Download as pdf or txt
Download as pdf or txt
You are on page 1of 52

Automation and Autonomy: Labour,

Capital and Machines in the Artificial


Intelligence Industry 1st ed. 2021
Edition Steinhoff
Visit to download the full and correct content document:
https://ebookmass.com/product/automation-and-autonomy-labour-capital-and-machin
es-in-the-artificial-intelligence-industry-1st-ed-2021-edition-steinhoff/
MARX, ENGELS, AND MARXISMS

Automation
and Autonomy
Labour, Capital and Machines in
the Artificial Intelligence Industry

James Steinhoff
Marx, Engels, and Marxisms

Series Editors
Marcello Musto, York University, Toronto, ON, Canada
Terrell Carver, University of Bristol, Bristol, UK
The Marx renaissance is underway on a global scale. Wherever the critique
of capitalism re-emerges, there is an intellectual and political demand for
new, critical engagements with Marxism. The peer-reviewed series Marx,
Engels and Marxisms (edited by Marcello Musto & Terrell Carver, with
Babak Amini, Francesca Antonini, Paula Rauhala & Kohei Saito as Assis-
tant Editors) publishes monographs, edited volumes, critical editions,
reprints of old texts, as well as translations of books already published
in other languages. Our volumes come from a wide range of political
perspectives, subject matters, academic disciplines and geographical areas,
producing an eclectic and informative collection that appeals to a diverse
and international audience. Our main areas of focus include: the oeuvre
of Marx and Engels, Marxist authors and traditions of the 19th and 20th
centuries, labour and social movements, Marxist analyses of contemporary
issues, and reception of Marxism in the world.

More information about this series at


http://www.palgrave.com/gp/series/14812
James Steinhoff

Automation
and Autonomy
Labour, Capital and Machines in the Artificial
Intelligence Industry
James Steinhoff
University of Toronto
Toronto, Ontario, Canada

ISSN 2524-7123 ISSN 2524-7131 (electronic)


Marx, Engels, and Marxisms
ISBN 978-3-030-71688-2 ISBN 978-3-030-71689-9 (eBook)
https://doi.org/10.1007/978-3-030-71689-9

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

Cover image: © Xphi Dech Pha Ti/EyeEm/Getty Image

This Palgrave Macmillan imprint is published by the registered company Springer Nature
Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
A.M.S.G.
Series Editors’ Foreword

Titles Published
1. Terrell Carver & Daniel Blank, A Political History of the Editions
of Marx and Engels’s “German Ideology” Manuscripts, 2014.
2. Terrell Carver & Daniel Blank, Marx and Engels’s “German Ideol-
ogy” Manuscripts: Presentation and Analysis of the “Feuerbach
chapter,” 2014.
3. Alfonso Maurizio Iacono, The History and Theory of Fetishism,
2015.
4. Paresh Chattopadhyay, Marx’s Associated Mode of Production: A
Critique of Marxism, 2016.
5. Domenico Losurdo, Class Struggle: A Political and Philosophical
History, 2016.
6. Frederick Harry Pitts, Critiquing Capitalism Today: New Ways to
Read Marx, 2017.
7. Ranabir Samaddar, Karl Marx and the Postcolonial Age, 2017.
8. George Comninel, Alienation and Emancipation in the Work of
Karl Marx, 2018.
9. Jean-Numa Ducange & Razmig Keucheyan (Eds.), The End of the
Democratic State: Nicos Poulantzas, a Marxism for the 21st Century,
2018.
10. Robert X. Ware, Marx on Emancipation and Socialist Goals:
Retrieving Marx for the Future, 2018.

vii
viii SERIES EDITORS’ FOREWORD

11. Xavier LaFrance & Charles Post (Eds.), Case Studies in the Origins
of Capitalism, 2018.
12. John Gregson, Marxism, Ethics, and Politics: The Work of Alasdair
MacIntyre, 2018.
13. Vladimir Puzone & Luis Felipe Miguel (Eds.), The Brazilian
Left in the 21st Century: Conflict and Conciliation in Peripheral
Capitalism, 2019.
14. James Muldoon & Gaard Kets (Eds.), The German Revolution and
Political Theory, 2019.
15. Michael Brie, Rediscovering Lenin: Dialectics of Revolution and
Metaphysics of Domination, 2019.
16. August H. Nimtz, Marxism versus Liberalism: Comparative Real-
Time Political Analysis, 2019.
17. Gustavo Moura de Cavalcanti Mello and Mauricio de Souza Saba-
dini (Eds.), Financial Speculation and Fictitious Profits: A Marxist
Analysis, 2019.
18. Shaibal Gupta, Marcello Musto & Babak Amini (Eds), Karl Marx’s
Life, Ideas, and Influences: A Critical Examination on the Bicente-
nary, 2019.
19. Igor Shoikhedbrod, Revisiting Marx’s Critique of Liberalism:
Rethinking Justice, Legality, and Rights, 2019.
20. Juan Pablo Rodríguez, Resisting Neoliberal Capitalism in Chile:
The Possibility of Social Critique, 2019.
21. Kaan Kangal, Friedrich Engels and the Dialectics of Nature, 2020.
22. Victor Wallis, Socialist Practice: Histories and Theories, 2020.
23. Alfonso Maurizio Iacono, The Bourgeois and the Savage: A Marxian
Critique of the Image of the Isolated Individual in Defoe, Turgot and
Smith, 2020.
24. Terrell Carver, Engels before Marx, 2020.
25. Jean-Numa Ducange, Jules Guesde: The Birth of Socialism and
Marxism in France, 2020.
26. Antonio Oliva, Ivan Novara & Angel Oliva (Eds.), Marx and
Contemporary Critical Theory: The Philosophy of Real Abstraction.
27. Francesco Biagi, Henri Lefebvre’s Critical Theory of Space.
28. Stefano Petrucciani, The Ideas of Karl Marx: A Critical Introduc-
tion.
29. Terrell Carver, The Life and Thought of Friedrich Engels, 30th
Anniversary Edition.
SERIES EDITORS’ FOREWORD ix

30. Giuseppe Vacca, Alternative Modernities: Antonio Gramsci’s Twen-


tieth Century.
31. Kevin B. Anderson, Kieran Durkin & Heather Brown (Eds.),
Raya Dunayevskaya’s Intersectional Marxism: Race, Gender, and the
Dialectics of Liberation.
32. Marco Di Maggio, The Rise and Fall of Communist Parties in
France and Italy.
33. Ryuji Sasaki, A New Introduction to Karl Marx: New Materialism,
Critique of Political Economy, and the Concept of Metabolism.
34. Kohei Saito (Ed.), Reexamining Engels’s Legacy in the 21st Century.
35. Paresh Chattopadhyay, Socialism in Marx’s Capital: Towards a De-
alienated World.
36. Marcello Musto, Karl Marx’s Writings on Alienation.
37. Michael Brie & Jörn Schütrumpf, Rosa Luxemburg: A Revolu-
tionary Marxist at the Limits of Marxism.

Titles Forthcoming
Miguel Vedda, Siegfried Kracauer, or, The Allegories of Improvisations
Gianfranco Ragona & Monica Quirico, Frontier Socialism: Self-
organisation and Anti-capitalism
Vesa Oittinen, Marx’s Russian Moment
Kolja Lindner, Marx, Marxism and the Question of Eurocentrism
Jean-Numa Ducange & Elisa Marcobelli (Eds.), Selected Writings of Jean
Jaures: On Socialism, Pacifism and Marxism
Adriana Petra, Intellectuals and Communist Culture: Itineraries, Problems
and Debates in Post-war Argentina
George C. Comninel, The Feudal Foundations of Modern Europe
James Steinhoff, Automation and Autonomy: Labour, Capital and
Machines in the Artificial Intelligence Industry
Spencer A. Leonard, Marx, the India Question, and the Crisis of
Cosmopolitanism
Joe Collins, Applying Marx’s Capital to the 21st century
Levy del Aguila Marchena, Communism, Political Power and Personal
Freedom in Marx
Jeong Seongjin, Korean Capitalism in the 21st Century: Marxist Analysis
and Alternatives
Marcello Mustè, Marxism and Philosophy of Praxis: An Italian Perspective
from Labriola to Gramsci
x SERIES EDITORS’ FOREWORD

Satoshi Matsui, Normative Theories of Liberalism and Socialism: Marxist


Analysis of Values
Shannon Brincat, Dialectical Dialogues in Contemporary World Politics: A
Meeting of Traditions in Global Comparative Philosophy
Stefano Petrucciani, Theodor W. Adorno’s Philosophy, Society, and Aesthetics
Francesca Antonini, Reassessing Marx’s Eighteenth Brumaire: Dictatorship,
State, and Revolution
Thomas Kemple, Capital after Classical Sociology: The Faustian Lives of
Social Theory
Tsuyoshi Yuki, Socialism, Markets and the Critique of Money: The Theory
of “Labour Note”
V Geetha, Bhimrao Ramji Ambedkar and the Question of Socialism in
India
Xavier Vigna, A Political History of Factories in France: The Workers’
Insubordination of 1968
Attila Melegh, Anti-Migrant Populism in Eastern Europe and Hungary:
A Marxist Analysis
Marie-Cecile Bouju, A Political History of the Publishing Houses of the
French Communist Party
Gustavo Moura de Cavalcanti Mello & Henrique Pereira Braga (Eds.),
Wealth and Poverty in Contemporary Brazilian Capitalism
Peter McMylor, Graeme Kirkpatrick & Simin Fadaee (Eds.), Marxism,
Religion, and Emancipatory Politics
Mauro Buccheri, Radical Humanism for the Left: The Quest for Meaning
in Late Capitalism
Rémy Herrera, Confronting Mainstream Economics to Overcome Capi-
talism
Tamás Krausz, Eszter Bartha (Eds.), Socialist Experiences in Eastern
Europe: A Hungarian Perspective
Martin Cortés, Marxism, Time and Politics: On the Autonomy of the
Political
João Antonio de Paula, Huga da Gama Cerqueira, Eduardo da Motta
e Albuquer & Leonardo de Deus, Marxian Economics for the 21st
Century: Revaluating Marx’s Critique of Political Economy
Zhi Li, The Concept of the Individual in the Thought of Karl Marx
Lelio Demichelis, Marx, Alienation and Techno-capitalism
Dong-Min Rieu, A Mathematical Approach to Marxian Value Theory:
Time, Money, and Labor Productivity
SERIES EDITORS’ FOREWORD xi

Salvatore Prinzi, Representation, Expression, and Institution: The Philos-


ophy of Merleau-Ponty and Castoriadis
Agon Hamza, Slavoj Žižek and the Reconstruction of Marxism
Kei Ehara (Ed.), Japanese Discourse on the Marxian Theory of Finance
Éric Aunoble, French Views on the Russian Revolution
Elisa Marcobelli, Internationalism Toward Diplomatic Crisis: The Second
International and French, German and Italian Socialists
Paolo Favilli, Historiography and Marxism: Innovations in Mid-Century
Italy
Terrell Carver, Smail Rapic (Eds.), Friedrich Engels for the 21st Century:
Perspectives and Problems
Juan Dal Maso, Hegemony and Class: Three Essays on Trotsky, Gramsci and
Marxism
Patrizia Dogliani, A Political History of the International Union of
Socialist Youth
Alexandros Chrysis, The Marx of Communism: Setting Limits in the Realm
of Communism
Stephen Maher, Corporate Capitalism and the Integral State: General
Electric and a Century of American Power
Paul Raekstad, Karl Marx’s Realist Critique of Capitalism: Freedom,
Alienation, and Socialism
Alexis Cukier, Democratic Work: Radical Democracy and the Future of
Labour
Christoph Henning, Theories of Alienation: From Rousseau to the Present
Daniel Egan, Capitalism, War, and Revolution: A Marxist Analysis
Genevieve Ritchie, Sara Carpenter & Shahrzad Mojab (Eds.), Marxism
and Migration
Emanuela Conversano, Capital from Afar: Anthropology and Critique of
Political Economy in the Late Marx
Marcello Musto, Rethinking Alternatives with Marx
Vincenzo Mele, City and Modernity in George Simmel and Walter
Benjamin: Fragments of Metropolis
David Norman Smith, Self-Emancipation: Marx’s Unfinished Theory of the
Working Class
Ronaldo Munck, Rethinking Development: Marxist Perspectives
Preface

2020–2021 was a weird time to transform a dissertation into a book. The


uprising of Black Lives Matter and the COVID-19 pandemic are only
two factors that made the intentionally narrow focus of this project, on
artificial intelligence and Marxism, sometimes feel divorced from matters
of immediate social concern. Yet, one book can only be about so many
things and, I suspect, the capitalist use of machines will continue to be a
determining factor in most people’s lives for a while to come.

Toronto, Canada James Steinhoff

xiii
Acknowledgements

Before this text was a book it was a dissertation. For the writing of
that dissertation I was fortunate enough to be supervised by Nick Dyer-
Witheford. This book would never have come to be without his guid-
ance. I’d also like to thank Marcello Guarini and Jeff Noonan for forma-
tive experiences studying the philosophy of AI and Marx, respectively.
Special thanks go to my FIMS friends and colleagues Victoria O’Meara,
Ryan Mack, Tyler Wagner, Chandell Gosse, Daniel Goyette, Vincent
Manzerolle, Atle Mikkola Kjøsen and Indranil Chakraborty as well as my
postdoctoral supervisor Anissa Tanweer.
I also want to express my gratitude to an old circle of friends. In subtle
ways, this book came to be on the basis of years of bizarre discussions and
artistic endeavours with Bohdan Pidskalny, Kyle Lefaive, Joshua Olsen,
Nicholas Manuel, Trevor Pepper, Martin Schiller, Caleb Farrugia, Colin
Wysman, Nate Gelinas and James Oltean-Lepp.
I’m grateful to Rob Simone, who produced some fine diagrams for
this book, and to Sean Landis, whose stellar editorial work improved the
quality of what you are about to read by several orders of magnitude. Of
course, any errors or inadequacies remain my own.
This book draws on research kindly supported by the Social Sciences
and Humanities Research Council of Canada and the Harry Bridges
Center for Labor Studies at the University of Washington.

xv
Contents

1 Introduction: Automation, Autonomy and Artificial


Intelligence 1
Your Means of Production 1
Revolutions 2
AI in the Real World 4
Machinery and Marxists 5
The Central Argument 7
Computing Machinery 9
Recursion 13
What This Book Is Not 15
Chapter Outline 16
Appendix 1: List of Interviewees 19
Bibliography 20
2 Labour, Capital, Machine: Marxist Theory
and Technology 29
Introduction 29
Political Economy 30
Marx on Value and Labour 33
Marx on Machines 40
The Fragment on Machines 45
Marxism(s) 46
Soviet Marxism 48
Western Marxism 50

xvii
xviii CONTENTS

Labour Process Theory 51


The New Reading of Marx 55
Cybernetic Capitalism 59
Conclusion 64
Bibliography 65
3 Post-Operaismo and the New Autonomy of Immaterial
Labour 75
Introduction 75
From Operaismo to Post-Operaismo 76
Post-Operaismo 78
Immaterial Labour Theory 84
Human–Machine Hybridization 86
Abstract Cooperation 90
New Autonomy from Capital 92
The Technological Argument for New Autonomy 94
Conclusion 95
Bibliography 95
4 Industrializing Intelligence: A Political Economic
History of the AI Industry 99
Introduction 99
The Historical Context 100
The Advent of AI Research 104
The AI Winter 109
Expert Systems: The First Era of the AI Industry 110
Strategic Computing: AI and the State Part I 113
The Decline of Expert Systems 116
The Rise of Machine Learning 119
Deep Learning: The Second Era of the AI Industry 122
Conclusion 124
Bibliography 125
5 Machine Learning and Fixed Capital: The
Contemporary AI Industry 133
Introduction 133
Charting the AI Industry 135
AI Capital Composition 138
AI Tech Giants 139
AI Dinosaurs 142
CONTENTS xix

AI Startups 143
AI Think Tanks 144
National AI Strategies: AI and the State Part II 145
AI Capital Concentration 148
Open-Source AI, Clouds, AI Chips 149
Labour in the AI Industry 151
Labour Composition: Race and Gender 155
AI Labour Organization 157
Conclusion 159
Bibliography 159
6 A Dark Art: The Machine Learning Labour Process 171
Introduction 171
The Machine Learning Labour Process 172
Stage 1: Data Processing 173
Stage 2: Model Building 174
Stage 3: Deployment 176
The Commodity Form of AI 177
Empirical Control 182
AI as Automation 187
The Automation of AI Work 189
Automated Machine Learning 193
Synthetic Automation 195
Other Forms of Automation in Machine Learning 199
Conclusion 200
Bibliography 200
7 New Autonomy and Work in the AI Industry 207
Introduction 207
AI Work and Human–Machine Hybridization 208
AI Work and Abstract Cooperation 213
AI Work and New Autonomy 215
Autonomy for What? 218
Conclusion 222
Bibliography 222
8 Conclusion: Harry Braverman Overdrive 227
Introduction 227
Theoretical Synthesis 228
Automation on Steroids 229
xx CONTENTS

Optimism and Agency 233


Conclusion 235
Bibliography 239

Index 241
List of Figures

Fig. 6.1 Waterfall methodology (Adapted from Royce 1970) 182


Fig. 6.2 Agile methodology (Adapted from Kuruppu 2019) 184
Fig. 6.3 Scrum methodology (Adapted from Weaver-Johnson 2017) 185

xxi
CHAPTER 1

Introduction: Automation, Autonomy


and Artificial Intelligence

Your Means of Production


A friend sent me a meme in 2016.1 “You have been visited by the
Time Travelling Terminator of Full Automation” read bold text above
a legion of skeletal anthropoid robots assembled in front of the red
banner of the USSR. The Terminator’s statement was captioned: “Your
means of production… give them to me now! And capitalism will collapse
under the weight of its own contradictions”. This utterance puns on
Arnold Schwarzenegger’s famed line, delivered nude in the 1991 James
Cameron film Terminator 2: Judgment Day: “I need your clothes, your
boots and your motorcycle”. While the T-800 model Terminator played
by Schwarzenegger merely wants garments so it can blend into human
society, the Time Travelling Terminator of Full Automation has a slightly
larger demand: the totality of machines, tools and facilities for producing
goods and services. This is the very same demand made historically by
Marxists and communists who would seize the means of production
from their capitalist owners. In the Terminator mythos, the eponymous
android is a creation of Skynet; an artificial intelligence produced by the
technology corporation Cyberdyne Systems. In Terminator 2, a captured
Terminator is reprogrammed against Skynet, and sent back in time from a

1 The original meme cannot be included here due to the indeterminate copyright status
of memes.

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


Switzerland AG 2021
J. Steinhoff, Automation and Autonomy, Marx, Engels, and Marxisms,
https://doi.org/10.1007/978-3-030-71689-9_1
2 J. STEINHOFF

dystopian future in which machines are exterminating the last remnants of


humanity, to 1995 Los Angeles to defend the pre-teen delinquent John
Connor, the future leader of the human resistance. In the meme, automa-
tion, a technique by which capitalists aim to increase the productivity of
labour, is turned against capital and threatens to destroy it.
How did Marxism and a pop culture android end up in one meme?
In a roundabout way, this book aims to answer that question. Marxist
thought has a long and complicated relationship with technology. Yet, it
has engaged relatively little with artificial intelligence (AI), especially in its
contemporary form called machine learning (ML). The first goal of this
book is to contribute to a Marxist understanding of AI. AI, as it exists in
the world today is, I argue, best understood as an automation technology.
The primary context for understanding AI is not academia nor science
fiction, but industry—specifically, the emergent AI industry, a small but
influential subset of the broader tech industry. Since its inception, AI has
been implicated in the capitalist mode of production that now covers the
world.
The second goal of this book is, to borrow from Raniero Panzieri
(1965), to “defy all kinds of mystical ideas” about labour. Marxist
thinkers have extended and adapted Marx’s work to the increasingly
computational capitalism that has evolved since his time. One particularly
influential line of thought has been the theory of “immaterial labour”
(Hardt and Negri 2000). On the basis of a study of the AI industry and
labour therein, we will assess the adequacy of immaterial labour theory
for explaining the nature of contemporary high-technology work. This
introductory chapter outlines the central argument in the pages to come.
But first, we should answer a suite of more general questions. Why draw
on Marx to think about AI? Why look at the AI industry to think about
Marxist theory? What can writings from one hundred and fifty years ago
tell us about purportedly intelligent machines? Simply put, two central
and deeply integrated elements of Marx’s analysis continue to hold sway
over our lives today: capital and machines.

Revolutions
Marx conducted his analysis of capital in the wake of what is commonly
known as the industrial revolution, in which steam-powered, machine-
based manufacturing became the predominant form of industry in Europe
and parts of North America, somewhere between 1760 and 1840
1 INTRODUCTION: AUTOMATION, AUTONOMY … 3

(Toynbee 2011 [1887]).2 Marx published the first volume of Capital


in 1867. A second industrial revolution, characterized primarily by the
proliferation of electricity, took place from around 1870 to the First
World War in 1914 (Landes 2003 [1969]). This was followed by a
third, characterized by the proliferation of digital networks and infor-
mation technologies, after the Second World War (Rifkin 2011). Since
the mid-2010s, we have purportedly been living amid a fourth industrial
revolution. According to Klaus Schwab (2017), founder and executive
chairman of the World Economic Forum, the fourth industrial revolu-
tion is defined by a “confluence of emerging technology breakthroughs,
covering wide-ranging fields such as artificial intelligence (AI), robotics,
the internet of things (IoT), autonomous vehicles, 3D printing, nanotech-
nology, biotechnology, materials science, energy storage and quantum
computing” (7). The economists Erik Brynjolfsson and Andrew McAfee
(2014) argue that after a long first machine age following the first indus-
trial revolution, we are now entering a “second machine age” defined
by the same array of technologies referred to by Schwab. Brynjolfsson
and McAfee (2017) suggest that of these technologies, AI will be the
most significant because, like electricity or the combustion engine, it
is a general-purpose technology with nearly limitless applications (3–4).
Perhaps unsurprisingly, AI luminaries agree.
Professor, entrepreneur and venture capitalist Andrew Ng has called
AI the new electricity. He asserts that AI is going to remake the world
“[j]ust as electricity transformed almost everything 100 years ago” (Lynch
2017). This sentiment is widespread. The consulting firm Accenture
proclaims that AI is “the future of growth” (Purdy and Daugherty 2016).
However, it is important to note that the history of AI is characterized by
waves of excessive hype, followed by troughs of disappointment. What-
ever its future may be, since 2015 AI has been riding high on a wave
of enthusiasm. The consulting firm Gartner (2018) put the approach to
AI known as deep learning at the apex of its hype cycle—the “peak of
inflated expectations”—but AI ebullience continues unabated.

2 The notion of the industrial revolution has been justly criticized for oversimplifying
complex social processes, attributing deterministic power to technology and overlooking
the gradual nature of historical change (De Vries 1994; Žmolek 2013). I use the concept
here only as a means for summarizing technological milieus and to produce an overview
of the technological evolution of capital.
4 J. STEINHOFF

AI in the Real World


A degree of this enthusiasm is warranted. AI is having real effects on
the world. AI is now an integral component of all kinds of digital tech-
nologies. So-called smart speakers, like Amazon Echo and Google Home,
use AI to recognize human speech. Police use facial recognition AI
from companies like Clearview AI and iOmniscient to automatically scan
surveillance video for persons of interest. AI is being used to automate
all kinds of business processes, from advertising to logistics to human
resources. Autonomous freight trucks, powered by AI from companies
like Waymo and TuSimple, now operate in limited areas (with human
overseers) and automated retail stores, such as Amazon Go, use AI to
dispense with (some) retail employees. While it is not yet replacing human
workers en masse, AI is widely used to conduct sophisticated analytics and
make predictions anywhere large quantities of digital data are available.
A distinct industry is emerging around AI within the broader tech-
nology industry. In 2016, for the first time the top five largest companies
in the world by market value all came from the same country and industry:
US big tech (Mosco 2017, 65). The same five were still on top in
2019: Apple ($961.3 B), Microsoft ($946.5 B), Amazon ($916.1 B),
Alphabet/Google ($863.2 B) and Facebook ($512 B), with the Chinese
technology corporation Alibaba in sixth place ($480.8 B) (Murphy et al.
2019).3 All of these companies are now deeply involved in the production
of AI.
Beyond industry, in the virtual worlds of games, AI systems have won
a series of unexpected victories over human experts. In 2015, Deep-
Mind’s AI system AlphaGo, trained on massive amounts of data from
expert human gameplay, defeated the world champion of the ancient
strategy game Go. Two years later, AlphaGo was defeated by its successor,
AlphaGo Zero, which drew on no human gameplay data and instead
learned the game by playing itself repeatedly in simulations (Silver et al.
2017). Such tales of the increasing power and autonomy of AI systems
have led public figures including Bill Gates, Elon Musk and the late
Stephen Hawking to describe AI as an existential threat to humanity. In
less eschatological terms, concerns have been widely voiced over the possi-
bility of widespread technological unemployment driven by a new wave of

3 All quantities of money are in United States dollars (USD) unless otherwise noted.
1 INTRODUCTION: AUTOMATION, AUTONOMY … 5

AI-powered automation. Concerns about automation have been amplified


by the COVID-19 pandemic (ongoing at the time of writing in 2021),
which is compelling many businesses to seek new ways to minimize labour
costs (The Economist 2020).
Predictions vary widely in terms of what the social and economic effects
of AI will be and when they will happen. However, across the political
spectrum, and from a variety of sectors ranging from academia to govern-
ment to industry, a vague consensus is emerging that, sooner or later,
capitalist economies are going to have to address the question of what
to do when machines are able to perform much of society’s necessary
labour. For this reason, among others, the analyses conducted by Marx in
the late 1800s remain relevant. It is true that Marx’s work is historically
specific; he was writing about the particularly new industrializing capital
of his time and place. However, at the same time his analysis is concerned
with the persistent features and “structures that lay at the heart of every
particular capitalism” (Heinrich 2007, 4). One of these features, since
even before industrialization, is machinery.

Machinery and Marxists


Marx argued that capitalism is distinguished from previous modes of
production by its systemic production and reincorporation of surplus-
value. Machines are capital’s primary means of controlling and enhancing
the productivity of the labour on which it relies to produce surplus-value.
Marx showed how capital is inherently driven, by competition and class
struggle, to ceaseless technological revolution founded on the capture
of skills and knowledge from workers and their emulation in machines.
Capital tends towards an increasingly machinic state. Marx thus described
machinery as a “most powerful weapon” for capital (1990, 562). Today
we typically refer to the process by which capital becomes more machinic
as automation.
The term automation was not coined until after the Second World War.
In an account of production at Ford Motor Company for the magazine
American Machinist, Le Grand (1948) defined automation as “the art of
applying mechanical devices to manipulate work pieces into and out of
equipment, turn parts between operations … and to perform these tasks
in timed sequence with the production equipment so that the line can
6 J. STEINHOFF

be wholly or partially under push-button control at strategic stations”.4


Four years later, management theorist John Diebold (1952) popularized
the notion of automation as “denoting both automatic operation and
the process of making things automatic. In the latter sense it includes
several areas of industrial activity such as product and process re-design,
the theory of communication and control, and the design of machinery”
(ix). An adequate history of automation would require its own volume(s),
but the interested reader can refer to the excellent and neglected work of
Ramin Ramtin (1991), which I draw on throughout this book.
Ramtin distinguishes automation from mechanization. Mechanization
involves the introduction of thermodynamic machines into (primarily
manual) labour processes, which Marx witnessed in his time. Automation
involves the implementation of manual or cognitive labour processes with
information technology. For Ramtin (1991) automation is defined by the
“systemic application of the principle of feedback” or the use by a system
of its output as input (60). While simple forms of feedback have been
utilized since antiquity, according to Ramtin, automation proper did not
exist until the advent of software computing technology which “can be
programmed to perform an unlimited number of tasks” (Ramtin 1991,
50). Automation is distinguished from mechanization by this flexibility.
By harnessing feedback, computing technology can perform many tasks.
Capital, which is continually driven towards an increasingly machinic
state, has a complementary relationship with the computer’s capacity to
harness feedback.
Thinkers since Marx have developed and revised his framework to
analyse the evolution of capital throughout the subsequent industrial
and technological revolutions. Marxist thinkers call the era around the
second industrial revolution Fordism, which finds its most famous expres-
sion in the assembly line and its deskilled “mass worker” (Wright 2002).
The digital third industrial revolution has been described by Marxists
and others as a transition to post-Fordism, which refers to how capital
sought to overcome the organized power of the mass worker through the
deployment of information technologies (Amin 1994, 1–34).

4 According to Yost (2017, 293) the term “automation” was first uttered by Ford
Motor Company vice president for manufacturing Delmar S. Harder, but Le Grand was
the first to publish the term.
1 INTRODUCTION: AUTOMATION, AUTONOMY … 7

The school of thought called post-operaismo (post-workerism) has been


highly influential in the analysis of post-Fordism, but it disputes the
complementary relationship between capital and computers suggested
above. Post-operaismo thinkers argue that the proliferation of computers
upsets the technological balance of power, in favour of labour. Previous
kinds of work are transformed into what they call “immaterial labour”
(Lazzarato 1996) which possesses new powers which “cannot be
contained by the capitalist modes of control” (Hardt and Negri 2009,
143). In the early days of the internet, Antonio Negri (1996), a
prominent figure in post-operaismo, asked: “[i]s this the third industrial
revolution or the time of transition to communism?” (156).
While information technology has yet to catapult the world into
communism, post-operaismo continues to advance its immaterial labour
theory with little revision. In this respect, Hardt and Negri’s Assembly
(2017) differs little from Empire (2000).5 Immaterial labour theory was
conceived around the birth of the public internet, but is it still useful
in the era of capitalist machine learning? Is it possible that information
technology is increasing the power of labour against capital, even as new
streams of capital coagulate around an artificially intelligent attractor and
big technology firms achieve previously unheard of power and wealth?

The Central Argument


Post-operaismo thinkers celebrate the proliferation of information tech-
nologies because they see them as increasing the capacities of the
labouring class for self-organized immaterial labour. Empowered by tech-
nology, immaterial labourers, the argument goes, can obtain an increasing
autonomy from capital. Eventually, they can throw off capital completely
and transition to a new autonomous mode of production. The claim for
a new sense of autonomy is the central focus of this book. I will assess

5 The past decade has seen the rise of techno-optimist left accelerationist thinkers, influ-
enced by immaterial labour theory (Williams and Srnicek 2014; Mason 2016; Bastani
2019). These thinkers attribute similar revolutionary significance to fourth industrial
revolution technologies as post-operaismo did to those of the third. Their calls for full
automation and the end of work have been extolled by no less than Negri (2014) himself.
I focus primarily on post-operaismo throughout the book, but consider left accelerationism
briefly in the conclusion.
8 J. STEINHOFF

it via a case study of the AI industry. According to post-operaismo, high-


tech workers, such as the data scientists and machine-learning engineers in
the AI industry, should be immaterial labourers par excellence. However,
if we examine the work they do, we find little evidence that these
labourers possess the attributes accorded to them by post-operaismo—
especially autonomy from capital. On the contrary, AI work presents us
with yet another example of the fragmentation, deskilling and automa-
tion of labour processes long familiar to Marxist labour process studies.
I argue that the AI industry demonstrates the increasing autonomy of
capital from labour and not the other way around.
Central to my argument is an empirical study of work in the AI
industry based on interviews that I conducted in 2017–2018 with
workers (labour) and management (representatives of capital) in the AI
industry. I interviewed a variety of technical workers, including data
scientists and machine-learning engineers, as well as several CEOs.6 For
anonymized profiles of the interviewees, see the Appendix. Post-operaismo
tends to make sweeping claims about the economy as a whole so it is
useful to test its claims by drilling down into a specific sector.
The empirical study is interpreted through a reading of Marx which
draws on labour process theory (LPT) and value-form analysis or the
New Reading of Marx (NRM). The crux of the study is an analysis of
the machine learning labour process, or the concrete series of actions
that go into producing AI commodities (Chapter 6). Labour and capital
encounter one another in the labour process and their encounters are
mediated by technology. How labour processes are structured and how
they change can tell us about relations between labour and capital. The
labour process is where the new autonomy of immaterial labour posited
by post-operaismo theorists should be evident.
Analysis of the labour process alone is, however, not enough. Labour
processes are structured according to the broader dynamics of an industry,
from class struggle to inter-capitalist competition, and must therefore be
situated in that context. To get such a higher order perspective I also
conduct a political economy analysis of the AI industry (Chapter 5).

6 There are two unfortunate lacunae in my interviews. Women are entirely unrepre-
sented. The difficulty I had recruiting women reflects the gender imbalance in the industry
discussed in Chapter 5. Ghost workers, whose essential contributions to AI production
are discussed also in Chapter 5, are also unrepresented. This is simply because, at the time
I conducted interviews, I was unaware of the significance of their work.
1 INTRODUCTION: AUTOMATION, AUTONOMY … 9

Further, since any given industry is a product of the social and material
conditions that came before it, an attempt at historicizing is also neces-
sary. This book therefore also includes a rough history of how AI became
an industry (Chapter 4). A complete chapter breakdown is provided at
the end of this chapter. But first we need to determine exactly what kind
of machine AI is.

Computing Machinery
Artificial intelligence is so infused with hype that it is important to begin
discussing it from the most mundane level. AI is software, intended to
model some aspect(s) of intelligence (loosely defined), which runs on the
hardware of a digital computer. AI is not equivalent to robotics. Robots
may incorporate AI, but this is not necessarily the case. Robots have
bodies of some type while AI systems are software programs. A precise
definition of AI is hard to come by because there is no consensus on how
to define intelligence nor on how it should be implemented in machines
(Legg and Hutter 2007; Wang 2008). One early AI researcher defined
AI as “making a machine behave in ways that would be called intelli-
gent if a human were so behaving” (McCarthy et al. 1955). Another
defines AI as “making machines intelligent, and intelligence is that quality
that enables an entity to function appropriately and with foresight in
its environment” (Nilsson 2010, xiii). The problem of definition has
been further complicated by a phenomenon, often called the AI effect,
whereby once a problem in AI is solved, it is no longer regarded as
requiring intelligence. AI pioneer and first AI historian Pamela McCor-
duck (2004 [1979]) notes that “every time somebody figured out how
to make a computer do something—play good checkers, solve simple but
relatively informal problems—there was a chorus of critics to say, ‘that’s
not thinking’” (204).
There is one famous way of defining AI which requires mention. The
mathematician, computer scientist and cryptographer Alan Turing (1950)
proposed it first in his now famous thought experiment “the Imitation
Game”. Often called the Turing Test today, this game involved a subject
engaging in textual dialogue with a machine and a human, both of which
were hidden from view. The subject would attempt to determine which
was human and which was machine, and if he/she could not, the machine
would win the game. The Turing Test posits that when an interlocutor
cannot distinguish between machine and human, the question of machine
10 J. STEINHOFF

intelligence becomes moot. Turing’s test is thus not concerned with what
the inner workings of the machine are, nor how (or if) they produce
intelligence at all. Turing’s approach, which centres on the appearance of
intelligence, rather than its technical definition, is philosophically inter-
esting, but it is not very useful when talking about AI that exists in the
world today, most of which we do not engage with linguistically.
Instead, let us assemble a working definition of AI that applies to all its
possible applications. For Wang (2008), intelligence is “adaptation with
insufficient knowledge and resources” which means that an intelligent
system “is finite, works in real-time, is open to novel tasks, and learns
from experience” (371). Kaplan (2016) defines the “the essence of intel-
ligence” as the “ability to make appropriate generalizations in a timely
fashion based on limited data. The broader the domain of application,
the quicker conclusions are drawn with minimal information, the more
intelligent the behavior” (5–6). By positing criteria of temporal finitude,
finite data and generality of application, the definitions offered by Kaplan
and Wang aim to distinguish AI from mere computation. Their definitions
usefully highlight the fact that human intelligence functions with limited
data all the time, except in formal situations with “perfect information”
like Chess (Mycielski 1992, 42). They also draw on the intuition that
a program which could theoretically solve a difficult problem if it were
given 4,000,000 years to conduct an exhaustive search through every
possible solution does not possess intelligence in the same sense that we
attribute to humans.
These definitions also usefully raise the issue of generality. True intelli-
gence, they suggest, has general applicability, while mere programs have a
single function. Artificial general intelligence (AGI) is the name given to
hypothetical AI with human-like capacities “to transfer learning from one
domain to other domains” (Muehlhauser 2013).7 While AGI research
remains highly speculative, there are at least 45 active AGI research
projects across academia and industry (Baum 2017). Actually existing
AI is, by contrast, called “narrow” because each system is designed to
accomplish a particular task (Johnson et al. 2016, 4246). Just as a word
processor is not useful for generating 3D models, facial recognition AI

7 Yann LeCun, the deep learning expert, makes the compelling point that human intel-
ligence is “nowhere near general” because there are many operations it is very poor at or
incapable of (quoted in Macaulay 2020). Regardless, we can consider human intelligence
relatively general.
1 INTRODUCTION: AUTOMATION, AUTONOMY … 11

is not going to make stock market predictions or interpret speech. All


references to AI in this book, unless otherwise noted, refer to actually
existing narrow AI.
I suggest a working definition of AI as the array of computational tech-
niques intended to model some aspect(s) of intelligence (loosely defined)
which are capable of overcoming barriers of finite time, data and resources
and are possibly capable of transferring learning from one domain to
others. As a heuristic, the more domains a program can function in,
and the less time, data and resources it requires, the more “intelligence”
it exhibits. This is an imperfect formulation, but it will suffice for the
purposes of this book. As we will see, the divide between AI and computa-
tion is not a hard and fast one.8 Even the most advanced AI involves, to a
large degree, computing techniques which require much more resources,
such as data, than human intelligence. On this heuristic continuum of
intelligence, there is no existing AI that is anywhere near the human.
Actually existing AI does not, in any widely accepted sense, think. It
is worth spending some time elaborating this. Even if we attempt to
distinguish AI from mere computation, as a program, what AI does is
compute. Computation is the execution of a process which “proceed[s]
from initially given objects, called inputs, according to a fixed set of rules,
called a program, procedure, or algorithm, through series of steps and
arrive[s] at the end of these steps with a final result, the output” (Soare
1996, 286). Computation can only proceed if the algorithm governing
it is “precise and definite, with each successive step clearly determined”
(Soare 1996, 286). As anyone who has tried coding will know, the
slightest error can render a program useless, because a computer follows
all instructions precisely, lacking anything like common sense or reflec-
tion. Computation occurs “without any understanding of the significance
or meaning of the task” (Carter 2007, 55).9

8 One hard boundary we should establish is between actually existing and fictional AI.
While we will not spend time on the philosophical aspects of AI, it is important that AI
not be conceived as sentient or conscious machinery. For the purposes of this book, it is
irrelevant whether AI ever could be truly intelligent, conscious, moral or anything else.
We are interested in what sorts of uses actually existing AI is being put to in the absence
of such qualities.
9 Unfortunately this is not the place to explore whether or not the human mind–
brain is a computational system. For more on computationalism, see Piccinini (2015) and
Negarestani (2018).
Another random document with
no related content on Scribd:
si commode! Vous pouvez de Vienne envoyer un projectile dans
Kaltenleutgeben avec un fusil calibre 12; là, en moins d’une heure, par des
trains étonnamment lents, vous pouvez oublier dans des collines boisées et
dans des sentiers ombragés la clarté fatigante et la grosse chaleur de la ville;
vous pouvez trouver un air pur et frais, le concert des oiseaux, le repos et le
calme du paradis.
A proximité de Vienne, il existe une multitude d’autres sanatoriums tous
plus charmants les uns que les autres. Vienne est située au centre d’un
magnifique amphithéâtre de montagnes couvertes de lacs et de forêts;
vraiment aucune autre ville ne jouit d’une position aussi heureuse.
Le sanatorium de Hochberghaus représente un bâtiment de grande
dimension planté solitairement au sommet d’une montagne couverte de
forêts épaisses. C’est un établissement pour l’appétit, et les gens qui ont
perdu leur appétit viennent là pour le retrouver. A mon arrivée, le
professeur Haimberger me prit à part dans son cabinet de consultation et me
posa les questions suivantes:
—Il est six heures. Quand avez-vous pris votre dernier repas?
—A midi.
—Qu’avez-vous mangé?
—Presque rien.
—Qu’y avait-il sur la table?
—Les mets habituels.
—Côtelettes, poulets, légumes, etc.?
—Oui, mais ne m’en parlez pas, cela m’exaspère.
—En êtes-vous dégoûté?
—Oh! oui, effroyablement: je voudrais qu’on ne m’en parlât plus jamais.
—La simple vue de la nourriture vous écœure, n’est-ce pas!
—Plus que cela, elle me révolte.
Le docteur réfléchit un instant, puis prenant un long menu, il le
parcourut lentement des yeux.
—Je pense, dit-il, que vous désirez plutôt manger des... mais, tenez,
choisissez vous-même.
Je jetai un coup d’œil sur cette liste et mon estomac se souleva
d’horreur: de toutes ses combinaisons barbares à jamais inventées, celle-là
était certainement la plus atroce. En tête du menu je lus: gras-double
coriace, avancé, assaisonné à l’ail; au milieu de ce menu, mes yeux
tombèrent sur: jeune chat, vieux chat, chat en salmis; au bas je lus: bottes de
matelot frottées au suif et servies crues.
Je ne pus m’empêcher de dire au docteur:
—Je ne comprends pas que vous preniez en plaisantant un cas aussi
sérieux que le mien. Je suis venu ici pour retrouver de l’appétit et non pour
perdre le peu qui me reste.
Il me répondit gravement:
—Je ne plaisante pas; pourquoi plaisanterais-je?
—Mais je ne puis manger ces horreurs!
—Pourquoi pas? me demanda-t-il avec une naïveté admirable.
—Pourquoi pas? parce que, docteur, depuis des mois je ne puis rien
supporter de plus substantiel que des omelettes et des flans. Vos plats
innommables...
—Oh! vous arriverez à les aimer. Ils sont très bons, il faut que vous en
mangiez; c’est le règlement de mon établissement et je ne tolérerai aucune
infraction.
—Dans ce cas, docteur, répondis-je en souriant, vous ne m’empêcherez
pas de partir, je pense; je m’en vais.
Il parut froissé et me dit sur un ton qui changea l’aspect des choses:
—J’espère bien que vous ne me causeriez pas ce préjudice. Je vous ai
accepté dans mon établissement en toute confiance; vous ne me ferez pas
pareil affront; ce sanatorium est mon gagne-pain; si vous partiez d’ici avec
le triste appétit que vous avez actuellement, on le saurait, et on ne
manquerait pas de dire que ma cure n’a pas réussi pour vous; les gens en
concluraient que les autres cas doivent ressembler au vôtre. Vous ne partirez
pas; vous ne me ferez pas cet affront!
Je lui présentai des excuses et lui promis de rester.
—A la bonne heure, j’étais sûr que vous ne partiriez pas et que vous ne
retireriez pas le pain de la bouche de mes enfants.
—Votre famille mange donc ces mets abominables?
—Ma famille! naturellement non.
—Ah! et vous?
—Moi non plus.
—Ah! je commence à comprendre. Vous êtes comme les médecins qui se
gardent bien de suivre leurs ordonnances.
—Il ne s’agit pas de cela. Vous n’avez pas mangé depuis six heures,
voulez-vous souper maintenant ou plus tard?
—Je n’ai pas faim. Mais peu m’importe l’heure, je voudrais seulement
ne plus penser à manger. C’est à peu près l’heure à laquelle je prends mon
souper, et la régularité m’est recommandée par tous les médecins, soit, je
vais essayer de grignoter un peu, mais j’aurais besoin d’un stimulant.
Le docteur me tendit son odieux menu.
—Choisissez, à moins que vous ne préfériez prendre votre repas plus
tard.
—Oh! voyez-vous, montrez-moi le chemin de ma chambre, je vous prie,
je ne puis me soumettre à votre règlement.
—Attendez un instant avant de prendre une décision définitive; voici un
autre règlement: si vous choisissez immédiatement, on vous apportera de
suite les plats que vous désignerez sur le menu; mais si vous différez, il
vous faudra attendre mon bon plaisir. Vous ne pouvez manger un plat de ce
menu sans ma permission.
—Dans ce cas montrez-moi ma chambre et envoyez la cuisinière se
coucher, car rien ne presse pour moi.
Le professeur me fit monter l’escalier et me désigna un appartement très
confortable se composant d’un petit salon, d’une chambre à coucher, d’une
salle de bains.
Les fenêtres donnaient sur un paysage verdoyant sur des prés, des
vallons, et des collines couvertes de forêts. Dans le salon j’aperçus
beaucoup de livres sur une étagère. Le professeur me dit qu’il allait me
laisser seul et il ajouta:
—Fumez, lisez tant qu’il vous plaira, buvez autant d’eau que vous
voulez. Quand vous sentirez la faim, sonnez et commandez; je déciderai
alors si on doit ou non exécuter vos ordres. Votre cas est mauvais et très
enraciné; je crois que les quatorze premiers plats du menu sont trop délicats
et ne vous conviendraient pas. Je vous demande comme grande faveur de
vous en abstenir et de ne pas les commander.
—Rassurez-vous, et ne vous mettez pas martel en tête à ce sujet. Avec
moi vous allez faire des économies; l’idée d’amuser l’appétit d’un homme
malade avec ce menu à peine digne d’une buse est une pure insanité.
Je lui fis cette observation avec amertume, car je me sentais outré de ces
procédés barbares qui équivalaient à un assassinat d’un nouveau genre. Le
docteur parut mécontent, mais non offensé; il posa le menu sur la commode
à la tête de mon lit et à portée de ma main, et ajouta:
—Votre cas n’est pas un des pires que j’aie rencontrés, pourtant je le
considère comme grave, et il exige un traitement énergique. Je vous
demande donc de restreindre votre appétit et de ne considérer aucun plat
avant le numéro 15 du menu.
Il me quitta; je commençai à me déshabiller, car j’étais éreinté et je
tombais de sommeil. Je dormis 15 heures et me réveillai bien reposé à 10
heures du matin. Le café de Vienne! Je pensai dès mon réveil à cette
jouissance incomparable, à ce délicieux café à côté duquel tous les autres
cafés d’Europe et des hôtels américains sont de la détestable lavasse. Je
sonnai et commandai un café avec du pain viennois, autre délicieuse
invention. Le garçon me parla à travers le guichet pratiqué dans la porte, et
me dit de consulter le menu. Je le congédiai immédiatement. Après mon
bain je m’habillai et me préparai à faire un tour de promenade; je trouvai la
porte fermée, j’étais bel et bien bloqué. Je sonnai: le garçon arriva et
m’expliqua que, d’après le règlement, les patients ne pouvaient sortir avant
le premier repas. Je n’avais jamais éprouvé auparavant l’envie de sortir,
mais cette fois c’était différent. Rien ne vous donne envie de prendre l’air
comme la réclusion.
A deux heures de l’après-midi je n’avais rien mangé depuis 26 heures;
j’avais bien quelquefois senti la faim, mais cette fois je m’aperçus que
j’éprouvais une véritable fringale avec un grand F.
Pourtant, je ne me sentais pas assez affamé pour affronter l’odieux menu.
Pour faire bonne contenance j’essayai de lire et de fumer; les livres à ma
disposition ne parlaient que de naufrages, de gens perdus dans le désert,
enfouis sous des éboulements de mines; de gens mourant de faim dans des
villes assiégées; les plats les plus immondes avec lesquels les hommes
affamés assouvissent leur faim en pareil cas défilèrent devant mes yeux. Au
début ces récits me soulevèrent le cœur; au bout de quelques heures, ils me
produisirent une impression moins répugnante, et je me surpris même à
claquer des lèvres en pensant à ces mets odieux. Après 45 heures de jeûne,
je tirai vigoureusement la sonnette et je commandai le second plat du menu:
une espèce de compote composée de caviar et de goudron. On me le refusa
encore. Pendant les 15 heures qui suivirent, je tirai la sonnette et
commandai successivement un plat en suivant la liste de haut en bas. Toutes
les fois j’éprouvai un refus: en arrivant au nº 15 je me sentis rassuré et mon
cœur se mit à battre de plus en plus vite à mesure que mon espoir
augmentait. Je n’avais pas mangé depuis 60 heures, mais je me sentais sûr
de la victoire; je commandai donc le nº 15: jeune poulet bouilli dans l’œuf,
six douzaines servies chaudes et sentant bon! Un quart d’heure après ce plat
arrivait escorté du docteur qui se frottait les mains de joie. Il me dit avec
excitation:
—En voilà une cure, une vraie cure! je savais bien que j’y arriverais, ma
méthode ne manque jamais le but; eh! bien, vous avez retrouvé votre
appétit, vous le reconnaissez; dites-le moi pour me faire plaisir.
—Apportez-moi vos saletés maintenant; je puis manger tout ce qui
figure sur votre menu.
—Voilà qui est superbe! je savais bien que mon système est infaillible.
Comment trouvez-vous ces oiseaux?
—Je n’ai jamais rien mangé d’aussi délicieux et pourtant en principe je
n’aime pas la volaille. Voyez-vous, ne m’interrompez pas, ne me faites pas
perdre une seule bouchée.
—La cure est parfaite, reprit le docteur, tout danger a disparu
maintenant; laissez de côté la volaille, je vais vous faire apporter un bifteck.
Le bifteck vint; gros à ne pas tenir dans un panier, avec des pommes de
terre, du pain viennois et du café. Je mangeai à pleines dents ce repas bien
digne de tous les sacrifices que j’avais faits pour le mériter et je laissai
tomber des larmes de reconnaissance dans la sauce. C’étaient des larmes de
reconnaissance à l’égard du docteur qui venait de me rendre une petite dose
du sens commun sorti de mon esprit depuis tant d’années.

CHAPITRE II
Il y a trente ans de cela, le professeur Haimberger faisait un long voyage
sur un bateau à voiles; il y avait 15 passagers à bord. Le service courant de
table était ainsi réglé: à 7 heures du matin une tasse de mauvais café prise
au lit; à 9 heures déjeuner se composant de: mauvais café avec lait
conservé, petit pain graisseux, friture, poisson salé. A 1 heure lunch avec:
langue froide, jambon, bœuf conservé, petit pain graisseux. A 5 heures,
dîner avec soupe épaisse aux pois, poisson salé, bœuf conservé chaud servi
avec choucroute, porc bouilli garni de fèves, pudding. De 9 à 11 h. du soir
souper consistant en thé, lait conservé, langue froide, jambon, pickles,
biscuits de mer, huîtres aux vinaigres, pieds de cochon au vinaigre, os
grillés, lapin sauté.
A la fin de la première semaine, personne ne mangeait plus à bord, tout
le monde ne faisait plus que grignoter. Les passagers venaient à table sans
appétit uniquement pour tuer le temps et parce qu’ils savaient qu’il faut
manger à des heures régulières. Ce menu grossier et monotone les écœurait;
chaque jour, on les voyait errer sur le navire sans appétit, bâillant à se
décrocher l’estomac, de mauvaise humeur, maussades et silencieux. Parmi
eux se trouvaient trois dyspeptiques avérés, au bout de trois semaines ils
étaient devenus des ombres; il y avait aussi un invalide cloué sur son lit qui
ne mangeait que du riz bouilli et ne pouvait supporter la nourriture ordinaire
du bord. Survint un naufrage; les passagers furent recueillis dans les canots
de sauvetage et durent se contenter de la portion congrue qui s’impose en
pareil cas. A mesure que les provisions s’épuisaient, les appétits
augmentaient. Lorsqu’il ne resta plus que du jambon cru et que la ration de
chacun fut réduite à deux onces par jour, tous les passagers avaient un
appétit merveilleux. Au bout de 15 jours les dyspeptiques, l’invalide et les
passagères les plus difficiles mâchaient avec délice les bottes des matelots
et se plaignaient qu’on ne leur en donnât pas à discrétion. Pourtant,
quelques jours auparavant, ces mêmes passagers ne pouvaient supporter le
fastidieux bœuf salé, la choucroute et les autres crudités qu’on leur servait à
bord. Un navire anglais les recueillit. Au bout de 10 jours les 15 passagers
étaient en aussi bon état qu’au moment du naufrage.
—Leur mésaventure ne leur a causé aucun tort, dit le professeur, le
remarquez-vous?
—Oui.
—Vous en rendez-vous bien compte?
—Mais oui, je le crois.
—Vous paraissez hésiter et ne pas bien comprendre l’importance de ce
fait; je vous répète qu’aucun d’eux n’a souffert de cette mésaventure.
—Je le comprends bien et je le trouve étonnant.
—Pas le moins du monde étonnant; rien de plus naturel; ils étaient
soumis à une cure d’appétit naturelle, la meilleure et la plus sage au monde.
—Est-ce cet exemple qui vous a suggéré l’idée de votre traitement?
—Vous l’avez dit.
—Vous avez donné à ces gens une leçon bien précieuse.
—En effet, en ce qui concerne la santé et beaucoup d’autres choses, la
moyenne des gens ne connaissent que ce qui se fait autour d’eux, et ils
obéissent uniquement à de stupides préjugés; ils sont incapables de
coordonner trois ou quatre faits et d’en tirer des conclusions; incapables
d’observer par eux-mêmes, ils n’agissent que par routine et par imitation. Si
ceux que nous appelons les animaux inférieurs étaient aussi stupides que
l’homme, ils disparaîtraient tous de la terre en moins d’une année.
—Vos passagers n’ont donc pas profité de la leçon?
—Pas le moins du monde. A bord du navire anglais ils reprirent leurs
repas réguliers et bientôt on les vit grignoter de nouveau sans appétit,
dégoûtés de la nourriture, de mauvaise humeur, l’air piteux, bâillant, jurant
et grognant tout le long du jour.
—Ah! je comprends maintenant: votre théorie est...
—Bien simple. Ne mangez que quand vous aurez faim; si vous n’avez
pas de goût pour la nourriture, réjouissez-vous et ne mangez que quand
vous sentez une faim violente, alors vous éprouverez une grande jouissance
et vous vous porterez merveilleusement.
—Dois-je observer une certaine régularité pour les heures de mes repas?
—Lorsque vous êtes en train de vaincre un mauvais appétit, la régularité
n’est pas nécessaire; mais elle ne vous fera pas de mal aussi longtemps que
votre appétit restera bon. Dès que votre appétit diminue, appliquez de
nouveau le grand remède, c’est-à-dire le jeûne, long ou court suivant les
besoins du cas.
—La meilleure nourriture, je suppose, est la plus saine?
—Toutes les nourritures sont saines, du moins suffisamment saines pour
les gens qui en font usage. Que la nourriture soit raffinée ou grossière, elle
aura toujours bon goût, et elle rassasiera son homme à condition qu’il
surveille son appétit et qu’il s’impose un léger jeûne toutes les fois qu’il
sent son appétit faiblir. Nansen était habitué à des menus très soignés, mais
il ne souffrit nullement le jour où, à tous ses repas, il dut manger de la
viande d’ours et cela pendant des mois; parce que son appétit se voyait
toujours tenu en éveil par la difficulté de trouver régulièrement de la viande
d’ours.
—Mais les docteurs ordonnent une nourriture délicate et recherchée à
leurs malades.
—Ils ne peuvent faire autrement; car le malade est imbu d’idées
préconçues et de superstitions, et ne consentirait jamais à se mettre à la
diète. Il se croirait perdu s’il jeûnait.
—Cela l’affaiblirait, n’est-ce pas?
—Pas le moins du monde. Voyez plutôt les malades de notre naufrage:
ils vécurent pendant 15 jours d’une pincée de jambon cru et de bottes de
matelot à sucer; ils furent bien un peu affaiblis, mais ces privations ne leur
firent aucun mal; elles les préparèrent à manger de bon cœur des plats plus
substantiels, et leur valurent une robuste santé. Mais ils ne surent pas tirer
profit de cet enseignement et ils restèrent des malades. Connaissez-vous le
truc qu’emploient les docteurs de sanatorium?
—Quel est-il?
—Mon système repose sur la diète: cure de raisins, bains, cure de boue,
tout cela se vaut. Le raisin, le bain et la boue ne sont là que pour la frime; la
véritable cure s’opère par une diète subrepticement appliquée. Le malade
est habitué à quatre repas, à des heures tardives; examinez avec moi le
régime qu’il doit suivre au sanatorium: il se lève à 6 heures du matin,
mange un œuf, se promène deux heures avec les autres malades, boit
lentement un verre d’un liquide filtré qui sent la pourriture. Il se promène
ensuite deux heures, mais seul; si vous lui parlez il vous répond d’un air
préoccupé: «Mon eau! je promène mon eau pour la digérer; je vous en prie,
ne m’interrompez pas»; et il continue à arpenter le terrain.
Il mange ensuite une feuille de rose sucrée, s’étend dans sa chambre
pendant des heures dans le silence et la solitude les plus complets; il ne doit
ni lire, ni fumer. Le docteur vient lui ausculter le cœur, lui tâter le pouls, la
poitrine, le dos et l’estomac en écoutant tous ses battements dans un
flageolet de deux sous; puis il fait préparer le bain du malade d’un demi-
degré réaumur plus froid qu’hier. Après le bain, un autre œuf, un verre du
même liquide infect à 3 ou 4 heures de l’après-midi, puis promenade
solennelle suivant le rite non moins solennel. A six heures, dîner: un
beignet et une tasse de thé. Repromenade.
A 8 1/2, souper; à 9 h., lit. Songez un peu à six semaines de ce régime! Il
y a de quoi affamer un homme et le mettre en splendide forme. N’importe
où, à Londres, à New-York, à Jéricho, ce régime produirait le même effet.
—Combien faut-il de temps pour mettre ici une personne en forme?
—Régulièrement il faudrait un jour ou deux, mais en fait il faut compter
de une à six semaines, suivant le caractère et la mentalité du patient.
—Comment cela?
—Vous voyez là-bas cette foule de jeunes femmes qui jouent au foot-
ball, boxent et sautent par-dessus des barrières? Elles sont ici depuis 6 ou 7
semaines; à leur arrivée elles ressemblaient à des spectres; elles avaient
pour habitude de grignoter des friandises et des sucreries à heures fixes
quatre fois par jour et ne se sentaient d’appétit pour rien. Je leur posai des
questions et les enfermai dans leurs chambres; les plus frêles furent mises à
la diète pendant 9 ou 10 heures, les autres pendant 12 ou 15 heures. Bientôt
elles me supplièrent, disant qu’elles souffraient énormément de nausées, de
mal de tête. Il fallait les voir manger après cette réclusion! Je les laissai
ensuite circuler dans la maison et prendre les 4 repas auxquels elles étaient
habituées.
Au bout de 2 jours je dus intervenir, car leur appétit diminuait. Je fis
sauter un repas; cela les rétablit vite; puis elles reprirent les quatre repas; je
leur demandai de supprimer d’elles-mêmes un repas sans attendre mon
intervention. Pendant 15 jours, ces personnes luttèrent, mais ne purent y
arriver, car l’énergie leur manquait; mais maintenant elles ont acquis cette
énergie et elles se portent à merveille; d’elles-mêmes, de temps à autre,
elles suppriment un repas; elles pourraient en toute sécurité retourner
guéries chez elles, mais elles ne se sentent pas encore assez sûres d’elles-
mêmes, et elles préfèrent attendre un peu.
—Tous les cas ne sont pas identiques?
—Naturellement. Parfois un homme apprend ce truc en une semaine et il
sait régler son appétit en supprimant fréquemment un repas sans s’en
apercevoir.
—Mais pourquoi supprimer un repas tout entier au lieu d’une partie d’un
repas?
—C’est un moyen mesquin et incomplet. Si votre estomac ne fait pas
entendre un appel vigoureux, il vaut mieux le laisser tranquille et lui
accorder un repos complet. Certaines personnes peuvent prendre plus de
repas que les autres et s’en trouvent bien. Il y a autant d’espèces d’appétit
que d’espèces de gens. Je vais vous montrer tout à l’heure un homme qui
avait pour habitude de grignoter huit fois par jour. Je l’ai réduit maintenant
à 6 repas par jour et il se porte à merveille. Combien de repas par jour
prenez-vous?
—Autrefois, il y a 22 ans, je prenais un repas et demi; pendant ces deux
dernières années, 2 et demi: café et petit pain à 9 heures, lunch à 1 heure,
dîner à 8 heures.
—Autrefois vous preniez un repas et demi, c’est-à-dire café et petit pain
à 9 heures, dîner le soir, et rien entre, n’est-ce pas?
—Parfaitement.
—Pourquoi avez-vous ajouté un repas?
—Pour obéir à ma famille, qui s’imaginait que j’allais me tuer.
—Vous trouviez suffisant un repas et demi par jour pendant 22 ans?
—Très suffisant.
—Votre mauvais état de santé actuel est dû à ce repas supplémentaire.
Supprimez-le; vous mangez plus souvent que votre estomac ne vous le
demande, cette nourriture se perd au lieu de vous profiter. Vous mangez
moins maintenant en un jour à vos deux repas et demi qu’autrefois à un
repas et demi.
—C’est très vrai, beaucoup moins, car autrefois mon dîner était très
substantiel.
—Mettez-vous à un seul repas par jour, le dîner, pendant quelques jours,
jusqu’à ce que vous éprouviez un appétit vigoureux, régulier et bien assis;
ensuite adoptez en permanence un repas et demi par jour et n’écoutez pas
les balivernes de votre famille. Quand vous éprouvez un malaise,
particulièrement une atteinte de fièvre, ne mangez rien du tout pendant 24
heures; cela vous guérira et la diète triomphera du mal de tête le plus
opiniâtre. Je ne connais pas de douleur de tête qui puisse résister à la diète
absolue pendant 24 heures.
—C’est vrai, j’ai constaté ce fait bien des fois.
EXTRAIT DU TIMES DE LONDRES EN
1904
CHAPITRE PREMIER
Correspondance du Times.—Chicago, 1ᵉʳ avril 1904.
Je reprends ma conversation téléphonique interrompue hier.
Depuis plusieurs heures cette vaste cité ne cesse de parler de l’étrange
incident mentionné dans mon dernier rapport.
Pour répondre à votre désir, je vais vous conter ce roman depuis son
origine jusqu’à son dénouement. Par un concours de circonstances fortuites,
j’ai moi-même joué un rôle personnel dans une partie de ce drame. La scène
se passe à Vienne le 31 mars 1898 à 1 heure du matin. J’avais assisté dans
la soirée à une réunion mondaine; vers minuit je m’en allai en compagnie
des attachés militaires anglais, italien et américain; nous décidâmes de nous
rendre chez le lieutenant Hillyer, attaché militaire américain, pour terminer
la soirée et fumer ensemble quelques cigarettes.
En arrivant, nous trouvâmes plusieurs visiteurs qui nous avaient
devancés: le jeune Chepannik, M. K. son bailleur de fonds, M. W.,
secrétaire de ce dernier, et le lieutenant Clayton, appartenant à l’armée des
Etats-Unis. A ce moment, la guerre était imminente entre l’Espagne et notre
pays, et le lieutenant Clayton venait d’arriver en Europe chargé d’une
mission militaire.
Je connaissais bien le jeune Chepannik et ses deux amis, et j’avais
rencontré M. Clayton une ou deux fois à West-Point, quand il était à l’école
des cadets. Ce dernier passait pour un officier capable, plutôt emporté de
caractère et qui avait son franc-parler.
Cette réunion intime offrait un peu un but d’affaire: il s’agissait
d’examiner la possibilité d’adapter le télélectroscope au service de l’armée.
Ce projet semble étrange aujourd’hui, mais il n’en est pas moins vrai qu’à
cette époque personne ne prenait au sérieux cette invention, à l’exception
toutefois de son auteur. Même le bailleur de fonds de l’inventeur considérait
cet instrument comme un curieux et bizarre jouet. Il en était tellement
persuadé qu’il venait d’ajourner à la fin de ce siècle la mise en circulation
publique de son invention, en la cédant à bail pour deux ans à un syndicat
qui devait l’exploiter à l’exposition de Paris.
Au moment où nous entrions au fumoir, le lieutenant Clayton et
Chepannik discutaient chaudement en allemand sur le télélectroscope.
Clayton disait:
—Maintenant, vous connaissez ma façon de penser; et il accompagna
cette déclaration d’un vigoureux coup de poing sur la table.
—Mais je n’en fais aucun cas, répliqua le jeune inventeur, sur un ton
aussi calme que blessant.
Clayton se tourna vers M. K. et reprit:
—Je me demande pourquoi vous gaspillez votre argent à répandre ce
jouet. A mon avis, jamais il ne rendra le moindre service à l’humanité.
—C’est possible, c’est possible, mais je ne regrette pas l’argent que j’ai
hasardé; pour ma part je le considère comme un simple jouet, mais
Chepannik en augure mieux et je le connais assez pour croire qu’il voit plus
juste que moi (avec ou sans son télélectroscope).
Cette répartie ne fit qu’irriter Clayton davantage; il affirma de nouveau
sa conviction que cette invention ne rendrait jamais à l’humanité le moindre
service. Posant sur la table un farthing anglais, il ajouta:
—Prenez cette pièce, monsieur K., et emportez-la; si jamais le
télélectroscope rend quelque service à un être humain (j’entends un service
réel), vous m’adresserez par la poste cette pièce, et je retirerai alors ce que
j’ai dit. Est-ce entendu?
—Entendu; et, ce disant, M. K. mit le farthing dans sa poche.
M. Clayton se tourna alors vers Chepannik et se mit à l’invectiver sur un
ton plutôt injurieux. Chepannik lui répondit vertement et lui administra un
coup de poing. Il s’ensuivit un pugilat, et les attachés durent séparer les
deux combattants.
La scène maintenant se passe à Chicago, en automne 1901. Dès que le
syndicat de Paris eut exploité le télélectroscope, cet appareil fut livré à
l’usage public et s’adapta bientôt au système téléphonique du monde entier.
Le téléphone «à distance illimitée» venait d’atteindre son dernier
perfectionnement, et chacun pouvait suivre les événements quotidiens du
globe; deux témoins séparés par n’importe quelle distance pouvaient
échanger leurs vues sur les faits mondiaux.
Chepannik venait d’arriver à Chicago; Clayton, alors capitaine, était au
ministère de la Guerre. Ces deux individus rouvrirent leur querelle de 1898;
à trois reprises différentes ils en vinrent aux coups et furent séparés par des
témoins. Un beau jour, Chepannik disparut pendant deux mois; ses amis
supposèrent d’abord qu’il était parti en voyage d’excursion, et qu’il leur
donnerait bientôt de ses nouvelles, mais ils ne reçurent rien de lui; ils
pensèrent alors qu’il était retourné en Europe; mais le temps passait, et on
ne savait toujours rien de Chepannik; personne ne s’en inquiéta, car, comme
beaucoup d’inventeurs et de poètes, il pouvait, au gré de son caprice, aller
et venir sans avertir personne.
C’est maintenant que se passe la tragédie. Le 29 décembre, dans un
compartiment sombre et retiré de la cave située sous la maison du capitaine
Clayton, un cadavre était découvert par une des servantes de ce dernier. Les
amis du défunt reconnurent sans hésitation le cadavre de Chepannik; cet
homme avait trouvé une mort violente.
Clayton fut arrêté, cité en justice, et accusé de ce meurtre. Toutes les
preuves les plus indiscutables étaient contre lui: Clayton le reconnaissait
lui-même. Il n’en jura pas moins qu’il n’avait pas commis ce crime et qu’il
lui était complètement étranger.
Les lecteurs s’en souviennent, il fut condamné à mort; ses nombreux et
puissants amis firent l’impossible pour le sauver, car personne parmi eux ne
mettait en doute la sincérité de son serment. Moi-même, je cherchai à le
sauver, car, depuis que je le fréquentais avec suite, j’avais acquis la
conviction qu’il était incapable d’attirer un ennemi dans un coin pour
l’assassiner. Au cours de 1902 et de 1903, le gouverneur lui accorda
plusieurs sursis; finalement le jour de l’exécution fut fixé au 31 mars. Le
gouverneur se trouvait dans une situation des plus embarrassantes, car il
était le propre oncle de la femme de Clayton. Le mariage des deux jeunes
gens avait eu lieu en 1899; Clayton était alors âgé de 34 ans et la jeune fille
de 33; de leur union très heureuse naquit une petite fille, qui avait alors trois
ans. Un sentiment de pitié pour la jeune mère et pour son enfant contint
d’abord les protestations; mais cela ne pouvait durer, car en Amérique la
politique a main sur tout; bientôt les ennemis politiques du gouverneur
signalèrent à l’attention publique les sursis successifs qu’il accordait, et ils
réclamèrent que justice se fît.
Les partisans mêmes du gouverneur s’inquiétèrent, et parurent nerveux;
ils vinrent plusieurs fois le trouver à Springfield, et eurent avec lui de longs
entretiens.
Le gouverneur se trouvait pris entre deux feux; d’un côté sa nièce le
suppliait de pardonner à son mari; de l’autre ses partisans lui demandaient
de se souvenir qu’il était le premier magistrat de l’Etat et qu’il fallait en
finir avec l’exécution de Clayton.
Cédant à la voix du devoir, le gouverneur donna sa parole que
l’exécution ne serait plus reculée; ceci se passait il y a deux semaines. Mrs
Clayton l’implora de nouveau, disant:
—Maintenant que vous venez de donner votre parole, mon dernier espoir
s’évanouit, car je sais que vous ne reviendrez jamais sur votre promesse.
Vous avez fait tout votre possible pour sauver John; je n’ai rien à vous
reprocher; vous l’aimez, vous m’aimez aussi, et si vous pouviez le sauver
vous le feriez. Il ne me reste plus qu’à adoucir les quelques jours qui lui
restent à vivre avant que la nuit éternelle ne vienne assombrir ma vie. Oh!
vous serez avec moi, ce jour terrible; vous ne me quitterez pas, n’est-ce
pas?
—Je vous le promets, pauvre enfant, je resterai auprès de vous.
Sur l’ordre du gouverneur, Clayton reçut tous les adoucissements
possibles dans sa prison; sa femme et son enfant passaient les journées
auprès de lui; moi, je lui tenais compagnie la nuit. On le fit sortir de l’étroite
cellule qu’il occupait depuis le commencement de sa réclusion pour lui
donner les appartements spacieux et confortables du gardien-chef. Il ne
pouvait s’empêcher de penser à l’infortuné inventeur, lâchement assassiné,
et aux joies que lui procurerait le télélectroscope s’il en possédait un dans sa
prison. Son désir fut exaucé.
Mis en communication avec le poste téléphonique international, il put
ainsi chaque jour et chaque nuit appeler un coin différent du globe, prendre
part à sa vie, jouir de ses spectacles étranges, parler avec ses habitants; en
un mot, s’imaginer, grâce à ce merveilleux instrument, qu’il était aussi libre
que les oiseaux de l’air. Cette illusion lui faisait oublier un instant sa prison,
ses chaînes et ses barreaux macabres. Il parlait rarement, et je me gardais
bien de l’interrompre lorsqu’il était absorbé par son appareil.
Je restais assis dans son petit salon, je lisais, je fumais, et les soirées me
parurent calmes, reposantes et plutôt agréables. De temps à autre je
l’entendais crier: «Donnez-moi Yeddo; donnez-moi Hong-Kong; donnez-
moi Melbourne.» Je continuais à lire, à fumer confortablement, tandis qu’il
était en communication avec ces pays lointains où le soleil commençait à
poindre à l’horizon, et où les travailleurs se rendaient à leur travail
quotidien. Parfois, la conversation qui arrivait de ces régions lointaines
m’intéressait, et j’écoutais.
Hier, l’instrument resta silencieux; c’était plutôt naturel, car l’exécution
devait avoir lieu le lendemain.
Cette veille fatale fut consacrée aux larmes, aux sanglots et aux adieux.
Le gouverneur, la femme du prisonnier et son enfant restèrent à la prison
jusqu’à onze heures et quart du soir et les scènes auxquelles j’assistai me
désolèrent. L’exécution devait avoir lieu à 4 heures du matin. Quelques
minutes après onze heures, des coups de marteau interrompirent le silence
de la nuit. Une grande clarté se produisit au dehors et l’enfant se mit à crier:
«Qu’est-ce que c’est, papa?» Courant à la fenêtre avant qu’on ait pu
l’arrêter, la petite fille battit des mains et appela sa mère: «Oh! viens vite,
maman, viens voir la jolie chose qu’ils préparent!» La pauvre mère savait
ce qui se passait; elle s’évanouit. Le gibet se dressait sous les fenêtres du
prisonnier.
L’infortunée fut emportée plus morte que vive dans ses appartements;
Clayton et moi nous restâmes seuls, pensifs, songeurs et immobiles comme
des statues.
La nuit était froide et lugubre, car l’hiver venait de faire sa réapparition
momentanée, comme il arrive souvent au début du printemps. Le ciel était
sombre et sans étoiles, et un vent violent soufflait du lac. Le silence de la
pièce que nous occupions était si complet que, par contraste, tous les bruits
du dehors prenaient des proportions exagérées: ces bruits s’harmonisaient
avec la situation et avec notre état d’âme: le tonnerre retentissait, le vent
soufflait en tourbillons et avec des gémissements lugubres; de temps à
autre, une giboulée de grésil venait cingler les vitres, et au dehors, sans
discontinuer, les constructeurs de gibet martelaient leurs sourds coups de
masses. Au bout d’un certain intervalle, un nouveau son arriva jusqu’à
nous, à peine perceptible au milieu des mugissements de la tempête. Une
cloche sonnait minuit. Nous entendîmes avec la même anxiété la même
cloche marquer de sa voix lugubre un, deux, trois coups. Cette fois nous
retînmes notre respiration; Clayton n’avait plus que 60 minutes à vivre!
Clayton se leva et, s’approchant de la fenêtre, il leva les yeux vers le ciel
sombre; écoutant les bourrasques de vent et de pluie, il murmura: «Est-ce là
la dernière vision terrestre d’un homme qui va mourir?» Puis il ajouta:
«Non! il faut que je voie encore le soleil; il me faut du soleil. Un instant
après, il cria fiévreusement: «Chine, donnez-moi la Chine, Péking!»
Etrangement ému je pensai en moi-même: «Dire qu’un être humain est
capable d’accomplir cet inimaginable prodige de changer l’hiver en été, la
nuit en jour, la tempête en calme, de donner l’illusion de la liberté à un
prisonnier dans sa cellule, et de verser des rayons de soleil d’une clarté
aveuglante sur un pauvre diable qui va mourir dans une profonde
obscurité!»
Je prêtai l’oreille.
—Quelle clarté! quelle lumière! quelle splendeur! est-ce bien Péking?
—Oui.
—Quelle heure est-il?
—Midi.
—Que signifient cette foule et ces costumes de gala? cette profusion de
riches couleurs et ce luxe somptueux? Pourquoi toutes ces splendeurs et
cette clarté aveuglante?
—On va couronner notre nouvel empereur.
—Mais, je croyais que cette cérémonie avait eu lieu hier?
—Aujourd’hui, pour nous, représente hier pour vous.
—C’est vrai, j’ai l’esprit confus ces derniers jours pour certaines raisons
particulières.......—Est-ce le commencement de la procession?
—Non, mais seulement les préparatifs deux heures avant.
—Encore deux heures à attendre!
—Oui; pourquoi soupirez-vous?
—Parce que j’aimerais assister à toute la procession.
—Et pourquoi ne le pourriez-vous pas?
—Il faut que je sorte immédiatement!
—Vous avez un rendez-vous?
Après une pause:—Oui.
—Après une autre pause: Que vois-je dans ce splendide pavillon?
—C’est la famille impériale et toutes les têtes couronnées du monde
entier venues pour la cérémonie.
—Qui sont ces personnes dans les deux pavillons de droite et de gauche?
—A droite, ce sont les ambassadeurs, leur famille et leur suite; à gauche,
des étrangers de marque.
—Vous seriez bien bon de...
Boum! à travers la tempête de pluie et de vent, la même cloche qui tout à
l’heure venait de sonner une demi-heure. La porte s’ouvrit, le gouverneur,
la jeune femme et l’enfant entrèrent; l’infortunée portait des vêtements de
deuil. Elle s’effondra en sanglotant sur la poitrine de son mari. Ne pouvant
supporter ce spectacle lamentable, je me retirai dans la chambre voisine et
fermai la porte. Je restai là silencieux, attendant, prêtant l’oreille aux
grincements et au fracas de la tempête. Au bout de quelques minutes, qui
me parurent un siècle, un certain mouvement se produisit dans le salon où
se tenait le condamné; je compris que le clergyman, le sheriff et la garde
venaient d’entrer.
Après quelques mots échangés à voix basse, un grand silence se fit;
j’entendis un murmure, une prière, des sanglots, puis des pas—le départ
pour le gibet.
L’enfant cria d’une voix douce: «Ne pleure pas, maman, on nous rendra
bientôt papa, et nous le ramènerons à la maison.»
La porte se referma. Ils étaient partis. Je me sentis honteux. De tous les
amis du condamné j’étais le seul à ne pas l’accompagner jusqu’au gibet;
sans courage, sans volonté, j’arpentais la pièce en luttant contre moi-même
et en me répétant que j’allais le suivre. Mais nous sommes ce que nous
sommes, et nous ne pouvons changer notre nature. Je restai là, sur place. Je
repris ma marche nerveuse et agitée au travers de la pièce et soulevai la
fenêtre, poussé par cette étrange fascination qu’exercent sur l’esprit humain
la terreur et l’effroi, et je plongeai du regard dans la cour. A la clarté
aveuglante des lampes électriques, j’aperçus le petit groupe des témoins
privilégiés, la jeune femme sanglotant contre la poitrine de son oncle: le
condamné se tenait debout sur l’échafaud, la corde passée autour du cou, les
bras liés derrière le dos, la tête couverte d’un bonnet noir; le sheriff, à ses
côtés, avait une main appuyée sur la bascule, le clergyman se tenait devant
lui, la tête découverte et son livre à la main:
—«Je suis la Résurrection et la Vie.»
Je tournai la tête, ne pouvant en entendre ni en voir davantage.
Machinalement et sans m’en rendre compte, mes yeux se portèrent sur
l’étrange appareil, et voici que Péking et la procession du Fils du Ciel
m’apparurent. Une seconde plus tard je me penchai à la fenêtre, essoufflé,
suffoquant, essayant de parler, sentant qu’il le fallait absolument. J’entendis
ces paroles du clergyman: «Que Dieu ait pitié de votre âme. Amen.»
Le sheriff fit tomber le bonnet noir sur le visage du condamné et posa la
main sur la bascule. Rassemblant toutes mes forces je m’écriai:
—Arrêtez, pour l’amour de Dieu, arrêtez, cet homme est innocent. Venez
voir ici Chepannik en chair et en os.
Trois minutes après, le gouverneur prenait ma place à la fenêtre, et
ordonnait:
—Faites tomber ses liens et rendez-lui la liberté.
Trois minutes plus tard, le salon du condamné était envahi. Inutile de
vous décrire cette scène, une véritable orgie de joie.
Un messager porta un mot à Chepannik au pavillon. En apprenant cette
nouvelle, il fut saisi d’un effroi rétrospectif. Il se précipita à l’extrémité du
fil pour parler à Clayton, au gouverneur et à tous les autres; la jeune femme,
ne sachant comment lui exprimer sa profonde reconnaissance pour avoir
sauvé son mari, lui envoya une foule de baisers à 12.000 milles de distance.
Tous les télélectrophonoscopes du globe furent mis en communication et
fonctionnèrent, et pendant bon nombre d’heures, les rois et les reines de
plusieurs royaumes (quelques reporters aussi, bien entendu) conversèrent
avec Chepannik et le comblèrent de louanges. Les sociétés scientifiques,
qui jusqu’alors avaient omis de le nommer membre honoraire,
s’empressèrent de lui conférer cette dignité.
Comment expliquer son absence et sa disparition de parmi nous? La
chose est bien simple: pour se soustraire à la renommée universelle du
monde, et à l’engouement général du public qui ne lui laissait ni trêve ni
repos, il avait laissé pousser sa barbe, avait coiffé son nez de verres de
couleur, s’était déguisé, et, sous un faux nom, il parcourait le monde
incognito.
Tel est le récit du drame qui, au printemps de 1898, commença à Vienne
par une querelle de peu d’importance et qui, au printemps de 1904, faillit se
terminer par une tragédie.

CHAPITRE II
Correspondance du Times.—Chicago, 5 avril 1904.
Aujourd’hui, par le chemin de fer électrique, arriva de Vienne une
enveloppe à l’adresse du capitaine Clayton; elle contenait un farthing
anglais. Le réceptionnaire de cette enveloppe parut quelque peu ému. Il
demanda la communication avec Vienne et se trouva face à face avec M.
K.; il lui dit: «Je n’ai pas besoin de vous en raconter bien long. Ma
physionomie vous en dit assez. Ma femme a le farthing; rassurez-vous, elle
ne le jettera pas au vent».

CHAPITRE III
—Correspondance du Times.—Chicago, 23 avril 1904.—Maintenant que
l’affaire Clayton a suivi son cours et qu’elle est terminée, je vais la résumer.
La nouvelle de la délivrance romanesque de Clayton, qui, on peut le dire,
échappa miraculeusement à une mort infâme, plongea toute cette région
dans le ravissement et la joie, pendant les neufs jours proverbiaux; mais,
après cela, le désenchantement inévitable survint et les gens se prirent à
réfléchir, pensant: «Pourtant, un homme a été tué et Clayton l’a tué.»
D’autres répondirent: «C’est vrai, nous avons perdu de vue cet important
détail, l’excitation a fait dévier notre jugement de la ligne droite.»
A l’unanimité on décida qu’il fallait de nouveau traduire Clayton en
justice; des mesures furent prises en conséquence et on renvoya l’affaire
devant la cour de Washington; car, en Amérique, d’après le nouveau
paragraphe additionnel à la constitution de 89, les seconds procès ne sont
plus affaires d’Etat, mais deviennent affaires nationales; ils doivent en
conséquence être jugés devant la cour suprême des Etats-Unis. Les juges
furent donc convoqués à Chicago. La session s’ouvrit avant-hier avec le
decorum habituel, très impressionnant. Les neuf juges, présidés par leur
nouveau président, firent leur apparition en robe noire. En ouvrant la
séance, le président s’exprima ainsi:
«A mon avis ce cas est des plus simples; le prisonnier que vous voyez à
la barre fut accusé du meurtre du nommé Chepannik; jugé avec impartialité,
il a été justement condamné à mort pour le meurtre de ce même Chepannik.
Il devient évident maintenant que le nommé Chepannik n’a nullement été
assassiné. Les décisions des tribunaux français dans l’affaire Dreyfus
établissent sans contredit que les jugements des tribunaux restent
permanents et ne peuvent être révisés. Nous devons respecter et admettre ce
précédent, car tout l’édifice de la jurisprudence repose sur des précédents.
Le prisonnier que vous voyez à la barre a donc été justement et
impartialement condamné à mort pour le meurtre du nommé Chepannik; à
mon avis, il n’y a qu’une façon de résoudre ce cas: il doit être pendu».
Le juge Crawford objecta:
—Mais, Excellence, on lui a pardonné devant le gibet.
—Ce pardon n’est pas valable et ne peut subsister, car il s’appliquait au
fait d’avoir tué Chepannik; or il ne l’a pas tué. Un homme ne peut recevoir
le pardon d’un crime qu’il n’a pas commis; ce serait une absurdité.
—Mais Excellence, il a tué un homme.
—Ce détail reste complètement étranger à la question, et ne doit
nullement nous inquiéter. La cour ne doit pas juger ce crime avant que le
prisonnier n’ait expié l’autre.
Le juge Hallec continua:
—Si nous ordonnons l’exécution du prisonnier, Excellence, nous
provoquerons une erreur judiciaire, car le gouverneur le graciera de
nouveau.
—Il n’en aura pas le pouvoir. Il ne peut gracier un homme pour un crime
qu’il n’a pas commis. Comme je le faisais remarquer tout à l’heure, ce
serait une absurdité.
Après un moment de réflexion, le juge Wadsworth ajouta:
—Plusieurs de mes collègues et moi, arrivons à conclure, Excellence,
que nous commettrions une erreur en faisant pendre le prisonnier pour avoir
tué Chepannik, alors que tout prouve qu’il ne l’a pas tué.
Au contraire, il est prouvé qu’il a tué Chepannik. Le précédent français
établit que nous devons nous en tenir au jugement de la cour.
—Mais Chepannik est encore en vie.
—Dreyfus aussi.
A la fin il fut reconnu impossible d’ignorer le précédent français. Une
seule chose restait à faire: Clayton fut livré au bourreau. Cette décision
provoqua une excitation générale l’Etat tout entier se leva comme un seul

You might also like