Download the full version of the textbook now at textbookfull.

com

Methodological Issues in Psychology :

Concept, Method, and Measurement 1st Edition
David Trafimow

https://textbookfull.com/product/methodological-
issues-in-psychology-concept-method-and-
measurement-1st-edition-david-trafimow/

Explore and download more textbook at https://textbookfull.com

 Recommended digital products (PDF, EPUB, MOBI) that
you can download immediately if you are interested.

Biota Grow 2C gather 2C cook Loucas

https://textbookfull.com/product/biota-grow-2c-gather-2c-cook-loucas/

textbookfull.com

Equity and justice in developmental science : theoretical

and methodological issues 1st Edition Horn

https://textbookfull.com/product/equity-and-justice-in-developmental-
science-theoretical-and-methodological-issues-1st-edition-horn/

textbookfull.com

Decent Work: Concept, Theory and Measurement 1st Edition

Nausheen Nizami

https://textbookfull.com/product/decent-work-concept-theory-and-
measurement-1st-edition-nausheen-nizami/

textbookfull.com

Progress in Optomechatronic Technologies Proceedings of

International Symposium on Optomechatronic 2018 Amalia
Martínez-García
https://textbookfull.com/product/progress-in-optomechatronic-
technologies-proceedings-of-international-symposium-on-
optomechatronic-2018-amalia-martinez-garcia/
textbookfull.com
Microbial Resources From Functional Existence in Nature to
Applications Ipek Kurtböke (Eds.)

https://textbookfull.com/product/microbial-resources-from-functional-
existence-in-nature-to-applications-ipek-kurtboke-eds/

textbookfull.com

Elite Schooling and Social Inequality: Privilege and Power

in Ireland's Top Private Schools 1st Edition Aline
Courtois (Auth.)
https://textbookfull.com/product/elite-schooling-and-social-
inequality-privilege-and-power-in-irelands-top-private-schools-1st-
edition-aline-courtois-auth/
textbookfull.com

Communication Technologies for Vehicles 10th International

Workshop Nets4Cars Nets4Trains Nets4Aircraft 2016 San
Sebastián Spain June 6 7 2016 Proceedings 1st Edition
Jaizki Mendizabal
https://textbookfull.com/product/communication-technologies-for-
vehicles-10th-international-workshop-nets4cars-nets4trains-
nets4aircraft-2016-san-sebastian-spain-june-6-7-2016-proceedings-1st-
edition-jaizki-mendizabal/
textbookfull.com

Over-the-Counter Natural Cures: Take Charge of Your Health

in 30 Days with 10 Lifesaving Supplements for under $10
Shane Ellison
https://textbookfull.com/product/over-the-counter-natural-cures-take-
charge-of-your-health-in-30-days-with-10-lifesaving-supplements-for-
under-10-shane-ellison/
textbookfull.com

Statistics for Technology A Course in Applied Statistics

Third Edition Chatfield

https://textbookfull.com/product/statistics-for-technology-a-course-
in-applied-statistics-third-edition-chatfield/

textbookfull.com
The Cambridge Companion To Literature And Food J. Michelle
Coghlan

https://textbookfull.com/product/the-cambridge-companion-to-
literature-and-food-j-michelle-coghlan/

textbookfull.com
Methodological Issues in Psychology

Methodological Issues in Psychology is a comprehensive text that challenges current practice in

the discipline and provides solutions that are more useful in contemporary research, both basic
and applied.
This book begins by equipping the readers with the underlying foundation pertaining to
basic philosophical issues addressing theory verification or falsification, distinguishing different
levels of theorizing, or hypothesizing, and the assumptions necessary to negotiate between
these levels. It goes on to specifically focus on statistical and inferential hypotheses including
chapters on how to dramatically improve statistical and inferential practices and how to
address the replication crisis. Advances to be featured include the author’s own inventions,
the a priori procedure and gain-probability diagrams, and a chapter about mediation analyses,
which explains why such analyses are much weaker than typically assumed. The book also
provides an introductory chapter on classical measurement theory and expands to new concepts
in subsequent chapters. The final measurement chapter addresses the ubiquitous problem of
small effect sizes in psychology and provides recommendations that directly contradict typical
thinking and teaching in psychology, but with the consequence that researchers can enjoy
dramatically improved effect sizes.
Methodological Issues in Psychology is an invaluable asset for students and researchers of
psychology. It will also be of vital interest to social science researchers and students in areas
such as management, marketing, sociology, and experimental philosophy.

David Trafimow is a Distinguished Achievement Professor of psychology at New Mexico

State University, a Fellow of the Association for Psychological Science, Executive Editor of
the Journal of General Psychology, and for Basic and Applied Social Psychology. He received
his PhD in psychology from the University of Illinois at Urbana-Champaign in 1993. His
current research interests include attribution, attitudes, cross-cultural research, ethics, morality,
philosophy and philosophy of science, methodology, potential performance theory, the a priori
procedure, and gain-probability diagrams.
Methodological Issues in Psychology
Concept, Method, and Measurement

David Trafimow
First published 2024
by Routledge
605 Third Avenue, New York, NY 10158
and by Routledge
4 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2024 David Trafimow
The right of David Trafimow to be identified as author of this work has been asserted in
accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988.
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any
form or by any electronic, mechanical, or other means, now known or hereafter invented,
including photocopying and recording, or in any information storage or retrieval system,
without permission in writing from the publishers.
Trademark notice: Product or corporate names may be trademarks or registered trademarks,
and are used only for identification and explanation without intent to infringe.
Library of Congress Cataloging-in-Publication Data
Names: Trafimow, David, author.
Title: Methodological issues in psychology : concept, method, and
measurement / David Trafimow.
Description: New York, NY : Routledge, 2023. | Includes bibliographical
references and index.
Identifiers: LCCN 2023007447 (print) | LCCN 2023007448 (ebook) | ISBN
9781032429816 (hardback) | ISBN 9781032429786 (paperback) | ISBN
9781003365167 (ebook)
Subjects: LCSH: Psychology--Research--Methodology.
Classification: LCC BF76.5 .T727 2023 (print) | LCC BF76.5 (ebook) | DDC
150.72--dc23/eng/20230415
LC record available at https://lccn.loc.gov/2023007447
LC ebook record available at https://lccn.loc.gov/2023007448
ISBN: 978-1-032-42981-6 (hbk)
ISBN: 978-1-032-42978-6 (pbk)
ISBN: 978-1-003-36516-7 (ebk)
DOI: 10.4324/9781003365167
Typeset in Times New Roman
by Deanta Global Publishing Services, Chennai, India
Several people were influential in my writing this book. I thank Bas
Verplanken who first suggested the idea decades ago, and it has been in
the back of my mind ever since. I also thank Michael Hyman for helping
me spread scientifically revolutionary ideas to top business journals, a
desirable expansion. And Tonghui Wang and the mathematics group
have been fantastic at helping me translate conceptual ideas into precise
mathematical formulation. Finally, and most of all, I thank my wife
Sabine Trafimow who not only tolerates, but even encourages, my extreme
focus on improving the social and behavioral sciences.
Contents

Acknowledgments ix
Preface x

PART I
General Methodological Issues 1

1 A Philosophical Foundation 3

2 The Reality underneath the Reality: Examples from the Hard Sciences 18

3 The TASI Taxonomy and Implications 30

4 Why We Should Not Engage Null Hypothesis Significance Testing 47

5 How to Think about Replicating Findings 62

6 The A Priori Procedure (APP) 74

7 Gain-Probability Diagrams 89

8 The Unfortunate Dependence of Much Social Science on Mediation Analysis 106

PART II
Measurement Issues 123

9 The Classical Theory and Implications 125

10 Potential Performance Theory 144

11 Auxiliary Validity 161

viii Contents

12 Unit Validity and Why Units Matter 171

13 A Tripartite Parsing of Variance 182

14 Shocking Measurement Implications 195

References 209
Index 217
Acknowledgments

There are many people to thank for their influence on this book. Most of all, my wife, Sabine,
who provided all the beautiful artwork as well as constant encouragement. In addition, I thank
two wonderful friends, Mike and Stacey. Mike did a wonderful job providing high resolution
scanning and aid transmitting huge files to the publisher. Stacey was a consistent source of
support. In addition, I thank the many people who have collaborated with me on research and
thereby aided me in thinking through some of the ideas presented in the book.
Then too, others have contributed less directly but nevertheless importantly. My parents,
Alice and Jordan, provided much of my education and especially the lesson that the questions
one asks may be as, or even more, important than the answers. I also thank my brother, Jonnie,
and sister, Janet, for the advice, support and help provided over the years. I have also benefitted
from my sister-in-law, Elsa, and my brother-in-law Uli who has had innumerable conversations
with me about my research and helped me sharpen my thinking. My friends Butch, Doris, and
Joy also provided encouragement. Finally, my acknowledgments would be incomplete without
mentioning my good friend Bas, who has been encouraging me to write this book for years and
is already after me to write another book.
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
Preface

I have long felt that the social and behavioral sciences are not doing well, and not doing nearly as
well as they ought to do. To dramatize the point—and I am an opera fanatic, so I love drama—
imagine being in graduate school 150 years ago, when psychology was starting to become a
formal science. Suppose someone asked you to make a prediction about how much progress
there would be in psychology by the year 2023. When I imagine myself in this scenario, my
strong belief is that I would have predicted much more progress than has been attained. That is,
I would not only have been wrong, but way wrong!
All would be well if I were way wrong in the good direction, the direction of underpredicting
progress. But being way wrong in the bad direction, the direction of overpredicting progress, is
intolerable even though my wrongness is completely hypothetical. Of course, it is easy to make
excuses, and I have heard many of them.

• Psychology is too multi-causal.

• There is no way to observe mental processes.
• Psychology is too young a science.
• Etcetera.

Although one or more of the bullet-listed excuses may sound reasonable, and may even be true,
I do not buy any of them. Making excuses is too easy. Why not instead try to figure out how to
make impressive progress despite the obstacles?
And that brings us to this book. My main premise is that the social and behavioral sciences
are not nearly as impressive as they should be after all this time, an issue explained in detail in
Chapter 2. We not only ought to do better, but can do better, and the goal of this book is to push
the social and behavioral sciences in the direction of improved progress. Hence, I do not rehash
what other methodology textbooks say—these have not sufficiently stimulated progress. Rather,
this book is very much about how we can do better than what we are doing, and you will find
many places where I disagree with traditional or textbook thinking and suggest superior alterna-
tive thinking. Put simply, this is a book about what I think, not about methodological traditions,
and so there is no attempt at methodological coverage. This is a selfish book in the sense that it
is about what I want to say for improving research, not about what anyone else has said.
Consequently, this was an easy book to write as most of it is from my own mind. There was
little reason for external research, which is what usually causes writing a book to be a difficult
process. There is nothing like a labor of love to render writing easy. I hope you gain as much
from reading the book as I did from writing it.
David Trafimow
January 11, 2023
Part I

General Methodological Issues

1 A Philosophical Foundation

Imagine the theory that attitudes—the extent to which people like or dislike performing behav-
iors—cause behaviors (e.g., Fishbein, 1980; Fishbein & Ajzen, 1975; 2010). The theory might
be true or false. The obvious course of action is to test the theory in some way by comparing

DOI: 10.4324/9781003365167-2
4 General Methodological Issues

or contrasting it against observations. To use a convenient term that means the same thing, an
attitude researcher would wish to submit the theory to an empirical test (an observation or set
of observations that bear on the theory). If the results of the empirical test are in line with the
theory, then the theory is supported; but if the results of the empirical test contradict the theory,
then the theory is not supported.
For now, let us postpone questioning how a researcher traverses the distance from the theory
to an empirical test, and instead focus on the strength with which the findings obtained from
the empirical test justify conclusions concerning the theory. To continue with attitudes, imagine
that a researcher randomly assigns participants to read or not read an essay in favor of wearing
seatbelts. The researcher intends that reading the pro seat belt essay will increase positive atti-
tudes towards wearing seatbelts relative to not reading it. In psychology terminology, the essay
group—the group that gets the treatment—is called the experimental group or experimental
condition, whereas the other group is called the control group or control condition. The predic-
tion is that participants in the experimental condition will wear seatbelts more than participants
in the control condition.
Suppose that the researcher’s prediction pans out; participants in the experimental condition
do wear seatbelts more than participants in the control condition. Is the researcher justified in
concluding that the theory is true?
Unfortunately, this happy conclusion is premature. Perhaps the prediction panned out for a
reason other than that the essay caused attitude change that, in turn, caused behavior change.
For example, perhaps the researcher simply got lucky and the participants in the experimental
condition just happened to be more prone to wearing seatbelts than participants in the control
condition, even notwithstanding any influence on the part of the essay. Or perhaps the essay
worked by influencing something other than attitudes, such as affect towards the behavior (the
essay made people happier about wearing seatbelts). In general, the fact that an empirical test
pans out in the desired way is insufficient for concluding that the theory is true. The best one can
conclude is that the empirical test supports, but does not prove, the theory.
Philosophers sometimes use logic to make this point. Consider the following syllogism (con-
clusion drawn from two premises).

Syllogism 1
If the theory is true, then the prediction should occur. (Premise 1)
The prediction occurs. (Premise 2)
Therefore, the theory is true. (Conclusion)

Of course, the syllogism is flawed because, as we have seen, the prediction could have been
confirmed for a reason other than the theory, thereby compromising the ability of the confirmed
prediction to prove the theory. Philosophers often term syllogisms that are problematic in this
way invalid. A valid syllogism is one where if the premises are true, the conclusion must be true
(e.g., Syllogism 2, coming up soon); and an invalid syllogism is one where even giving the truth
of the premises fails to prove the conclusion (e.g., Syllogism 1). In addition, a syllogism can be
sound or unsound. A syllogism is sound if (a) it is valid and (b) the premises are true. Otherwise,
the syllogism is unsound. Syllogism 1 is both invalid and unsound. An example of a valid (good
logic) but unsound (a false premise) syllogism is the following: if the moon exists, it is made of
green cheese; the moon exists; therefore, the moon is made of green cheese.
But now let us switch courses and imagine that the prediction does not pan out; the researcher
observes no difference in seatbelt wearing between participants in the experimental condition
 A Philosophical Foundation 5

and participants in the control condition. Does the failed prediction validly prove the theory
false? Interestingly, from the point of view of strict logic, it does, though we will see later that
strict logic has limitations. To see this, let us rewrite the foregoing syllogism but adjusting for
the fact that we are now supposing that the prediction fails.

Syllogism 2
If the theory is true, then the prediction should occur. (Premise 1)
The prediction does not occur. (Premise 2)
Therefore, the theory is not true. (Conclusion)

In contrast to Syllogism 1, where assuming the truth of the two premises fails to validly prove
the truth of the conclusion, in Syllogism 2, the truth of the two premises does prove the truth of
the conclusion. And so, we have an asymmetry where predictions that pan out do not prove the
theory true, but predictions that fail prove the theory false.
Lest the invalidity of Syllogism 1 and the validity of Syllogism 2 are elusive to the reader, it
may be more intuitive to think in terms of more concrete entities. Consider a rewritten Syllogism
1, that has the same form as the original Syllogism 1, but in terms of rain and sidewalks instead
of in terms of theories and predictions.

Rewritten Syllogism 1
If it is raining, then the sidewalks are wet. (Premise 1)
The sidewalks are wet. (Premise 2)
Therefore, it is raining. (Conclusion)

The problem with Syllogism 1 is obvious immediately: someone might have left the sprinkler
system on, thereby causing the sidewalks to be wet, even in the absence of rain. An easy way to
check the validity of any syllogism is to change the entities to rain and sidewalks but keeping
the form the same. Let us now move to Syllogism 2 and rewrite it similarly.

Rewritten Syllogism 2
If it is raining, then the sidewalks are wet. (Premise 1)
The sidewalks are not wet. (Premise 2)
Therefore, it is not raining. (Conclusion)

There is clearly no way for both premises to be true and the conclusion to be false; thus,
Syllogism 2 is valid. As a test, consider a contrary argument that perhaps a tarp is covering the
sidewalks thereby keeping them dry even when it is raining. But the problem with the contrary
argument is that if a tarp is covering the sidewalks, then Premise 1 is false. And if one of the
premises is false, then we do not have a fair test of validity which, to remind the reader, is that
if the premises are true, then the conclusion must be true.
Now that we see that Syllogism 1 is invalid, even when rewritten to feature rain and side-
walks, and that Syllogism 2 is valid, even when similarly rewritten, what follows? To reiterate,
there is an asymmetry: it is not possible to prove theories with confirmed predictions, but it is
possible to disprove theories with disconfirmed predictions. This asymmetry suggests that it
might be better, or at least more definitive, to have predictions fail to pan out than have them
succeed. After all, a successful prediction is insufficient to draw a strong conclusion, such as
that the theory is true. However, a failed prediction is sufficient to draw a strong conclusion, that
6 General Methodological Issues

the theory is false. Theory disconfirmation is logically possible whereas theory confirmation
is not, and it is better to do that which is possible than that which is not possible. Thus, many
philosophers, most notably Karl Popper (e.g., 1963; 1972), have argued that it is better to try to
disconfirm theories than to try to confirm them.
However, if it is only possible to disconfirm theories, and not possible to confirm them, one
might wonder how it is possible to make progress in science. Progress may seem impossible
if there is no way to prove theories true. Popper suggested an interesting solution. According
to Popper, science can progress by disconfirming theories and replacing them with better theo-
ries. The scientific program, then, is one where researchers propose theories and subject them
to empirical tests. If the predictions pan out, no conclusion is drawn, though the theory can be
said to have at least withstood a test. But failed predictions disconfirm theories, which can be
replaced with better theories, to the betterment of science. Sometimes philosophers claim that
failed predictions falsify theories (prove them false), and the larger program of falsifying theo-
ries to replace them with better ones can be considered a falsificationist program or philosophy
of science. In summary, the asymmetry the syllogisms illustrate implies that falsifying theories
is better than verifying them because the former is logically possible whereas the latter is not.
Popper’s falsificationist program implies something interesting about scientific theories that
has had a strong influence in psychology, including at the level of introductory textbooks. Put
simply, if the goal of a research scientist ought to be to falsify theories, so they can be replaced
with better ones, an implication is that the theories must be capable of being shown wrong in
the first place. In a word, theories must be falsifiable. Theories that are not falsifiable cannot
be shown wrong, by definition, and so the falsificationist program cannot proceed. In contrast,
theories that are falsifiable allow the falsificationist program to proceed. For this reason, the
methodology chapter of many psychology textbooks includes some sort of statement about how
theories must be falsifiable to be of value to psychology (or any field that considers itself to be
a science). Unfortunately, although many textbooks extol the importance of having falsifiable
theories, they do not explain the underlying justification for the claim, particularly the asym-
metry illustrated by Syllogism 1 and Syllogism 2.
What would be an example of an unfalsifiable theory? A typical example would be the theory
that God created the world. It is difficult to imagine a way to subject this theory to an empiri-
cal test that would have a chance of disconfirming the theory. The falsificationist argument
would not be that the theory is wrong, only that it is unfalsifiable and consequently of no use to
scientists.
There is one last point to be made. Many philosophers distinguish between theories that have
been falsified, theories that have not been falsified but are falsifiable in principle, and theories
that are not falsifiable in principle. It is only the theories that are not falsifiable in principle that
are not considered useful in science. The theory that God created the world is considered unfal-
sifiable in principle because it is difficult to imagine a way to submit that theory to an empirical
test. However, the foregoing arguments need not be the last word.

Contrary Arguments
We have thus far only considered strict logic, but there is more to science than that. For exam-
ple, Kuhn (1962) invoked the notion of incommensurability, to be explained shortly, to argue
that there is no way to have crucial (definitive) experiments to falsify theories. Aside from
Kuhn, a limitation is that logic is strongly subject to the nature of the premises, with different
premises supporting different conclusions. Another limitation is that the premises might not be
true, at least not without modification. In that case, as we have seen, syllogisms with at least one
 A Philosophical Foundation 7

false premise are unsound, even if logically valid. The subsequent subsections address each of
these issues, in turn.

Kuhnian Incommensurability

Put briefly, Kuhn (1962) argued that there is a prescientific phase where there is no dominant
paradigm, that is, no dominant way to perform research. Once there is a dominant paradigm,
then researchers use that paradigm to solve all sorts of problems, many of which may not have
been considered by the original inventors of the paradigm. Kuhn termed this “normal science”
where researchers solve small puzzles but do not question foundational assumptions. However,
even with a dominant paradigm, there may be one or more cases where an empirical prediction
fails to pan out. Such empirical anomalies may pile up as experiments continue to be performed.
At some point, researchers may propose a different theory to handle the anomalies, and the
new theory may be in competition with the old theory. The obvious course of action would be
for researchers to test the two theories against each other, to determine the best one. However,
Kuhn argued that researchers cannot take this obvious course of action. The reason is that either
different theories use different words or, even if they use the same words, the words have differ-
ent meanings in the competing theories. Because of the different meanings, there is no way to
directly compare the theories, and no way to subject them to competitive testing.
To see this in detail, let us consider the use of the word ‘mass’ from a Newtonian (1642–1727)
and Einsteinian (1879–1955) perspective. An important characteristic of mass, for Newton, is
that it does not change with velocity. For example, the same rocket ship traveling through space
would have the same mass regardless of whether it would be traveling at 5,000 miles per hour or
50,000 miles per hour. In contrast, for Einstein, mass does change depending on velocity, with
more velocity implying more mass. So, a rocket ship traveling at 50,000 miles per hour would
have slightly more mass than a rocket ship traveling at 5,000 miles per hour. Thus, the word
has different meanings for Newton and Einstein, and the two theories are, to use Kuhn’s (1962)
famous word, incommensurable. That is, despite using the same words, they are not speak-
ing the same language, and there is no way either theory can be used to understand the other.
There is no way to relate crucial terms, such as mass, as used in one theory, to the other theory.
Without the ability to relate crucial terms across theories, there is no way to design definitive
ways to test the theories against each other. And if there is no way to perform definitive empiri-
cal tests, there is no way to falsify either theory. (A related complication, glossed over here for
the sake of brevity, but explained in the Appendix on Definitions at the end of this chapter, is
the lack of explicit definitions.)
Because of the impossibility of falsifying either theory, Kuhn (1962) argued that the theory
that comes out on top does so largely due to sociological factors. As aficionados of the older
theory retire or die, they are replaced with new scientists who are more likely to have been
indoctrinated in the newer paradigm or who are otherwise more open to it. Thus, what Kuhn
considered revolutionary science is largely a matter of a process where new paradigms are
developed to handle empirical anomalies, the new paradigms are incommensurable with older
ones thereby precluding empirical tests from being definitive, and sociological factors largely
determine the winning paradigm. In this Kuhnian scheme, there is no way for the falsificationist
program to work.
However, it is possible to criticize Kuhnian incommensurability. Consider an experiment
where two clocks are synchronized. One clock is placed in an airplane which, after taking off,
is moving considerably faster than the clock that remains on the ground. According to Newton,
there is no reason for the two clocks to be become out of synch during the flight. However,
8 General Methodological Issues

according to Einstein, greater velocity implies a slower passage of time. Therefore, the clock
that was on an airplane should be slightly behind the other clock due to its faster velocity.
The moral of this empirical story is that even if there is incommensurability at the theoreti-
cal level, there need not be incommensurability at the empirical level. A Newtonian scientist
and an Einsteinian scientist may disagree about the meaning of mass, but they can agree on
clock readings. Hence, it is possible to perform definitive empirical tests despite theoretical
incommensurability.
Or consider another test, concerning the precession of the perihelion of Mercury (how the
point of closest approach of Mercury to the sun changes). The value is different for Newton and
Einstein, and Einstein’s value is closer to the empirically determined one (see Einstein, 1961,
for details). Although the theories may be incommensurable, distance and angular distance
measurements are not incommensurable, any more than time measurements are incommensura-
ble in the clocks example. Aficionados of different theories can agree on measurements despite
their theoretical disagreements. Thus, theoretical incommensurability need not translate into
empirical incommensurability. In turn, if theories are empirically commensurable, regardless
of theoretical incommensurability, then falsification is possible after all. Therefore, Kuhnian
incommensurability constitutes an insufficient reason to reject the falsificationist program. The
following two subsections will be more problematic for the falsificationist program.

The Form of the Major Premise

Consider again the first premise in Syllogisms 1 and 2, sometimes termed the major prem-
ise because it is a general statement: “If the theory is true, then the prediction should occur.”
However, there is no law requiring that the major premise be stated in that form, as I pointed out
in a recent article (Trafimow, 2020). We might state the major premise in a different way: “If
the theory is not true, then the prediction should not occur.” And if we change the major premise
in this way, it is easy to show that Syllogism 1, which used to be invalid, now becomes valid.
And Syllogism 2, which used to be valid, now becomes invalid. In turn, because the previous
argument made in favor of a falsificationist philosophy of science depended on the invalidity of
Syllogism 1, and the validity of Syllogism 2, reversing these is inconvenient, to say the least.
Let us commence with Syllogism 1, but with the new major premise.

Syllogism 1 With the New Major Premise

If the theory is not true, then the prediction should not occur. (Premise 1)
The prediction occurs. (Premise 2)
Therefore, the theory is true. (Conclusion)

Perhaps surprisingly, when we change the form of the major premise so that falsity of the theory
implies that the empirical prediction fails, the entire syllogism that verifies the theory suddenly
becomes valid. We have the opposite of that which was argued previously because now theory
confirmation is logically possible.
And we can play a similar game with Syllogism 2. However, substituting the new major
premise for the old one makes what used to be a valid syllogism now invalid.

Syllogism 2 with the New Major Premise

If the theory is not true, then the prediction should not occur. (Premise 1)
The prediction does not occur. (Premise 2)
Therefore, the theory is not true. (Conclusion)
 A Philosophical Foundation 9

If a prediction fails, the old Syllogism 2 shows that the failure disconfirms the theory. And this
demonstration is logically valid. In contrast, the new Syllogism 2, with the new major premise,
is not logically valid. It does not provide a sufficient reason to reject theories.
Well, then, if stating the major premise in one way supports falsifying theories as more
logical than verifying them but stating the major premise in a different way supports verifying
theories as more logical than falsifying them, it places the researcher who wishes to be logical
in a quandary. Is it better to try to falsify theories or is it better to verify them? Both are equally
logical or illogical, depending on how the major premise is stated.
Worse yet, the issue of how the major premise is stated imposes only one sort of problem.
Matters become more complex when we consider syllogism soundness, and the necessity to
consider auxiliary assumptions.

Auxiliary Assumptions

Let us return to our theory, that attitudes—the extent to which people like or dislike perform-
ing behaviors—cause behaviors. We have thus far ignored the issue of how one traverses the
distance between the theory and an empirical test. It is now convenient to address that issue.
As will become clear eventually, addressing that issue has a knock-on effect of influencing our
syllogisms, which, in turn, influences how we should think about falsification and verification.
To alert the reader, this will be a long subsection, but a crucial one that will figure strongly in
subsequent chapters.
To commence, consider that the focal construct in the theory—attitude—cannot be directly
observed. There is no known way to look at a person and see that person’s attitude. Thus, atti-
tude is a nonobservational term. In contrast, in the case of our hypothetical experiment involv-
ing an essay designed to influence attitudes, it is possible to observe the essay. Likewise, as a
generality, behavior cannot be observed, though a single behavior, such as wearing a seatbelt,
can be observed. Thus, attitude and behavior are nonobservational terms that can be distin-
guished from the essay designed to manipulate attitude or the single way a researcher measures
behavior, both of which can be observed. Stated more generally, theories contain nonobserva-
tional terms (e.g., attitude and behavior), whereas empirical predictions contain observational
terms (the essay and the single behavior measure can be observed). Another way to say the
same thing, that will be convenient throughout this book, is to term the empirical prediction an
empirical hypothesis. Using the new term, we could say that theories contain nonobservational
terms, empirical hypotheses contain observational terms, and the hope is that the observational
terms in the empirical hypothesis have something to do with the nonobservational terms in the
theory. With respect to attitudes and behaviors, the hope is that the essay really does manipulate
attitudes and that the measure of the extent to which people wear seatbelts really does measure
behavior. To foreshadow, the issue of measurement will be the focus of Part II of this book.
Of course, the distinction between observational terms and nonobservational terms is an
oversimplification because there are various degrees to which it is possible to observe enti-
ties. To illustrate, consider that germ theory includes the notion of germs, which could not be
observed prior to the invention of microscopes, but could be, and were, observed afterwards.
The extent to which germs should be considered nonobservational (because they cannot be
observed directly) or observational (because they can be observed with microscopes) is not
completely clear. Nevertheless, the ambiguity is not important for present purposes, though
it will become important in subsequent chapters. For now, it is sufficient that terms in empiri-
cal hypotheses are more observational than terms in theories. Because it is awkward to speak
in terms of entities being more observational, or less observational, I will continue to refer to
Visit https://textbookfull.com
now to explore a rich
collection of eBooks, textbook
and enjoy exciting offers!
10 General Methodological Issues

theoretical terms as nonobservational and empirical terms as observational, but please keep on
mind that these are relative terms.
Well, then, if theories contain nonobservational terms and empirical hypotheses contain
observational terms, how do we traverse the distance from theories to empirical hypotheses?
Obviously, something needs to be added to the system that connects nonobservational terms
in theories to observational terms in empirical hypotheses. That entity is termed auxiliary
assumptions.
Returning to attitudes and an essay designed to increase attitude positivity towards wearing
seatbelts, why should we believe that the essay does that which it was designed to do? Let us
consider two examples and your own intuitions.
For the first example, consider the following sentences that might be part of a larger essay.

There is much evidence that wearing seatbelts can prevent injuries and can reduce the prob-
ability of being killed in a car accident. Statistics indicate that wearing seatbelts reduces
the probability of injury in a car accident by more than 50%. Statistics also indicate that
wearing seatbelts reduces the probability of death by 67%.

The second example is similar in form, but with an obvious change.

There is much evidence that wearing hats can prevent injuries and can reduce the probabil-
ity of being killed in a car accident. Statistics indicate that wearing hats reduces the prob-
ability of injury in a car accident by more than 50%. Statistics also indicate that wearing
hats reduces the probability of death by 67%.

As you doubtless have noticed, the difference between the two essays is that the first one fea-
tures wearing seatbelts whereas the second one features wearing hats. Both essays provide the
same statistics, which I made up for the examples.
Which essay do you believe is more likely to influence attitudes towards wearing seatbelts?
Probably, you would choose the first essay over the second essay. A reason is that the first essay
features wearing seatbelts, which is relevant to attitudes towards wearing seatbelts, whereas the
second essay features wearing hats, which is irrelevant to attitudes towards wearing seatbelts.
Using more general language, the first essay includes the mention of the observational term of
importance in the empirical hypothesis, but the second essay does not. Therefore, the first essay
is more likely to change the relevant attitude than is the second essay. And we now arrive at an
auxiliary assumption, that the first essay is relevant to the type of attitude we wish to manipu-
late but the second essay is not relevant to the type of attitude we wish to manipulate. It bears
emphasis that the theory, itself, does not say anything about essays or about essay relevance;
this is an assumption that we must add to the theory to arrive at the empirical hypothesis. It is
an auxiliary assumption that connects the theoretical term—attitude—with the essay that can
be observed.
Of course, the assumption of relevance is not the only auxiliary assumption. There are many
more. To see this, let us return to the first essay and consider reasons why it might not work,
despite its relevance. One potential problem is that people might not care about injury and death
statistics, and if they do not care about these statistics, there is little reason for them to change
their attitudes. Thus, another auxiliary assumption is that participants care about injury and
death statistics.
 A Philosophical Foundation 11

Another potential problem might be that even if participants care about injury and death sta-
tistics, they might not believe that the statistics apply to them. Thus, even if the essay convinces
participants that people ought to wear seatbelts, in general, that general prescription does not
apply to them specifically. In turn, if the general prescription does not apply to them specifically,
then they need not wear seatbelts. Thus, another auxiliary assumption is that participants will
see the prescription in the essay as applying to them.
Worse yet, even if participants are convinced that they ought, themselves, to wear seatbelts,
they might not be interested in doing what they ought to do. Their attitudes—what they like or
dislike—might be completely uninfluenced by what they ought to do. Thus, yet another auxil-
iary assumption is that people’s attitudes become consistent with what they feel they ought to
do.
Finally, there are seeming trivial issues, such as the assumption that the researcher hands
participants the correct forms. Or if the experiment is performed via computer or cellphone, that
the program causes the correct forms to appear on participants’ computer or cellphone screens.
Obviously, if participants do not get the correct forms, the experiment is unlikely to be success-
ful. An auxiliary assumption, though often an implicit one, is that participants get the correct
forms.
We have seen that many auxiliary assumptions, often unstated, are necessary to connect
nonobservational terms in theories with observational terms in empirical hypotheses. But that is
not the only function auxiliary assumptions serve. Another function is that they set initial condi-
tions. For example, consider that to perform the experiment properly, it is necessary to randomly
assign participants to conditions. We will explore the statistical implications of random assign-
ment in a subsequent chapter. For now, however, it is sufficient that random assignment provides
reason to believe that the experimental and control conditions are approximately equivalent
with respect to causally relevant factors. One of these might be initial attitudes towards wearing
seatbelts. Although we would expect that different people would differ in their initial attitudes,
we would at least hope that the mean initial attitudes in the two randomly assigned conditions
would be approximately equal. If that is so, then the anticipated difference in mean seatbelt
wearing between the two conditions, if obtained, could not reasonably be attributed to an initial
difference in the two conditions. The assumption of initial equality increases our confidence that
the anticipated effect, if obtained, is due to reading the essay versus not reading it. In contrast,
suppose we were to doubt that the process of randomly assigning participants to conditions suc-
cessfully induces an initial equivalence of mean attitudes in the two conditions. In that case, we
would be wary of concluding that the anticipated effect, if obtained, would be due to reading
the essay versus not reading it. Thus, although our focus will be more on the ability of auxiliary
assumptions to bridge the gap between nonobservational terms in theories and observational
terms in empirical hypotheses than on setting initial conditions, both matter and matter a lot.
Once we understand that empirical hypotheses come not just from theory, but from auxiliary
assumptions too, a crucial implication is that Syllogism 2, the one that is used to justify the
falsificationist program, is unsound despite being logically valid. To see why, consider this syl-
logism again, copied for you below.

Syllogism 2
If the theory is true, then the prediction should occur. (Premise 1)
The prediction does not occur. (Premise 2)
Therefore, the theory is not true. (Conclusion)
12 General Methodological Issues

Our understanding of the importance of auxiliary assumptions renders Premise 1, the major
premise, untrue. Rather, the premise should be amended to say that if the theory is true and the
auxiliary assumptions leading to the empirical hypothesis are true, then the prediction should
occur. But once we amend Premise 1 accordingly, the conclusion no longer follows. Rather, the
best we can conclude from a failed prediction is that either the theory is false or at least one aux-
iliary assumption is false. This is an immense distance away from what had seemed a definitive
conclusion that the theory is false. In turn, if we cannot definitively conclude that the theory is
false, then we cannot carry through the falsificationist program.
To nail down this point, let us rewrite Syllogism 2 with the amended Premise 1 and the
amended conclusion.

Amended Syllogism 2 with Auxiliary Assumptions

If the theory is true and the auxiliary assumptions are true,
then the prediction should occur. (Premise 1)
The prediction does not occur. (Premise 2)
Therefore, the theory is not true or at least one auxiliary
assumption is not true. (Conclusion)

The good news about the amended Syllogism 2 is that it is logically valid; if Premise 1 and
Premise 2 are true, then the conclusion must be true. However, the bad news, as we have seen, is
that the conclusion is insufficient for falsificationist needs. To make the falsificationist program
work, it is necessary to be able to falsify the theory so it can be replaced with a better theory.
However, the conclusion is insufficient to falsify the theory because an auxiliary assumption
could be to blame for the failed prediction.
Before saying more about the implications of auxiliary assumptions for falsification, it is
worth inserting a quick note that auxiliary assumptions are a problem for verification as well
as for falsification. Just as a failed prediction can be blamed either on the theory or on at least
one auxiliary assumption, a successful prediction can be credited to either the theory or at least
one auxiliary assumption (Trafimow, 2017a). An important consequence is that a successful
prediction does not unambiguously support the theory because one or more powerful auxiliary
assumptions might be responsible. Therefore, the necessity to have auxiliary assumptions ren-
ders absolute falsifiability (or absolute verifiability) impossible. The best we can do is talk about
some sort of ‘reasonable’ falsifiability that would be limited by our confidence in the correctness
of the relevant auxiliary assumptions. The impossibility of absolute proof, either way, is widely
accepted (e.g., Duhem, 1954; Lakatos, 1978; Meehl, 1990; Trafimow, 2009).
We have seen that widening our discussion to include auxiliary assumptions, exemplified in
the Amended Syllogism 2, creates a problem for the falsificationist program because there is
now no way to unambiguously disprove theories. However, there is another problem that has
not received much attention but may be equally or even more important. To see the problem,
recall an earlier point that theories must be falsifiable, in principle, to be of scientific value. A
theory that is not falsifiable cannot be eliminated to make room for, and be replaced by, another
theory. An example of such a theory, as we saw earlier, is the creation theory, that God created
the world.
However, an imaginative consideration of auxiliary assumptions suggests that we should,
perhaps, rethink whether the creation theory is unfalsifiable. To set up the forthcoming argu-
ment, consider a common belief among scientists in the 1850s, that theories about the chemical
composition of stars are unfalsifiable, in principle. Why? Because there is no way to travel to
 A Philosophical Foundation 13

a star and gather samples of star material for chemical analysis. Soon after, however, the sci-
ence of spectroscopy was invented, and it became possible to test theories about the chemical
composition of stars based on the light stars emit. Moreover, the history of science is replete
with similar examples of seemingly unfalsifiable theories that were later subjected to convinc-
ing experimental tests and were shown to be reasonably falsified (but not falsified absolutely
because it could have been that one or more auxiliary assumptions were to blame).
Let us now return to the creation theory: God created the world. I will now argue that although
the theory is currently unfalsifiable, it is not unfalsifiable in principle under the umbrella of rea-
sonable falsifiability. To see this, imagine that a devotee invents a prayer that causes God to
appear and answer questions truthfully. The devotee asks God the following question: “God,
did you create the world?” God retorts: “Of course I did not create the world; if I had created the
world, I would have done a much better job!” God’s retort, if it were to happen, would falsify
the creation theory. The falsification is not absolute because we can question various auxiliary
assumptions, such as whether the being that answered the prayer really is God; whether the
being, even if it really is God, wanted to avoid responsibility for having done a poor job with
creation by denying having been responsible; and so on. Nevertheless, within the limits of the
auxiliary assumptions, there would be some degree of reasonable falsification, and so the crea-
tion theory is reasonably falsifiable, despite the seeming impossibility of testing it. Hence, crit-
ics of the creation theory who claim it is not falsifiable, even in principle, are on shaky grounds.
This is not to say that the creation theory is a good theory, but at least the falsifiability argument
is insufficient for discrediting it.
There is a reason for having emphasized that the creation theory is falsifiable, in principle,
despite many pronouncements that it is not. There is a fashion in psychology that ought to be
changed. Inevitably, if a theory survives long enough, someone will declare it unfalsifiable.
However, the creation example shows that it is practically impossible to have a theory that is
unfalsifiable, in principle. This is because there is no way of knowing whether a clever scientist
will invent the requisite auxiliary assumptions that allow the allegedly unfalsifiable theory to be
subjected to convincing empirical testing. We saw this in physics with the example of theories
of the chemical composition of stars. The invention of spectroscopy provided the necessary
auxiliary assumptions to allow convincing (though not absolute) tests of theories of the chemi-
cal composition of stars. The creation example shows how, in principle, it is practically always
possible to invent the requisite auxiliary assumptions to test theories under the rubric of reason-
able falsification.

Psychology Examples of Unfalsifiable Theories that Were Falsified

There is one psychology theory that has been deemed unfalsifiable in more textbooks than any
other, and that theory is Freud’s psychodynamic theory (1955; 1959). Why is this theory alleged
to be unfalsifiable? The answer is that Freud’s stages of psychosexual development are replete
with statements about unconscious processes that are typically pronounced untestable. After all,
how do we know there even is an unconscious, and even if we did know that, how can we know
what goes on there when there is no way for participants to tell us?
And yet, I was involved in what I consider a reasonably definitive test of this allegedly unfal-
sifiable theory. To understand the test, it is first necessary to understand something about psy-
chosexual development, according to Freud (see Sarnoff, 1976, for an accessible description).
Without going through all five of Freud’s stages, the crucial point is that parents have much
power over young children and much less power as children grow older, especially when they
14 General Methodological Issues

enter adolescence. And there are two reasons. The first reason is that very young children are at
a severe disadvantage compared to their parents with respect to both physical and mental matu-
rity. The second reason is that it takes time to develop effective defense mechanisms against
one’s own unconscious feelings of guilt that are tied to previous parental behavior. Young chil-
dren have not had time to develop such defense mechanisms, whereas they are well developed
by adolescence. Both because of the issue of development of physical and mental maturity, and
because of the issue of development of defense mechanisms, young children should mostly do
what their parents think they should do, whereas adolescents should be better able to do what
they want to do. In social psychology terminology, the behavior of young children should be
influenced mostly by subjective norms (what one thinks important others, such as parents, think
they should do) and the behavior of adolescents should be influenced mostly by attitudes (what
people like or want to do). I and some students and colleagues tested the idea by measuring
children’s attitudes, subjective norms, and behavioral intentions (what people intend to do) for
thirty behaviors (Trafimow et al., 2002). If Freud is correct, subjective norms should best pre-
dict behavioral intentions for young children, whereas attitudes should best predict behavioral
intentions for older children and adolescents. In contradiction to Freud, attitudes best predicted
behavioral intentions no matter the age of the children, nor did the extent of the prediction
depend on age. Thus, despite the vaunted impossibility, in principle, of testing Freud’s theory,
we not only tested it but falsified it too.
Another example concerns the theory of reasoned action (e.g., Fishbein, 1980). According to
Greve (2001), the theory is not falsifiable. For example, Fishbein (1980) argued against separate
cognitive (thinking) and affective (feeling) components of attitudes, thereby potentially rendering
the theory unfalsifiable because of the lack of specification of the nature of attitudes. However,
I and a colleague were nevertheless able to provide a strong test of the theory (Trafimow &
Sheeran, 1998). We made an auxiliary assumption that it would be possible for participants to
distinguish between cognitive and affective beliefs about the potential consequences of behav-
iors that cause attitudes. During attitude formation, cognitive beliefs are compared or contrasted,
thereby resulting in the formation of a cognition about the behavior; and affective beliefs are
compared or contrasted, thereby resulting in the formation of an affect about the behavior. The
consequence of these processes is the formation of associations between cognitive beliefs and
other cognitive beliefs, between affective beliefs and other affective beliefs, but not between
cognitive and affective beliefs. Thus, the retrieval of a cognitive belief should stimulate the
participant to follow an association to another cognitive belief, whereas the retrieval of an affec-
tive belief should stimulate the participant to follow an association to another affective belief. In
contrast, because there is a lack of associations between cognitive and affective beliefs, it should
be unlikely that a participant would retrieve an affective belief after a cognitive belief, or a cog-
nitive belief after an affective belief. We tested this reasoning by simply having people write
down their beliefs about behaviors and looking at the order of the retrieved items. If Fishbein is
correct in assuming a lack of distinction between the two kinds of beliefs (or between cognition
and affect), then the order of retrieval should be random. In contrast, according to our reasoning,
because of the pattern of associations among the beliefs, retrieved items should exhibit a clus-
tered order, where cognitive beliefs tend to be retrieved together and affective beliefs tend to be
retrieved together. The data strongly supported the latter prediction, thereby falsifying the lack
of distinction between cognition and affect assumed by the theory. Interestingly, a later version
of the theory incorporated the distinction (Ajzen & Fishbein, 2005).
Thus, two examples of how psychology theories that had been evaluated as unfalsifiable, in
principle, not only turned out to be falsifiable, but were falsified too. Of course, these successful
Random documents with unrelated
content Scribd suggests to you:
The Project Gutenberg eBook of D'un pays
lointain: Miracles; Visages de femmes;
Anecdotes
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: D'un pays lointain: Miracles; Visages de femmes; Anecdotes

Author: Remy de Gourmont

Release date: March 24, 2021 [eBook #64920]

Language: French

Credits: Laurent Vogel and the Online Distributed Proofreading Team

at https://www.pgdp.net (This file was produced from
images generously made available by The Internet
Archive/Canadian Libraries)

*** START OF THE PROJECT GUTENBERG EBOOK D'UN PAYS

LOINTAIN: MIRACLES; VISAGES DE FEMMES; ANECDOTES ***
 REMY DE GOURMONT

D’un
Pays Lointain
MIRACLES — VISAGES DE FEMMES
ANECDOTES

SIXIÈME ÉDITION

PARIS
MERCVRE DE FRANCE
XXVI, RVE DE CONDÉ, XXVI

MCMXXII

DU MÊME AUTEUR

Roman, Théâtre, Poèmes

SIXTINE.
 Le Fantôme. Le Château singulier. Théâtre
LE PÉLERIN DU SILENCE.
muet. Le Livre des Litanies. Pages retrouvées.
LES CHEVAUX DE DIOMÈDE.
D’UN PAYS LOINTAIN.
LE SONGE D’UNE FEMME.
LILITH, suivi de THÉODAT.
UNE NUIT AU LUXEMBOURG.
UN CŒUR VIRGINAL. Couverture de G. d’Espagnat.
COULEURS, suivi de CHOSES ANCIENNES.
HISTOIRES MAGIQUES.
DIVERTISSEMENTS, poésies complètes, 1912.

Critique, Littérature

LE LATIN MYSTIQUE (Etude sur la poésie latine du moyen-âge)

(Crès, éditeur).
LE LIVRE DES MASQUES (Ier et IIe), gloses et documents sur les
écrivains d’hier et d’aujourd’hui, avec 53 portraits par F.
Vallotton.
LA CULTURE DES IDÉES.
LE CHEMIN DE VELOURS. Nouvelles dissociations d’idées.
LE PROBLÈME DU STYLE. Questions d’Art, de Littérature et de
Grammaire.
PHYSIQUE DE L’AMOUR. Essai sur l’instinct sexuel.
ÉPILOGUES. Réflexions sur la vie, 1895-1898 ; 1899-1901 (2e
série) ; 1902-1901 (3e série) ; 1905-1912 (volume
complémentaire) ; 4 vol.
ESTHÉTIQUE DE LA LANGUE FRANÇAISE, édition revue, corrigée et
augmentée.
PROMENADES LITTÉRAIRES (1re, 2e, 3e, 4e et 5e séries) ; 5 vol.
PROMENADES PHILOSOPHIQUES (1re, 2e et 3e séries) ; 3 vol.
DIALOGUES DES AMATEURS SUR LES CHOSES DU TEMPS (Épilogues, 4e
série, 1905-1907).
NOUVEAUX DIALOGUES DES AMATEURS SUR LES CHOSES DU TEMPS
(Épilogues, 5e série, 1907-1910).
DANTE, BÉATRICE ET LA POÉSIE AMOUREUSE.
PENDANT L’ORAGE.
LETTRES A L’AMAZONE.
PENDANT LA GUERRE.
LETTRES D’UN SATYRE.
LETTRES A SIXTINE.

PAGES CHOISIES, avec un portrait.

JUSTIFICATION DU TIRAGE

Droits de traduction et de reproduction réservés pour tous pays.

 PROLOGUE

D’UN PAYS LOINTAIN

— D’où viens-tu?
— D’un pays lointain. Je suis né dans une maison noire surgie du
milieu d’une plaine grise, autour de laquelle un cercle de lumière
étincelait, pareil aux gloires où s’écrivent les traits sévères d’une
vierge de vitrail ; mais ce halo d’espérance et de bénédiction ne
ceignait que du néant, du gris et du noir. Mon père et ma mère,
comme tous les habitants de ce pays lointain, étaient aveugles ;
seuls, quelques enfants voyaient : si l’on s’en apercevait, on leur
crevait les yeux, — pour les rendre conformes. J’avais un frère, on
lui creva les yeux ; j’avais une sœur, on lui creva les yeux.
Pendant l’opération, pratiquée par un excellent prêtre, aimé de
tous et surtout du Seigneur, ma mère disait : « C’est un petit
moment à passer, mes chéris ; j’ai subi cela aussi, moi, à votre âge,
et je n’en suis pas morte. Allons, un peu de courage! » Elle
promettait des confitures, du sucre et des gâteaux à la fleur
d’oranger.
Mon père, qui était né aveugle, parla plus longuement. Il dit,
avec une rude tendresse : « Petits sauvages, vous n’avez donc
aucun sentiment des convenances? Ces gamins veulent se
distinguer! Ces gamins ne veulent pas faire comme tout le monde!
Alors, vous consentez à être ridicules, c’est-à-dire à éprouver des
sensations — et, de là, des sentiments ou des idées — inconnues et,
par conséquent, méprisées des autres hommes? Réfléchissez bien.
Si vous gardez vos yeux, cette source incongrue — à ce que l’on dit,
— de pensées vaines et de dangereux désirs, on vous poussera du
coude avec dédain, on vous marchera sur les pieds, on vous donnera
des coups de genou, par mégarde, on s’ameutera contre vous, on
vous tirera les cheveux et on dansera la sarabande autour de la bête
curieuse. Ah! vous vous préparez une jolie existence!…
— Mais ils ne refusent pas de se laisser crever les yeux!
interrompit ma mère. N’est-ce pas, mes chéris?
— Ils ne refusent pas? Je l’espère bien, mais je dois les prévenir
de ce qu’ils vont gagner à perdre le plus méprisable des sens, — et
de ce qu’ils perdraient à le conserver. Mes enfants, je puis vous
énumérer, avec ma double autorité d’aveugle et de père, les joies
d’un être privé de la vue : la première joie est une joie intime et
profondément satisfactoire, la joie de la répulsion surmontée, du
devoir accompli ; en second lieu, vous ressentirez un plaisir
d’orgueil, mais d’orgueil permis, le plaisir d’être absolument pareil à
tous vos petits camarades, le plaisir de vivre parmi des égaux ; ce
plaisir vous accompagnera durant toute votre vie, enfin, châtrés de
la vue, vous aurez conquis la paix qui naît de l’incuriosité ; après de
calmes jeux, de douces études de paisibles amours, de bons repas,
de propices digestions, vous vous endormirez dans la certitude de
n’être jamais sortis du droit chemin, de n’avoir jamais cueilli aucune
fleur, de n’avoir jamais contemplé le ciel, ni la nuit, quand — dit-on
— il s’orne du regard attristé des séraphins, ni le jour, quand le
Soleil, ce maître abominable du sang et des sèves, réchauffe
l’impureté des instincts…
Ma mère interrompit encore une fois :
— Comment voulez-vous, mon ami, que des enfants
comprennent de telles pensées? Mettez-vous à leur portée. Et puis,
tout cela est dangereux. En parlant ainsi, vous leur apprenez à
raisonner…
 — Oui, mon amie, dit mon père ; cela pourrait, peut-être, leur
apprendre à raisonner. Parfois, la connaissance trop précise du bien
pousse les curiosités à retourner l’étoffe, — geste dont proviennent
nécessairement les plus grands malheurs. Aussi, je me tais.
L’excellent prêtre souriait et se contentait d’approuver de la tête,
car il n’avait plus assez d’intelligence pour parler lui-même. En
dehors de ses formules et de ses opérations, le vieux magicien
n’était capable que des mots et des mouvements dictés par l’instinct
de la conservation. Sa mémoire rituelle commençait même à
s’affaiblir : il oubliait des verbes essentiels dans le prononcé des
exorcismes et quand il remettait le péché — fort rare, il est vrai, en
ce pays, — de « tentative intellectuelle », « effort pour
comprendre », il lui arrivait de ne pas exiger du pénitent, après
l’absolution, le serment sacramentel : « Serviam. — Je suis l’esclave
éternel. »
Même faite par d’aussi débiles mains, l’opération réussit. Mon
frère et ma sœur sont demeurés là-bas, « dans le pays lointain ».
— Mais toi?
— Moi, j’étais intelligent et hypocrite. Jamais personne ne se
douta que j’y voyais. J’enfermais mes impressions, mes joies, mes
désirs, sous une triple serrure, dans mon crâne, invincible coffret, et
un jour…
— Et un jour?
— … Je m’enfuis. Je traversais la plaine grise et, ayant marché
longtemps, j’entrai dans une forêt lumineuse, dont chaque arbre
ressemblait à une femme, la chevelure parée de diamants et le cou
imagé de perles. On respirait dans cette forêt un air si violemment
imprégné des odeurs de la vie que j’en eus mal à la tête ; mes
doigts se crispaient au chatouillement des hautes herbes ; mon
cœur chantait si fort que tout mon corps en tremblait. Enfin,
j’étendis les mains, embrassant, comme Apollon, les genoux d’un
des arbres-femmes. Ce contact m’apaisa, mais je tombai sur le côté
et je m’endormis.
 Le lendemain, je continuai mon voyage, et j’arrivai ici. D’abord, je
ne m’aperçus guère que j’avais changé de pays : les hommes
avaient, il est vrai, les yeux ouverts, mais ils semblaient ne se servir
de leur vue que pour se guider matériellement à travers la vie ;
depuis, j’ai rencontré quelques voyants véritables.
J’oubliais de vous dire qu’en traversant la forêt lumineuse, j’y
cueillis… devinez quoi?
— Une fleur rare?
— Oui, une âme! Au matin, avant de quitter la forêt sacrée qui
avait abrité ma lassitude et protégé mon sommeil, je me promenai
quelques instants sous les branches tombantes, mais toutes étaient
plus hautes que mon bras levé, et je désespérais d’emporter même
le souvenir d’une feuille. O feuillage, qui étais pour moi aussi vivant
et aussi parfumé qu’une chevelure d’amour, je te regardais onduler
au-dessus de ma tête aussi loin de ma main que l’aile des oiseaux
ou la neige violette des nuages matinals. J’allais obéir (une force me
traînait) et m’éloigner seul, sans le témoin que je voulais ; j’étais
déjà sur la lisière et je voyais le vaste horizon et, là-bas, où les deux
cercles se joignent, la cime obscure d’une autre forêt, lorsqu’une
branche fleurie de petits cœurs roses s’abaissa vers moi, comme un
geste de pitié. Je cueillis la branche où tremblait la grappe des petits
cœurs roses et je continuai ma route.
Arrivé au but de mon voyage, je choisis une maison afin d’abriter
la branche fleurie, comme ses sœurs m’avaient abrité moi-même,
car j’ai toujours aimé la culture du sentiment ; c’est une occupation
pleine de grâce et qui ne demande que la bonne volonté d’un
jardinier soigneux : au milieu du jardin, il y a une fontaine où l’on
peut se laver les doigts, quand ils sont tachés de sang.
Je plaçai donc ma branche fleurie de petits cœurs roses dans un
vase de majolique plein de sable d’or, et le vase sur une cheminée,
primitif autel ; à gauche du vase, j’inclinai les Damnées, de Filiger,
pour me remémorer la méchanceté des Dieux, et à droite, la Vigne
abandonnée, où de Groux a écrit l’inutilité du Sacrifice.
 Ensuite, j’allai étudier les formes de la vie, apprendre selon quelle
mode, riaient, s’ennuyaient ou pleuraient les hommes de mon
temps. Ils s’ennuyaient surtout, leur capacité passionnelle étant fort
médiocre et leur force nerveuse si fugitive qu’un désir ou un rêve
suffisait souvent à l’épuiser toute. Je constatai encore qu’ils
s’ennuyaient sans dignité, avec de petits gémissements de chien à la
chaîne et de vaines colères contre les astucieux et contre les forts
dont les jouissances irritaient leur impuissance originelle. Leur
consolation était de penser à l’avenir, de prédire des temps
meilleurs, de se vautrer dans les joies futures et de regarder la lune
avec des verres de couleur.
J’étais las de tant d’inoffensives niaiseries, quand je rencontrai
Armelle, vase plus beau que mon vase de majolique et d’où sortait
une fleur d’or ocellée de bleu. C’était une créature aussi étourdie
qu’un oiseau, aussi timide, mais qui se laissa prendre avec la main.
Elle n’avait notion ni de bien, ni de mal, ni de beau, ni de laid, d’une
sensibilité tout animale, sans pudeur et sans trouble dans l’amour.
Nous eûmes d’abord des rencontres furtives, des intimités
illusoires, dont elle aggravait la vanité par l’aveu de ses regrets et
l’implorante langueur de ses attitudes. A dessein, je prolongeais la
période du désir ; j’aimais l’impatience d’Armelle et son geste, sur la
berge, de se vouloir jeter à l’eau. Toute femme est vierge pour celui
qui ne l’a pas possédée, car la virginité n’est pas autre chose que de
l’inconnu, peut-être de l’inconnu plus obscur, — et je restais au seuil
du mystère, quoique le gardien n’en fût aucunement farouche.
Je désirais aussi, par ces jeux dilatoires, exaspérer la bête et
qu’elle bondît dans le cirque, au jour de la fête, avec des élans
sauvages et toute la violence d’une nature contrariée et aiguillonnée
— mais je fus trompé.
L’ayant menée en ma maison, je lui expliquai l’autel familier que
j’avais ordonné avec de précieuses décorations autour de la branche
fleurie de petits cœurs roses. Mon air grave et même un peu
hiératique étonnait sa candeur animale habituée à de moins
solennels prolégomènes ; elle s’approcha et ouvrit tout grands ses
beaux yeux bleu d’amour. Les images, sévères acolytes, ne captaient
pas son regard ; elle le fixait sans distraction sur la branche fleurie
de petits cœurs roses. Je me taisais, feignant même de m’astreindre
à effacer la poussière qui troublait un des coins de la glace ; alors sa
curiosité s’enhardit et elle toucha du doigt un des petits cœurs
roses ; elle ressemblait à une chatte qui veut jouer ; toute la grappe
trembla et un des petits cœurs roses tomba dans le sable d’or :
assurée que j’avais détourné la tête et oublieuse de la glace qui me
disait tout, Armelle prit le petit cœur rose et le mangea.
J’étais allé m’asseoir à l’autre bout de la chambre. Armelle vint à
moi et je la voyais s’avancer toute pâle d’amour ; son attitude
d’oiseau voltigeant s’était transformée en la grâce splendide d’un
cygne qui se meut sur un canal avec une fierté royale ; ses
mouvements se voyaient à peine sous sa robe traînante et ses bras
tombaient le long de son corps comme des tiges brisées par un coup
de vent.
Elle vint à moi et s’agenouilla, me baisant les mains ; puis elle
pleura silencieusement. La douleur ne contrariait pas la pureté de sa
face extasiée et les transparentes larmes qui roulaient sur ses joues
semblaient les perles détachées d’un chapelet de sourires.
Je me penchai sur elle et je la baisai au front, doucement ;
quelques perles tombèrent encore de ses yeux souriants, quelques
perles et la croix — et un grand soupir annonça que le cœur
d’Armelle s’était soulagé, grain à grain, de tout le chapelet des
douleurs suprêmes et des joies infinies.
Son corps s’affaissa sur mes pieds, sa tête s’arrêta sur mes
genoux et ses bras tombaient vraiment comme des tiges massacrées
par l’orage.
Armelle était morte.
Je compris que les petits cœurs roses étaient de merveilleuses
hosties contenant chacune une âme et je compris aussi qu’en se
communiant avec un de ces petits cœurs roses, Armelle s’était
empoisonnée… Les âmes sont de terribles poisons.
 LIVRE I

MIRACLES

PHOCAS

A Octave Mirbeau.

Le préteur donna lui-même les instructions les plus précises au

décurion chargé d’arrêter Phocas. Ce magistrat, nommé Aurélius,
était un homme grave, probe et intelligent ; excellent jurisconsulte, il
n’abusait point de sa science, ni des codes, ni des édits pour écraser
d’une rigueur uniforme et traditionnelle les criminels cités à son
tribunal ; tout au contraire, profitant de la liberté qu’avaient alors les
juges de décider selon leur conscience, il aimait à oublier
l’impérative dureté des lois pénales, — et plus d’une fois on
l’entendit condamner à une notable amende d’avares et inflexibles
riches « coupables selon lui de ne pas s’être laissé voler, attendu que
le voleur était dans le besoin le plus extrême et qu’il y a un certain
degré de misère qui autorise celui qui n’a rien à prendre à celui qui
possède tout. » De tels jugements paraîtraient aujourd’hui fort
scandaleux et notre moralité raffinée s’en indignerait ; mais au IVe
siècle, à Sinope, dans la province de Pont où se passe cette histoire,
les hommes, dénués de grands principes, acceptaient volontiers la
justice telle que la comprenait Aurélius ; vexés, mais convaincus que
de laisser mourir de faim une créature humaine, ou de l’étrangler de
ses propres mains, c’est un crime égal, ils payaient l’amende, puis,
pour éviter d’être volés justement, ils faisaient, de leur propre
volonté, la part des pauvres.
Les idées chrétiennes avaient pénétré peu à peu à Sinope,
comme dans une grande partie de l’Empire romain, mais pas encore
sous leur véritable nom ; ce nom était toujours détesté, et on y
professait pour la religion nouvelle une horreur mêlée de crainte ;
seules, devançant les dogmes, la justice et la pitié, mendiantes
boiteuses, avaient franchi les murs de la ville et murmuré tout bas
de singulières paroles que le peuple se répétait avec surprise.
De vrais chrétiens, instruits de la naissance, de la mort et de la
résurrection du Nazaréen, il n’y en avait guère, à Sinope, que dans
les faubourgs, parmi les tisserands, et, dans la campagne, parmi les
paysans et les esclaves des grands domaines ; on disait que le
principal d’entre eux, le plus instruit et, par conséquent, le plus
dangereux, était un nommé Phocas, jardinier de son état, homme
libre, qui cultivait un petit enclos et en vendait les produits aux
portes de la ville.
Donc, par une étrange contradiction, le peuple, qui aimait la
justice, haïssait ceux qui étaient les vivants exemplaires de la justice,
et Aurélius lui-même, le juge secourable, entrait en colère et jurait
par les dieux infernaux dès que l’on prononçait devant lui le nom de
Chrétien. Sur ces entrefaites, des édits arrivèrent qui ordonnaient la
recherche et la condamnation de tout sectateur de l’idée nouvelle.
Aurélius lut les édits que lui envoyait le préfet de la province et, pour
la première fois de sa vie, il fut joyeux d’avoir lu un édit impérial.
Ayant fait venir Amasius, le chef de la décurie de soldats que l’on
employait à la recherche des criminels, il lui commanda de s’emparer
de Phocas et de l’amener à Sinope, mort ou vif.
 Les instructions portaient, rédigées sur des tablettes de cire :
« Phocas, chrétien, contempteur des Dieux, ennemi de l’empereur et
du peuple romain. Bandit redoutable et conspirateur astucieux, chef
d’une bande de cruels coquins, il est encore un magicien des plus
experts ; il connaît l’art incroyable de tuer à distance, soit par
d’effroyables combinaisons d’éléments, soit par des signes, soit par
une entente secrète avec les Génies inférieurs. Vous vous
approcherez de lui prudemment et en usant de ruse ; il y va peut-
être de votre vie, mais il y va sûrement du salut de la République. »
Amasius médita ces instructions, choisit quelques légionnaires
résolus, épaves des guerres barbares, et la petite troupe se mit en
marche. Elle allait un peu au hasard, car — la police, en ces temps,
était sommaire — on ignorait l’endroit précis où conspirait Phocas en
arrosant ses salades. Cela devait être là-bas, au fond d’un vallon qui
creusait parmi la forêt une clairière de verdure ; on irait là, tout
d’abord, et on s’informerait près des bûcherons.
Dans l’imagination d’Amasius, brave décurion qui avait occis plus
de Goths qu’il n’avait de dents dans les mâchoires, Phocas se cachait
en une ténébreuse caverne, en quelque inaccessible repaire, et il
augurait que la quête serait difficile et pénible ; mais la saison était
belle, les hommes décidés : « On en sera quitte, songeait-il, pour
dormir quelques nuits en plein air, sous la protection de la déesse
aux douze mamelles. »
Ils partirent de grand matin et, ayant suivi un ruisseau qui
coupait en deux la forêt de Sinope, ils se trouvèrent, un peu avant
midi, en face d’une petite cabane couverte de roseaux, derrière
laquelle paraissait s’étendre un agréable jardin. Amasius n’eut aucun
soupçon ; il cogna à la porte et demanda l’hospitalité.
La porte s’ouvrit et parut un homme vêtu, tel qu’un paysan,
d’une tunique courte qui laissait les jambes nues à partir des
genoux ; ses cheveux étaient ras et sa barbe longue ; il avait l’air las
et doux ; ses yeux, sous des paupières tombantes, étaient bleus et
un peu vagues. L’homme semblait avoir une cinquantaine d’années,
mais son âme, certes, était toute jeune, car il manifesta une grande
joie, de ce que la Providence lui envoyait des étrangers :
 — Entrez, entrez! Comment? Des soldats? Les Goths sont-ils
revenus?
— Non, dit Amasius, mais nous cherchons un bandit plus féroce
que les fils des Amales, un chrétien, un contempteur des dieux (il
récitait son instruction), un magicien, qui connaît l’art incroyable de
tuer à distance…
Il n’y a pas de magicien par ici, dit Phocas, mais le pays est plein
de voleurs. Ils n’attendent même pas que mes salades soient
poussées pour me les arracher. Cela me donne double besogne, il
faut que je recommence mes semis, — mais, que voulez-vous? s’ils
me prennent mes salades, c’est qu’ils en ont besoin, plus besoin que
moi, peut-être, — et d’ailleurs, je leur pardonne et je leur donne ce
qu’ils me dérobent.
— Vous êtes trop indulgent, dit Amasius, et l’empereur, qui est
juste, a résolu de punir le chef de ces coquins, car il doit être leur
chef, mes instructions le portent.
— Quel est son nom? demanda Phocas.
— Son nom?
Il consulta ses tablettes :
— Phocas.
— Phocas! dit le pauvre jardinier, mais je le connais, il se tient
tout près d’ici. C’est un chrétien.
— Mes instructions le portent, dit Amasius.
— C’est bien lui, dit Phocas, — un chrétien absolu, un chrétien
farouche, un contempteur des dieux! Je vous l’amènerai moi-même,
avant le coucher du soleil. Vous tombez bien! Phocas! Ne soyez pas
inquiets, il vous appartient, il est entre vos mains. Mais en attendant,
puisque vous êtes mes hôtes, je vous dois toute l’hospitalité et
d’abord le repas. Du pain, des légumes de mon jardin, — ce que
Phocas en a laissé.
— C’est Phocas qui vous vole vos salades? demanda Amasius.
— Lui-même.
 — Nous ne le ménagerons pas.
— Je l’espère bien, dit Phocas.
Phocas continua :
— Et, pour les hôtes, je détiens là, enfouie sous terre, une
amphore de vin d’Asie… Moi je n’en bois jamais, l’eau du ruisseau
est si bonne…
— Nous la boirons! dirent les soldats.
— Je l’espère bien, dit Phocas.

Les soldats et le jardinier se mirent à table. Phocas, sur l’instance

d’Amasius but un peu de vin, et alors sa joie s’exalta :
— Que je vous aime, mes amis, s’écria-t-il, vous et tous mes
frères, tous les hommes! Souvent, quand je me repose de mon
labeur, quand mes laitues, arrosées, s’endorment, comme de bonnes
petites créatures, dans la paix du soir, souvent je rêve au bonheur
futur de l’humanité, fille de Dieu, et aussi au bonheur immédiat que
trouverait en lui-même chacun de nous, s’il vivait en amour, en
justice et en charité. Aimez-vous les uns les autres. Si votre frère a
froid, donnez-lui place à votre foyer ; s’il a faim, qu’il puisse s’asseoir
à votre table ; s’il est ignorant, instruisez-le ; s’il est méchant,
forcez-le d’être bon, en étant bon pour lui… Les temps vont changer.
Je vois venir un siècle, tout vêtu de blanc, comme un ciel matinal ; il
vient sur la mer, et les vagues s’apaisent, et les grands oiseaux qui
planent sur les eaux volent autour de lui et lui font un cortège
d’amour… Il vient, je le vois! Il a les yeux clairs d’un messager de
bonne nouvelle, il chante un cantique d’allégresse ; le battement de
ses ailes a une vertu pacifiante… Il vient, je le vois! L’archange
lumineux aborde parmi nous… Aimez, aimez, soyez implacables à
force d’aimer! Aimez les hommes malgré eux, aimez-les tant que
votre amour les dompte, les transforme, et les refaçonne à l’image
de Celui qui, pouvant tout, choisit de mourir…
Les soldats, sans bien comprendre, étaient émus ; Amasius aurait
voulu entendre encore cette parole d’amour, plus enivrante que le
vin d’Asie ; mais, fidèle au mot d’ordre, il songeait aussi à Phocas,
l’abominable bandit, et il fit l’effort de dire :
— Maître, je reviendrai te voir, car ton discours m’a remué
comme jamais je ne le fus par les plus belles harangues. Je ne
t’oublierai pas… J’ai entendu parler d’un philosophe nommé Socrate
ou Platon, je ne sais plus, que mon centurion vénère comme un
dieu… Tu seras mon Socrate… Oh! que tes paroles m’ont fait de
bien… Jamais je n’avais entendu de pareilles choses…
Il se tut ; puis faisant un nouvel effort :
— Et ce Phocas?
Le pauvre jardinier se leva et dit :
— Je suis Phocas.
— Toi? Maître, le vin d’Asie t’a-t-il fait tourner la tête?
— Je suis Phocas.
Par des tablettes, par une plaque de bronze qui lui affirmait, pour
son courage en des temps de peste, la reconnaissance de la ville
d’Antioche, Phocas prouva qu’il était Phocas.
Convaincu, Amasius murmura quelques paroles de mépris pour la
sottise du préteur Aurélius, — puis il emmena Phocas, et la nuit
n’était guère avancée quand ils entrèrent dans Sinope.

Dès le lendemain matin, Phocas fut jugé. Le peuple, prévenu,

accourait en grande foule ; à la vue du bandit, du chrétien, de
l’impie qui haïssait les dieux, il poussa de joyeux cris :
— A mort! A mort! criait le peuple.
Aurélius, après quelques menues tortures et un court
interrogatoire, où Phocas avait avoué son crime d’être chrétien,
proféra la sentence :
— Aux bêtes!
Et le peuple répéta :
— Aux bêtes, le chrétien! Aux bêtes, aux bêtes!
 Peu après midi, le cirque fut ouvert et Phocas parut dans l’arène.
Sans souci des hurlements de la foule heureuse, sans songer aux
fauves ni aux taureaux, il cria d’une voix forte :
— Je suis chrétien!
Puis il s’agenouilla et attendit, en priant.
Ce fut un taureau qui sortit de l’ergastule.
La bête fonça sur sa proie, la transperça d’un coup de corne, la
fit sauter en l’air, puis s’éloigna.
Phocas retomba au milieu d’une pluie de sang. Il n’était même
pas évanoui et, comprimant son ventre d’où sortaient ses entrailles,
il put se remettre à genoux et continuer sa prière.
A ce moment, il aperçut, près de la porte de l’ergastule, Amasius
et ses soldats qui avaient été postés là, l’épée au poing, pour
chasser la victime au centre de l’arène, si elle cherchait à fuir vers
les caves ; il reconnut ses amis, et, rassemblant ses forces, se
souleva pour leur envoyer, d’une main lourde, un signe d’amour et
un signe d’adieu.

Les soldats, qu’un désir de gloire et de mystère avait touchés, se

consultèrent un instant ; puis, tous, d’un bond, coururent à Phocas,
en criant :
— Nous sommes les fils de Phocas! Nous sommes chrétiens!
Ce fut une belle fête et dont le peuple de Sinope se souvint
longtemps, car on lâcha des lions et des panthères, et, au lieu d’une
victime, il y en eut une douzaine : les yeux des femmes burent du
sang.
 LA MÉTAMORPHOSE DE DIANE

Quand il vit la lune pâlir et trembler dans le ciel pur, voile égarée
sur le bleu des mers, Héliodore eut peur d’un tel présage et, se
dressant, les bras levés, il prononça des mots conjuratoires.
En vain. Les dieux fuyaient, oreilles sourdes ; et, de leurs lèvres
si éloquentes et si riches en sagesse, il ne tombait plus dans le
sanctuaire que des oracles brisés par d’invisibles et nouvelles
foudres.
Héliodore reprit sa place sur le banc de pierres, au seuil du
temple. Le vent du soir était triste comme un adieu ; on n’entendait
d’autres bruits que le sanglot des roseaux ; il pleura comme les
roseaux, tout uni d’amour au deuil des choses et des dieux.

Il pleura longtemps, puis il s’endormit, à ce seuil, toujours

gardien et toujours prêtre : des cris le réveillèrent et des lueurs de
torches. Des gens s’avançaient, petits, demi-nus, ceints de cuirs mal
grattés, avec de longs cheveux huilés, aux mains des épieux et des
branches de pin qui flambaient et fumaient dans la nuit. Le chef
laissa tomber son épieu sur la tête d’Héliodore, et le prêtre, lié de
courroies, fut jeté parmi les sanglots des roseaux ; ensuite, on pilla
avec soin le sanctuaire de Diane aux genoux blancs.
Ces barbares avaient un pouvoir destructeur vraiment divin ; ce
que les hommes avaient mis des siècles à construire, ils le
démolirent en quelques heures de nuit, et, tous les ors enlevés et
chargés sur des chariots, ils s’excitèrent par dérision à traîner hors
du temple l’Artémis inviolée dont le marbre, par sa candeur
surhumaine, étonnait la piété des pèlerins. Ils voulurent encore, sans
doute pour être agréables à leur dieu particulier, et croyant anéantir
son indestructible grâce, morceler l’effigie de la déesse blanche,
mais l’effigie voulut demeurer intacte, et les barbares s’éloignèrent,
lassés d’un sacrilège inutile.
Alors Héliodore rompit ses liens et se leva lamentable, d’entre les
sanglots des roseaux ; le jour nouveau naissait ; ayant lavé la vase
qui lui cloîtrait les yeux, il vit l’horreur de la dévastation impie et la
Vierge, son amour, couchée en travers du sentier, comme un cadavre
laissé là après le meurtre et après le stupre nocturne.
Il se laissa tomber près de la déesse et, ayant baisé ses pieds, il
s’évanouit.

« Marbre pur, marbre de grâce,

Genoux fiers,
Hanches où nulle main n’écrivit jamais son désir,
Crèche où nul enfant n’a dormi,
Source où l’oiseau n’est pas venu boire,
Ventre inaccessible,
Neiges éternelles,
Bras qui n’ont daigné accoler que le tronc sacré des chênes,
Mains qui n’ont caressé que les flancs des chiens blancs,
Seins qui n’ont palpité que de l’agonie des biches,
Bouche d’orgueil,
Marbre pur, marbre de grâce! »

Héliodore en son sommeil, balbutiait ces litanies, et, à chaque

invocation, il ajoutait un pardon, une supplication, l’expression de sa
honte, de son désespoir, de son amour.
« Pardonne-moi, Diane Artémis! Tu m’avais choisi comme gardien
et je n’ai pas su éloigner de toi les voleurs! Tu m’avais choisi comme
prêtre et je n’ai pas su te préserver du sacrilège. »
 Quand Héliodore eut ainsi prié, en toute simplicité et en toute
humilité, il lui sembla que la déesse se levait et se penchait vers lui,
et il lui sembla que la bouche d’orgueil et de grâce disait :
« Je te pardonne, Héliodore, car tu m’aurais donné ta vie, si
j’avais voulu de la vie ; mais les barbares te l’ont laissée par mon
ordre, afin que tu sois témoin d’un miracle tel que les hommes n’en
ont pas encore vu de pareil.
» Les dieux sont anciens, Héliodore, tu le sais ; mais, si anciens,
ils ont eu une naissance et ils doivent tous mourir. L’heure est venue
de leur mort. Les dieux meurent, au moment où je te parle, mais ils
ne meurent pas comme des hommes ; ils meurent comme des
dieux, leur essence permane et va revivre en de nouvelles formes.
» Ces changements sont nécessaires pour leur propre gloire et
pour la joie des hommes ; quand les dieux sont trop vieux ils
n’inspirent plus ni la terreur, ni l’amour ; ils deviennent indifférents
aux âmes familières et aux cœurs distraits ; les hommes, ces
éternels prisonniers, n’ont plus confiance en l’échelle de grâce, ils
ont peur qu’elle ne rompe sous leurs pieds, ils n’osent plus monter
au ciel : alors, retombés dans la tristesse de leur nature, ils rampent,
comme aux premiers jours du monde, dans le marécage obscur de
l’animalité.
» Il faut des échelles nouvelles ; c’est pourquoi des arbres ont
été abattus dans la forêt de l’infini.
» Dors, Héliodore. Quand tu te réveilleras, toi qui m’aimas telle
que je fus, tu m’aimeras telle que je serai, et, par l’échelle nouvelle,
tu monteras si haut que tu en auras le vertige. »
Diane se tut et Héliodore crut voir, s’en allant vers le temple, une
femme vêtue d’une blanche robe traînante, toute semée d’étoiles
bleues ; autour de sa tête, il y avait une lueur de soleil et, de ses
mains étendues, des rayons très doux tombaient vers la terre. Elle
entra dans le temple.
 Héliodore dormit encore ; quand il se réveilla, il vit que le temple
avait été restauré selon un art nouveau : partout, sur la blancheur
des murs, on avait peint des figures inconnues, des nimbes, des
agneaux et des lettres grecques appelées thau.
Il se leva et entra dans le sanctuaire, dont il se croyait toujours le
gardien et le prêtre, mais, ivre sans doute d’un si long sommeil, il ne
reconnaissait ni les trésors, vases, lampes, encensoirs, pourtant
remis à leur place tels qu’avant le pillage, ni la physionomie des
fidèles, ni l’effigie sacrée qui se dressait toujours sous le même dais
de soie et de perles, — et il restait debout, tout surpris, lorsque la
voix de son rêve sonna encore en son cœur :
« Héliodore, reconnais-moi, et aime-moi comme tu aimas Diane.
Je suis la toujours Vierge ; approche-toi : si tu me dis quelques
paroles d’amour, tu comprendras, car c’est l’amour qui fait tout
comprendre. Viens, Héliodore, et mets le pied au premier échelon de
l’échelle. »
Les fidèles chantaient :

Ave, semper virgo,

Ave, scala cœli.

Héliodore mêla sa voix à celle du chœur, et il aperçut aussitôt,

dressée devant lui, une échelle nouvelle faite avec les plus précieux
bois fauchés dans la forêt de l’infini. D’un élan il monta aux plus
hauts échelons ; il monta si haut qu’il en eut le vertige, si haut qu’il
comprit les mystères éternels et la loi qui veut que tout ce qui
change ne change qu’en forme et non pas en essence.
 RÉGELINDE

C’était au temps que les providentiels Barbares venaient de

libérer l’Europe de la tradition romaine. Les Goths fécondaient la
paresseuse Espagne. Une autre beauté surgissait d’entre les
décombres des temples vains. Des Aphrodites morcelées comme
jadis des pierres jetées par Deucalion, une humanité nouvelle
naissait au monde, rayonnante de force et de naïveté, ingénue et
violente — et de la poussière des Cérès broyées aux lourdes meules
habituées à la docilité du grain les hommes du Nord pétrissaient un
pain inconnu qui donnait aux mâles le mystère de la volonté et aux
femmes le mystère de la grâce.
Régelinde était fille de roi.
Joyau! l’écrin se ferma sur elle le jour de sa naissance et ne se
rouvrit plus. Elle vécut dans le palais et dans les jardins royaux,
unique, seule de son rang et seule de son essence, aussi unique que
l’améthyste taillée en coupe où son père n’avait bu qu’une fois, y
buvant mêlé à du vin noir le sang frais d’un tributaire rebelle au
tribut.
Vêtue d’une robe blanche stellée de croix de jais, avec au col la
bande de pourpre et au doigt la bague d’argent des fiançailles
secrètes, elle passait en silence et les officiers se taisaient sur son
passage et s’inclinaient, les yeux voilés de la main gauche, selon la
mode orientale apportée à Hispal par Isidore, fils de Grégoire,
médecin du roi et homme docte.
Nul jamais n’adressait la parole à Régelinde que son père,
Resçaon, Majorien l’évêque et sa nourrice Ipa ; aucune de ses
quarante esclaves n’eût osé toucher au bas de sa robe sans un ordre
de ses yeux ou un signe de son doigt : Régelinde était fille de roi.
 Princesse! et adorée muettement par la gent du palais, comme
une émanation, comme une incarnation d’Iscratène, le Soleil boréal,
comme Iscratène elle-même, l’Astre féminin qui pendant six mois
aime les hommes et pendant six mois les hait.
Mais Resçaon était chrétien, baptisé dans les neiges par Abbas le
martyr qui, pour ondoyer l’enfant, maître du Septentrion, avait fait
fondre à la chaleur de sa main un morceau de glace coupé en forme
de croissant de lune, — et Régelinde, chrétienne, ne se croyait pas
Iscratène, mais la fille privilégiée du Dieu vivant.
Humble aux pieds de Majorien l’évêque, acceptant ses dires
pénitentiels, humble en face de son père, l’oreille ouverte à ses
conseils, elle retrouvait dans la solitude l’orgueil d’être l’unique
Régelinde et la joie d’être aimée par Celui devant qui les rois ne sont
que de la poussière et les évêques de la cendre : Dieu aimait
Resçaon, Dieu aimait Majorien, Dieu aimait Ipa, — mais Dieu
n’aimait pas Ipa, Majorien ni Resçaon comme il aimait Régelinde : et
c’était vrai, aussi vrai qu’il y a sept planètes dans le firmament, aussi
vrai que le tonnerre est une clameur du ciel, un avertissement
d’avoir à pleurer nos péchés.
Or, un matin, Resçaon appela sa fille et lui annonça la venue du
prince des fiançailles secrètes. Le courrier arrivé dans la nuit le
précédait de six jours de marche : qu’elle se préparât donc à
recevoir comme seigneur le jeune roi d’Hippone, Saran, celui qui
portait au doigt une bague d’argent toute pareille à la bague de
Régelinde.
Saran! son rêve était allé souvent vers Hippone et vers Saran ; et
même, à force de penser à lui, lorsque la nourrice lui contait
l’histoire des fiançailles secrètes, parfois elle se l’était figuré : tel à
peu près et aussi superbe que Zinthe, le chef des Archers bleus, qui
avait un zigzag de foudre tatoué sur le front, aussi superbe, l’œil
aussi froidement doux, mais plus royal.
Saran! elle allait donc devenir femme!
Régelinde médita ce mystère et comme elle était très pure, ce fut
en vain. Sans doute, le lendemain des noces, au lieu de la robe
blanche stellée de croix de jais, elle revêtirait la robe de pourpre et,
quand elle deviendrait mère, la robe de sinople frangée d’or rouge,
si c’était un fils, frangée de lin, si c’était une fille, — mais comment
deviendrait-elle mère?
Interrogée, Ipa répondit, en levant au ciel ses yeux gris :
— « Iscratène, ma mère, Christ, mon sauveur, vous entendez ce
qu’elle demande? »
Ce fut tout. Alors Régelinde commanda qu’on fit venir et qu’on
laissât seul avec elle Isidore, fils de Grégoire.
Médecin du roi et maître du cérémonial, Isidore était magicien. Il
avait étudié sous les plus savants, à Thèbes, à Chrysopolis, à
Alexandrie, enfin, à Erythrée, la ville des sables rouges, dont les
habitants conversent librement avec les démons et dont le prince,
Hucar, trois fois ressuscité, use de plus de femmes en un jour qu’il
n’y a de grains de raisins dans une vigne royale.
Isidore entra. Il n’était ni jeune ni vieux, mais il paraissait fort vif,
doué d’une surnaturelle santé.
— « Princesse Régelinde, celui que tu enfermes avec toi, vierge,
doit être un vieillard. »
Isidore s’affaissa soudain comme sous un fardeau de siècles, et
Régelinde parla :
— « Enseigne-moi la science des générations. Dis-moi comment
le Père engendra le Fils ; dis-moi quelles sont les conjugaisons des
astres. Nomme-moi les principes, les causes et les moyens. Quel est
le père des ægipans et quelle est leur mère? Apprends-moi les
normes et les ambigénies, la généalogie des semblables et celle des
disparates, la création de l’homme et celle de l’ibre, celle du
musmon et celle de l’ange ; j’écoute. »
— « Je me tairai, répondit Isidore, fils de Grégoire ; mais
regarde. »
Et l’infinité des mondes se déroulant dans les espaces, tels que
les anneaux d’une chaîne prodigieuse, Régelinde vit les générations