant in what
regards the vaccination case. It seems clear that, here, only the Rs that are feasible

55
For instance, Alexy (2010), p. 8; Petersen (2017), p. 38.
56
With a good example for this claim, Sartor (2010), p. 198.
57
Which seems to be, thus, the main reason say that necessity is the strongest test among all given
by proportionality. On this test as the strongest one, Klatt and Meister (2012), p. 90; Petersen
(2017), p. 100. Specifically on the problems posed by Pareto optimality in balancing, Chapman
(2013), p. 267.
58
On this inverse realizability, Sieckmann (2018), p. 5; Alexy (2017), p. 37.
32 D. Duarte

can enter the necessity test: it has no place in necessity an imaginary (but unfeasible)
R that fully eliminates the spread of the disease (not known yet).59
(lxxiii) Within the present conflict between P1 and P2 only two Rs are empirically feasible: R2 and
R3; necessity is confined to them; in the rare cases where all solutions have no feasibility
constraints (perhaps, public health against the freedom of assembly, where for the sake of
the former any R limiting the latter seems possible), necessity (theoretically) includes all
alternative Rs.

A final observation regarding necessity is still required: this second proportion-

ality test is the sole one where a comparison among different Rs takes place. It was
already seen that in suitability only the R chosen by the decision-maker (R2) is being
tested: it strictly regards the assessment of whether such R (and no other than this
one) is appropriate to satisfy its end (here P2). Nonetheless, the same applies to
proportionality in the narrow sense. Given that only one R surpasses the necessity
test (at least by choice), it is just this one R that will be the object of the following
test.60 In the final proportionality stage, the point is just to give an answer to the
question of whether the necessary R is also “balanced” or not.
(lxxiv) In general terms, necessity leads to one of two outcomes: (i) one R is necessary and the
others are unnecessary; and (ii) more than one Rs are necessary (whenever gains and losses
are different and ratios are equal or unequal); obviously, necessity is irrelevant when only
one R is feasible.
(lxxv) However, when more than one Rs are necessary, the decision-maker will choose one R, the
R that moves to proportionality in the narrow sense; therefore, notwithstanding what
happens in the necessity test, in the next one only a specific R is being tested.

Accordingly, when R2 surpasses the necessity test, it follows that R2 is compar-

atively a better option than R3. However, R2 has to face now its final test: propor-
tionality in the narrow sense or, what is the same, the two laws of balancing.
Although the content of these laws has already been presented, three other aspects
need to be seen: (i) their connection with indifference curves; (ii) their connection
with optimality, and (iii) their eliminatory character. In what regards the first, the
decisive observation is that indifference curves have nothing to do with balancing:
indifference curves show points of equal trade-off between Ps and, hence, are only
pertinent when more than one Rs are being compared (in necessity).
(lxxvi) Regarding the satisfaction of P1 in the y axis and of P2 in the x axis, indifference curves
show the points where both Ps have the same cumulative satisfaction; for instance, a R with
27 (128) of gains in P2 and 27 (128) of losses (¼ 21 [2] of gains) in P1 and a R with 24

59
This is R2 and R3. As probably noted, R1 (inaction) has been discarded. The reason is simple:
considering the threat the disease is for human life, R1 would bring into the analysis the norm
(P) conferring the right to be alive. However, since such move implies abandoning the model of two
Ps, it would change the matrix of the present exercise.
60
Then, while in necessity two Rs (or more) are compared, in proportionality in the narrow sense
only one R is on the table. For this reason, in the latter the contraposition is strictly realized between
principles (effective place of balancing). It is difficult to understand, therefore, the connection
between balancing and necessity that is mentioned by Sieckmann (2012), p. 93, or the so called
“balancing necessity” described by Sartor (2010), p. 203.
From Constitutional Discretion to the Positivist Weight Formula 33

(16) of gains in P2 and 24 (16) of losses (¼ 24 [16] of gains) in P1 (both with a cumulative
satisfaction of 28 [256]);
(lxxvii) There might be as many indifference curves as we want; their difference rests on the fact
that the further to the right the line is the higher the cumulative satisfaction (s) of both P1
and P2 is (s[P1, P2]); though, they only show that Rs standing on different points of the
same curve express an equal trade-off; thus, if these curves can be relevant in necessity,
they are irrelevant for balancing:61

What is effectively relevant for balancing, inversely, is the curve with the
constant of direct proportionality expressed by each of the laws of balancing: as
seen before, “the more the losses, the more the gains (substantive) or the more the
certainty (epistemic)”. As it is visible, these laws require some constant to be obeyed,
which means that its curve expresses a specific ratio of satisfaction and
unsatisfaction of the two Ps at hand (or one P and certainty in the epistemic law).
Therefore, this curve does not show anything relevant for the comparison of
alternative Rs; differently, it focuses on the position one R stands regarding the Ps
in conflict.62 For proportionality in the narrow sense, this is what really matters.

61
Indifference is usually represented by a convex curve bending to the left (Sieckmann 2012, p. 90;
Rivers 2007, p. 174). However, this convexity signals the decreasing marginal rate of substitution
(when the same degree of higher satisfaction of one P does not substitute the same degree of lower
satisfaction of the other), which implies that whenever this marginal rate is constant (perfect
substitutes) the curve of indifference becomes linear (Varian 2005, p. 39; Banerjee 2015, p. 42).
Given that this feature (perfect or imperfect substitution) is irrelevant for the present purposes
(conflicts of principles do not necessarily imply the marginal rate of substitution to decrease; it is a
strict contingent effect), nothing prevents indifference to be linearly represented. On conflicts
between principles with convexity, Sartor (2010), p. 193. On indifference, Steiner (1994), p. 156;
Kolmar (2017), p. 150.
62
Evidently, we may compare as many Rs as we want with the constant of direct proportionality.
The main point is that, differently from indifference, where the information given by its curve is
useful for the comparison between Rs, with the constant of those laws the information given by its
34 D. Duarte

(lxxviii) The curve with the constant of direct proportionality is obtained from k ¼ yx, where k is the
constant and, under the double triadic scale, y and x are the maximum satisfaction of P1
(256) and P2 (256); thus, since here k ¼ 256
256 ¼ 1, the constant is 2 and the curve has a
0
0
slope of -2 .
(lxxix) Since “the more the losses the more the gains (or certainty)” is exactly the same as “the
less the gains the more the gains (or certainty)”, this curve can be sketched with gains in
both axis; evidently, this means that below the curve a R is disproportional, but propor-
tional above:

Regarding the second. It seems then that proportionality in the narrow sense has
no connection with optimality (only necessity has).63 Given that “ought implies can”
(proportionality only works within the realm of the feasible), the point of Pareto
optimality (matching the feasible optimal solution) can be lower than the constant
required by both laws of balancing.64 When the point of Pareto optimality is in such
position, it follows that an optimal R (surpassed necessity) can still be dispropor-
tional (if its gains in P2 are lower than its losses in P1). Thus, only one conclusion
follows: optimality plays no role in balancing. It only deals with Rs in comparison
and disproportionality is still possible with an already optimal R.
(lxxx) In the next figure, the point of Pareto optimality (matching the feasible optimal solution) is
below the constant demanded by the laws of balancing (because the optimal R has a ratio
smaller than 20); when P2 prevails (below southwest $ northeast dashed line), R2, even

curve is only useful for evaluating how much one specific R is satisfying and unsatisfying the
conflicting Ps.
63
Differently, Sieckmann (2012), p. 93; Eriksen (2006), p. 83.
64
Which means that, for the cases where the conflict between two Ps is (empirically) compatible
with all the Rs under the inverse satisfiability of both Ps (as in [lxxiii]), the constant is on the curve of
possible cumulative satisfaction of both Ps (this is, 28 [256]). Consequently, in these specific cases, a
R above is impossible, a R on the curve is optimal and proportional in the narrow sense. However,
below, besides non-optimal (already disapproved in the necessity test), it would never be propor-
tional in the narrow sense.
Exploring the Variety of Random
Documents with Different Content
Gammer Gurton's Needle.
The equally renowned "Gammer Gurton's Needle," was acted sixteen
years after "Ralph Roister Doister," at Christ's College, Cambridge. It
is usually attributed to John Still (born 1543) a member of Christ's,
Master of Arts in 1565, and later Master of that College, Vice-
Chancellor of the University, and finally Bishop of Bath and Wells
(died 1608). As Vice-Chancellor, Still was a stickler for Latin plays at
Cambridge, which were more educational but not so popular as
dramas in English. The plot turns on the loss of a needle by old
Gammer Gurton, the suspicion, raised by a wag, that another old
woman has stolen it; the search for the needle; combats about the
needle, and the final discovery of that implement in the seat of a
man's breeches. A sturdy beggar, Diccon, is "the Vice," and sets
Gammer Gurton and another gammer to a scolding match. Hodge, a
servant, with his broad dialect, and insistent demand for the needle,
that a large and unseemly hole which ventilates his breeches may
instantly be patched, has perhaps the most comic part, and when
somebody slaps Hodge and drives the needle (which had stuck in his
breeches), into a safe part of his person, the joy of a Cambridge
audience knew no limits. The play is thoroughly rustic, the language
is of an amazing breadth, and no doubt the drama made abundant
mirth among the Cantab wits. Members of the sister University,
where poets have been rare in comparison with these glories of
Cambridge, need not covet Still, unless he wrote the famous drinking
song in the Second Act, "Back and Side go bare, go bare!"
The Bishop of Bath and Wells probably looked back with mingled
feelings on the jolly, noisy achievement of his youth, which has
made him immortal, for all have heard of "Gammer Gurton's
Needle". It is written in rhyming lines of from fourteen to sixteen
syllables.

"Gorboduc."
"The Gammer," though low, is lively; not so is "Gorboduc"; it is a
tragedy of unspeakable dullness composed in blank verse which has
no merit except that of regularity, the sense usually, though not
always, ending at the close of each line. The author, Thomas
Sackville, later Lord Buckhurst and Earl of Dorset, and High
Treasurer under James VI and I, was born at Buckhurst, Sussex, in
1536. His grandmother was aunt of Anne Boleyn, so he was a
second cousin of Queen Elizabeth. At the Inner Temple, as a young
man, he met Thomas Norton, and the pair composed "Gorboduc,"
which was acted in the Inner Temple in 1561. The authors were
inspired by no other Muse than that of Seneca, the moral
philosopher, Roman tragedian, and tutor of the Emperor Nero. The
play tells how Gorboduc, a mythical King of Britain, abdicated, and,
dividing his realm into two parts, gave the country north of the
Humber to the younger, and the portion south of the Humber to the
elder of his two sons, Ferrex and Porrex. Each had a kind of tutor,
and each had a favourite. They were both discontented, the younger
slew the elder son, and the mother of both avenges the elder on the
younger of her children. The result was national ruin, in which
"Fergus Duke of Albany" (apparently King of Scotland is meant) took
an active part. There are very long speeches, no action; a
messenger brings the news of the distressing occurrences, and a
Chorus moralizes on them. Carried away by grief when his wife
murders his surviving boy, Gorboduc pronounces the name of
Eubulus with the penultimate syllable short, and expires with
decency behind the scenes. Eubulus then utters a political forecast in
more than a hundred lines, and the drama concludes.
"Gorboduc" was printed in 1565: translations of Seneca's plays were
also being written: George Gascoigne translated a piece named
"Jocasta" (the wife of Œdipus) from the Italian, and a prose comedy,
"The Supposes" from Ariosto. This great Italian poet and his
countrymen adapted to Italian manners the plots and characters
which the ancient comic dramatists of Rome, Terence and Plautus,
derived from late Greek comedy of everyday life. Thus an element of
orderliness in comedy was introduced in England from adaptations of
Italian adaptations of Roman copies of late Greek plays. Such stock
characters as the austere father, the spendthrift son, the cunning
servant, the boastful soldier, the nurse, soft of heart and loose of
tongue, invaded the comedy of France, and, to a slighter degree,
that of England.
Meanwhile Richard Edwards produced a curious Interlude of a
classical nature, "Damon and Pythias," the characters being Greek,
Sicilian and English—a dash of buffoonery is mixed with very
lamentable matter. The Drama was formless, unable to attain
definite shape, till some twenty-five years had passed when we
reach the date of the immediate predecessors of Shakespeare, such
as Marlowe, Greene, Lyly, Peele, and the other University young men
about town. The influences of the old waggish or controversial
Interludes, of the Senecan school of stiffness, and of translations or
imitations of Italian comedies, were seething in the cauldron of the
age.

CHAPTER XVII.

WYATT AND SURREY. GASCOIGNE. SACKVILLE.

The names of Sir Thomas Wyatt (1503-1542) and of Henry Howard,

Earl of Surrey (1517-1547), are for ever memorable in English
poetry, not so much for what they actually achieved as for what they
attempted. They abstained from allegory, still lingering in its unlovely
dotage, and from doggerel. They wrote of themselves and their own
loves, joys, and sorrows, but though their verse is concerned with
their personal emotions, these are treated in a conventional way,
borrowed from continental poetry. They turned to the Italian
sonneteers, especially to Petrarch, and saw afar the dawning of the
"Pléiade," the company of French reformers of poetic style and
language, Ronsard, du Bellay, and the rest, or at least of Mellin de
Saint-Gelais, their predecessor. But both Wyatt and Surrey died
young, Wyatt by an unfortunate chance, Surrey as a victim of the
jealous tyranny of Henry VIII. The two young poets thus live
together in men's memories like the Bion and Moschus of Greece:
theirs is "unfulfilled renown".
Wyatt, of a Yorkshire family, was son of Sir Henry Wyatt, of Allington
in Kent, a man who had strange vicissitude of fortune in the reigns
of Richard III and Henry VII. Thomas went very early to St. John's
College, Cambridge, married at 17, was a glory of the Court of Henry
VIII, went on diplomatic missions to Italy (Venice, Ferrara, Bologna,
Florence, and Rome), studied Italian literature, was now in favour
and now in prison, and made love, with more or less of earnestness,
to Anne Boleyn, being fortunate in escaping from the doom of her
admirers when Henry VIII took her life. Favoured by Henry's
minister, Thomas Cromwell, but detested, and accused of diplomatic
misdeeds by Bishop Bonner, Wyatt defended himself with a success
then very rare, retired from Court and wrote satires and poems on
the advantages of retirement; paraphrased the Seven Penitential
Psalms, and died of a fever caught from fatigue and travel, in
October, 1542, lamented in verse by Surrey.
The reader of his sonnets, the earliest in English, is amazed to find
that we have travelled through so many centuries of the life of
English poetry, and only reached lame lines that can scarcely be
scanned. Since Chaucer the art of verse had become very dim,
perhaps in consequence of the transitional state of the language, the
obsolescence of the sound of the final e, and the Anglicizing of the
sounds of borrowed French words by throwing back the accent (as
in honour for honour, virtue for virtue). Wyatt, when he began to
write sonnets, put accents in strange places, and counted syllables
on his fingers, content if he could reckon ten of them, in a line. To
rhyme "aggrieved" to "wearied," is like the tramp's effort to make
"workhouse" rhyme with "sorrow". The young student in a novel of
Henri Murger's reads only the rhymes in sonnets. If we study in that
way Wyatt's sonnet "The Lover Waxeth Wiser," we find that the last
words in the first eight lines are
 aggrieved
last
past
wearied
buried
fast
haste
stirred.
He usually tried to keep to the Petrarchian arrangement of rhymes in
the first eight lines a b b a a b b a, but, contrary to Italian rule, his
last two lines were always a rhyming couplet, as in Shakespeare's
"Sonnets," in which the Petrarchian model is wholly disregarded. The
sonnet thus ends with an emphatic clench, usually moral, while in
the Italian sonnet the last six lines resemble the withdrawal of the
wave of the first eight lines.
The sonnet, with its concision and its technical difficulties, afforded
excellent practice to poets who endeavoured to bring delicacy and
order into the chaos and coarseness of verse as written by Skelton
and his contemporaries. But a good sonnet is among the rarest of
good things, and the mere technical difficulties once overcome,
men's minds may turn out sonnets of no value with the rapidity of
machine work. The stock character of this kind of poetry, the Lover,
with his strange far-fetched conceit in his almost metaphysical
refinements, is apt to become as tedious as the old figures of
allegory; however, he was a novelty. Wyatt improved with practice in
sonnet-making, though such rhymes as "mountains" "fountains,"
"plains," "remains," are a stumbling-block to the modern reader. But
his "And wilt thou leave me thus?" and "Forget not yet the tried
intent," with their brief refrains are immortal lyrics, heralding the
music of the age of Elizabeth.
His epigrams are not the stinging wasps of verse commonly called
epigrams, but are brief poems in the manner of the epigrams of the
Greek Anthology. The satires on the Court, based on Italian poems,
and including a form of the "Town and Country Mouse," are not in
Skelton's violent way, but the work of a gentle man, and the poems
in rhyme royal, seven line stanzas, with six syllables to the line, are
charming novelties.

The Earl of Surrey.

The date of Surrey's birth is uncertain: it was four or five years after
the battle of Flodden (1513), in which his grandfather—"an auld
decrepit carle in a chariot—" was victorious over the fiery James IV.
The title Earl of Surrey is a courtesy title, borne by the poet as son
of the Duke of Norfolk. He was at least a dozen years younger than
his friend Wyatt, and was a lively young courtier, who was made a
Knight of the Garter in 1541. He married very early, in 1532, and his
famous passion for fair Geraldine may have been merely poetical—
the usual story about Geraldine and the magic mirror is derived from
a novel of 1554. About 1542 he was imprisoned for a matter of a
duel, a challenge at least, and in 1543 went about London at night
breaking windows with a stone-bow. He wrote a poem in which he
gravely maintains that he was merely punishing the wicked city for
her sins. Again released from prison he saw some fighting in France,
and, returning, patronized a poet named Churchyard, who later wept
unmelodiously above his early tomb. Early in 1546 Surrey had the
worse of a battle with the French near Boulogne, was superseded by
the Earl of Hertford, and, in January, 1547, was accused of a sort of
heraldic high treason (quartering the arms of Edward the Confessor,
who, of course, had never heard of armorial bearings), and
executed, shortly before the death of the tyrant, Henry VIII.
Surrey's versification, especially in the sonnet, is much superior to
that of Wyatt, but he is less apt to keep to the rules of rhyme, in the
first eight lines; indeed he writes in the form of Shakespeare's
sonnets. His "Prisoned in Windsor" is a pleasant picture of a young
gallant's life, who takes his eye off the ball at Tennis to watch the
ladies in the dedans: hunts, tilts, and makes friends. The moral
poems in lines of fourteen feet are of no great merit, but Surrey's
translation of the Second Book of the Æneid is the first English
example of blank verse, borrowed from Italian practice. The lines are
stiff and hard; and the main merit is the novelty, the first birth of the
measure that was to become, in forty years, "Marlowe's mighty line".

Tottel's Miscellany.
The poems of Wyatt and Surrey were not published till long after the
deaths of the authors, when they appeared, with many other pieces,
in "Tottel's Miscellany". Other writers represented there are Nicholas
Grimald, with his jog-trot metre, the "poulter's" or poulterer's
measure of from twelve to fourteen syllables to the dozen—so were
eggs sold by a custom of the trade. Surrey's retainer, Thomas
Churchyard, a man very busy with sword and pen, was also a writer
in the "Miscellany"; and indeed was a literary hack-of-all-work. There
came, after the brief gleam of sunshine that fell on Wyatt and
Surrey, another generation of wooden versifiers and translators, with
whose names, Tusser the bucolic, Phaer, Golding, Googe, and
Whetstone, it is hardly necessary to fill the page and burden the
memory. They may be studied by the curious, but they wrought no
deliverance. To generations which possess superabundance of
versifiers and no great poets, these barren years are a kind of
consolation. For reasons not to be discovered there are such periods
in the literary life of all nations, as in England between Pope and
Cowper.
The versifiers in "Tottel's Miscellany" keep harping unmelodiously on
the strings of Surrey and Wyatt, many of their pieces are
complimentary addresses to ladies, or laments on the deaths of
friends. Poor conceits are twisted and tormented; there is hardly any
promise of advance; we scarcely hear any of the bird-like musical
notes with which the later part of the reign of Elizabeth sang so
wondrously.

Gascoigne.
George Gascoigne (1525 (?)-1577) was an interesting character. He
was a Cambridge man, a member of the Society of Gray's Inn, a
poet who, like Scott, composed his verses in the saddle: a Member
of Parliament who was opposed as "a common rhymer... noted for
manslaughter... a notorious Ruffian," and even a spy, certainly he
owed debts, and was disinherited by his father. He wrote on
woodmanship, but was apt to forget to shoot at the deer that came
within range of his cross-bow. As a captain in the Low Countries he
and his command were surprised and taken by the Spaniards: he
came home, published his Posies (1575) and, he says, got not a
penny by the venture: he then wrote "The Steel Glass," a kind of
satire, the mirror of the age, in blank verse, and next wrote in
common ballad measure the long and amazingly prosaic "Complaint
of Philomene".
In 1572 Gascoigne published "A Hundred Sundry Flowers, bound up
in one small Posy". The long title sets forth that some of the flowers
were culled in the gardens of Euripides, Ovid, Petrarch, and Ariosto,
others are from English orchards. The native flowers are the sweeter
and more fair. While our poets were turning into stiff measures the
sonnets of Italy, Gascoigne could write so naturally and melodiously
his own English, as in his "Lullaby of a Lover".
Sing lullaby, as women do,
Wherewith they bring their babes to rest,
And lullaby can I sing too,
As womanly as can the best.
Beneath the stiff borrowed phrases and metres there was always
this native and tuneful spirit of unsophisticated song.
In 1575 he was a maker of words for the Masques at Leicester's
famous reception of Elizabeth at Kenilworth (see the novel of that
name, where Scott calmly introduces Shakespeare as already a
successful dramatist). He satirized drunkards: we have already seen
that he translated a tragedy, "Jocasta," from the Italian; he wrote a
love story in rhyme of a personal kind, and his brief "Instructions" is
the earliest English work, in no way indebted to Aristotle, on the Art
of Poetry. As he also translated, we have seen, a comedy from the
Italian, and a prose tale, a kind of work later fashionable, Gascoigne
may be regarded as an intrepid explorer in many fields of literature.
"He first beat the path to that perfection which our best poets have
aspired to since his departure," says Nash (1589). "He brake the ice
for our quainter poets that now write," says Tofte (1615). But the
path as trodden by this pioneer continued to be rough. Gascoigne
was an example of the versatility and literary ambition which many
young gentlemen displayed in the age of Elizabeth; mingling poetry
and study and serious thought with their gallant adventures in love,
diplomacy, war, and travel.
His "Certayne Notes of Instruction concerning the making of Verse in
English" is a very brief pamphlet. He quotes "my master, Chaucer"
against alliterative "thunder in Rym, Ram, Ruff," but mentions no
other poet. Be original, he says, if you sing of a lady do not applaud
her "crystal eye" or "cherry lip," which Spenser did not disdain, for
these things are trite and obvious. The great matter is "to avoid the
uncomely customs of common writers," says this "common rhymer".
Do not use "obscure and dark phrases in a pleasant sonnet". Do not
wander out of your "Poulters measure" metre into lines of thirteen
syllables. Give every word its natural emphasis: do not make
treasure into treasure. Chaucer is to be followed as a master of
prosody. You should write:—
"I understand your meaning by your eye,"
not,
"Your meaning I understand by your eye",
"The more monosyllables that you use, the truer Englishman you
shall seem".
There follows advice on the caesura, and all this counsel shows that,
in the early years of Elizabeth, versification was at a very low ebb.
In practice, Gascoigne did not always shine. There are few passages
of interest in the stiff blank verse of his "Steel Glass" (the mirror that
does not flatter). The best passage, and it is very good, describes
the labourer,
Behold him, priests, and though he stink of sweat,
Disdain him not, for shall I tell you what?
Such climb to heaven before the shaven crowns,
because the labourers
feed with fruits of their great pains
Both king and knight and priests in cloister pent.
It would be cruel to quote "Philomene," no stall-ballad creeps more
tardily on a longer road than Gascoigne in his tale of her who sings,
in a later poet's words,
Who hath remembered thee, who hath forgotten?
They have all forgotten, oh summer swallow,
But the world shall end when I forget.

Sackville.
The poetry of Thomas Sackville (1536-1608) is not to be found in his
dull tragedy, "Gorboduc," but in his contributions to a vast and once
popular collection, "The Mirror for Magistrates". This work is
intended to admonish men in power by rhymed histories of the falls
of English peers and princes. This was the plan of Chaucer's Monk,
in "The Monk's Tale," which that sound critic, the Host, could not
long endure. The model was Boccaccio's work on "The Falls of
Princes," Englished by Lydgate. The enterprise started by Baldwin
and others in 1554-1559, suggests a dread lest English verse should
return to Lydgate in the den of Giant Despair, and take up with
sepulchral solemnity the tale of tragedies from the darkest days of
the unfortunate ancient Britons. A mammoth compilation was
gradually evolved, for doleful matter was not far to seek, but
Sackville's two contributions, the "Induction," and the "Complaint of
Buckingham"—the Buckingham executed under Richard III,—alone
concern us.
In the "Induction" the poet describes the gloom of winter, and, in
the mediaeval way, dwells long on the constellations. As he muses,
he is met by a very deplorable female form—
With doleful shrieks that echoed in the sky.
She proclaims herself to be Sorrow, a goddess, and guides Sackville
"to the grisly lake" of Avernus, over which no fowl may fly and live.
A number of rueful figures of allegory are encountered, Dread,
Revenge, Misery, Care, Old Age, and Sleep, and these are drawn
with abundant vigour and variety. The stanza on Sleep gives the
measure of the versification, which is rapid, concise, various,
sustained, and in its music heralds the arrival of Spenser.
The body's rest, the quiet of the heart,
The travail's care, the still night's frere was he,
And of our life on earth the better part,
Reiver of sight, and yet in whom we see
Things oft that tide, and oft that never be,
Without respect, esteeming equally
King Croesus' pomp and Irus' poverty.
One stanza in the description of the home of the dead seems to
have been suggested by famous lines in the Eleventh Book of the
"Odyssey".
The "Induction" ends with the appearance of the spirit of
Buckingham, who not only tells his own tragedy at great length, and
in full historical detail, but introduces several other ancient tragedies,
those of Cyrus, Cambyses, Brutus, Cassius, Besseus, Alexander the
Great, Clitus, Phalaris, Pheræus, Camillus, and Hannibal. From these
fallen princes we drop to
One John Milton, Sheriff of Shropshire then,
who arrested Buckingham, and to
A man of mine, called Humphrey Banastaire,
who betrayed his master. Banastaire is then cursed in eleven
stanzas. "May Banastaire live to the age of eighty, and then be tried
for theft. May his eldest son expire in a pig-sty; his second son be
strangled in a puddle, and his daughter be smitten by leprosy."
It cannot be denied that this tragedy, including as it does the murder
of the Princes in the Tower, is rather too rich in terrible components,
and does not, especially when Banastaire is being dealt with, affect
us in the same measure as Dante's pictures of the Inferno. On the
whole it is the manner, not the matter, of Sackville that contains
more than mere promise: his management of the stanza and of the
music of the line is far in advance of anything that had come from
an English pen since the death of Chaucer. As for the gloom and
horror, these were congenial to a people which, since the burning of
the Maid of France (1431), had seen an endless sequence of
violence, murder, martyrdoms, and massacres of peers, Princes,
Queens, Bishops, and humble folk.

CHAPTER XVIII.

PROSE OF THE RENAISSANCE.

A great, indeed an inestimable influence in literature at this juncture,

was that of the long-forgotten Greek language, Greek poetry, and
Greek philosophy. When Erasmus, who then had little Greek, arrived
in England and visited Oxford (1499), he found there Grocyn,
Linacre, and Colet, who had acquired Greek on the continent; and,
with Sir Thomas More, were already competent classical scholars.
But their Greek learning was mainly turned into the channel of
theology, the study of the sources of Christian doctrine, the New
Testament, the Greek Fathers; and they were attracted by the
philosophy of Plato which appeared to "utter a Christian voice" much
more clearly than do the writings of the idol of the Middle Ages,
Aristotle.
Greek, however, does not visibly affect the poetic literature of
England much, before the date of Spenser, about 1580. The violent
times of Henry VIII and Mary Tudor were not favourable to severe
study and exquisite appreciation of the Greek genius, a most
desirable corrective of the prolixity of mediaevalism, and of the
English passion for horrors in stage plays. To most people knowledge
of the contents of the Greek classics came through translations, and
these translations, as in the case of the historian Thucydides, were
done from French versions, while Plato was read through Italian
commentators, much influenced by Plato's disciples in early Christian
times, the Neoplatonists, dreamers of beautiful dreams concerning
things that cannot be uttered.
Study produced also a very wide acquaintance with Greek mythology
—Shakespeare's humblest characters have heard of many a Grecian
fable—yet the spirit, the exquisite balance, and the refinement of the
Greek genius, hardly affected our authors. We may detect it in
More's (1478-1535) "Utopia," where the adventurers carry with them
to "Nowhere" a "pretty fardel," or parcel, of the cheap neat Greek
books printed by Aldus. The fancied State of Utopia, with its
comfortable communism and perfect freedom in religion, is derived
from the "Republic" of Plato, and in religion is more liberal than, in
his later work, "The Laws," he would have permitted it to be. But the
"Utopia," written in Latin, was meant for the learned.
Though the "Utopia" was published in 1516, and became famous in
Europe, it did not reach unlearned English readers till an English
translation, by Ralph Robynson, appeared in 1551. They now had
More's eloquent advocacy of communism before them as regulated
in his imaginary state, with a Six Hours' Day, universal training of
men and women for war, and habit of assassinating the leaders of
hostile nations. There is tolerance of all religions which accept a
deity and the immortality of the soul: atheists are disqualified for
public offices.
In his English works on religious and social controversy, which are
little read, More is not only a Catholic and a Conservative, but in
discussion is given to abusive and violent language which would
have horrified the courteous Plato, the urbane Aristotle, and that
model of a devout and ardent student, and perfect gentleman, Pico
della Mirandola, whose Life More gave in English. On both sides the
controversialists of the Reformation delighted in violent personal
abuse, in some Greek orators they found examples of that art. The
first effect of Greek in England, by producing a new Biblical criticism
and an attack on the foundations of the mediaeval Church, was to
"bring not peace but a sword," the wars of religion.

Elyot.
No man did more for the intelligence of Greek than Sir Thomas Elyot
(1499 1546)1 author of "The Governour," a long treatise, on the
education of a gentleman, and on the nature of forms of
government. Elyot bubbles over with Greek, and translates such
passages of Homer as he quotes into English verse, the alternate
lines rhyming. He is of the Greek opinion that a gentleman should be
taught, if he has a taste for art, to draw, paint, and execute works in
sculpture, not as a base professional artist, but as an amateur.[1]
Elyot would have a boy, at 7 years old, begin with Greek, learning it
through Latin, which he picks up, with French, in conversation.
Grammars of Greek are now almost innumerable. Grammar, he says
with much truth, "if it be made too long and exquisite to the learner,
in a manner mortifieth his courage. And by that time he cometh to
the most sweet and pleasant reading of old authors, the spark of
fervent desire of learning is soon quenched with the burden of
grammar." Elyot would start his pupil as early as possible with what
will interest a child, Æsop's Fables in Greek, and then pass to Lucian,
who is amusing as well as elegant. "But I fear me to be too long
from noble Homer, from whom as from a fountain proceeded all
eloquence and learning." Throughout, Elyot wishes first to interest
the pupil; but where, he asks, is he to find qualified schoolmasters?
They were as cruel as in the days of St. Augustine, and while Elyot's
system of education, in sports as well as in books, is free and
joyous, like that of Gargantua in Rabelais, little boys were suffering
the horrors described by Agrippa d'Aubigné in his Memoirs. Elyot
translated works of Isocrates, Plutarch, and others, wrote a medical
work "The Castle of Health," was clerk of the Privy Council, and went
on various diplomatic missions. Elyot was not a professional
instructor of youth: he was, it seems, educated privately, and of
neither university; what pleases us in him is his unstaled zest for
learning, his fresh enthusiasm.
The best English of the age and the most durable is that of Thomas
Cranmer (1489-1556) as we read it in the Liturgy of the Church of
England, while much of the merit of King James's Authorized Version
of the Bible rests on the foundation of Miles Coverdale's translation
(1488-1568). How easy it is to translate the Bible into English which
is not a marvel of diction and rhythm, we are too frequently
reminded by the Revised Version.

Ascham.
Roger Ascham (1515-1568) was a Yorkshire man of the middle
classes, who lived by his learning, and did not find that it paid him
as well as he wished. Going early to St. John's College, Cambridge,
he was a pupil of the famous Sir John Cheke, who introduced the
English way of pronouncing Greek. It is certainly wrong—no people
pronounce the vowels as we do; but if Cheke resisted the
pronunciation of the modern Greeks, perhaps he is not much to be
blamed. Ascham obtained a Fellowship and a Readership in Greek,
the Fellowship he lost when he married: he did not long retain his
tutorship to the Princess Elizabeth; as secretary to an ambassador in
Germany he continued to teach Greek to his chief; and in his letters,
Latin or English, we find him often in straits for money and begging
for assistance. Camden, writing under James I, says that he lost
money at dicing, and in his attack on gambling, in his "Toxophilus," a
dialogue on Archery (1545), Ascham shows a rather unholy
knowledge of all the tricks on the dice-board. Probably he had paid
for his education. He contemplated a work on the noble sport of
cock fighting, on which, of course, there was betting, and perhaps
Ascham was not in all respects so severe a Puritan as in his
unworthy attacks on that noblest of romances, "The Morte d'Arthur".
Sir Lancelot is a better gentleman than many who were to be met at
a cock fight. Ascham had little sympathy with the Italian influences
that were so potent in Elizabethan literature. Italy was certainly
profligate and luxurious,
An Englishman that is Italianate
Doth quickly prove a devil incarnate,
was an English translation of an Italian proverb. Ascham, like his
contemporaries, was nothing if not patriotic. The bow of yew and
the grey goose shaft had won many a victory over Scots and French,
as in "Toxophilus," Ascham reminds these peoples; therefore he
desired that archery should be universally practised. But the
harquebus, a musket lighter than the heavy hand gun of the
fifteenth century, was already, in disciplined hands, more than a
match for the bow.
"Toxophilus," to our age, appears pedantic. We have endless
classical examples, and learn that the Trojans drew the bow-string
only to the breast, not the ear (which is true), while they used iron
arrow-heads as against the bronze arrow-heads of the Greeks, a fact
not so certain. When he does come to practice, Ascham's teaching in
archery is reckoned sound and good. His ideas are summed up in
the prayer that the English
Through Christ, King Henry, the Book, and the Bow
May all manner of enemies quite overthrow.
In writing English, Ascham was all for plain English. Foreign words
Anglicized make such a mixture "as if you put malmsey and sack,
red wine and white, ale and beer, all in one pot". Yet he advocates in
his "School Master," published after his death, a yet more
unhallowed blend, the use of Greek measures in English verse. "Our
English tongue in avoiding barbarous rhyming may as well receive
right quantity of syllables as either Greek or Latin." (He means
"quantity" as opposed to accent, as if one said carpenter.) As an
example he quotes Mr. Watson's rendering of the third line of the
"Odyssey" into two English hexameters
All travellers do gladly report great praise of Ulysses,
For that he knew many men's manners and saw many cities.
Obviously if we are to say "men's manners," making "man" in
"manners" long, we must not make "vellers" in "travellers" short, as
Mr. Watson does. We are reduced to
Gladly report great praise of Ulysses do the travellers.
This absurd manner of imitating Greek measures in English was
upheld, twenty years later, by Gabriel Harvey, who, for a moment,
nearly corrupted the practice of Spenser, the most naturally musical
of poets. Ascham's own prose style is unaffected, not corrupted by
eccentricities, but not harmonious. A new perfection, a false
perfection, was to be sought later, through the antitheses,
alliterations, and pedantic wit of Lyly's "Euphues!"

Lyly's Euphues.
The prose of Ascham was clear and was plain, disdaining decoration
and far-fetched gorgeous phrases. But for the gorgeous and the
exotic, the taste of the Elizabethan Age was pronounced, as we see
in the strange over-gaudy costumes of the period, the various ruffs,
the jewelled velvets and silks, worn by men and women. A like
dressing for thoughts was demanded, and the supply was provided
by John Lyly, whose plays are to be mentioned later. Lyly was born a
Kentish man (1554?); Magdalen, in Oxford, was his college; his
plays, acted by the boys of the Chapel Royal and St. Paul's, are of
1584-1594. But he made his mark earlier, as a prose writer, in his
"Euphues, the Anatomy of Wit" (1579), and the sequel, "Euphues
and his England" (1580). The style became a fashion, a fashion
which affected even those who, like Sidney, were in would-be revolt
against it. Lyly, like all writers of the periods just before and after
him, was copious in classical allusions. He was not the first to hunt
in all directions, especially in fictitious natural history, for similes, and
needless decorations; but he hunted further and more assiduously:
emphatically his style is that of the unresting Bird of Paradise. Every
sentence is a thing bristling with points and antitheses and
alliterations. The first part of the book was a kind of novel; two
friends, at Naples, woo the same woman, quarrel, write long letters,
and the question of education, in the wide sense in which the
Renaissance understood education, is always prominent. There is
endless conversation and discussion of life, love, and learning,
always in the same style of fantastic decoration and allusion: all
continued when Euphues arrives in England, all conveying general
information not verified by experiment. "I have read that the bull,
being tied to a fig tree, loseth his strength; that a whole herd of
deer stand at the gaze if they smell a sweet apple"; facts on which
the cattle-breeder or the hunter would not, if well advised, rely. This
was the kind of science against which Bacon uprose. But Lyly
appealed, in his Dedication, and with success, "To the Ladies and
Gentlewomen of England," who found in the book a kind of love-
story, much philosophizing on that dear theme; and a pleasurable
example of a new way of being witty and romantic. Lyly was the
chief cause of the difficulty in telling a plain tale plainly which besets
the minor writers of the age of Elizabeth.
Before approaching the chief prose writers of Elizabeth's time, we
must turn aside to her greatest poet, and his friend, to Spenser and
Sir Philip Sidney, and to the Drama.

Sidney.
Spenser did not more surely attain immortality by his verse than Sir
Philip Sidney (1554-1586) by his life, writings, and character. He was
one of those who, as Plato says, are born good, exemplars of natural
charm and excellence. He is the ideal gentleman of the type which
Spenser professed to educate by the examples of his virtuous
knights, brave, pious, courteous, and just. The son of Sir Henry
Sidney and nephew of Elizabeth's Leicester, Philip Sidney was born
into the Court, but was not of it; his heart was set on other things
than pleasure, splendour, flattery, and promotion. Educated at
Shrewsbury School, he went to Christ Church at 14, being already
the friend of the noble Fulke Greville, who, however, went from
Shrewsbury to Cambridge. In 1572 he was attached to the English
embassy in France, and, on the night of the Bartholomew massacre
was sheltered in the house of his future father-in-law, Walsingham.
Till 1575 he travelled, chiefly in Germany, and made the
acquaintance of his constant correspondent and adviser, Languet,
whom he celebrates as a shepherd of the Ister, and as his own
religious Mentor. In Venice his portrait was painted by Veronese; at
Vienna he perfected himself in horsemanship under Pugliano, whose
enthusiasm he describes so amusingly in his "Defence of Poesie". For
a man so earnest as Sidney was, he had a fine sense of humour.
Returning to England in 1575, he, like Gascoigne, was with Elizabeth
at the famous pastimes at Kenilworth, now best known through
Scott's novel, "Kenilworth". Afterwards, at the house of the Earl of
Essex, he met the Earl's daughter, Penelope, later Lady Rich, the
Stella of his sonnets. Essex desired their marriage, but fate decided
otherwise. In 1577 Sidney went, a young diplomatist, to the
Emperor and the German Princes, and later, was obliged to attend
the Court, while his mind was set on adventures beyond the Atlantic;
on failing in that, he trifled with the idea of introducing Greek metres
into English poetry. In 1579, he quarrelled with the Earl of Oxford in
the tennis court. A duel was not permitted, but as Sidney also gave
Elizabeth his opinion about her distasteful flirtation with the odious
Duc d'Anjou, the worst of the bad Valois Princes, he retired to
Wilton, the house of his sister, Lady Pembroke, and there wrote the
pastoral romance, "Arcadia".
He was recalled to Court, sat in Parliament for Kent, and in 1583
parried a daughter of Walsingham. He was forbidden to join Drake's
American expedition of 1585, in fact he was always thwarted in his
desire for action and for such deeds of chivalry as the conditions of
his age permitted—they leaned somewhat to piracy and filibustering.
At length, as Governor of Flushing, while Leicester commanded the
forces engaged against Spain in the Low Countries, he fell in a
cavalry charge against a superior force at Zutphen. His leg was
broken by a musket bullet from the Spanish trenches: it was now
that he handed the cup of water that was at his lips to the soldier
whose need was greater than his. He lingered for some weeks, and
died on 17 October, 1586.
The beautiful character of Sidney cannot be more strongly attested
than by the agony of grief exhibited, at his death, by the handsome
and wicked Master of Gray. He was about to be sent on the Scottish
embassy to plead for the life of Mary Stuart, while his desire was to
be fighting under Sidney's banner. He expresses, in a touching letter,
the sudden revulsion of his nature from his wonted treacheries; and,
contrary to the falsehood of tradition, he did not betray, but, to his
own loss, did his best to save the Queen whose cause he had
previously deserted.
As a poet, Sidney, whose works were all published after his death, is
best remembered for the sonnets of Astrophel to Stella, Lady Rich.
There is a controversy as to whether these are mere exercises in
gallant but "platonic" love-verse, or whether they reveal a true
passion, as Charles Lamb maintained. The sonnet in which he says
that he has found his fortune too late, and has lost what he had
unwittingly won,
O punisht eyes
That I had been more foolish or more wise,
seems to set forth a truly tragic situation. Perhaps only poets can be
the critics in such a case as this of Sidney.
The sonnets vary much in poetic value; some are written in
Alexandrines, a metre not consonant with the traditions of the
English Muse.

Sidney's "Defence of Poesie."

Readers who fail to find brilliant merit in English literary poetry
between Chaucer and Spenser may not be ill-pleased to note that Sir
Philip Sidney was strong on their side. Acquainted as he was with
the poetry of Greece, Rome, Italy, and France, he could see nothing
to admire in the efforts and experiments of such writers as Occleve,
Lydgate, Hawes, Googe, Churchyard, and Turbervile. His "Defence of
Poesie" (or, according to the title of the first edition (1595), his
"Apologie for Poesie") was elicited by the unauthorized dedication to
himself of Stephen Gosson's "School of Abuse". Gosson was a young
Oxford man who had tried his hand as a playwright, and been
disgusted, he says, by the disorders of the playhouses, where his
comedy and morality may have been hooted. He therefore tried to
make himself notorious, or he expressed his penitence, by assailing
poets who deal in the silly conceits of Lyly's "Euphues".
"The scarab flies over many a sweet flower and lights in a cow-
shard... it is the manner of swine to forsake the fair fields and
wallow in the mire: and the whole practice of poets, either with
fables to show their abuses, or with plain terms to unfold their
mischief, discover their shame, discredit themselves, and disperse
their poison through the world". Gosson chooses Virgil as one of his
terrible examples, and whether he is a genuine or a hypocritical
puritan, or a mere fribble in search of notoriety, he made a mistake
when he thought to find a patron or a butt in Sidney, who does not
advertise Gosson's name in the "Defence of Poesie".
After a general defence of poetry furnished with precedents drawn
from every quarter, even from the respect paid to their minstrels by
the Irish, Sidney defines the final end of poetry as being "to lead
and draw us to as high a perfection as our degenerate souls, made
worse by their clay lodgings, can be capable of...." If poetry does not
always attain this end, "it is not the fault of the art, but that by few
men that art can be accomplished". He quotes Aristotle's "Poetics" to
the effect that poetry is more philosophical and more serious than
philosophy. Nothing in history is so noble but that "the poet may, if
he list, make it his own, beautifying it both for further teaching, and
more delighting, as it please him, having all, from Dante's heaven to
his hell, under the authority of his pen". Here Sidney seems to differ
from Scott, who regarded some examples of human fortunes, for
example in the case of Mary Stuart, as beyond the range of the
poetic art. But Sidney, foreseeing the objection, adds, "I speak of the
art, not of the artificer". Sidney then discusses the various Kinds of
poetry. As to the Comedy, "naughty play-makers and stage-keepers
have made it justly odious,"—so far he sides with the Puritans of his
time. In speaking of the lyric, he says: "I must confess mine own
barbarousness; I never heard the old song of Percy and Douglas"
("Chevy Chase"), "that I found not my heart moved more than with
a trumpet". Indeed the true spirit of poetry did dwell, disregarded by
wits and courtiers, in the popular poetry and the ballads. But poetry,
he knows not why, finds, in our time, a hard welcome in England: "I
think the very earth laments it, and therefore decks our soil with
fewer laurels than it was accustomed, for heretofore poets have in
England also flourished". If poets are not esteemed it is because
they do not deserve esteem, for we are "taking upon us to be poets
in despite of Pallas," invita Minerva. Our would-be poets are
destitute of genius—which was very true. "Chaucer undoubtedly did
excellently in his 'Troilus and Cressida': of whom truly I know not
whether to marvel more either that he, in that misty time, could see
so clearly, or that we, in this clear age, go so stumblingly after him."
What ailed Sidney's age was lack of terseness and clearness. Most
poets did not know what they would be at; they were confused by
the tumult of religion, the loss of old ideals, the language in
transition, the tyranny of the misunderstood classics, the constant
effort to imitate Greece, Rome, France, and Italy. They could not yet
see life and literature steadily, and see them whole. Sidney found
little that "had poetical sinews," except in Chaucer; parts of "The
Mirror for Magistrates," the Earl of Surrey's lyrics, and Spenser's
"'Shepherd's Calendar' hath much poetry in his 'Eclogues,' indeed
worthy the reading, if. I be not deceived. That same framing of his
style to an old rustic language I cannot allow..."
Sidney then banters the absurdities of the lawless stage, of the
alliterative writers, of the seekers after unnatural history, like Lyly in
his "Euphues," and of the love poets. "If I were a mistress never
would they persuade me that they were in love, so coldly they apply
fiery speeches," "swelling phrases" learned from books.
It was poetry, not the English poets of his age, that Sidney
defended, and he might well marvel at our modern zeal which
devotes time and scholarship to a chaos of tentative experiments by
men who wished to be poets without possessing the poetic genius.
Sidney's best poems and his "Defence of Poesie" retain their
freshness; but that book of his which was most popular suffers from
the changes of time and taste. At most periods prose fiction is more
welcome to human nature than poetry or criticism. Sidney's book
"The Countess of Pembroke's Arcadia," is a novel, written by the
author at Wilton, when, as we saw, he was neither in favour at Court
nor permitted to risk himself in adventures on sea or land. The book
was to Sidney what "The Faery Queen" was to Spenser, a wilderness
of delights of his own creation, a retreat into a world of fantasy. He
wrote it in sheets read, or sent as soon as finished, to his sister, the
Countess of Pembroke; the book was meant for her, not for the
world. Not long after his death, an unauthorized copy was published
(1590), and unauthorized edition followed, and the general delight in
the romance is attested by its constant reissues.
The author did not construct any regular plot, he allowed his fancy
to wander among the shipwrecks and piratical adventures of the late
Greek romances; and in an Arcadia which never existed, and a
Laconia most unhistorical. But the high and chivalrous ideals of the
author, in his rural prose idylls, as in his battles and combats; the
truth and constancy of his lovers; the beauty of his descriptions,
made this mixture of the Spanish heroic romances that infatuated
Don Quixote with the Arcadian pastorals, the delight of four
generations. Milton blamed the captive Charles I for copying the
beautiful and appropriate prayer of the captive Pamela, long after
Shakespeare had interwoven with the story of King Lear, Sidney's
tale of the blind King of Paphlagonia.
In its new mode "The Arcadia" was to four generations what
Malory's "Morte Arthur" had been in its day. As late as 1660, we find
Sir George Mackenzie imitating the "Arcadia" in his heroic and
historic romance, "Aretina," where Argyll and Montrose play their
parts. Indeed the "Arcadia" was a fruitful parent of the interminable
heroic French romances which Major Bellenden laughs at in "Old
Mortality," and from which Scott did not disdain to borrow a
description in "Ivanhoe". It is indeed curious to compare Sidney's
description of an Amazon (Book I, Chap, XII.) with an actual
representation of a genuine Amazon by a Hittite artist, discovered on
the stone work of a gate at Boghaz Keui. That lady-warrior wears a
corslet of scale armour, while Sidney's has a doublet of sky-coloured
satin, covered with plates of gold. Her feet are shod in crimson
velvet buskins, while the massive legs of the real Amazon are naked.
The contrast of fact and fancy are violent, of course, throughout the
romance. The style is less conceited than that of "Euphues," and is
always noble, but the long sentences and overabundance of
parentheses are not in accordance with modern taste. The profusion
of love-passages and of martial adventures, "with notable images of
virtues, vices, or what else," and the poetic if uncurbed fancies, were
what the world demanded from a novel, and what Sidney gave in
the Arcadia, with many lyrics, and imitations of the amœbean verse
of the shepherds of Theocritus.

Spenser.
After two centuries of verse that was tuneless or tentative, the
second great English poet came, Edmund Spenser (1552?-1599). We
know from his "Prothalamion" that Spenser was born in London—
my most kyndly Nurse,
That to me gave this Lifes first native sourse,
Though from another place I take my name,
An house of auncient fame—
that is, the House of the Spencers of Althorp who are in the ancestry
of the Duke of Marlborough's Churchills.
Spenser was certainly their kinsman, in what degree is unknown, but
his own family must have been poor. He was educated at Merchant
Taylors' School, was aided by the munificent Robert Nowell, and
obtained a Sizarship (corresponding to the old Oxford servitorship),
at Pembroke Hall, Cambridge (1569). Here he made two friends,
Gabriel Harvey, a true friend, if a rather pedantic don (the Hobbinol
of his "Shepherd's Calendar"), and E. Kirke, the E. K. who furnished
the notes explanatory of old English words in that poem. Spenser
also gained the good graces of Grindal, then Bishop of London, later
Primate, a puritan, who fell into Elizabeth's disgrace, and is
applauded as Algrind by Spenser in the "Shepherd's Calendar".
Spenser's youth was passed in an England disturbed by the claims of
the captive Mary Stuart to the Crown; by the rebellion of her
adherents in the North; by the papal excommunication of Elizabeth,
and by the pretensions of the extreme puritan exiles who, driven
abroad by the Marian persecution, had imbibed at Geneva the
doctrines of Calvin. In their attacks on the English Bishops they out-
wearied even the successors of Calvin in Geneva, who regarded
them as men not to be satisfied by any concessions; "a sect of
perilous consequence who would have no king but a presbytery,"
said Elizabeth. Here were all the elements which caused Elizabeth's
cruel persecution of Catholics, the long struggle of the puritans
under Elizabeth and James I, the wars under Charles I, and the
strife with Spain and Catholic Ireland. In the words of James VI, it
was "a world-wolter," and Spenser, as a poor young man, eager to
make his fortune, had to swim as best he might in the cross-currents
of this troublesome world. He never enjoyed the peaceful leisure of
a Tennyson or a Wordsworth; he had to play an active part in
strenuous and most unhappy affairs.
His nature, too, was divided. With all his love of pleasure and of
beauty he leaned, though not virulently, towards the puritan party,
and, as a good patriot, loathed and detested Rome.
It is probable that, when a freshman at the age of 17, he contributed
to a Miscellany, Van der Noodt's "Theatre of Worldlings" (1569),
translations in blank verse of certain sonnets of the French poet
Joachim du Bellay, and of Petrarch. These, re-cast into the form of
sonnets, recur in a volume of Spenser's, of 1591.
After taking his Master's degree (1576) Spenser visited Lancashire,
and if his words as Colin Clout in the "Shepherd's Calendar" be
autobiographical, lost his heart to a lady whom he calls Rosalind,
"the widow's daughter of the glen". According to Gabriel Harvey she
"christened him her Signior Pegaso," though neither his poetry nor
his wooing won her from her cruelty. Many years later he still writes
of her with chivalrous affection, so, like Scott, he had his heart
broken and cleverly pieced again.
By 1579 Spenser was in London, a literary retainer or protégé of
Elizabeth's favourite, the Earl of Leicester; while he also enjoyed the
friendship of Leicester's nephew, Sir Philip Sidney, the Flower of
Chivalry, himself a poet, and the best beloved man of his time. Now
(1579) Spenser published, and dedicated to Sidney, his "Shepherd's
Calendar," a set of twelve eclogues or pastoral poems, one for each
month. The pastoral had wandered far from the rural beauty of
Theocritus, and, in the hands of Mantuan and Clement Marot, had
become a vehicle for allegory, and even of Protestant
argumentation. Spenser does not stray far into party and puritanic
politics, but they are not unknown to his shepherds. In January, as
Colin Clout, he bewails the coldness of Rosalind,
She laughs the songs that Colin Clout doth make,
which is carrying cruelty very far. February is occupied with a rustic
dispute between youth and age: the metre is one of the measures of
the "Lay of the Last Minstrel":—
Who will not suffer the stormy time,
Where will he live tyll the lustry prime?
(Shepherd's Calendar, Feb., 11. 15, 16.)
They burn'd the chapel for very rage
And cursed Lord Cranstoun's Goblin-page.
(Lay of the Last Minstrel, C. II., Stanza, 33).
March, with the dialogue of Willie and Thomalin about the strange
bird, Love, is adapted from the Greek of Bion in a most pleasant
manner, and April contains a melodious song of fair Eliza, a Maiden
Queen; which probably procured Spenser's presentation to Elizabeth.
The great variety of melodious verse of which Spenser was already a
perfect master is, for us, perhaps the chief merit of his pastorals.
Through life Spenser keeps up the shepherd's mask, and Raleigh, in
his verse, is "The Shepherd of Ocean". The rival Protestant and
Catholic clergy also appear as shepherds, good or bad, while in
another eclogue the perfect poet, Cuddie, complains, like Theocritus,
of public indifference, and is advised to sing of redoubted knights:
and, indeed, Spenser had already conceived the idea of his knightly
romantic poem "The Faery Queen," and was ambitious to excel his
model, Ariosto. In this Harvey discouraged him; "Hobgoblin" must
not "run away with the garland from Apollo".
Fortunately Spenser followed his own genius, and, though he dallied
with the fashion for wedding Greek measures to English words, as in
the English hexameters of Watson and Harvey, he dropped many
projects at which he had glanced, and was constant to his "Faery
Queen".
The manuscript of that great poem must have been the companion
of Spenser in many strange wanderings,
In savage soil far from Parnassus Mount,
as he says. He was attached, as we have seen, in 1578, to the
household of Leicester, and may have gone on a mission of his to
France. To be patronized by Leicester was to risk incurring the
enmity of Burleigh. The long rivalry between Elizabeth's brilliant and
wavering favourite—who once so nearly brought her into a plight
almost as bad as that of Mary Stuart—and her sagacious counsellor,
Sir William Cecil (Lord Burleigh)—who now and again saved his
Queen "as by fire"—might have furnished Spenser with a high theme
for a poetic allegory. But chance had made him Leicester's man, not
Burleigh's man, so that he never won the fortune for which he
sought. Who, indeed, would seek fortune in Ireland? Spenser did,
accompanying Lord Grey of Wilton to an isle more than commonly
distressful.
To the natural hatred between the Irish and their English invaders
was now added the fury of religious rancour. Rebellion after rebellion
was punished by horrible reprisals. Lord Grey is notorious for his
massacre of six hundred disarmed Italian and Spanish filibusters at
Smerwick (November, 1580), and the poet of the "Faery Queen" was
present at this abominable deed. It was neither without precedent
nor imitation. Seventy years later David Leslie, urged on by a
preacher, massacred the remnant of Montrose's Irish contingent at
Dunaverty. Spenser himself in his most Interesting "View of the
Present State of Ireland" says concerning the foreign prisoners,
"there was no other way but to make that short way with them
which was made". He defends Grey's ruthless policy; he had made
Ireland "ready for reformation" when he was recalled, on the charge
of being "a bloody man" who had left the country in ashes (1582).
Grey was pursued by the clamour of a horrified people, that is, he
was Spenser's Sir Arthegal, molested by the Blatant Beast, the
public. The idea of the public is a Blatant Beast is borrowed from
Plato.
It was in the service of Grey, and in a land laid waste, that Spenser,
acting as Grey's secretary during the horrors of the war in Munster,
wrote part of the "Faery Queen". He held public posts, was Clerk of
Decrees, and Clerk of the Council of Munster, he received 3000 acres
of land, and a ruinous castle of the Desmond family, Kilcolman,
between Mallow and Limerick (1586).
Unhappy was his fortune, but, in absence from London, he had the
advantage of being beyond the influences of the critical literary
society of the capital with its reviews in form of pamphlets, its
satires, jealousies, and quarrels. There is a record of a conversation
of 1584 (published in 1606) in which Spenser described to his
friends the aim and scope of the "Faery Queen". Each virtue was to
be incarnate in a knight, whose adventures should teach it by
example. In a letter to Raleigh, whom he met in Ireland, Spenser
says that Prince Arthur (as in the first Canto) is to be a perfect
exemplar of "the twelve private virtues". The Faery Queen herself is,
first, Glory in general and next Gloriana, the royal and "most virtuous
and beautiful" Queen Elizabeth, who also appears as Belphœbe. He
is to begin in the middle, before telling how knights, ladies, dwarfs,
and a palmer bearing an infant with bloody hands came seeking
adventures to a festival of the Faery Queen. "Many other adventures
are intermeddled."
The "Faery Queen" is not, and does not aim at being an epic. It is
without beginning, middle, or end, for the last six books were not
written, or the manuscript perished when Spenser was driven from
Kilcolman.
The original scheme is that of the "Morte d'Arthur," moralized, and
intermingled with allegory. The poem is an allegorical romance
adapted to the state of England, Ireland, and the Continent under
Elizabeth, and to the war of the Reformation against the dragon of
Rome and the Scarlet Woman of the Seven Hills, the seeming fair
and inwardly filthy Duessa, who is occasionally meant for Mary
Stuart. Such unity as the poem possesses is given by the conflict of
Good, as Spenser understood it, against Evil, private and public, the
vices, and the Church of Rome. The Red Cross Knight wears the
armour which St. Paul describes, and in which Bunyan equipped
Christian and Greatheart.
There are people, says Spenser, who prefer to have Virtue
"sermoned at large, as they use". But while Spenser insists on being
taken as a moral preacher in his way, his true ideal is Beauty, and it
is the gleam of Beauty that he follows as he wanders with knights
and ladies through enchanted forests, and "awtres dire". Like the
knights in the "Morte d'Arthur" he "rides at adventure"; in every
page a new adventure opens, and leads to others endlessly, through
conflicts with Saracens,—Sansfoy, Sansloy, Sansjoy,—with the wily
Magician, Archimage, and his glamour; with Despair, in a wonderful
passage; with dragons and dragonettes, with Acrasia and all the
charms of her abode of wanton bliss, which is depicted with great
enthusiasm (Book II, Canto XII). This canto is remote indeed from
the puritan taste, despite its moral ending
Let Gryll be Gryll, and have his hoggish mind,
But let us hence depart, whilst weather serves and wind.
The whole is derived, in the last resort, from the palace of Circe in
the Tenth book of the "Odyssey," and it is curious to compare the
severe and classic charm of the Greek with the boundless luxury of
the Italian Renaissance in Spenser.
The "Faery Queen," indeed, despite the moral intention, which is
perfectly sincere, is the very Lotusland of poetry. It is a garden of
endless varieties of delight, endless but not prolix, for there is a
perpetual change of scene and of characters and nothing is constant
but the long and ever-varying music of the verse, Spenser's own
measure, in which each stanza is a poem, while the strong stream of
melody carries the half-dreaming reader down the enchanted river,
and forth into the fairy seas.
The Spenserian measure with the Alexandrine that ends the stanza
may not be the best vehicle for narrative. But Spenser's stream does
flow from the mountains of Lotusland, and the air of Lotusland
occasionally lulls the vigilance of the poet as well as of the the
reader. The stanza (Book VI, Canto X) which opens
One day, as they all three together went
To the greene wood to gather strawberries,
There chaunst to them'a dangerous accident:
A Tigre forth out of the wood did rise,
narrates an accident as unexpected as dangerous! We cannot but be
reminded of the "Swiss Family Robinson," and when Spenser makes
Sir Calidore kill the tiger and cut off its head with a shepherd's
crook, he is plainly overcome by "drowsihead".[2]
It is true that Spenser soon lost hold of his main allegory, and
allegorized the moving events and some of the personages of his
time. The gods, in Euripides, make a false Helen of clouds and
sunbeams and for her the Trojans and Achæans war and die. So, in
Spenser's poem, the witch makes a false Florimel of snow, informed
by "a wicked spright" with burning eyes for the destruction of
mankind, and the false Florimel is another form of the white witch,
Mary Stuart. The affairs of Ireland, France, "Belge," and Spain
appear in knightly or magical disguise in the procession of dissolving
views; a pageant of the rivers of Ireland and England anticipates
Drayton's "Polyolbion": the romance becomes, like "Piers Plowman,"
a farrago of all that is in the poet's mind.
Of Spenser, Ben Jonson might have said, as of Shakespeare,
Sufflaminandus erat, "he needed to have the drag put on". Like
Pindar in youth, "he sowed from the sack, not from the hand". His
archaic words and unsuccessful imitations of archaic words annoyed
the critics of his time more than they vex us. If he "writ no
language," "writ the language of no time," as Ben Jonson said, the
"Iliad" and "Odyssey," too, are in the language of no time, represent
no one dialect that ever was actually spoken. But Spenser was
writing about no actual time: his own age is confused with the fairy
age of chivalry, and the ages of the "Morte d'Arthur," and of Greek
mythology. With Spenser we are "out of space, out of time," and of
his adoration of Chaucer, his ancient words keep us in mind. That
great and noble effort towards perfection, the spirit of chivalry, was
his ideal; and in Sir Philip he saw the last of the gentle and perfect
knights. To the flattery of Elizabeth we must submit: she needed it
all if to her subjects she was to, stand for England and their love of
England.
Spenser's blemishes are of his age; no pure and perfect work of
immaculate art could arise in a poetry which was only emerging
from a kind of chaos, too much learning being the successor of too
much ignorance, and a divine genius being left at large with no
control from sane and temperate criticism.
Somewhat eclipsed by the new star of Elizabeth's fresh favourite,
Essex, Raleigh visited his Irish lands in 1589, met Spenser, read the
"Faery Queen" in manuscript, and brought "Colin Clout Home again".
The poem of that name (1591) while full of sugared compliments to
Elizabeth, is also touched with satire of her new courtiers. Sidney
was dead, Leicester was dead, Burleigh "hated poetry and painting".
The first part of the "Faery Queen" (1590) had made Spenser
famous, but had won him no prize of Court favour save a small
pension.
His "Mother Hubberd's Tale of the Ape and the Fox" may have been
written earlier and now was published; in this the satire is much
more keen; the poet finds even "the Comic Stage defaced and
vulgarized, in his 'Tears of the Muses,' where "our pleasant Willy that
is dead of late," cannot conceivably be Shakespeare—the silence of
John Lyly may be intended.
When Spenser returned to Ireland a collection of his miscellaneous
poems was published, containing, among other things, "Mother
Hubberd's Tale," "The Tears of the Muses," "The Ruines of Rome"
(sonnets from the French of Joachim du Bellay).
The "Ruines of Time," dedicated to "Sidney's sister, Pembroke's
mother," Lady Pembroke, begins with a vision of the genius of the
ruined Roman city, Verulam, and in a far-off way reminds us of the
Anglo-Saxon poem on the Ruined City. There is a lament for the fall
of ancient empires, and the sorrows of the House of Dudley.
Spenser's mood was that of melancholy and disappointment,
presently cheered by his marriage with Elizabeth Boyle. From his
love came his sonnets, and his matchless "Epithalamion," his "love-
learned song". If the "Faery Queen," and all else that Spenser did
were lost, the "Epithalamion" and the "Prothalamion" would win for
him the crown of the chief of English poets before Shakespeare. The
marriage occurred in June, 1594: then troubles with the Irish whom
he had supplanted, or some other cause, sent him to England, with
the last three books of his romance. The affair of Duessa's treatment
caused James VI to remonstrate through Bower, the English
ambassador to Holyrood, and though the poet was not punished, his
designs may not have been advanced. He now published his Hymns
to Love and Beauty, Earthly and Heavenly, the latter under the
influence of Plato, and his "Prothalamion" for the Ladies Elizabeth
and Katherine Somerset. These splendid poems were his swan-song;
Ireland called him, and in October, 1598, the natives whom he had
despoiled drove him from Kilcolman, which they burned. Spenser
died, a ruined man, in Westminster (16 January, 1599), Essex paid
for his funeral, he lies in Westminster Abbey.
As Hephæstus, when he fashioned the arms of Achilles, melted
bronze and gold and silver in his furnace, so Spenser combined the
wealth of Greece and Italy, France, Rome, and England in the great
crucible of his genius. In the "Epithalamium," for example, we find a
translation of four lines from a sonnet of Ronsard, mingling with
notes from Theocritus and the Song of Songs, with all the beautiful
things of all the creeds. It would, perhaps, be unfair to call the style
of Spenser, as it appears in the "Faery Queen," "Corinthian". Yet the
metal in which he works is like that "Corinthian bronze" formed, at
the conflagration of the city, from the molten gold and silver and
copper of the sacred vessels and images of the gods. The spoils of
all old poetry are mingled with his own. He has been called "the
poets' poet"; his successors have taken from him his very tones. As
has been said well, when Spenser writes—
Scarcely had Phœbus in the glowing East
Yet harnessëd his fiery-footed team,
that is Shakespeare, the Shakespeare of "Romeo and Juliet".
And taking usury of time forepast
Fit for such ladies and such lovely knights,
that is Shakespeare again, the Shakespeare of the Sonnets.
Many an Angel's voice
Singing before the eternal Majesty
For their triune triplicities on high:
that is the younger voice of Milton.
And ever and anon the rosy red
Flasht thro' her face,
one might fancy the unmistakable note and accent of Tennyson.[3]
English poetry fell with the neglect of Spenser, who was buried and
forgotten from the middle of the seventeenth century till Thomson
revived his measures in the middle of the eighteenth, and English
poetry came fully to her own again when the magic book of Spenser
was opened by Keats.
 [1] A well-known diplomatist of Queen Elizabeth, Harry Killigrew,
is said to have been "a Holbein in oils".
[2] On this and on the more than mediaeval size of "The Faery
Queen," see Mr. Mackail's "Springs of Helicon," pp. 132-28.
[3] Mackail, "Springs of Helicon," pp. 90, 91.

CHAPTER XIX.

THE ELIZABETHAN STAGE AND PLAYWRIGHTS.

The rejoicing age of Elizabeth was fond of "variety entertainments".

The Court Masques, such as those of Lyly, and George Peele's
"Arraignment of Paris," abounded in songs, music, and dancing, and
were expensively furnished. The Universities had their own amateur
authors and performers. The "children" of St. Paul's and other
schools acted so naturally that, as we read in "Hamlet," they became
serious rivals of the professional actors.[1] "An aery of children, little
eyases, that cry out on the top of question, and are most
tyrannically clapped for it, these are now the fashion". Polonius
indicates the many sorts of plays, "tragedy, comedy, pastoral,
pastoral-comical, historical-pastoral, tragical-historical, tragical-
comical, historical-pastoral, scene individual, or poem unlimited.
Seneca cannot be too heavy or Plautus too light." From authors of
the heavy Senecan school came blank verse: "the light people"
continued, when Shakespeare wrote "Love's Labour's Lost," to
employ rhymes in many measures; till Peele, and above all Marlowe,
introduced a more free and varied and accomplished blank verse.
The general taste turned from many imitations of the ponderous
Seneca to plays of more freedom, but even moralities and interludes
of the old sort continued to be played in the age of the
Shakespearean drama.
There were countless troops of players, vagabonds in the eyes of the
law—those who held no licence from a noble (as "the Earl of