The heart'!/mathnnatic.

s has ilS mwms, ofmhich linguistic reason knoir>s FERNANDO ZALAMEA

nothing. Just bke Stephm, i n]qyce's A Portrail of the Artist as a Young Man.
-.oho must confront the 'unl.d heart eflife' and choose wlu:thu to aooi d
it or rathl'T to immnJe him.re!fin i t , the phikJsopher efmathematics
must con.fronJ the 'wild heart' ofmathematics. Synthai Synthetic Philosophy of
Philosorh>
or
Contemporary Mathematics
Contc111pir.UY
Mathemiit

U RB
ANO
\UC

ww-..urbanomic:.com
www.sequcnccprcss.com
 FERNANDO ZALAMEA

Synthetic Philosophy
of Contemporary
Mathematics

Translated by
ZACHARY LUKE FRASER

b
URBANOMIC
 Published in 2012 by

URBANOMIC SEQ.DENCE PRESS

THE OLD LEMONADE FACTORY 36 ORCHARD S TREET
WINDSOR Q.UARRY NEW YORK
FALMOUTH TRll 3EX NY 10002
UNITED KINGDOM UNITED STATES

Originally published in Spanish as

Filosefia Sintetica de las Matematicas Contemportineas
©Editorial Universidad Nacional de Colombia, 2009
This translation © Sequence Press

All rights reserved.

No part of this book may be reproduced or transmitted in any form or by

any means, electronic or mechanical, including photocopying, recording or
any other information storage or retrieval system, without prior permission
in writing from the publisher.

LIBRARY OF CONGRESS CONTROL NUMBER

2012948496

BRITISH LIBRARY CATALOGUING-IN-PUBLICATION DATA

A full record of this book is available

from the British Library

ISBN 978-0-9567750-1-6

Copy editor: Daniel Berchenko

Printed and bound in the UK by
the MPG Books Group, Bodmin and Kings Lynn

www.urbanomic .com
www.sequencepress.com
 CONTENTS

Introduction: Options Traditionally Available to Mathematical

Philosophy, and a Prospectus of the Essay 3

Part 1: The General Environment of Contemporary

Mathematics 19
Chapter 1: The Specificity of Modem and
Contemporary Mathematics 21
Chapter 2: Advanced Mathematics in the Tracts of
Mathematical Philosophy: A Bibliographical Survey 49
Chapter 3: Toward a Synthetic Philosophy of
Contemporary Mathematics 109

Part 2: Case Studies 131

Chapter 4: Grothendieck: Forms of High
Mathematical Creativity 133
Chapter 5: Eidal Mathematics: Serre, Langlands,
Lawvere, Shelah 173
Chapter 6: Q.uiddital Mathematics: Atiyah, Lax,
Connes, Kontsevich 205
Chapter T Archeal Mathematics: Freyd, Simpson,
Zilber, Gromov 239

Part 3: Synthetic Sketches 267

Chapter 8: Fragments of a Transitory Ontology 269
Chapter 9: Comparative Epistemology
and Sheafification 295
Chapter 10: Phenomenology of
Mathematical Creativity 327
Chapter 11: Mathematics and Cultural Circulation 353

Index of Names 377

All that we call invention, discovery in the highest sense ef
the word, is the meaningful application and the putting
into practice efa very originalfeeling eftruth, which, over
a long and secret period ifformation, leads unexpectedly,
with lightning speed, to somefertile intuition. [ .. ] It is a
.

synthesis efworld and spirit that effers the most sublime

certainty efthe eternal harmony efexistence.
GOETHE,
'
WILHELM MEISTER S APPRENTICESHIP (1829)

Poetry points to the enigmas efnature and aims to resolve

them by way efthe imagination; philosophy points to the
enigmas efreason and tries to resolve them by way efwords.
GOETHE, POSTHUMOUS FRAGMENTS

The most important thing, nevertheless, continues to be

the contemporary, because it is what most clearly reflects
itselfin us, and us in it.
GOETHE, NOTEBOOKS ON MORPHOLOGY (1822)

Mathematicians are a bit like Frenchmen: when something

is said to them, they translate it into their own language,
and straight away it becomes something else entirely.
GOETHE, POSTHUMOUS FRAGMENTS
Introduction
O PT I O N S TRAD I T I O NALLY AVAI LABL E TO

MATHEMATICAL P H I LO S O PHY, AND A

PROSPECTUS OF T H E E S SAY

Drawing on the four maxims of Goethe's that we have

placed as epigraphs, we would like to explain, here, the
general focus of this Synthetic Philosophy ef Contemporary
Mathematics. The four terms in the title have, for us,
certain well-defined orientations: 'synthetic' points to
the connective, relational environment of mathematical
creation and to a veiled reality that contrasts with inven­
tion; 'philosophy', to the reflective exercise of reason on
reason itself; 'contemporary', to the space of knowledge
elaborated, broadly speaking, between i950 and today;
'mathematics', to the broad scope of arithmetical, alge­
braic, geometrical and topological constructions - going
beyond merely logical or set-theoretical registers. Now,
in setting the stage in this way, we have immediately
indicated what this essay is not: it is from the outset clear
that it will not be a treatise on the 'analytic philosophy
of the foundations of mathematics in the first half of the
twentieth century'. Since the great majority of works in
mathematical philosophy (chapter 2) fall exclusively
within the subbranch those quotation marks encapsulate,
perhaps we can emphasize the interest that lies in an essay
like this one, whose visible spectrum will turn out to be

3
 SYNTHETIC PHILOSOPHY

virtually orthogonal to the one usually treated in reflections

on mathematical thought.
These pages seek to defend four central theses. The
first postulates that the conjuncture 'contemporary math­
ematics' deserves to be investigated with utmost care,
and that the modes of doing advanced mathematics
cannot be reduced (chapters I, 3) to either those of set
theory and mathematical logic, or those of elementary
mathematics. In this investigation, we hope to introduce
the reader to a broad spectrum efmathematical achievements
in the contemporary context, which might have otherwise
remained inaccessible. The second thesis says that to really
see, even in part, what is happening within contemporary
mathematics (chapters 4-7), we are practically forced to
expand the scope of our vision and discover the new prob­
lematics at stake, undetected by 'normal' or 'traditional'
currents in the philosophy of mathematics (chapters 2-3).
The third thesis proposes that a turn toward a synthetic
understanding of mathematics (chapters 3, 8-n) - one
that is largely reinforced in the mathematical theory of
categories (chapters 3-7) - allows us to observe important
dialectical tensions in mathematical activity, which tend
to be obscured, and sometimes altogether erased, by the
usual analytic understanding. The fourth thesis asserts
that we must reestablish a vital pendular weaving between
mathematical creativity and critical reflection - something
that was indispensable for Plato, Leibniz, Pascal and

4
 I NTRODUCTION

Peirce - and that, on the one hand, many present-day

mathematical constructions afford useful and original
perspectives on certain philosophical problematics of the
past (chapters 8-n), while, on the other hand, certain
fundamental philosophical insolubilia fuel great creative
forces in mathematics (chapters 3-7, 10) .
The methods utilized in this work include describing a
particular state of affairs ('contemporary mathematics':
chapters 4-7) , reflecting on this description ('synthetic
philosophy': chapters 1, 3, 8-n), and contrasting this two­
fold description and reflection with other related aspects
('mathematical philosophy', theory of culture, creativity:
chapters 2, 8-n) . We hope to make the hypotheses under­
lying those descriptions, reflections and comparisons
explicit throughout. Note that our survey's main endeavor
has been to try to observe mathematical movements on
their own terms, and that its filter of cultural reorganiza­
tion has been articulated only a posteriori, so as to try to
reflect as faithfully as possible those complex, and often
elusive, movements of mathematics.
The problems that mathematics have posed for philo­
sophical reflection have always been varied and complex.
Since the beginnings of both disciplines in the Greek
world, the advances of mathematical technique have
perpetually provoked philosophical reflections of a fun­
damental nature. The privileged.frontier of mathematics
- the fluctuating, intermediary warp between the possible

5
 O F CONTEMPORARY MATHE MATICS

(hypothesis), the actual (comparisons) and the necessary

(demonstrations), the bridge between human inventive­
ness and a real, independent world - has spawned all
kinds of alternative positions regarding what mathematics
'is', what its objects are, and how it knows what it knows.
The ontologi,cal 'what', the pursuit of the objects studied by
mathematics, and the epistemologi,cal 'how', which concerns
the way in which those objects should be studied, currently
dominate the landscape of the philosophy of mathematics
(Shapiro's square, figure I, p.10) . But curiously, the
'when' and the 'why', which could allow mathematical
philosophy to form stronger alliances with historical
and phenomenological perspectives (chapters 10, 11),
have mostly vanished into the horizon, at least within
the Anglo-Saxon spectrum. This situation, however, must
of necessity be a passing one, since there do not seem to
be any intrinsic reasons for reducing the philosophy of
mathematics to the philosophy of mathematical language.
Everything rather points toward a far broader spectrum
of pendular practices, irreducible to imagi,nation, reason or
experience, through which the conceptual evolution of
the discipline will outstrip the sophisticated grammatical
discussions that have been promoted by the analysis of
language.
The traditional problems of the philosophy of math­
ematics have parceled themselves out around certain
great dualities that have incessantly left their mark upon

6
 I NTRODUCTION

the development of philosophical reflection. Perhaps

the unavoidable crux of the entire problematic lies in
the deep feeling of wonder and astonishment that has
always been a product of the 'unreasonable applicability'
of mathematics to the real world. How can mathematics,
this extraordinary human invention, grant us so precise
a knowledge of the external world? Tue responses given
have been numerous, carefully argued and, frequently,
convincing. On the one hand, ontologi,cal realism has pos­
tulated that the objects studied by mathematics (whatever
they are: ideas, forms, spaces, structures, etc.) lie buried
in the real world, independently of our perception, while
ontowgi,cal idealism has suggested that mathematical objects
are mere mental constructions. A realist stance thereby
simplifies our supposed access to the real, while imposing
strong constraints on the world (existential, formal, and
structural constraints, etc.); in contrast, an idealist stance
dismisses the world, sparing it from reliance on dubious
organizational scaffoldings, but it faces the problem of
mathematics' applicability head-on. Epistemologi,cal realism,
on the other hand, has postulated (independently of any
ontological position) that mathematical knowledge is not
arbitrary and that its truth values are indices of a certain
real stability, while epistemologi,cal idealism has regarded
truth values as mere man-made mediations, which do
not need to be propped up by any real correlate. An
idealist stance again secures for itself a greater plasticity,

7
 SYNTHETIC P H ILOSOPHY

with greater possibilities of access to the mathematical

imagination, but it encounters serious difficulties at the
junction of the imaginary and the real; a realist stance
helps to understand mathematical thought's material suc­
cess, but it places rigid restrictions on its creative liberty.
Interlaced with these basic, primary polarities, several
other important and traditional dualities have found
themselves in mathematical philosophy's spotlight. The
necessity or contingency of mathematics, the universality
or particularity of its objects and methods, the unity or
multiplicity of mathematical thought, the interiority or
exteriority of the discipline, the naturalness or artificiality of
its constructions - each could count on having defenders
and detractors of every sort. The status accorded to the
correlations between physics and mathematics has always
depended on the position one takes (whether consciously
or unconsciously) with respect to the preceding alterna­
tives. At the opposite extremes of the pendulum, we may
situate, for example, a necessary, universal, unique and
natural mathematics, very close to strongly realist posi­
tions, and a contingent, particular, multiple and artificial
mathematics, coming very close, here, to the idealist
extreme. But the vast intermediate range between these
oscillations of the pendulum is ultimately what merits the
most careful observation.' One of the principal objectives
i An excellent overview of the entire pluralistic explosion ofphilosophies of mathematics
can be found in G. Lolli, Filosefia delta matematica: L'ereditti de! novecento (Bologna: ii

8
 I NTRODUCTION

of the present work is to demonstrate that - beyond a

binary yes/no alternation - certain mixtures are vital for
obtaining a thorough and accurate understanding of what
it is to do mathematics, with respect to both its general
and global structuration and many of its highly detailed,
particular and local constructions.
In his excellent monograph, lhinking about Math­
ematics, Shapiro has made good use of a few of the
aforementioned dualities in order to trace out a brilliant
landscape of current philosophy of mathematics. 2 Restrict­
ing himself to the Anglo-Saxon world,3 Shapiro goes on
to classify several prominent bodies of work in virtue of
their various realist or idealist stances (figure 1, based on
Shapiro's text) .4

Mulino, 2002). Lolli detects at least fourteen distinct currents (nominalism, realism,
Platonism, the phenomenological tradition, naturalism, logicism, formalism, the
semiotic tradition, constructivism, structuralism, deductivism, fallibilism, empiricism,
schematism), in addition to a 'spontaneous philosophy' of mathematicians.

S. Shapiro, 'Ihinking about Mathematics: '!he Philosophy ofMathematics (Oxford: Oxford

University Press, 2000).

3 The restriction is not, however, explicit, and Shapiro commits the common Anglo­
Saxon sin of believing that anything that has not been published in English does
not form part of the landscape of knowledge. The identification of 'knowledge' with
'publication in English' has left outside of the philosophy of mathematics one who,
to our understanding, is perhaps the greatest philosopher of 'real mathematics' in the
twentieth century: Albert Lautman. For a discussion of 'real mathematics' (Hardy,
Corfield) and the work of Lautman, see chapters 1-3.

4 Shapiro, 'Ihinkingahout Mathematics, 32-3. Shapiro calls realism in truth-value the vision
according to which 'mathematical statements have objective truth values, independent
of the minds, languages, conventions, and so on of mathematicians' (ibid., 29). To
simplify, we will here give the name 'epistemological realism' to this realism in truth value.

9
 OF CONTEMPORARY MATHEMATICS

EPISTEMOLOGY
.... ... . . ..... .. ..... ........ .. ..... .. ..... ..
.
. .
; Realism ; Idealism
. .
· · · · · · · · · · · · · · · · · · · · · · �-----�-----1
Maddy
0
N Realism Resnick Tennant
T
0 Shapiro
...................... i----+--1
L
0
G Chihara Dummet
Y Idealism
Hellman Field

Figure 1. Contemp(}1'(lry tendencies inphilosophy efmathematics, according to Shapiro.

Aside from the differential details of the works included

in the square above - a few of which we will compare
with the results of our own investigations in part 3 of
this essay - what we are interested in here is the biparti­
tion that Shapiro elaborates. There is no place in this
diagram for an ontological position between realism and
idealism, nor for an epistemological mixture of the two
polarities. Is this because such mediations are philosophi­
cally inconsequential or inconsistent, or simply because
they have been eliminated in order to 'better' map the
landscape? One of our intentions in this essay will be to
show that those mediations are not only consistent from
a philosophical point of view (following Plato, Peirce and
Lautman: chapter 3) , but also indispensable from the point
of view of contemporary mathematics. Shapiro's square

10
 I NTRODUCTION

therefore turns out to be only an ideal binary limit of a

far more complicated, real state of affairs;s in an extended
square, several new cells would appear, opening it onto
tertiary frontiers.
Benacerraf's famous dilemma likewise presents itself
by way of a dual alternative: either we coherently adopt a
realism at once ontological and epistemological, and then
find ourselves faced with difficult problems about how
we might have knowledge of mathematical objects that
do not originate in our own acts of invention and cannot
be experientially perceived in nature; or we adopt a more
flexible idealist epistemology and then find ourselves faced
with other, equally difficult problems, as we inquire into
the profound harmony between mathematics and the
external world. But this either-or dilemma would not have
to be considered as such if we could take stock of other
intermediate positions between realism and idealism. We
believe, in fact, that mathematics in its entirety produces
illuminating examples of mediations between real and
ideal configurations, and it does so from the most varied
and complementary points of view (chapters I, 4-7) .
Considered, as it usually is, from classical and dualistic
perspectives, Benacerraf's Dilemma should be viewed with
caution; however, considered within a broader metalogic,
attentive to the dynamic evolution of mathematics - with
5 Just as, similarly, classical logic should in reality be understood as an ideal limit of
intuitionistic logic. Cf. Caicedo's results, discussed in chapter 8.

11
 SYNTHETIC PHI LOSOPHY

its progressive osmoses and transferences between the

real and the ideal·- the dilemma falls apart, since there
is no longer any reason to adopt dual exclusions of the
either-or variety (chapters 8, 9) .
It is important, here, to point out the dubious worth
of fixing one's ontology and epistemology in advance,
adopting them a priori before any observation of the
mathematical universe, and presuming to impose certain
rigid partitions upon the latter. At the very least, such
an adoption of philosophical presuppositions prior to
even setting eyes on the mathematical world has limited
our perspective and has led to the perception of a rigid,
static and eternal mathematics, a perception that has little
or nothing to do with the real mathematics that is being
done every day. Instead, a living mathematics, in incessant
evolution, should be considered the basic presupposition
of any subsequent philosophical consideration whatsoever.
The study of the continuities, obstructions, transfers and
invariants involved in doing mathematics should then - and
only then - become an object of philosophical reflection.
The elaboration of a transitory ontology and epistemology,
better matched to the incessant transit of mathematics, is
the order of the day. 6 The peerless strength of mathemat­
ics lies precisely in its exceptional protean capacity, a
6 Alain Badiou explores this idea in his Court trait/ d'ontologie transitoire (Paris: Seuil,
1998). [Tr. N. Madarasz as Briefings on Existence: A Short Treatise on TransitlYrJ Ontology
(Albany, NY: SUNY Press, 2006). J Part 3 of the present study makes further inroads
into that 'transitory philosophy' that, we believe, mathematics demands.

12
 I NTRODUCTION

remarkable transformative richness that has rarely been

philosophically assimilated.
One of the traditional problems that mathematical
philosophy has had to confront, in this respect, has to do
with the general place of mathematics within culture as a
whole. Here, a dualistic reading is again burdened with
immediate problems: If mathematics is understood as an
evolutive forge, internal to contingent human creativity,?
the problem arises as to how we can explain its appar­
ently necessary character and its cumulative stability; if,
instead, mathematics is understood as the study of certain
forms and schemas that are independent of its cultural
environment,8 there arises the problem as to how we
can explain the markedly historical character of math­
ematical 'discoveries'. In practice, a middle way between
both options seems far better adjusted to the reality of
doing mathematics (see, in particular, the meditations
on Grothendieck in chapter 4): a fluctuating, evolving
activity, full of new possibilities, springing from disparate
cultural realms, but always managing to construct precise
invariants for reason behind the many relative obstructions
that the mathematical imagination is always encounter­
ing. What drives both mathematical creativity and its
subsequent normalization is a to and fro that tightly

7 This is the case, for example, in R. L. Wilder, Mathematics as a Cultural System ( Oxford:
Pergamon Press, 1981).

8 M. Resnik, Mathematics as a Science ifPatterns (Oxford: Oxford University Press, 199 7).

13
 OF CONTEMPORARY MATHE MATICS

weaves together certain sites ofpure possibility and certain

necessary invariants, within well-defined contexts.
Without this back-andforth between obstructions and
invariants, mathematics cannot be understood. The
wish to reduce, a priori, the doing of mathematics to one
side of the balance or the other is, perhaps, one of the
major, basic errors committed by certain philosophers
of mathematics. The transit between the possible, the
actual and the necessary is a strength specific to mathemat­
ics, and one that cannot be neglected. To consider that
transit as a weakness, and to therefore try to eliminate
it, by reducing it either to contingent or to necessary
circumstances (another version of an either-or exclusion),
is an unfortunate consequence of having taken sides in
advance, before observing the complex modal universe of
mathematics. In fact, as we shall demonstrate in part 2
of this book, on the basis of the case studies of part 2, in
mathematics, discovery (of necessary structural schemas)
is just as indispensable as invention (of languages and
possible models) . The tight mathematical weave between
the real and the ideal cannot be reduced to just one of its
polarities, and it therefore deserves to be observed through
a conjunction of complementary philosophical points of
view. We believe that any reduction at all, or any preemp­
tive taking of sides, simply impedes the contemplation
of the specificities of mathematical transit.

14
 I NTRODUCTION

Both Wilder and Resnik, to point tojust one complementary

polarity, have much to offer us. A hypothesis (chapter i),
a program (chapter 3) and a few detailed case studies
(chapters 4-7) will prepare us with some outlines for a
synthesis (chapters 8-11) by which various central aspects
of complementary perspectives, like those of Wilder
and Resnik, can come to be 'glued' together in a unitary
whole. We may point out that one of the essential and
basic motivations of this work is the desire to elaborate, in
order to reflect on mathematics, a sort of sheafthat would
allow us to reintegrate and 'glue together' certain comple­
mentary philosophical viewpoints. As will become clear
in part 2, the notion of a mathematical sheaf is probably
the fundamental distinguishing concept around which
the elaboration of contemporary mathematics, with new
impetus, begins, with all of its extraordinary instruments
of structuration, geometrization, gluing, transfer and uni­
versalization - and so the attempt to look at mathematics
from a sheafefequally complex perspectives turns out to be a
rather natural one. To achieve this, we will have to delimit
certain 'coherence conditions' between complementary
philosophical perspectives (chapters i, 3) in order to then
proceed with a few sketches of 'sheaving' or of 'structural
synthesis' (chapters 8-n).

15
 SYNTHETIC P H I LOSOPHY

ACKNOWLE D G E M E NTS

To Pierre Cassou-Nogues, Marco Panza and Jose Fer­

reir6s, who, with their invitation to Lille in 2005, helped
me to confirm continental philosophy's crucial role in
achieving a greater comprehension of advanced math­
ematics, and also to define the singular interest that
this monograph might gather in a field rather flattened
by analytic philosophy. To Carlos Cardona, who, with
his brilliant doctoral thesis on Wittgenstein and Godel,
provoked me to contradict him, thereby giving rise to the
solid kernels of this work. To my students in L6gica IV
( 2006), who, during this text's gestation, stoically bore
the brunt of some of its most abstract assaults. To my
colleges and participants in my Seminario de Filosofia
Matematica, whose constructive criticisms gave rise to
various lines of thought inscribed in this text. To Juan Jose
Botero, who, with his invitation to present as a plenary
speaker in El Primer Congreso Colombiano de Filosofia
(2006), opened an uncommon space to me, one that
other colleagues (mathematicians, philosophers, cultural
scholars) carefully ignored. To Andres Villaveces, who
proposed some remarkable clarifications in the course
of writing this text, and who unconditionally supported
me with his always-magnificent enthusiasm. To Xavier
Caicedo, who produced a profound and encouraging
review of the monograph, supporting my (unsuccessful)

16
 INTRODUCTION

candidacy for El Premio Ensayo Cientlfico Esteban de

Terreros ( 2008). To Javier de Lorenzo, whose friendship
and generosity came very close to placing this text with an
improbable commercial publishing house in his country.
To Alexander Cruz, Magda Gonzalez, Epifania Lozano,
Alejandro Mardn and Arnold Oostra, who corrected
many of the work's glitches and improved several of its
paragraphs. To La Editorial Universidad Nacional, who
granted me the opportunity of entering, by contest, its
fine collection, Obra Selecta.
My greatest thanks go to Zachary Luke Fraser, who
not only worked carefully on a daunting translation, but
in many paragraphs improved on the original. Finally,
heartfelt thanks go to my editors at Urbanomic, Robin
Mackay and Reza Negarestani, whose great vision is
opening a wide range of much-needed doors for new
forms of contemporary thought.

17
 PART ONE

The General Environment of

Contemporary Mathematics
C HAPT E R 1

T H E S P E C I F I C ITY OF M O D E R N AND

CO NTE M PO RARY MAT H E MATICS

It is well known that mathematics is presently enjoying

something of a boom. Even conservative estimates suggest
that the discipline has produced many more theorems in
the last three decades than it has in its entire preceding
history, a history stretching back more than two thousand
years (including the very fruitful nineteenth and twentieth
centuries, and all the way up to the i97os). The great
innovative concepts of modern mathematics - which
we owe to Galois, Riemann and Hilbert, to cite only the
three major foundational figures - have been multiplied
and enriched thanks to the contributions of a veritable
pleiad of exceptional mathematicians over the last fifty
years. Proofs of apparently unattainable theorems - like
Fermat's Last Theorem, or the Poincare conjecture -
have been obtained, to the surprise of the mathematical
community itself, thanks to the unrelenting struggle
of mathematicians who knew how to carefully harness
the profound explorations already undertaken by their
colleagues. The boom in mathematical publications and
reviews appears unstoppable, with an entire, flourish­
ing academic 'industry' behind it; though the excessive

21
 SYNTHETIC PHILOSOPHY

number of publications could be taken to discredit their

quality (it is tempting to say that publication should be
cause for penalty rather than promotion), the immense
liveliness of mathematics is made manifest in the frenetic
activity of publishing houses. Meanwhile, mathemat­
ics' relations with physics again seem to have found a
moment of grace, as the two are profoundly interlaced
in the study of superstrings, quantizations, and complex
cosmological models.
Strangely, however, the philosophy efmathematics has
rarely taken stock of the genuine explosion that mathemat­
ics has witnessed in the last fifty years (chapter 2). Two
reasons may be given for this: firstly, the view that, in
spite of the advance and evolution of mathematics itself,
mathematics' methods and the types of objects it studies
remain invariable; secondly, the simple myopia before
the new techniques and results, because of a certain
professional incapacity to observe the new thematics
at stake. In practice, in fact, there seems to be a mutual
feedback between these two tendencies; on the one hand,
the conviction that set theory (and with variants of first­
order logic) already provide sufficient material for the
philosophy of mathematics to shore up an unwillingness to
explore other environments of mathematical knowledge;
on the other hand, the inherent difficulty involved in the
advances of modern mathematics (from the second half
of the nineteenth century through the first half of the

22
 MODE RN AND CONTE M PO RARY MATHE MATICS

twentieth) , and, a fortiori, those ofcontemporary mathemat­

ics (from the middle of the twentieth century onwards),
can be avoided by hiding behind the supposed ontological
and epistemological invariability of the discipline. This
deliberate neglect of the discipline's (technical, thematic,
creative) landscape is a situation that would be seen as
scandalous in the philosophy of other scientific disciplines,9
but to which the philosophy of mathematics seems able
to resign itself, with a restrictive security and without the
least bit of modesty.
Two great extrapolations - equivocal by our lights,
as we shall attempt to demonstrate throughout this essay
- support the idea, ubiquitous in the philosophy of math­
ematics, according to which it is unnecessary to observe
the current advances of the discipline. On the one hand,
the objects and methods of elementary mathematics and
advanced mathematics are considered not to essentially
differ from one another; on the other hand, the develop­
ment of mathematics is presupposed to have a markedly
necessary character and an absolute background. If, from
an epistemological and ontological point of view, the
exploration of the Pythagorean Theorem offers nothing

9 A philosophy of physics that does not take stock of the technical advances in physics,
for example, would be unthinkable. B. d'Espagnat, Le reel voili. Analyse des cancepts
quantiques (Paris: Fayard, 1994), for instance, performs an admirable philosophical
study of quantum physics, in which the notable technical advances of the discipline
are carefully observed, and in which it is demonstrated that, in order to understand
quantum physics, new ontological and epistemological approaches, adapted to the new
methods and objects of knowledge, are required.

23
 OF CONTEMPO RARY MATHEMATICS

different than the exploration of Fermat's, then making the

effort to (philosophically) understand all the instruments
of algebraic and complex variable geometry that opened
a way to proving Fermat's Theorem would, of course, be
pointless. If, from a historical and metaphysical point of
view, the evolution of mathematics is considered not to
give rise to new types of 'entities', then it would be equally
absurd to try to entangle oneself in the complexities of
contemporary mathematical creativity. We nevertheless
believe that these two ubiquitous suppositions - that there
is no distinction between elementary and advanced math­
ematics; that there is no duality of transits and invariants
in mathematics - are only valid, in part, in determinate,
restrictive contexts, and we consider the extrapolations
of these suppositions into the 'real' totality of mathemat­
ics (and contemporary mathematics, in particular), to
constitute a profound methodological error.
Following David Corfield, we will call 'real mathemat­
ics' the warp of advanced mathematical knowledge that
mathematicians encounter daily in their work,10 a warp that
can be seen as perfectly real from several points of view:
as a stable object of investigation for a broad community,
as an assemblage of knowledge with a visible influence

10 D. Corfield, TowardsaPhilasophyefRea/Mathematics(Cambridge: Cambridge University

Press, 2003). As Corfield has pointed out, Hardy, in his polemical A Mathematician's
Apology, called 'real mathematics' the mathematics constructed by figures like Fermat,
Euler, Gauss, Abel and Riemann (G.H. Hardy, A Mathematician 'sApolagy [Cambridge:
Cambridge University Press, 1940], 59-60; cited in Corfield, 2).

24
 MO DERN AND CONTEMPORARY MATHEMATICS

on the practice of the discipline, and as a framework that

can be effectively contrasted with the physical world.
Both elementary mathematics and set theory, the object
of extensive considerations in analytic philosophy, are
thus but slight fragments of 'real' mathematics. They
considerably extend the scope of classical mathematics
(midseventeenth to midnineteenth century), but in view
of the totality (figure 2, overleaf) , it must be observed
that, nowadays, modern and contemporary mathematics
largely make up the core of the discipline. Let us take as
a basic supposition the insufficiently appreciated fact that
apprehending the totality of mathematical production
with greater fidelity and technical precision may turn out
to be of great relevance for philosophy.

25
 SYNTHETIC P H I LOSOPHY

mathematics

advanced (='real')

contemporary

modern

elementary

classical

Figure !!. Correlations between the areas of mathematics: elementary, advanced,

classical, modern, contemporary.

The boundaries that allow us to distinguish the afore­

mentioned areas are clearly historical, since leading
mathematical research becomes progressively more complex
throughout its evolution. Nevertheless, the boundaries
can also be associated with certain types of mathematical
instruments, introduced by great mathematicians, whose
names still serve to characterize each epoch:

26
 M OD ERN AND CONTEMPORARY MATHE MATIC S

Classical mathematics (midseventeenth to midnine­

teenth centuries): sophisticated use of the infinite
(Pascal, Leibniz, Euler, Gauss);

Modern mathematics (midnineteenth to midtwen­

tieth centuries): sophisticated use of structural and
qualitative properties (Galois, Riemann, Hilbert);

Contemporary mathematics (midtwentieth century

to present): sophisticated use of the properties of
transference, reflection and gluing (Grothendieck,
Serre, Shelah).

In particular, there accumulated in modern mathematics

an enormous quantity of knowledge, which evolved and
went on to make up to the current body of mathematics:
set theory and mathematical logic, analytic and algebraic
number theory, abstract algebras, algebraic geometry,
functions of complex variables, measure and integration,
general and algebraic topology, functional analysis, differ­
ential varieties, qualitative theory of differential equations,
etc.11 Even if a series of important mathematical theorems
have succeeded in proving that any mathematical con­
struction can be represented inside a suitable set theory

n The Mathematical Subject Classification 2000 (MSC 2000) includes some sixty principal
entries in a tree that goes on to rapidly branch out. Above, we indicate only a few of
the initial indispensable entries in the tree.

27
 OF CONTEMPORARY MATHEMATI C S

(and that the enormous majority of mathematics can be

represented inside the Zermelo-Fraenkel theory, with its
underlying first-order classical logic), it is nevertheless
clear, within mathematical practice, that what value these
'facsimiles' have is merely logical, andfar removed from
their genuine mathematical value. We believe the fact that
mathematical constructions can be reduced, theoretically,
to set-theoretical constructions has been enlisted as yet
another prop by means of which, in the philosophy
of mathematics, one could, for so long, avoid a more
engaged inspection of 'real mathematics'. Nevertheless,
as we shall soon see, the structures at stake and the ways of
doing things differ dramatically between set theory and
other mathematical environments, and the ontology and
epistemology we propose should, consequently, differ as
well (to say nothing of history or 'metaphysics' - chapters
io-11). The possibility of reducing the demonstration of a
complex mathematical theorem to a series of purely set­
theoretical statements (a possibility existing only in theory
and never executed in practice, as soon as certain rather
basic thresholds are crossed) has, in the philosophy of
mathematics, been elevated to a fallacious extrapolation,
an extrapolation that has allowed certain philosophical
perspectives to shirk any investigation into the present
of 'real mathematics', beyond mathematical logic or
set theory.

28
 MOD ERN AND CONTEM PORARY MATHEMATIC S

The environment of advanced mathematics, already clearly

delimited by the middle of the twentieth century, found
an exceptional philosopher in Albert Lautman.12 For
Lautman, mathematics - beyond its ideal set-theoretical
reconstruction - hierarchizes itself into real environments
of dramatically varying complexity, where concepts and
examples are interlaced through processes that bring the
free and the saturated into structural counterpoint with
one another, and where many of the greatest mathemati­
cal creations emerge through the mediation of mixtures.
Entering into the vast conglomerate of the mathematics
of his time, Lautman was able to detect certain features
specific to advanced mathematics,13 features that do not
appear in elementary mathematics:

12 The work of Albert Lautman (1908-1944) deserves to be understood as the most incisive
philosophical work of the twentieth century that both situated itself within modern
mathematics and sought to.outline the hidden mechanisms of advanced mathematical
creativity, while synthesizing the structural and unitary interlacings of mathematical
knowledge. Lautman's writings, forgotten and little understood at present, have
resurfaced in a new French edition (A. Lautman, Les mathimatiques, !es !dies et le Riel
physique, [Paris: Vrin, 2006]), a recent English translation (tr. S. Duffy as Mathematics,
Ideas and the Physical Real [London: Continuum, 2ouJ) and in the first complete
translation of his works into another language (tr. F. Zalamea as Ensayos sobre la
diatectica, estructura y unidad de las matematicas modem.as [Bogota: Universidad Nacional
de Colombia, 2011]). For a critical presentation of Lautman's work, see my extensive
scholarly introduction to the Spanish edition. In the present essay, I aim to develop
Lautman's work somewhat, and extend its scope from mod.em mathematics (as known
to Lautman) to contemporary mathematics (which now lie before us).
13 The critical works attending to the multiplicity of advanced mathematical creations are
few in number, and so it's worth calling attention to a work so kindred to Lautman's as
that of Javier de Lorenzo, who has always been attentive to the deep strata and diverse
ramifications of modern mathematical invention. Among his works, see, in particular,
Introducci6n al estillo mathematico (Madrid: Tecnos, 1971); La matematica y el problema de
su historia (Madrid: Tecnos, 1977); El mtftodo axiamatico y sus creencias (Madrid: Tecnos,
1980) ; Filosefias de la matematicafin de .siglo XX (Valladolid: Universidad de Valladolid,
2:000 ). Lorenzo does not seem to be familiar with Lautman, nor does he mention him
in his writings.

29
 SYNTHETIC P H I LOSOPHY

1. a complex hierarchization of diverse mathemati­

cal theories, irreducible to one another, relative to
intermediary systems of deduction;

2. a richness of models, irreducible to merely linguistic

manipulations;

3. a unity of structural methods and conceptual polari­

ties, behind their effective multiplicity;

4. a dynamics of mathematical activity, contrasted

between the free and the saturated, attentive to divi­
sion and dialectics;

5. a theorematic interlacing of what is multiple on

one level with what is one on another, by means of
mixtures, ascents and descents.

We should contrast elementary mathematics - the privi­

leged focus of analytic philosophy - with the advanced
mathematical theories that make up the sweeping spec­
trum of modern mathematics. An often-repeated argu­
ment for the possibility of reducing the scope of inquiry
to that of elementary mathematics comes down to assur­
ing us that every mathematical proposition, since it is
a tautology, is equivalent to every other, and so, from
a philosophical perspective, it is enough to study the

30
 M OD ERN AND CONTEM PORARY MATHE MATI C S

spectrum of elementary propositions. For example, the

extremely anodyne '2+2=4' would, from a logical point
of view, be equivalent to the significant and revealing
Hahn-Banach Theorem (HB), since both propositions are
deducible from the Zermelo-Fraenkel system of axioms
(ZF). Nevertheless, the 'trivial' tautological equivalence
ZF I- HB <--> 2+2=4 is as far from exhausting the mathemati­
cal content of the theorems as it is from exhausting their
logical status. The equivalence effectively collapses as
soon as, instead of starting with ZF, we opt for intermedi­
ate axiomatic systems. In fact, Friedman and Simpson's
reverse mathematics'4 show that the basic propositions of
arithmetic (which, for example, Wittgenstein repeatedly
studies in his Lectures on the Foundations efMathematics)'5
show up on the lowest levels of mathematical develop­
ment (in the system RCA0, reduced to demonstrating
the existence of recursive sets), while the HB not only
requires more advanced instruments (a WKLo system
with weak forms of Konig's lemma), but is fully equiva­
lent to those instruments. To be precise, it turns out that
RCAoY::HB<->2+ 2=4, since we have RCA0f-HB�WKL0,
RCA01-2+2=4 and RCA0 Y:: WKL0•
The consequences of this state of affairs are obvious,
but they have not been sufficiently considered in the phi­
losophy of mathematics. First of all, it seems absurd to
14 S. G. Simpson, Suhsystems efSecund Order ArithmetU: (New York: Springer, 1999).
15 L. Wittgenstein, Lectures on the Foundations efMathematics, Cambridge 1939 (Chicago:
University of Chicago Press, 1989).

31
 OF CONTEMPORARY MATHEMATICS

compare pairs o f mathematical propositions with respect

to excessively powerful base systems. In the eyes efZF, all
demonstrable propositions are logically trivialized (as pairs
of equivalent tautologies) - not because the propositions
contain an identical l ogical (or mathematical) value in
themselves, but because the differences are not appreciated
by ZF. ZF is a sort of deductive absolute, in which both those
who study the set-theoretical universe and those who wish
to restrict themselves to elementary mathematics alone
can be quite comfortable; nevertheless, the intermediate
thresholds of deductive power in ZF (like the systems stud­
ied in reverse mathematics) constitute the genuinely relevant
environments from the point of view of 'real ' mathemat­
ics, with the multiple hierarchies and differences in which
one may carry out a mathematically productive study of
logical obstructions and transfers. Secondly, the idea of a
tautological mathematics, fully expressible within the nar­
row scope of elementary mathematics, seems untenable.
The moment we cross the complexity thresholds of system
RCA0 (and pass over into system ACA0, in which we can
prove the first important results of abstract algebra, like
the existence of maximal ideals in commutative rings), we
enter into a relative web of partial equiconsistencies where
the (supposedly stable and absolute) notion of tautology
is deprived of any real mathematical sense. Mathematics
goes on producing necessary theorems, but within vari­
able deductive contexts, whose oscillations and changes

32
 MODERN AND CONTEMPORARY MATHEMATICS

are fundamental to the expression of the theorems' true

mat hematical value. Thirdly, the vital presence, in the dis­
cipline, of certain logical and mathematical irreducibilities
becomes palpable. Mathematics' richness takes root in
its weave of demonstrations (the impossibility of evading
certain obstructions and the possibility of effecting certain
transfers), something that unfortunately disappears in the
light of extreme perspectives - whether from an absolutely
tautological perspective (ZF, where all is transferable) or
from elementary perspectives (subsystems of RCA0, where
all is obstruction).
The complex hierarchization of advanced mathematics
(point I, noted above) gives rise to a panoply of con­
structive scales, inverse correspondences and gradations
of every kind (particularly visible in Galois theory and
in the generalized theories of duality), which allow the
emergence of mathematical creativity to be studied with
greater.fidelity. The blooming and genesis of mathemati­
cal structures, hidden from a static, analytic approach,
are better seen from a dynamic perspective, in light of
which a problem, concept or construction is transformed
by the problem's partial solutions, the concept's refined
definitions, or the construction's sheaf of saturations and
decantations.16 In that eminently living and incessantly

16 The images of decantation, transfusion and distillation that recur throughout this work
indicate those creative gestures by way of which, as we shall see, mathematical ideas
or structures are 'poured' - sometimes with the help of others, as 'filters' - from one
register to another, often leaving behind, as a kind of 'sediment', features previously
thought to be integral to them.

33
 SYNTH ETIC PHI LOSOPHY

evolving field of thought that is mathematics, a profound

hierarchization is not only indispensable to, but is the very
engi,ne of creation. Any philosophy of mathematics that
fails to take stock of the complex, hierarchical richness
of advanced mathematics will be led to neglect not only
such delicate 'intermittencies of reason' (differences in
logical contrast) , but the even subtler 'intermittencies of
the heart' (differences in mathematical creativity) .
Among the features that distinguish modern from
elementary mathematics, Point Q likewise harbors robust
philosophical potential. Modern mathematics has pro­
duced, in all of its fields of action, really remarkable
conglomerates of models, extremely diverse and original,
with significant structural distinctions. It deals, in fact,
with semantic collections that greatly surpass the more
restrained syntactic theories that these collections help
shape, as can be seen, for example, when we compare the
explosive and uneven universe ofsimple finite groups with
the elementary axiomatization that underlies the theory.
Advanced mathematics contains a great semantic richness,
irreducible to merely grammatical considerations, though
afallacious extrapolation has presumed to identify the mak­
ing of mathematics with the making of certain grammatical
rules. Indeed, the supposed reduction of mathematical
thought to a deductive grammar is understandable from
the point of view of elementary mathematics, where
the models tend to be few and controlled, but it is a

34
 MOD ERN AND CONTE M PORARY MATHE MATICS

monumental trivialization to extrapolate that situation

to 'real ' mathematics, where classes bf models start to
behave in an altogether erratic fashion (see chapter 5,
on Shelah's works ) . The attempt to reduce mathematics
to grammar, in short, assumes a (fallacious ) reduction of
mathematics to elementary mathematics and then applies
the (plausible) identification of elementary mathematics
with finitary grammatical rules.
Tue richness of modern mathematics is, to a large
extent, rooted in the enormous diversity of structures and
models that have been constructed (or discovered - we
won't get into the question for now, though we believe
that both construction and discovery are indispensable; see
c,hapters 8, 9) . Structures of every sort have indelibly fur­
rowed the current landscape of mathematics, and a clear
and distinctive mark of advanced mathematics consists
in having to simultaneously consider multiple structures
in any comprehensive inspection of a mathematical phe­
nomenon. The phenomenon frequently demands to be
considered under complementary points of view, whereby
quite diverse arithmetical, algebraic, topological and
geometrical instruments crisscross each other. A funda­
mental characteristic of modern mathematics is its capacity
to operate transfusions between a multiplicity of appar­
ently discordant structures, exploiting remarkable sets of

35
 OF CONTEMPORARY MATH EMATICS

instruments that succeed in harmonizing the diversity. '7

Without variety, multiplicity and complexity, modern
mathematics would not have even been able to emerge;
and, as we shall see, without interlacing and unity, it would
not have been able to consolidate itself. The situation
is very different than that of elementary mathematics,
where structures are strictly determined - the integers,
the real plane, and little else - and, for that reason, are
unable to give rise to either a variability of models or a
fluxion between mathematical subdisciplines. This, again,
is a key observation: Though a restriction to elementary
mathematics might allow for a conflation of models and
language, and the elimination of semantics' variability
and fluxion, this sort of thing must be abandoned as
soon as we enter into advanced mathematics, where the
landscape is firmly governed by collections of (mathemati­
cal) structures and (physical) facts, often independently
of any syntactic or linguistic considerations.
The multiplicative and differential richness of modern
mathematics is accompanied by a complementary, pendu­
lar tendency toward the unitary and the integral (point 3,
indicated above). The dialectical tensions between the One
and the Multiple have found, in modern mathematics, a
fertile field of experimentation. The unity of mathematics
expresses itself, not only in virtue of a common base upon

17 Thus responding to the first epigraph by Goethe that appears at the beginning of
this study.

36
 MO DERN AND CONTE MPORARY MATHE MATICS

which the All is reconstituted (set theory) , but - before

all else - in the convergence of its methods and in the
transfusing of ideas from one to another of its various webs.
Tue penetration of algebraic methods into analysis, itself
subordinated to topology, the ubiquitous geometrization
of logic and the structural harmony of complex analysis
with arithmetic, are all examples in which mathematics'
global unity can be perceived in its local details. A pro­
found epistemological inversion shows how - contrary to
what we might think at first - an attentive observation of
practical diversity permits a later reintegration of the One
behind the Multiple. In fact, a full awareness of diversity
does not reduce to the disconnected, but rather turns
back to unity, whether in Peirce's pragmatism, Benjamin's
montage, Francastel's relay or Deleuze's difference. Simi­
larly - and with great technical precision, as we shall see
in chapters 4 and 7 modern mathematics seeks (and
-

finds) ways to concatenate a prolific multiplicity of levels

into great towers and unitary frameworks.
That reconstruction of the One behind the Multiple is
another of the fundamental marks allowing us to separate
elementary mathematics from advanced mathematics.
Elementary mathematics is 'one' from the outset, because
it has not yet managed to multiply or differentiate itself;
advanced mathematics, by contrast, having already passed
through explosively creative processes, has had to relearn
and reconstruct common ties and warps in the midst of

37
 SYNTHETIC PHILOSOPHY

diversity. The firmness and solidity produced by that

double weaving movement - differentiation/integration,
multiplication/unification - are virtues proper to advanced
mathematics, which can be only very faintly detected in
elementary mathematics. In fact, some of the great unitary
theories of contemporary mathematics - generalized
Galois theory, algebraic topology, category theory - are
trivialized on the elementary level, since the s tructures at
play fail to achieve enough differential richness to merit
any subsequent reintegration. It is factually impossible,
therefore, to claim to observe the same kinds of conceptual
movements in reasoning over tally marks as one finds
when, for example, one enters into the theory of class
fields. We believe that failures to understand or assume
this sort of distinction have done quite enough damage
to the philosophy of mathematics.
Immediately bound up in the weave between mul­
tiplicity and unity in modern mathematics, we find the
inescapable dynamism of doing mathematics (point 4).
Mathematics, developing as it has from the middle of the
nineteenth century up to the present day, has not ceased to
create new spaces for the understanding. A pendular pro­
cess - in which, on the one hand, meticulous saturations
within particular structures accumulate, and on the other,
the free behavior ofgeneric str uctures is set loose - allows
for the simultaneous contemplation of an uncommonly
precise spectrum of local obstructions/resolutions and

38
 MOD ERN AND CONTEMPORARY MATHEMATICS

a series of global organizational schemas. The dynamic

transit between the local and the global is one of the major
successes of modern mathematics, a transit that is hard
to perceive in elementary mathematics, in which a clear
preponderance of the local takes precedence. Again, there
seems to be an unwarranted extrapolation at work when
one presumes to take the eminently static, finished, stable
and 'smooth' character of elementary mathematics as char­
acteristic of all of mathematics in its entirety. Advanced
mathematics are, by contrast, essentially dynamic, open,
unstable, 'chaotic'. It is not by chance that, when one
asks mathematicians about the future of their discipline,
almost all of them leave the landscape completely open;
with a thousand forces pulling in different directions,
the 'geometry' of mathematical creativity is replete with
unpredictable singularities and vortices .
The back-and-forth between diverse perspectives (con­
ceptual, hypothetical, deductive, experimental ) , diverse
environments (arithmetical, algebraic, topological, geo­
metrical, etc.) and diverse levels of stratification within
each environment is one of the fundamental dynamic
features of modern mathematics. When that pendular
back-and-forth partially concretizes itself in theorematic
warps and interlacings, and when the transit of ascents and
descents between certain levels of stratifications - along
with a great arsenal of intermediate mixtures to guide the
transit - is systematized , we then find ourselves faced

39
 OF CONTEMPORARY MATHE MATICS

with (point 4) other peculiarities, specific t o advanced

mathematics . In _fact, at low levels of complexity, such
as those found in elementary mathematics, the eleva­
tions (ascents/descents) and intermediate constructions
(mixtures) naturally tend to trivialize away and vanish. It
was necessary, for example, for obstructions in infinitary
systems of linear equations and in classes of integral
equations to be confronted in order for the notion of a
Hilbert space, one of modern mathematics' most incisive
mixtures, to emerge, just as certain singularities in com­
plex variable functions had to be confronted for another
paradigmatically modern construction, the notion of
Riemann surfaces, to emerge. Similarly, Galois theory -
one of the great buttresses of mathematics' development,
with remarkable conceptual transfers into the most varied
mathematical domains - would be unthinkable had impor­
tant obstructions between webs of notions associated with
algebraic solutions and geometrical invariants not been
taken into account. In order to tackle problematics of
great complexity - stretched over highly ramified dialecti­
cal warps - modern mathematics finds itself obliged to
combine multiple mathematical perspectives, instruments
and bodies of knowledge, something that rarely happens
in the realms of elementary mathematics.

40
 MOD ERN AND CONTEMPORARY MATHE MATICS

Beyond points 1-5, which we have just discussed,18 and

which constitute an initial plane of separation between
elementary and modern mathematics (from the middle
of the nineteenth century to the middle of the twenti­
eth, as we have defined it), we believe that contemporary
mathematics (1950 to the present) incorporates additional
criteria that reinforce its specificity. Beyondconseroing those
distinctively modern characteristics (1-5),19 contemporary
mathematics bears new, distinctive elements, as compared
to elementary mathematics, among which we may point
out the following:

6. the structural impurity of arithmetic (Weil's conjec­

tures, Langlands's program, the theorems of Deligne,

Faltings and Wiles, etc.);

7. the systematic geometrization of all environments

of mathematics (sheaves, homologies, cobordisms,
geometrical logic, etc.);

18 Lautman's work (see note 12 and chapter 2) provides a great variety of technical
examples, concretizing the aforementioned tendencies, as well as other formulations
of points 1-5 .

19 There is no 'postmodern break' in mathematics. Following Rodriguez Magda, it is far

more appropriate to speak of transmodemity than of a dubious 'post'-modernity, when
seeking to characterize our age. (R M. Rodriguez Magda, Trammodernidad [Barcelona:
Anthropos, 2004]). In mathematics - and, in fact, in culture as a whole, as we have
remarked in our essay Razon de lajronterayjronteras dela razon (F. Zalamea, Razon de
lafrontera yjronteras de la razon: pensamiento de los limites en Peirce, Florenski, Marey, y
limitantes de la expresi6n en Lispector, Vieira da Silva, Tarkovski [Bogota: Universidad
Nacional de Colombia, 2010]) - continuous notions connected with traffic and
frontier are indispensable. The prefix 'trans-' seems, therefore, far more indicative of
our condition (and of the mathematical condition) than a premature 'post-'.

41
 SYNTHETIC PHILOSOPHY

8. the schematization, and the liberation from set­

theoretical, algebraic and topological restrictions
(groupoids, categories, schemas, topoi, motifs, etc.);

9. thejluxion and deformation of the usual boundaries

of mathematical stmctures (nonlinearity, noncom­
mutativity, nonelementarity, quantization, etc.);

io. the reflexivity of theories and models onto themselves

(classification theory, fixed-point theorems, monstrous
models, elementary/nonelementary classes, etc.).

Many of the major innovative works of the great con­

temporary mathematicians20 can be situated, grosso modo,
along the aforementioned lines, as we suggest in the
following table:21

20 The selection is, inevitably, personal, though the list indubitably includes some of the
fundamental figures of mathematics since i950. We only include in the table those
mathematicians whom we are studying in the second part of this essay. (The order of
appearance in the table corresponds to the order in which each author is studied in the
second part of our essay.) Other indispensable figures of contemporary mathematics do
not appear here (such as Borel, Chevalley, Dieudonne, Drinfeld, Eilenberg, Gelfand,
Margulis, Milnor, Smale, Thom, Thurston and Weil, to name just a few), since, in most
cases, we mention them only in passing, without dedicating a specific section to their
works .

.21 The marks indicate a clear preponderance of works along each line, and not mere
incursions that might be considered limited in comparison with the remainder of the
work of the mathematician in question. Grothendieck is clearly situated above all
other mathematicians of the last half century, as is faintly indicated by the.five marks
that serve to register the enormous presence of his work. The other marks should be
understood as merely indicative, though they are also adequately representative.

42
 M ODERN AND CONTEMPORARY MATHE MATICS

6 7 8 9 10

!orothendieck • • • • •

Serre • • •

• • •
Langlands

Lawvere • • •

Shelah • • • •

Atiyah • • •

Lax • •

Connes • • • •

Kontsevich • • •

Freyd • • •

Simpson • •

Gromov • • •

Zilber • • •

Figure 3. AJew great mathematici,ans and their contributions to the major lines ef
development efcontemporary mathematics.

Behind arithmetical m1xmg (6), geometrization ( 7 ) ,

schematization (8), structural fluxion (9) and reflexivity
( 10 ), we find some modes of conceptualization and con­
struction pertaining to contemporary mathematics that
are not in evidence (or that appear only in nuce) in the
period from i900-1950. An initial, fundamental inversion
consists in studying fragments of mathematics, by setting
out not from partial axiomatic descriptions (as in Hil­
bert's program) , but from classes of correlated structures.

43
 OF CONTEMPORARY MATHEMATICS

For both mathematical logic (with the unprecedented

blossoming of model theory) and pure mathematics (with
category theory), the objects studied by mathematics
are not only collections of axioms and their associated
models, but also, from an inverse perspective, classes of
structures and their associated logics (a point of view
that is indispensable for the emergence of abstract model
theory and generalized quantifiers, after Lindstrom) . In
cases where the class of structures is very extensive and
runs transversally through many fields of mathematics
(such as the intermediate categories between regular
categories and topoi), the breadth of perspective often
provides for new global theorems (synthetic hierarchiza­
tion, delimitation of frontiers, transference - as in Freyd's
representation theorems). In cases where the class arises
on the basis of certain particular structures and their
infinitesimal deformations (as in quantization), a precise
and profound apprehension of the class brings with it
local technical advances of a remarkable nature (analytic
decomposition, fluxion, asymptotic control - as in Perel­
man's proof of the Poincare conjecture). In either case,
however, mathematics explicitly comes back to precede
logic . The situation we are dealing with here is a basic one
(broadly prefigured by Peirce, to whom we shall return),
which may, in fact, have always subsisted in mathematical
practice, but which was, again and again, hidden from

44
 M OD ERN AND CONTEMPORARY MATHE MATI C S

the perspectives available to analytic philosophy or the

phil osophy of language in the twentieth century.
A second essential inversion has to do with contem­
porary mathematics' tremendous capacity to construct
incisive technical breaches of apparently insuperable
boundaries - nonelementaryclasses (Shelah), noncommu­
tative geometry (Connes), nonidempotent logic (Girard's
nonperennial, or linear, logic), etc. - going beyond the
normalized environments that had arisen naturally in the
discipline during the first half of the twentieth century.
Instead of progressing from a positive interior, in which
knowledge is accumulated, toward a negative, somehow
unknowable exterior, contemporary mathematics sets
itself within the determinate boundaries of the 'non-' from
the outset, and, setting out from those frontiers, goes on
to constructively explore new and astounding territories.
A third inversion consists in considering mathematical
mixtures, not as intermediary entities that are useful in
a deduction, but as proper, original entities, in which
the very construction of the discipline is at stake. There
is no space of contemporary mathematics that does not
find itself mongrelized by the most diverse techniques;
the mediatory ( 'trans-' ) condition, which in the first half
of the twentieth century could be seen as one step in
the path of demonstration, is today becoming the very
core of the discipline. The extraordinary combination
of arithmetical structuration, algebraic geometrization,

45
 SYNTHETIC PHI LOSOPHY

schematization, fluxion and reflexivity in Grothendieck's

work is a prime example, in which every instrument is
simply directed toward controlling the transit of certain
global mathematical conceptions through an enormous
spectrum of local environments.
We should note that, in every case, the oscillations and
inversion remarked upon do not appear in elementary
mathematics, and indeed cannot appear in the latter's
restricted fields of action. Moreover, before a sweeping set­
theoretical landscape such as ZF with first-order classical
logic, many of the aforementioned currents become 'non­
observables' of some sort. One ofthe basic deficiencies of
mathematical philosophy has consisted in not coupling its
philosophical instruments of observation with the environ­
ments observed , and in attempting to paint standardized
landscapes of the whole. Another methodology (one more
in tune with the development of contemporary mathemat­
ics ) could consist in observing certain environments of
mathematics through philosophical filters more adequate
to them - and then trying to synthetically glue the diverse
philosophical observations thus obtained . ( See chapter 3
for the pragmatic - in Peirce's sense - program that can
be sketched out in this way, and chapters 8-n for the
realization of certain partial gluings. )
If it is entirely natural to use the instruments of analytic
philosophy in order to view the set-theoretical universe,
with its underlying first-order classical logic, it seems

46
 M OD ERN AND CONTEMPORARY MATHEMATICS

rather wrongheaded to carry on with the same old instru­

ments in the study of other mathematical environments
(other environments that constitute the majority: set
theory occupies only a very limited space of mathemati­
cal investigation, as can be seen in the MSC .iwoo). As
we shall see later on, certain dialectical instruments are
indispensable for capturing schematization and fluxion,
just as it is only from a synthetic and relational perspective
that the great currents of contemporary structuration can
be understood , and only from a fully modal perspective
that the inexhaustible richness of the continuum can be
observed . The specificity of modern and contemporary
mathematicsforces us to keep changing our filters of philo­
sophical observation, and so we find the elaboration of a
new philosophical optics coming into play, one that - with
a pragmatic machinery of lenses to insure a minimum
of distortions - will allow us to survey the landscape of
'real' mathematics.

47
C HAPT E R 2

ADVAN C E D MAT H E MATI C S IN T H E TRACTS O F

MAT H E MATICAL P H I L O S O PHY!

A B I B L I O GRAPHI CAL S U RVEY

In this chapter we will review the reception that advanced

mathematics has enjoyed in mathematical philosophy. As
we shall see, the absences clearly outnumber the presences,
though there have been significant efforts to be open to
modern and contemporary mathematics. This chapter seeks
only to carve out a few bibliographic footholds in a global
descriptive landscape. In part 3 of this essay, we will come
back to several ofthe authors mentioned here, concentrating
on far more specific and constrained local problematics.
The first section, 'The P lace of Lautman', tries to
sum up some of the main contributions made by Albert
Lautman's work, as a paradigm for a philosophical per­
spective attentive to modern mathematics. The second
section, 'Approaching Real Mathematics', surveys, in
chronological order, a modest number of appearances
of advanced mathematics in philosophy since Lautman,
and the even slighter presence there of contemporary
mathematics. The third section, ' More P hilosophy, Less
Mathematics', explores the spectrum of modern and
contemporary mathematics that appears in certain tracts
of analytic philosophy of mathematics, and that is summed

49
 SYNTHETIC PHILOSOPHY

up in the OifordHandbook ofPhilosophy oJMathematics and

Logi,c, where a deepening of philosophical aspects takes
precedence over the advanced mathematics themselves,
which are neglected. This is, of course, a valid option, but
one that must be acknowledged as such: a choice that
results in a vast panorama being disregarded.
The heart of mathematics has its reasons of which lin­
guistic reason knows nothing. Just like Stephen, inJoyce's
A Portrait ofthe Artist as a 'Young Man, who must confront
the 'wild heart of life' and must choose whether to avoid
it or rather to immerse himself in it, the philosopher of
mathematics cannot avoid having to confront the 'wild
heart' of mathematics. She may elude it, if that is what she
wants, but her reflections would then fail to take stock of
many of mathematics' central aspects; in particular, an
understanding of mathematical creativity that neglects
advanced mathematics cannot be anything other than
a limited and s keletal understanding. What would we
say of a historian or philosopher of art who took it upon
himself to elaborate meticulous chromatic distinctions
while solely restricting himself to a set of mediocre paint­
ings, leaving aside, for example, the complex colorations
of Turner, Monet or Rothko? What would we make of a
literary critic who presumed to circumscribe the 'whole'
of literary invention while reducing it to the short story
or novella, without, for example, taking Proust or Musil
into consideration?

50
 A B I B LI OGRAPHICAL SURVEY

In the arts it would be unthinkable - almost atrocious - to

neglect the great creations of the gente while aspiring to
elaborate an aesthetics . In the philosophy of mathemat·
ics , however, it has apparently proved rather easy to skip
over the emblematic creations of advanced mathematics .
In the introduction and chapter I, we have indicated a
few of the reasons why such 'oblivion' has been so very
comfortable and so rarely disquieting: the belief that con·
templating the world of elementary mathematics equals
contemplating the world of advanced mathematics; the
standardization of perspectives based on philosophies
of language, the assumption that perceiving modern
and contemporary technical advances would not lead to
major changes in the philosophy of mathematics . As we
understand it, this has to do with preemptively taking up
positions in such a way that access to the 'real' world of
mathematics, in its continuous and contemporary devel­
opment, is impeded . Given that, as we have indicated
in chapter I, advanced mathematics hinges on peculiar
specificities that distinguish it from elementary mathe·
matics, to limit mathematical philosophy to elementary
mathematics - however philosophically sophisticated our
reasons for doing so might be - presupposes a dubious
reductionism. We will take up, in the first two sections of
this chapter, the works of a few philosophers of mathemat·
ics who have indeed attempted to access the 'wild heart'
of the discipline.

51
 OF CONTEMPORARY MATHEMATICS

2 . 1 T H E PLACE O F LAUTMAN

Albert Lautman perhaps came closer than any other

twentieth-century philosopher to understanding the
creative world of modern mathematics. The 'Essay on
Notions of Structure and Existence in Mathematics ' 22 is
the principal thesis for a doctorate in letters (philosophy) ,
defended by Lautman at the Sorbonne in 1937. The work,
dedicated to the memory of his friend and mentor Her­
brand, constitutes a genuine revolution, as much in the
ways ofdoing philosophy of mathematics, as in the depth
- the undercurrent - of the ideas set down and the hori­
zons anticipated . Lautman grafts together structural and
dynamical conceptions of mathematics, interlacing the
'life' of modern mathematics with a sweeping spectrum
of dialectical actions: the local and the global (chapter 1);
the intrinsic and the induced (chapter 2); the becom­
ing and the finished - closely tied to the ascent and the
descent of the understanding (chapter 3); essence and
existence (chapter 4); mixtures (chapter 5); the singular
and the regular (chapter 6). Lautman divides his thesis
into two large parts ( 'Schemas of Structure' and 'Schemas
of Genesis' ) so as to emphasize one of his fundamental

22 A. Lautman, Essai sur !es notions de structure et d'existence en mathhnatiques. I. Les schemas
de structure. II. Les schimas de genese (Paris: Hermann, 1938, 2 vols.). Republished in A.
Lautman, Essai sur !'unite des matMmatiques et divers ecrits (Paris: 10/18, 1977), 21-154.
Recently republished in A. Lautman, Lesmathimatiques /es idees et le reelphysique (Paris:
Vrin,2006). [Tr. S. DuffyasMathematics,IdeasandthePhysicalReal (London: Continuum,
2011), 87-193.]

52
 A B I BLIOGRAPH ICAL SURVEY

assertions regarding the mathematics of his epoch - that

modern mathematics has a structural character (a pre­
figuration of the Bourbaki group - Lautman was close
friends with Chevalley and Ehresmann) and consequently
mathematical creativity (the genesis of objects and con­
cepts) is interlaced with the structural decomposition of
many mathematical domains.
For the fir st time in the history of modern math­
ematical philosophy, a philosopher had conducted a
sustained, prefound and sweeping suroey efthegroundbreaking
mathematics efhis time; confronting its technical aspects
without ambiguity or circumlocution, and 'dividing' it
into basic concepts that he painstakingly explains to the
reader, Lautman presents a strikingly rich landscape of
the great inventive currents of modern mathematics. 23
Thus breaking with the usual forms of philosophical
exposition - which used to (and, unfortunately, still do)
keep the philosopher at a distance from real mathematics
- Lautman opens an extraordinary breach in an attempt
to seize upon the problematics of mathematical creativity.

23 What follows is a brief summary of the mathematical themes reviewed by Lautman in his
principal thesis. Chapter i: complex variable, partial differential equations, differential
geometry, topology, closed groups, functional approximations. Chapter 2 : differential
geometry, Riemannian geometry, algebraic topology. Chapter 3: Galois theory, fields
of classes, algebraic topology, Riemann surfaces. Chapter 4: mathematical logic, first­
order arithmetic, Herbrand fields, algebraic functions, fields of classes, representations
of groups. Chapter 5: Herbrand fields, Hilbert spaces, normal families of analytic
functions. Chapter 6: operators in Hilbert spaces, differential equations, modular
functions. Entering into the landscape that Lautman draws, the reader is then really
able tofeel the multiple modes and creative movements of modem mathematics, never
present in the elementary examples usually adduced in the philosophy of mathematics.

53
 SYNTHETIC PHILO.S OPHY

Breaking, too, with the common canons that arefunda­

mentally accepted_ in the works of his contemporaries (with
their epistemological or linguistic emphases), Lautman
seeks to repair the complex dialectico-hermeneutic tissue
that forms the backdrop of those works (with as many ties
to Plato as to Heidegger), far removed from the 'nai've'
Platonism that he repeatedly criticizes .
Attending to the most innovative mathematics of his
time, and open to a reinterpretation of the Ideas in their
original Platonic sense (as 'schemas of structure' that
organize the actual), Lautman exhibits the main lines
of support of mathematics' modern architectonic in his
'Essay on the Notions of Structure and of Existence in
Mathematics' . Dialectical oppositions, with their partial
saturations, and mixtures constructed to saturate struc­
tures, are linked to one another and to the underlying
living processes of mathematical technique. The unitary
interlacing of mathematical methods, through the ever­
permeable membranes ofthe discipline's various branches,
is a dynamical braiding in perpetual development. Math­
ematics, far from being merely one and eternal, is an
indissoluble ligature of contraries: it is one-multiple and
stable-evolutive. The richness of mathematics is largely
due to that elastic duplicity that permits, both technically
and theoretically, its natural transit between the ideal
and the real .

54
 A BIBLIOGRAPH ICAL SURVEY

In his complementary thesis for the doctorate of letters,

' Ess ay on the Unity of the Mathematical Sciences in
their Current Development ' , 24 Lautman explores the
profound unity of modern mathematics with more bril­
liant case studies . He believes he has detected that unity
in the burgeoning infiltration of algebra's structural and
finitary methods into all the domains of mathematics :
dimensional decompositions in the resolution of inte­
gral equations (chapter 1) , non-euclidean metrics and
discontinuous groups in the theory of analytic functions
(chapter 2), methods from noncommutative algebra in
differential equations (chapter 3), modular groups in
the theory of automorphic functions (chapter 4) . Laut­
man thereby stresses the tense union of the continuous­
discontinuous dialectic within modern mathematics , a
union that takes on aspects of 'imitation' or 'expression'
as it alternates between finite and infinite structures: imi­
tation when, in order to resolve a problem, one seeks to
trace a simple property of finite structures in the infinite;
expression when the emergence of a new infinite construc­
tion includes a representation of the finite domains that
prompted its emergence. The 'analogies of structure of
reciprocal adaptations' between the continuous and the
discontinuous are, for Lautman, among the basic engines
of mathematical creativity, a fact that seems to have been

24 Lautman, Essai sur !'unite des matlufmatiques et divers icrits, 155-202. Les ma.thbnatiques
!es idees et le reelphysique 81-124; translation, 45-83.

55
 OF CONTEMPORARY MATHE MATICS

borne out by the entire second half of the twentieth cen­

tury, both in the unitary entanglement of the methods of
algebraic geometry that brought about the demonstration
of Fermat's Theorem (Wiles), and in the geometrical and
topological adaptations that appear to be on the verge of
proving the Poincare conjecture (Perelman).
If, in his complementary thesis, Lautman's emphasis
primarily tends in the discrete --+ continuous direction
(the direction in which the tools of modern algebra help
to generate the concepts and constructions of analysis),
a study of the other direction, continuous --+ discrete,
can also be found in his reflections on analytical number
theory in 'New Investigations into the Dialectical Struc­
ture of Mathematics'. 25 In this brief pamphlet - the last
work Lautman published in his lifetime - he remarks on
how 'reflections on Plato and Heidegger ' are conjugated
with 'observations on the law of quadratic reciprocity
and the distribution of prime numbers', in an effort to
sustain, once again, one of his fundamental theses : to
show 'that that rapprochement between metaphysics and
mathematics is not contingent but necessary'. In the Hei­
deggerian transit between pre-ontological understanding
and ontic existence, Lautman finds, in channels internal
to philosophy, an important echo of his own reflections

25 A. Lautman, Nouvelles recherches sur la structure dialectique des mathfmatiques ( Paris:

Hermann, 1939). Republished in Lautman, Essai sur !'unit!..., 203-229. Les
mathhnatiques..., 125-221; translation, 197-219.

56
 A B I BLIOGRAPH ICAL SURVEY

regarding the transit of the structural and the existential

within modern mathematics.
In the second subsection of the first part of 'New
In stigations ', 'The Genesis of Mathematics out of the
ve
Dialectic', Lautman explicitly defines some of the fun­
damental terms that, in his thesis, are audible only in
whis pers : the pairs of dialectical notions (whole/part,
extrinsic/intrinsic, system/model, etc.), and the associated
dialectical ideas, which should be understood as partial
resolutions of oppositions between 'notions'. For example,
the understanding of the continuum as a saturation of
the discrete (the Cantorian completion of the real line)
is a Lautmanian 'idea' that partially responds to the
pair of 'notions', continuous/discrete, but it is clear that
there can likewise be many other alternative 'ideas' for
delimiting the notions at stake (like Brouwer's primordial
continuum, from which the discrete is detached, revers­
ing Cantor's process). In virtue of Lautman's synthetic
perception, mathematics exhibits all of its liveliness, and
the nonreductive richness of its technical, conceptual and
philosophical movements becomes obvious. And so the
harmonious concord of the Plural and the One, perhaps
the greatest of mathematics' 'miracles' shines forth.
The war years slowed the tempo of Lautman's writing
- he was deeply involved in military activities and in the
Resistance - but he nevertheless found time to try to lift
his work above the horror that surrounded him. In i939,

57
 SYNTHETIC PHILOSOPHY

during a memorable session of the French P hilosophy

Society, Lautman defended, alongside Jean Cavailles, the
theses the two friends had recently submitted . A transcript
of Lautman's intervention is preserved in 'Mathematical
Thought',26 where he insists on the structural character of
modern mathematics, and points to how contrary 'notions'
(local/global, form/matter, container/contained, etc.)
dwell within groups, number fields, Riemann surfaces
and many other constructions, and how 'the contraries
are not opposed to one another, but, rather, are capable
of composing with one another so as to constitute those
mixtures we call Mathematica'. At the end of his interven­
tion, in an homage to Plato and a return to the 'Iimaeus,
Lautman proposes an ambitious reconstruction of the
'theory of Ideas ' for mathematical philosophy, in three
great stages : a description of the inexhaustible richness
ofeffective mathematics; a hierarchization of mathemati­
cal geneses; a structural explanation of mathematics'
applicability to the sensible universe.
Lautman's last two works take up, in part, this final
task and thereby serve to bring his philosophical labor to
a coherent closure. Chapters from a monograph on the
philosophy of physics that Lautman was unable to com­
plete, 'Symmetry and Dissymmetry in Mathematics and in
Physics' and 'The Problem of'Iime', aggressively tackle the

26 A. Lautman, 'La pensCe mathCmatique', Bulletin de la Socitti FraTlfaise de Philosophif

XL, i946: 3-17.

58
 A B I B LIOGRAPH I CAL SURVEY

problem of interlacing the ideal with the real, by way of the

complex conceptual constructions of quantum mechan­
ics, statistical mechanics and general relativity. 27 Laut­
man establishes some remarkable correlations 'between
dissymmetrical symmetry in the sensible universe and
antisymmetric duality in the world of mathematics', dis­
c overs the potential of the recently developed lattice theory
(Birkhoff, Von Neumann, Glivenko) and meticulously
shows that any conception of time must simultaneously
account for both the global form of the entire universe
and its local evolutions . Seeking an explanation of time's
sensible duality (as both an oriented dimension and a fac­
tor of evolution), Lautman uncovers the duality's 'ideal'
roots in a profound and original structural investigation of
time's twofold behavior in differential equations. Raising
mathematics and physics to the point where they are seen
as higher-order 'notions' linked to symmetry (predomi­
nantly mathematical) and dissymmetry (predominantly
physical) , Lautman thus successfully completes the first
circumnavigation of his theory of 'ideas'.
Among Lautman's many contributions, both his stud­
ies of the mixtures of modern mathematics and his explica­
tion of the ideas/notions that facilitate the incarnation of
obstructions and resolutions in mathematical creativity are

27 A. Lautman, Symetrie et dissymetrie en mathhnatiqtJ£s et en physique (Paris: Hermann,

1946); Republished in Lautman, Essaisurl'uniti. .. , 231-80. Lesmathhnatiques... 265-99;
translation 229-62.

59
 O F CONTEMPORARY MATHEMATICS

ofparticular relevance. Mathematical mixtures are legion;

the ones most studied by Lautman include algebraic
topology, differential geometry, algebraic geometry, and
analytical number theory. These interlacings of a noun (the
main subdiscipline) with an adjective (the 'infiltrating'
subdiscipline) are just faint echoes of the real procedural
osmoses of modern mathematics; the rigid delimitations
of the past fade away, to emerge again as pliable folds
within a new classification that resembles not so much a
tree as a vast liquid surface over which information flows
between mobile nuclei of knowledge. Albert Lautman
is the only philosopher of modern mathematics who
has adequately emphasized and studied the advent of
mathematical mixtures in actuality, on the one hand, and
the 'ideas' and 'notions ' that allow the transit of those
mixtures to be understood in potentiality, on the other.
Given the inescapable importance of mixed constructions
in contemporary mathematics, it is of no surprise to say
that Lautman's philosophy of mathematics would be of
great worth to our time, if only it were better known. To
break with an imagined 'tautological' mathematics, the
dubious 'pure' invention of analytic philosophy, and to
open onto a contaminated, and much more real, mathemat·
ics, is the order of the day.
Lautman exalts the richness won by the introduction
of 'transcendent' analytical methods into number theory,

60
 A BIBLIOGRAPH I CAL SURVEY

and explains why mathematical creativity has a natural

tendency to require such mixtures and mediations:

the demonstration of certain results pertaining to

the integers depends upon the properties of cer­
tain analytic functions, because the structure of the
analytical means employed turns out to be already
in accordance with the structure of the sought after
arithmetical results.28

In fact, the great interest of the mixtures lies in their

capacity to partially reflect properties from one extreme
to another, and to serve as relays in the transmission of
information. 29 Whether they occur in a given structure
( Hilbert space) , in a collection of structures (Herbrand
ascending domains) or in a family of functions (Montel
normal families) , mixtures, on the one hand, imitate the

28 Lautman, Les matMmatiques... 245-6; translation, 208 {translation modified).

29 Francastel's reli (from the French relais, meaning 'relay') affords - for the work of
art - another mixture or junction of great value; one where the perceived, the real
and the imaginary are conjugated. 'The plastic sign, by being a place where elements
proceeding from these three categories encounter and interfere with one another, is
neither expressive (imaginary and individual) nor representative (real and imaginary),
but also figurative (unity of the laws of the brain's optical activity and those of the
techniques of elaboration of the sign as such).' (P. Francastel, La realidadfigurative
[1965] [Barcelona: Paid6s, 1988], 115). If we contrast a definition of the work of art as
a 'form that signifies itself' (Focillon) with a definition of the work of mathematics as
a 'structure that forms itself' (our extrapolation, motivated by Lautman), we can have
some intuition, once again, of the deep common ground underlying aesthetics and
mathematics. For a remarkable recuperation of a history of art that takes stock of the
complex and the differential, but that recomposes it all in a stratified and hierarchical
dialogue, attentive to the universal and to 'truth' (an eminently Lautmanian task), see
J. Thuiller, 'Ihiorie generalede l'histoirede !'art (Paris: Odile Jacob, 2003). For Focillon's
definition, and an extensive subsequent discussion, see p. 65.

61
 SYNTH ETIC PHILOSOPHY

structure of underlying domains and, on the other, serve

as partial building blocks for the structuration of higher
domains. Without this kind of sought-after contamination,
or premeditated alloying, contemporary mathematics
would be unthinkable. A remarkable result like the proof
of Fermat's Theorem (1994) was only possible as the
final exertion in a complex back-and-forth in which an
entire class of mathematical mixtures had intervened - a
problem concerning elliptical curves and modular forms
resolved by means of exhaustive interlacings between
algebraic geometry and complex analysis, involving the
zeta functions and their Galois representations.
'Mixtures' show up in Lautman's earliest surviving
manuscript - his posthumously published work on math­
ematical logic.3° At twenty-six, Lautman describes the
construction of Herbrand 'domains' and shows that 'the
Hilbertians succeeded in interposing an intermediary
schematic of individuals and fields considered not so
much for themselves as for the infinite consequences that
finitary calculations performed on them would permit' .31
Comparing that 'intermediary schematic' with Russell's
hierarchy of types and orders, Lautman points out that
'in both cases, we are faced with a structure whose ele­
ments are neither entirely arbitrary nor really constructed,

30 A. Lautman, 'ConsidCrations sur la logique mathematique', in Lautman,Essaisurl'unite. . . ,

305-15. Republished in Lautman, Les matMmatiques. . . , 39-46; translation 1-8.

31 Ibid., 315; 46; translation, 7-8 (translation modified) .

62
 A B I BLIOGRAPH I CAL SURVEY

but composed as a mixedform whose fertility is due to

its double nature'.32 This clear intuition of the mixed
forms of logic, in the thirties, when logic, by contrast,
tended to see itself as a 'pure form', is evidence of the
young philosopher's independence and acumen. Indeed,
it now seems obvious that those mixed forms of logic
were the underlying cause of the late twentieth-century
blossoming of mathematical logic, happily infiltrated by
algebraic, topological and geometrical methods. In that
sense, Lautman never presupposes an a priori logic, prior
to mathematics, but considers it as a constituent part
of doing mathematics, with a prescient sense of today's
pluralistic conception of logic, whereby a logical system
accommodates a collection of mathematical structures
rather than installing itself beneath it.
In a text on the 'method of division' in modern axi­
omatics (his first published article), Lautman already
moves to link the mention of mixtures with the great
philosophical tradition:

It is not Aristotelian logic, the logic of genera and

species, that intervenes here [i.e., in mathematical
creation], but the Platonic method of division, as it
is taught in the Sophist and the Philebus, for which the
unity of Being is a unity of composition and a point

32 Ibid, 315; 46; translation, 8 (translation modified).

63
 O F CONTEMPORARY MATHEMATICS

of departure for the search for the principles that are

united in the ideas.33

Lautman accentuates the dynamical interest of a mixture

'which tends to liberate the simple notions in which tha1
mixture participates', and thereby situates mathematica
creativity in a dialectic of liberation and composition. Ir
terms foreign to Lautman, but which situate his positior
on better-known terrain, on the one hand, the makin�
of mathematics divides a concept's content into defini·
tions (syntax) and derivations (grammar), and liberate�
its simple components; on the other hand, with modeh
(semantics) and transfers (pragmatics), it construct�
intermediary entities that resume the existence of thos{
simple threads, recomposing them into new concepts.
When the mixture succeeds in simultaneously combin·
ing a great simplicity and a strong power of reflection in
its components - as with Riemann surfaces or Hilber1
spaces, the subject of admiring and exemplary studie�
by Lautman - mathematical creation reaches perhap�
its greatest heights.
In Lautman's theses, the philosopher's entire reflec·
tive movement is propelled by a pendular contrasting
of complementary concepts (local/global, whole/part

33 A. Lautman, 'L'axiomatique et la methode de division', in Recherches philosophique:

VI, 1936-37' 191-203; republished in Lautman, Essai sur /'unite. . . , 291-304. Le:
mathbnatiques . .. , 69-80; translated in Lautman, Mathematics, Ideas... , 31-42.

64
 A B I BLIOGRAPHICAL SURVEY

extrinsic/intrinsic, continuous/discrete, etc. ) , but it is

in his 'New Investigations into the Dialectical Structure
of M athematics' that Lautman introduces the terms that
govern those dialectical interlacings. Lautman defines a
notion as one of the poles of a conceptual tension and an
idea as a partial resolution of that polarity. The concepts
of finitude, infinity, localization, globalization, calcula­
tion, modeling, continuity and discontinuity ( Lautman's
examples) are Lautmanian 'notions'. A few Lautmanian
'ideas' (our examples) would be the proposals according
to which the infinite is grasped as nonfinite ( the cardinal
skeleton) , the global as gluing of the local (compactness) ,
the model-theoretic as realization of the calculative ( set­
theoretical semantics) , or the continuous as completion
of the discrete ( the Cantorian line) .
The interest of 'notions' and 'ideas' is threefold: they
allow us tofilter out (liberate) unnecessary ornaments and
decant the grounds of certain mathematical frameworks;
they allow us to unify various apparently disparate con­
structions from the perspective of a 'higher' problematic
level; and they allow us to open the mathematical spectrum
to various options. Whether the mathematical landscape
is being filtered or unified (duality theorems in algebraic
topology and lattice theory) or being opened towards a
fuller scope of possibilities ( 'non-standard ideas', which
resolve in another way oppositions between fundamental
'notions': the infinite as the immeasurable in Robinson,

65
 SYNTHETIC PHILOSOPHY

the discrete as demarcation of a primordial continuum

in Brouwer, the calculative as system of coordinates for
the model-theoretic in Lindstrom), Lautmanian 'notions'
and 'ideas' let us roam transversally over the universe of
mathematics and explicate not only the breadth of that
universe, but its harmonious concord between the One
and the Many.
For Lautman, 'notions' and 'ideas' are situated on a
'higher' level, where the intellect is capable of imagining
the possibility efa problematic, which, nevertheless, acquires
its sense only through its immediate incarnation in real
mathematics. Lautman is conscious of how an a priori
thereby seems to be introduced into the philosophy of
mathematics, but he explains this as a mere 'urgency of
problems, prior to the discovery of solutions'. In fact, a
problem's 'priority' itself should be only considered as
such from a purely conceptual point of view, since, as
Lautman himself points out, the elements of a solution
are often found to have already been given in practice, only
later to incite the posing of a problem that incorporates
those data (none of which prevents, in a conceptual
reordering, the problem from ultimately preceding the
solution). In parallel with his strategy of apprehending
the global structure of a theory rather than predefining its
logical status, Lautman consistently situates mathemati­
cal logic as an activity within mathematics that should
not arbitrarily precede it and that should be situated

66
 A BIBLIOGRAPH ICAL SURVEY

on the same level as any other mathematical theory, thus

anticipating the current conception of logic, which has
been accepted in the wake of model theory. Follow­
ing Lautman, 'logic requires a mathematics in order to
exist', and it is in the weaving together of blended logical
schemata and their effective realizations that the force of
doing mathematics lies.
It is in the tension between a 'universal' (or 'generic' )
problematic and its 'concrete' (or 'effective' ) partial resolu­
tions, according to Lautman, that the better part of the
structural and unitary weaving of mathematics may take
root. As we will see in chapter 7, this is precisely the para­
digm proposed by the mathematical theory of categories.34
When Lautman looks at Poincare and Alexander's duality

34 Lautman never lived to know category theory, the rise of which began at the very
moment of his death (S. Eilenberg, S.Mac Lane, 'Natural isomorphisms in group
theory', Proc. Nat. Acad. Sci. 28, i942: 537-43; S. Eilenberg, S. Mac Lane, 'General
theory of natural equivalences', Tram. Amer. Math. Soc. 58, i945: 231-94). It is difficult
to know to what extent conversations with his friend Ehresmann - who introduced the
general theory of fiber spaces in the forties and promoted category theory in France
from the end of the fifties on - could have influenced, in its implicit depths, a conception
of mathematics so clearly recognizable (in retrospect) as categorical as Lautman's is.
Nevertheless, in the session of the French Philosophy Society in which CavaillCs and
Lautman defended their work, and in which Ehresmann participated, the latter already
pointed out precisely how a number of Lautman's philosophical conceptions should
be technically filtered and converted into equipment internal to mathematics itself: 'If I
have understood correctly, in the domain of a supramathematical dialectic, it would
not be possible to specify and investigate the nature of those relations between general
ideas. The philosopher could only make the urgency of the problem evident. It seems
to me that if we preoccupy ourselves with speaking about those general ideas, then we
are already, in a vague way, conceiving of the existence of certain relations between
those general ideas. From that moment, we can't then just stop in the middle of the
road; we must pose the problem, the genuinely mathematical problem, that consists in
explicitly formulating those general relations between the ideas in question. I believe
that a satisfactory solution can be given to that problem, regarding the relations between
the whole and its parts, the global and the local, the intrinsic and the extrinsic, etc. [ . . . ]
I believe that the general problems that Lautman poses can be stated in mathematical
terms, and I would add that we can't avoid stating them in mathematical terms.' The
entire rise of category theory effectively bears out Ehresmann's position.

67
 OF CONTE MPORARY MATHEMATICS

theorems, and describes how 'the structural investiga­

tion of a space. that receives a complex is reduced to
the structural investigation of that complex' ,35 when he
analyses the ascent toward a universal covering surface
and contemplates the hierarchy of isomorphisms 'between
the fundamental groups of different covering surfaces of
a given surface F and the subgroups of the fundamental
group F', when he mentions an inversion between Godel's
completeness theorem and Herbrand's theorem, which he
later extends to an alternation between form and matter
by way of certain mediating structures, or when - bolder
still - he asks if 'it is possible to describe, at the heart
of mathematics, a structure that would be something
like a initial sketch of the temporal form of sensible
phenomena' ,36 Lautman is in each case anticipating certain
techniques in categorical thought. These include functors
in algebraic topology, representable functors in variet­
ies, adjunctions in logic, and free allegories. In fact, by
'admitting the legitimacy of a theory of abstract structures,
independent of the objects linked to one another by those
structures', Lautman intuits a mathematics of structural
relations beyond a mathematics ofobjects which is to say, -

he prefigures the path of category theory.

35 Lautman, Les ma.tlufmatiques. . . , 201; translation, 162 (translation modified).

36 Ibid., 173; translation 132 (translation modified).

68
 A B I BLIOGRAPHI CAL SURVEY

The Lautmanian language of 'notions', 'ideas' and dia­

lectical hierarchies finds, in category· theory, a definite
technical basis. 'Notions' can be specified by means of
universal categorical constructions (diagrams, limits, free
objects), 'ideas' by means of elevations of classes of free
objects and functorial adjunctions, dialectical hierarchies
by means of scales of levels of natural transformations. In
this way, for example, Yoneda's Lemma technically expli­
cates the inevitable presence of the ideal in any thorough
consideration of mathematical reality (one of Lautman's
basic contentions), showing that every small category can
be immersed in a category of functors, where, in addition
to the representable functors that form a 'copy' of the
small category, there also inevitably appear additional
ideal functors ('presheaves'), that complete the universe.
What is at issue here is the ubiquitous appearance of
the 'ideal' whenever the capture of the 'real' is at stake,
a permanent and pervasive osmosis in every form of
mathematical creativity.
The majority of the schemas of structure and genesis
that Lautman studies in his principal thesis can be cat­
egorically specified and, most importantly, extended. For
example, the 'duality of local and global investigations'
is grafted onto a complex set of instruments of functo­
rial localizations and global reintegrations (Freyd-style
representation theorems), the 'duality of extrinsic and
intrinsic points of view' feeds into the power of a topos's

69
 SYNTHETIC PHILOSOPHY

internal logic (Lawvere-style geometrical logic), and the

'interest of the logical scheme of Galois theory' is extended
into a general theory of residuality (categorical Galois
connections in the style of Janelidze). So, when we see
how Lautman observes that 'certain affinities of logical
structure allow us to approach different mathematical
theories in terms of the fact that each offers a different
sketch of a solution for the same dialectical problem',
that 'we can speak of the participation of distinct math­
ematical theories in a common Dialectic that dominates
them', or that the 'indetermination of the Dialectic [ J . . .

simultaneously secures its exteriority', it seems natural

to situate his ideas in a categorical context, whether in
the weaving between abstract categories ('common Dia­
lectic') and concrete categories ('distinct mathematical
theories'), or in terms of free objects ('indetermination
of the Dialectic') whose extensive external applicability
throughout the entire spectrum of mathematics is precisely
a consequence of their schematic character.
The mutual enrichment of effective Mathematics and
the Dialectic (Lautman's capitalizations) is reflected in the
natural ascent and descent between Lautmanian notions
and ideas, on the one hand, and mixtures, on the other. In
fact, in ascending from the mixtures, the 'notions' and
'ideas' that allow us to situate the place of those mixtures
within an amplified dialectic are liberated; while descend­
ing from the 'notions', new mixtures are elaborated in

70
 A B I BLIOGRAPHICAL SURVEY

order to specify and incarnate the content of the 'ideas' at

stake . One of the great merits of Lautman's work consists
in its having shown how those processes of ascent and
descent must be indissolubly connected in the philosophy
of mathematics in extenso, just as they are in a Galois
correspondence in nuce.

' '
2 . 2 APPROAC H I N G REAL MAT H E MATI C S

In the following pages we will perform a brief survey of

the works of other authors who have tried to approach the
'heart' of 'real mathematics'. The survey will be chrono­
logical, and may be considered adequately representative,
but it is certainly not exhaustive. For each work, we will
indicate, firstly, what spectrum of mathematics it exam­
ines, and secondly, what global accounts obtain in light
of such an examination. As we shall see, these approaches
for the most part appear to concern classical mathematics
(P6lya, Lakatos, Kline, Wilder, Kitcher), though other
endeavors seek to examine modern (de Lorenzo, Mac
Lane, Tymoczko, Chatelet, Rota), and even contemporary
(Badiou, Maddy, Patras, Corfield), mathematics. We do
not find a comprehension of modern mathematics as
precise and broad as that achieved by Lautman in any of
the cases of which we are aware.

71
 OF CONTEMPORARY MATHEMATICS

George P6lya

P6lya's works constitute a mine of examples for bringing

the reader closer to the processes of discovery and inven­
tion (both processes being indispensable) at work within
classical and elementary mathematics- Mathematics and
Plausible Reasoning presents an important collection of
case studies that concern two major themes: analogical
and inductive constructions in mathematics, and modes
of probable inference-37 Volume I (Induction and Analogy
in Mathematics) explores classical analysis (primarily
concerning the figure of Euler, gloriously resurrected), the
geometry of solids, elementary number theory, the study
of maxima and minima, and certain elementary problems
from physics. P6lya carefully examines the weavings
between generalization and specialization, certain classes
of analogical hierarchies, the construction of the multiple
steps of a demonstration and conjecture confirmation.
He includes numerous, thoughtful exercises (with solu­
tions), which should be seen as a means for opening
the (philosophical or mathematical) reader's mind to a
nondogmatic understanding of mathematical practice.
Volume 2 (Patterns ef Plausible Inference) confronts the
problem of the plausibility of certain hypotheses from
which one deduces further statements (a sort of inverted
37 G. P6Iya) Mathematics andPlausible Reasoning, 2 vols. (Princeton: Princeton University
Press, 1954).

72
 A B I B LI OGRAPHICAL SURVEY

modus ponens, corresponding to Peirce's 'retroduction',

which P6lya, however, does not mention) . The problem
of plausible inference �A --> B, B .,.. .,.. A! consists in speci­
fying the conditions on the deduction A ----> B and on
the possible truth of B that must obtain in order for the
retroduction .,...,.. to A to be as plausible as possible. P6lya
goes on to tackle the progressive gradations of proof,
the small internal variations that allow us to overcome
the obstructions encountered in the solution process,
inventive chance, and the back-and-forth of hypotheses
and intermediate lemmata by which a demonstration
continuously takes shape. Through a close inspection of
classical mathematics, he detects a complex hierarchiza­
tion, which was later exploited in modern mathematics.
The other modern characteristics (semantic richness,
theorematic mixtures, structural unity) nevertheless fail
to appear within the classical horizon.
In Mathematical Discovery, P6lya restricts himself to
examples from elementary mathematics (basic geometrical
forms, numerical sums, Pascal's triangle) - though he
also includes a few classical references related to limits
and power series - to illustrate the gradual emergence
and concretization of mathematical ideas, from the vague
and apparently contradictory to the measured control of a
proof.38 By means of processes of figuration, superposition

38 G. P6lya, Mathematical Discovery (New York: Wiley, 1962).

73
 SYNTHETIC PHILOSOPHY

and amplification, P6lya shows how webs of auxiliary

notions and problems gradually take shape, converging
toward the solution of an initial problem, and how a
surprising interlacing of chance and discipline is often
found lurking behind various demonstrations. Numerous
examples and exercises with solutions once again bring
the reader closer to mathematical practice. This practi­
cal approach offers us a sense, however faint, of modern
mathematics' dynamical richness.

Imre Lakatos

Lakatos systematically introduces, into the philosophy of

mathematics, the method of conjectures and refutations
that Popper had applied to the philosophy of science as a
whole. In Proefs and Refutations: The Logi,c ofMathematical
Discovery, Lakatos explores the fluctuating mechanisms of
mathematical discovery, the changing norms of proofs, the
interlacing of counterexamples and lemmata in the con­
struction of a demonstration and the back-and-forth of a
mathematics understood as an experimental science.39 The
examples adduced are eminently classical and are treated
with patience and care: Euler's Polyhedron Theorem, Cau­
chy and the problems of uniform convergence, bounded
variation in the Riemann integral. Many dialectical
39 I. Lakatos, Pro'!ft and Refutations: '!he Logic ofMathematical Discovery (Cambridge:
Cambridge University Press, 1976). The book extends earlier articles from 1963-64.

74
 A BIBLIOGRAPHICAL SURVEY

forJUS explicit uses of the thesis-antithesis-synthesis

-

triad, expository games in Platonic dialogue, an incessant

back-and-forth between obstructions and resolution - run
through the work, which thereby discerns the emergence
of the dialectical dynamic that will govern the develop­
ment of modern mathematics.
Mathematics, Science and Epistemology posthumously
assembles Lakatos's articles on the philosophy of math­
ematics. 40 The spectrum observed is once again the envi­
ronment of classical mathematics (from the Greeks up
to Abel and Cauchy, on whom Lakatos's considerations
are centered), but the book also contains various com­
mentaries on the modern foundations (Russell, Tarski,
Godel), following the line preponderantly adopted by
twentieth-century philosophy of mathematics. Contem­
porary mathematics makes an appearance only in the
form of Robinson's nonstandard analysis, as a result of
Lakatos's interest in connecting it with a recuperation of
the infinitesimals utilized by Cauchy. Various hierarchies
are proposed, dealing with the steps of a proof (prefor­
mal, formal, postformal), and examples dealing with the
method of conjectures, proofs and refutations are refined.
The profound instruments of algebraic geometry which, by
the sixties, had already been constructed (by Grothendi­
eck), and which later led to proofs of the great theorems
40 I . Lakatos, Mathematics, Science and Epistemology; vol Q of Philosophical Papers (Cam·
bridge: Cambridge University Press, 1978).

75
 O F CONTEMPORARY MATHE MATICS

ofarithmetic, such as the Weil conjectures (Deligne, 1973),

go unmentioned..Instead, somewhat dubious speculations
are entertained regarding the undecidability of Fermat's
Theorem - yet another example of the distance between
the philosopher of mathematics and the mathematics of his
epoch, preventing him from taking anything like snapshots
of the mathematical thinking being forged around him.

Javier de Lorenzo

In his first monograph, Introducci6n al estilo matematico

�Introduction to Mathematical Style], De Lorenzo immedi­
ately shows himself to be awake to the modes of 'doing'
advanced mathematics.4' With creative verve, the author
confronts the great figures of modern mathematics (Cau­
chy, Abel, Galois,]acobi, Poincare, Hilbert, the Bourbaki
group, etc.) and argues that certainfragments of advanced
mathematics - group theory, real analysis, and abstract
geometries are his preferred examples - bring with them
distinct ways of seeing, of intuition, of handling operations
and even distinct methods of deduction, in each of their
conceptual, practical and formal contexts. De Lorenzo
points out how mathematics 'grows through contradis­
tinction, dialectically and not organically', and thereby
breaks with a traditional vision of mathematics, according

41 J. de Lorenzo, Introducci6n al estilo matem4Jico (Madrid: Tecnos, 1971).

76
 A B I BLIOGRAPHI CAL SURVEY

to which it grows by accumulation and progress in a verti­

cal ascent. He proposes instead a conceptual amplification
of the discipline, in which new realms interlace with one
another horizontally, without having to be situated one
on top of the other.
In La matematicay elproblema de su historia, De Lorenzo
2
postulates a radical historicity of doing mathematics. 4
Tue references to advanced mathematics are classified
in terms of three primary environments, within which,
according to De Lorenzo, the major ruptures and inver­
sions that gave rise to modern mathematics were forged:
the environment of 1827, in which the program for the
resolution of mathematical problems is inverted, setting
out 'from what seems most elusive in order to account for
why �problemsJ can or cannot be resolved', and in which
mathematics begins to feed on itself and its own limita­
tions; the environment of1875, in which the mathematical
tasks of the previous half century are unified (groups, sets)
or transfused from one register to another (geometrical
methods converted into algebraic or axiomatic methods),
generating the important constructions (Lie groups,
point-set topology, algebraic geometry, etc.) that drove
mathematics' development at the outset of the twentieth
century; the environment of 1939, in which the Bourbaki
group fixed the orientation of contemporary mathematics

42 ]. de Lorenzo, La matematicay elproblema de su historia (Madrid: Tecnos, 1977) .

77
 SYNTHETIC PH ILOSOPHY

around the notions of structure and morphism, inverted

the focus of mathematical research, and moved toward a
primordial search for relations between abstract structures
(algebras, topologies, orders, etc. ) . In this and other works
( see note i3), De Lorenzo also exhibits a subtle attention
to contemporary mathematics (with detailed citations
of Weil, Schwartz and Lawvere, for example) , though
modern mathematics remains his primary focus. 43 To sum
up, De Lorenzo argues that mathematical knowledge is
produced through very different contexts and branches,
following many tempos and rhythms. Incessant incorpora­
tions, transfers, osmoses, translations and representations
are afterward produced between the various environments
of mathematical knowledge; already-constructed notions
then give rise to new constructions by means of diverse
deformations and transfigurations.

43 Regarding Lawvere, for example, De Lorenzo points out - only seven years ( ! ) after
Lawvere introduced elementary topoi (i970) - that 'the interlacing of the theory of
categories with that of topoi, presheaves and algebraic geometry is showing itself to
be essential for the intentions of Lawvere and those working in the same direction, to
achieve a foundation, which he qualifies as "dialectical", for mathematical work, while
recognizing that such a foundation can only be of a descriptive character, achieving
in this way a revision of Heyting's intuitionistic logic as the one best adapted to topos
theory.' The investigation of the mathematical in progress ('is showing itself to be. . . ',
)
'those working.. .' not only surfaces in these unusual meditations of a historian and
philosopher, but is made in the mostfitting possible way, successfully detecting the
conceptual kernel of the situation: the interlacing of topoi with algebraic geometry and
with the underlying intuitionistic logic.

78
 A BIBLIOGRAPHICAL SURVEY

Raym ond L. Wilder

Mathematics as a Cultural System puts forward a valuable

and original conception of mathematics as a 'vectorial
system' in which various tendencies of mathematics
counterpose, superpose, interlace and consolidate with
one another, as if they were situated in a web of vecto­
rial operations. 44 Rather than understanding the realm
of mathematics according to the dispersive model of a
'tree', the vectorial system permits the introduction, with
greater finesse, of the fundamental ideas of directional­
ity, potentiality, normalization and singularity associ­
ated with vector fields. Wilder explores many examples
in classical mathematics (Leibniz, Fermat, Gauss and,
in particular, Desargues), and in modern mathematics
(Balzano, Lobachevski, Riemann, Hilbert), where a liv­
ing dialectic is established between potential fields (e.g.,
the resolution of algebraic equations), normal vectors
(e.g., ad hoc manipulations by radicals) and singulari­
ties (e.g., the 'ingenious' emergence of Galois). Wilder's
great knowledge of modem topology and algebra - he
is one of the few active mathematicians (together with
P6lya and Mac Lane) to appear in this chapter's biblio­
graphical survey - allows for a detailed demonstration of
the fact that mathematical 'reality' is a sort of changing

44 R. Wilder, Mathematics as a Cultural System (Oxford: Pergamon Press, t981).

79
 OF CONTEMPORARY MATHE MATICS

flux in the conceptual field of associated vectors, and

that various tendencies undergo constant modification
in accordance with their historical position in the web.
An evolution of collective mathematical intuition and a
search for invariants in that evolution allow us to see how
mathematical knowledge naturally modifies and stabilizes
itself, perduring despite its own plasticity.

Morris Kline

In Mathematics: The Loss of Certainty, the vision of a great

connoisseur of the history of mathematics takes flight,
accompanied by philosophical speculations that, however,
are quite a bit weaker.45 Kline proves to be particularly
attentive to four principal registers: Greek mathematics;
classical analysis (rise, development, disorder, founda­
tions, crisis, limitations); modern mathematics (reviewing
various works of Poincare, Weyl, Borel, Hilbert, von
Neumann, Stone, Dieudonne, etc.) ; the foundations of
mathematics (Cantor, Brouwer, Godel, etc.). Curiously,
for all his profound and extensive historical knowledge,
the reflections that such knowledge gives rise to are
debatable to say the least: an insistence on an 'illogical'
development of mathematics (in which errors, conceptual
shiftings and the recourse to intuition play a leading role),

45 M. Kline, Mathematics: The Loss efCertainty (New York: Oxford University Press, 1980).

80
 A BIBLIOGRAPH ICAL SURVEY

the perception of an 'unsatisfactory' state of mathematics,

the proclamation of an 'end to the Age of Reason', the
sense of a shattered multiplicity of mathematics with no
possibility for unification and the indication of a growing
isolation bringing about 'disasters' in the discipline. It is
surprising to find such a negative vision of mathematics
take shape in the i98os, when the discipline found itself
in full bloom. Once again, the preemptive occupation of a
philosophical position - Kline's postmodern predilection
for the supposed 'loss of certainty' - clouds the vision
and obscures the dynamic technical life that presents itself
all around the observer. If some of the critical points are
valuable (the place of error, multiplicity, relativity), to
carry them to the extreme and separate them from their
natural polar counterparts (proof, unity, universality)
brings about an excessive oscillation of the pendulum,
which impedes any detection of a far more complex
relational warp.

Philip Kitcher

TheNature efMathematicalKnowledge continues the focus

on episodes of classical mathematics in the tradition of
P6lya and Lakatos.46 The examples investigated include
Newton, Leibniz, Bernoulli, Euler, Cauchy, and an

46 P. Kitcher, TheNatureofMatherruiticoJKrww/,edge (New York: Oxford University Press, t983).

81
 SYNTHETIC PHILOSOPHY

extensive case study (chapter. 10) reviews the developmen t

of analysis (1650-1870). Modern mathematics shows up
much more intermittently, and the 'elementary' references
to Galois (reduced to the problem of the insolubility of
equations) and Riemann (with respect to the construction
of his integral) are symptomatic in that sense. Focused
fully on the classical spectrum, several of Kitcher's reflec­
tions prefigure with great acumen the complex web of
ideal constructions and operations that will appear in
modern mathematics, as well as its incessant evolution and
its coupling between conceptual fragments and real data.
Particularly sensitive to mathematical change, Kitcher
succeeds in evoking the dynamism of mathematics and
the discipline's unpredictable transit between the ideal
and the real as well as between the possible, the actual
and the necessary.

Thomas 'Ijmoczko

Tymoczko's work as editor of New Directions in the Phi­

losophy efMathematics helps to clearly explicate the two
great cascades into which the philosophy of mathematics
might flow, following analytic philosophy's many decades
of dominance_ 47 The first part of the book ('Challenging
Foundations') reminds us that philosophy has many

47 T. Tymoczko, ed.,NewDirectionsinthePhilmophy efMathemaJics (Boston: Birkhauser, 1986).

82
 A B I BLIOGRAPHICAL SURVEY

other themes to study within mathematics, aside from

foundations. 48 The second part ('Mathematical Practice')
p oints out that the philosopher should also be inclined
to observe mathematical practice, the evolutions of stan­
dards such as 'truth' and 'proof', the oscillation between
informal and rigorous proofs, and the complexity of the
mathematical architectonic. It is in the second part that
the articles coming closest to modern and contemporary
mathematics appear (Tymoczko on the four-color prob­
lem; Chaitin on computational complexity). But the texts
for the most part continue to evoke classical examples
(as Grabiner does, with respect to the development of
analysis in the eighteenth and nineteenth centuries).49
Tue sort of 'quasi-empiricism' that Tymoczko adopts indi­
cates that a deeper knowledge of mathematical practice
could help to resolve certain philosophical controversies
regarding realism and idealism, and that therefore (and

48 Paul Bernays, one of the great champions of the foundations of mathematics, already
pointed out in 1940, in a little· known review of Lautman's works, that 'it is to be said
in favor of Lautman's method that it is more suited than foundational discussions to
give to a philosopher an impression of the content and nature of modern mathematics.
Indeed it is worthwhile to emphasize that foundational problems by no means constitute
the only philosophically important aspect of mathematics' (P. Bernays, 'Reviews of
Albert Lautman', Journal '![Symbolic Logic 5, i940]: 22). This admirable display of
conscience by a genuine architect of the foundations of mathematics is something of
which too many philosophers of the foundations have stood in need.

49 The inclusion of an 'interlude' with two of P6lya's texts - written thirty years earlier
- is indicative of the meekness that has emerged in philosophy with respect to
approaching 'mathematical practice'. Of course, as is often the case in the Anglo-Saxon
academy, there is an obvious ignorance of anything not translated into English: to
speak of mathematical 'practice' without mentioning Lautman or De Lorenzo is
genuinely misguided, notwithstanding the ease with which this is done by anglophone
philosophers.

83
 OF CONTE MPORARY MATHEMATICS

this is one of the central foci of the present work) not

only more philosophy, but more mathematics, could be
of great assistance in resolving certain quandaries in
mathematical philosophy.

Saunders Mac Lane

Mathematics: Form and Function synopsizes the perspective

of an outstanding mathematician of the second half of
the twentieth century.so The main part of the monograph
- which should be seen more as a presentation, a bird's
eye view, of classical and modern mathematics, than as a
volume of mathematical philosophy - confronts head-on
the legacy of Galois and Riemann, and provides excellent
introductions to central themes in mathematics: groups,
algebraic structures, complex analysis, topology. Con·
temporary mathematics appears with respect to category
theory (Mac Lane was one of its founders) and sheaf
theory (a paradigm of contemporary methods) . Chapter
12, 'Tue Mathematical Network', explores the progres­
sive emergence of mathematical constructions (origins,
ideas, formal versions), and the incessant back-and-forth
between, on the one hand, themes, specialties, and sub­
divisions of mathematical knowledge, and, on the other,
transits, transformations and changes. For Mac Lane,

50 S. Mac Lane, Mathematics: Form and Function (New York: Springer, i!)86).

84
 A B I B LI OGRAPHICAL SURVEY

mathematical constructions arise by virtue of a network

of analogies, examples, proofs and shifts in perspective,
which let us encounter and define certain invariants amid
the change. If there is not an absolute 'truth', external to
the network, there nevertheless exist multiple gradations of
relevance, of correctness, approximation and illumination
inside the network. It has become one of mathematics'
central tasks to achieve the harmonious concord of those
gradations, to overcome multiple obstructions, to construct
new concepts with the residues.

Gian-Ca rlo Rota

Indiscrete Thoughts consists of an irreverent series of reflec­

tions, of great interest,s• by another leading mathemati­
cian of the second half of the twentieth century.52 It is an
uneven compilation, which includes anecdotes, historical
fragments, mathematical and philosophical reflections,
critical notes and brilliant, incendiary ideas. Above all,
and in the order of the compilation itself, Rota dedicates
a great deal of space to the biographies of mathematicians
(Artin, Lefschetz, Jacob Schwartz, Ulam) as creative

51 I am grateful here to the teachings of Alejandro Martln and Andres Villaveces, who
explained to me one memorable afternoon the importance of Rota's ideas, several
of which we will return to (by different routes) in part 3 of this book. F. Palombi,
La stella e l'intero. La ricerca di Gian-Carlo Rota tra matematica efenomenologia (Torino:
Boringhieri, 2003) presents several ideas of utmost relevance to our focus, and upon
on which we will later comment.
52 G.-C. Rota, Indiscrete 1houghtJ (Basel: Birkhiiuser, 1997) .

85
 SYNTHETIC PHILOSOPHY

individuals. For Rota, mathematics emerges in very spe­

cific vital and academic contexts (see the beautiful text
on 'The Lost Cafe'), giving rise to a dynamic, oscillating,
fluctuating discipline, with multiple concrete tensions,
indissolubly bound to personalities firmly situated in
place and time. The weaving between a generic mathemat­
ics and its particular incarnations, and the idea accord­
ing to which 'mathematics is nothing if not a historical
subject par excellence' (something that Jean Cavailles had
forcefully underscored half a century earlier) , underlie
the whole of Rota's thought and permeate some of his
most original conceptions: a 'primacy of identity', which
aims to define the 'essence' of an object as its very web of
factual superpositions, and which would help to replace
an obsolete mathematical ontology (the 'comedy of exis­
tence' of mathematical objects); a reappropriation of the
Husserlian notion of Fundierung (founding) in order to
rethink the mathematical transits between the factual
and the functional; a phenomenology of mathematics
open to forms of doing mathematics (beauty, varieties of
proof, imagination) usually neglected by the traditional
perspectives of mathematical philosophy.
The caustic and polemical article 'The Pernicious
Influence of Mathematics upon Philosophy' reveals
the excesses of a philosophy of mathematics oriented
toward formal juggling acts and bastardized by various
'myths' that have little to do with mathematical practice:

86
 A B I BLIOGRAPH I CAL SURVEY

the illusion of precision, axiomatic absolutism, the illu­

sion of permanence, conceptual reducibility. Paradoxical
as it may seem, Rota observes that analytic philosophy,
'perniciously influenced' by classical logic and by set
theory, has turned its back on and has abandoned high
mathematical creativity, be it geometrical, topological,
differential, algebraic or combinatorial, thereby estranging
itself from the real center of the discipline that helped it to
emerge. The philosophy of mathematics should therefore
turn back to examine, without prejudices and without
taking preestablished theoretical positions, the phenom­
enologi,cal spectrum of mathematical activity. Here, Rota's
reading - in three central articles on 'The Phenomenology
of Mathematical Truth', 'The Phenomenology of Math­
ematical Beauty', 'The Phenomenology of Mathematical
Proof', and, in four complementary texts, 'The Primacy
of Identity', 'Fundierung as a Logical Concept', 'Kant
and Husserl', and 'The Barber of Seville or the Useless
Precaution' - poses some vital problematics to which a
philosophy of mathematics aiming at a 'real' (in Corfield's
sense) understanding of the discipline should be open.
These include the emergence of mathematical creativity,
mathematics understood as the history of its problems,
the varieties of proof and the evolution of concepts, the
interlacings between the 'facts' of mathematics and their
constant functional reinterpretations, the superpositions
and nonreductive iterations of mathematical objects, and

87
 OF CONTEMPORARY MATHEMATICS

the ubiquitous transits between forms of analysis and

forms of synthesis. Rota's style - brief, distilled, caustic -
is not conducive to a systematic elaboration of his ideas,
but we will develop a few of them in part 2.

Alain Badiou

Being and Event53 offers a sophisticated example of how

to construct new philosophical meditations on the basis
of a patient observation of aspects of advanced math­
ematics_54 Badiou carefully explores Cohen's technique
offorcing - going beyond the mathematicians themselves
in the profundity and originality of his analysis - and
encounters one of the great contemporary supports by
means of which the Many and the Oone may be soundly
reintegrated. The investigation of the continuum hypoth­
esis, with its contrast between indiscernibility (Easton's
Theorem) and linguistic control (Godel's constructible
universe), exhibits certain oscillations of mathematical
thought in fine detaiL A profound ontologi,cal subversion
is suggested - the identification of 'mathematics' (the
science of pure multiplicities) and 'ontology' (the science

53 A. Badiou, L 'etre et l'fvlfnement (Paris: Seuil, 1988). (Tr. 0. Feltham as Being and Event
(London: Continuum, Qoo5).]
54 Badiou explicitly declares himself Lautman's admirer and heir. It is a unique case of
recognition and shared labor, even if the mathematical spectrum covered by Lautman
is much broader. Both Lautman and Badiou aim, however, to rethink and return to
Plato, setting out from the exigencies of contemporary thought.

88
 A BIBLIOGRAPHICAL SURVEY

of what is, insefar as it is), in virtue of the sheer force of

axiomatic set theory, which lets us name all the multiplici­
ties of mathematics and develop a (hierarchical, complex,
demonstrative) study of those multiplicities insofar as
they 'are'. Badiou's text includes a great many 'chronicles
of proofs' (the author's expression - i.e., proofs decon­
structed from the formal language and reconstructed in a
conceptual and philosophical language) which detail an
unusually broad landscape of modern and contemporary
set theory.
In his Short Treatise on Transitory Ontology,55 Badiou
continues his ontological 'subversion', so as to involve
an incisive re-envisioning of category theory and the
theory of elementary topoi. The construction of a dia­
logue between great figures of philosophy (Aristotle,
Plato, Descartes, Spinoza, Leibniz, Kant), contemporary
philosophers (Deleuze) , poets (Mallarme) and math­
ematicians both modern and contemporary (Cantor,
Godel, Cohen, Lawvere) is supremely original. We find
suggestions of the primacy of 'real' mathematics and a
consequent subordination of logic (topoi and associated
logics, classes of structures and associated logics, the
emergence of geometrical logic, an irreducible logical
weaving between the global and the local), which should

55 A. Badiou, Court troite d'ontologi,e tronsitoire (Paris: Seuil, 1998). [Tr. N. Madarasz as
Briefings on Existence: A Short Treatise on Transitory Ontology (Albany, NY: SUNY Press,
2006).)

89
 SYNTHETIC PHI LOSOPHY

bring about certain 'turns' in mathematical philosophy,

beyond analytic philosophy and the philosophy of lan­
guage. Badiou's nontrivial Platonic orientation (that is to
say, one not reduced to the 'external' existence of math­
ematical ideas and objects), an orientation that accords
with the 'condition of modern mathematics', is summed
up in three points: Mathematics is a thought (entailing,
against Wittgenstein's Tractatus, the existence of dynamic
processes that cannot be reduced to language); mathemat­
ics, like all thought, knows how to explore its boundaries
(undecidability, indiscernibility, genericity - entailing
the irreducibility of mathematics to a set of intuitions
or rules fixed in advance); mathematical questions of
existence refer only to the intelligible consistency of the
intelligible (entailing a marked indifference to 'ultimate'
foundations, and the adoption, instead, of a criterion of
'maximal extension' for all that is 'compossible', quite
similar to the richness of contemporary model theory) .
Mathematics - and ontology, with which it is identified
- is thereby understood as a sophisticated sheaf of meth­
ods and constructions for the systematic exploration of
the transitory.

Penelope Maddy

The contrast between the works of Badiou and those

of Maddy could not be greater, even though both refer

90
 A BIBLIOGRAPHICAL SURVEY

to the same mathematical spectrum: twentieth-century

set theory. In Realism in Mathematics, Maddy explores
descriptive set theory, the large cardinal axioms and the
continuum hypothesis, and she performs a detailed survey
of the contributions of the field's leading figures, from
Borel and Lusin to Martin, Moschovakis and Solovay.s6
Maddy shows that the richness of the set-theoretical
universe (new methods and models, new connections and
p erspectives, the possibility of obtaining verifiable con­
sequences) allows us to uphold a certain 'realism' - close
to some ofGodel's ideas - and to dismantle Benacerraf's
dilemma, since the notions of causality associated with the
dilemma lose their traction in set theory's sophisticated
relative consistency proofs. Though Maddy finds a certain
set-theoretical stability exactly where Badiou underscores
continuous transition above all else, we should point out
that both, specifically regarding the mathematics of their
time, succeed in proposing new questions and resolutions
for mathematical philosophy (the dissolution of Benacer­
raf's dilemma, the program for a transitory ontology). The
labor of the philosopher attentive to the mathematics of
her epoch is thus far from negligible.
In Naturalism in Mathematics, Maddy explores the
status of additional axioms for set theory, from the dou­
ble point of view of realism (the existence of objective

56 P. Maddy, Realism in Mathematics (Oxford: Oxford University Press, 1990) .

91
 OF CONTEMPORARY MATHE MATICS

universes of sets) and naturalism (the internal sufficiency

of mathematics and set theory, without need of extern al
justifications).57 Maddy reviews various axioms of great
mathematical interest (choice, constructibility, determi­
nacy, measurability, supercompactness, etc.), and set theo­
ry's major modern architects (Cantor, Dedekind, Zermelo,
Godel) appear extensively in her monograph, as do some
of its greatest contemporary practitioners (Cohen, Martin,
Moschovakis, Woodin, etc.). An emphatic observation of
practice runs through the entire text; a naturalist vision of
set theory is sustained through the direct contemplation
of how the set theoretical axioms emerge, are put to the
test and combined with one another inside mathematical
webs (being submitted to various combinatorial, deduc­
tive, conceptual and harmonic controls, until they are
either discarded or partially accepted).58 The search for
appropriate axioms and criteria of plausibility can thus
be seen as self-sufficient, without any need to invoke an
external ontology. (A brilliant example of such a method­
ology is presented in the final chapter of Maddy's book, in
studying the axiom of constructibility V=L and showing
that the axiom internally clashes with basic principles of
maximality, ubiquitous in mathematical practice.)

5 7 P. Maddy, Naturalism in Mathematics (Oxford: Oxford University Press, 1997).

58 Ifthe explicit term 'mathematics' appears in the titles of both of Maddy's monographs,
this is nevertheless restricted to set theory, a fragment of mathematical inquiry.

92
 A BIBLIOGRAPHICAL SURVEY

Gilles Chdtelet

Les enjeux du mobile: Mathimatique, physique, philoso­

phie �The stakes of the mobile: Mathematics, physics,
philosophyJs9 directly confronts the fundamental prob­
lems of mathematical thought's mobility, and of its natural
osmoses with physics and philosophy. Chatelet's text puts
several sui generis perspectives to work on the spectrum
of mathematical philosophy: an opening onto a sort of
primacy ef the visual in mathematical practice (thereby
bringing to bear part of Merleau-Ponty's general phenom­
enological program in the context of mathematics); 60 a
special sensitivity to the mobile emergence of mathematical
concepts and 'things', owing to a study of the gestures and
processes on the border of the virtual and the actual;
careful attention to and subtle analysis of the metaphori­
cal webs that accompany the doing of mathematics, and
govern its interlacings with physics and philosophy;
meticulous study, with detailed concrete cases, of the
modes of articulation of mathematical knowledge and of

59 G. Chiltelet, Les enjeux du mobile: MathbnatUµie, physique, philosophie (Paris: Seuil,

1993). (Tr. R. Shore & M. Zagha as Figuring Space: Philasophy, Mathematics and Physics
(Springer, 1999).]
60 On this recuperation of the diagram for the philosophy of mathematics, following the
clear French line of filiation - Lautman-Deleuze-Ch3.telet - see N. Batt, ed., Penserpar
/ediagramme: De Gill.esDeleuzea GillesChdte/et '!Morie-Litttirature-Enseignement 22 (Saint·
Denis: Presses Universitaires de Vincennes, 2004) and S. Duffy, ed., Virtual Mathematics:
7he Logic <ifDifference (Bolton: Clinamen Press, 2006). The latter compilation includes,
amid various articles dedicated to logic and mathematics in Deleuze, a posthumous
text of Chiltelet's (edited by Charles Alunni), 'Interlacing the Singularity, the Diagram
and the Metaphor'.

93
 SYNTHETIC PHILOSOPHY

its dialectical balances. The titles of the work's five chapters

are indicative of Chatelet's originality: 'The Enchantment
of the Virtual', 'The Screen, the Spectrum and the Pen­
dulum: Horizons ofAcceleration and Deceleration', 'The
Force of Ambiguity: Dialectical Balances', 'Grassmann's
Capture of the Extension: Geometry and Dialectic', and
'Electromagnetic Space'. Chatelet's array of examples
is concentrated in the modern period (Argand, Cauchy,
Poisson, Grassmann, Faraday, Maxwell, and Hamilton,
among others), but timeless interlacings recur as well
(Oresme, De Broglie) . In the introduction, Chatelet
quotes Andre Weil's lengthy explanation of the primordial
role that 'obscure analogies' play in mathematical investi­
gation - the threshold of creative penumbra that Chatelet
explores in approaching the 'gestures that inaugurate
dynasties of problems', the articulations and torsions
between reason and intuition, the 'rational capture of
allusions', and the structural and hierarchical deploy­
ment of the diagrams of thought. The fourth chapter is
something of a gem in the philosophy of mathematics.
Chatelet patiently reviews how Grassmann constructs
the 'synchronous emergence of the intuitive and the
discursive' in a living unity that is neither a priori nor a
posteriori, how the dialectic engenders new forms by way
of a careful hierarchy of scales in Grassmann's exterior
products, how Grassmann's very style leads to a natural
approach to the processes that enable the capture of

94
 A B I BLIOGRAPH ICAL SURVEY

self-reference ('comprehension of comprehension') and

how the apparent oppositions continuous/discrete and
equal/different consist influxes of mathematical inventive­
ness that serve to articulate its various, partial modes of
knowing (numbers, combinatorics, functions, extension
theory). Going further still, a magisterial thirty-page sec­
tion on Grassmann products explains, in vivid detail and
with the constant presence of diagrams, the great lines of
tension of Grassmann's system, which Chatelet explicates
in the first part of the chapter. The entire work constitutes
a major contribution to the philosophy of mathematics,
a contribution to which we will repeatedly return in the

third part of this study and that is, to our mind, the most
original work on the subject since Lautman's.

Frederic Patras

La pensee matMmatique contemporaine [Contemporary

mathematical thought] 6' provides an important leap for­
ward in the effort to approach contemporary mathematics.
The spectrum traversed is no longer the universe of set
theory - which, at the end of the day, is the customary
spectrum, for all of Badiou's originality and Maddy's
expertise - but includes genuinely mathematical aspects
(abstract algebra, algebraic geometry, topology, category

6I F. Patras, La pensee mathimatique contemporaine (Paris: PUF, 2001).

95
 OF CONTEMPORARY MATHE MATICS

theory) and incorporates the rise of modern mathematic:

(chapters I-4, with excellent introductions to Galois
to Dedekind's algebra and to the 'universal' Hilbert)
as well as aspects of the works of central figures of con.
temporary mathematics (chapters 5-8, on Bourbaki,
Lawvere, Grothendieck, Thom). Chapter 7, dedicated to
Grothendieck, is particularly valuable, owing to its sheer
singularity among treatises of mathematical philosophy.
It should be considered a monumental aberration that
a figure who, in all likelihood, is the most important
mathematician of the second half of the twentieth century
never seems to be seriously considered in 'mathematical
philosophy', and Patras seeks to put an end to this error.
The author shows that a comprehension efthe modes efemer­
gence ef mathematical creativity should constitute one ef the
indispensable tasks efmathematicalphilosophy, and indicates
that some of the great forces underlying Grothendieck's
work (aesthetic schematization, universal definition,
logical cleanliness, inventive 'innocence', 'listening' to
the 'voice of things', dialectical yin-yang) can help u s
understand the mathematical imagination as a form of
complex thought, in which multiple structural polarities
and bordering tensions interlace.

96
 A BIBLIOGRAPHICAL SURVEY

Da vid Garfield

From its polemical title onwards, Towards a Philosophy

efReal Mathematics aims to break the normative preju­
dices that the philosophy of mathematics makes use of,62
in particular the 'belief amongst philosophers to the
effect that the study of recent mainstream mathematics
is unnecessary'. 63 A lengthy introduction argues for the
value of a philosophical perspective oriented toward
nonelementary mathematics, and exhibits some of the
major problems that this approach encounters, but that
the 'foundationalist filter' still fails to detect: the status of
the structural borders of mathematics (beyond binarisms
and alternatives of the 'all-or-nothing' variety), the con­
nectivity of different mathematical theories, the evolution
of mathematical concepts, the contingency of mathematical
thought, and the progressive recursive richness of math­
ematical constructions. The subtitle of the introduc­
tion, 'A Role for History', indicates the path adopted by
Corfield - a junction of mathematics, philosophy and
history, in which current reflections on the discipline's
development take on a real relevance for the philosopher
of mathematics. Indeed, the text broaches various themes
from contemporary mathematics - automated proofs of

62 D. Corfield, Towards a Philosophy ofRealMathematics (Cambridge: Cambridge University

Press, 2003).
63 Ibid., Towards a Philosophy ofRealMathematics, 5.

97
 SYNTHETIC P H I LO SOPHY

theorems, modes of indeterminacy, theory of groupoi ds ,

n-categories - and elaborates an epistemological model
in which an intermingling of webs and hierarchies help s
to explicate the simultaneously multivalent and unitary
development of advanced mathematics. Chapters 2 and 3
deal with logical automata and serve to contrast the limits
of automatic proof with groundbreaking mathematical
creativity (chapter 4) , where the role of analogy turns o ut
to be indispensable for the invention of new concepts,
techniques and interpretations (with valuable examples
from Riemann, Dedekind, Weil, and Stone). Chapters 5
and 6 review problems of plausibility, uncertainty and
probability in mathematics (Bayesian theories) and in
science in general (quantum fields). Chapters 9 and 10
approach ongoing developments in mathematics (grou­
poids, n-categories) and the corresponding works of the
current investigating mathematicians (Brown, Baez),
concretely demonstrating how a mathematics can be
observed in utero from a philosophical point of view in
which certain traditional ontological and epistemological
obstacles have been dissolved. Chapters 7 and 8 focus
on the problem of the growth of mathematics (with an
appraisal and critique of Lakatos) , the importance of
opposed mathematical practices living together, and the
consequent necessity of not discarding from the philoso­
phy of mathematics the supposed residues of conceptions
of mathematics no longer in vogue.

98
 A B I BLIOGRAPHICAL SURVEY

Corfield tries to make the complex life of mathematics

heard (so that we may 'listen to the voice of things', as
Grothendieck would write in his Recoltes et semailles),
beyond which 'one can say with little fear of contradiction
that in today's philosophy of mathematics, it is the phi­
losophy that dictates the agenda'. According to Corfield,
a healthy inversion of perspectives, to the point where
a happy medium can be constructed, could help today's
philosophy of mathematics emulate the mental openness
of the great Russell by encouraging philosophers to:

L Believe that our current philosophy is not adequate

to make proper sense of contemporary mathematics;
2. Trust that some mathematicians can give us insight

into a better philosophical treatment;

3. Believe that the emerging picture will revitalize
philosophy. 64

Some of the examples studied by Corfield indicate how

fixing our attention on more mathematics (and not nece­
sarily more philosophy, as might narrowly be thought)
could help philosophy: the Hopf algebras at the heart of
the reasons for mathematics' applicability to quantum
physics, the groupoids that display novel interlacings
between symmetry (abstract equivalence) and asymmetry

64 Ibid., 270.

99
 OF CONTEMPORARY MATHE M ATICS

(noncommutativity), and the categorical languages o f

Makkai that eliminate poorly posed ontological ques­
tions. Altogether, the work supplies an interesting coun­
terweight to the dominant forces in the philosophy o f
mathematics, which are very attentive to language but far
removed from 'real' mathematics. The text concludes with
an important plea for today's philosophy of mathematics:
'Mathematics has been and remains a superb resource for
philosophers. Let's not waste it.'65

65 Ibid., 270.

100
 A BIBLIOGRAPHICAL SURVEY

2 . 3 M O R E P H I L O S O PHY, L E S S MAT H E MATI C S

We have indicated, in sections 2 . 1 and 2 . 2, how vari­

ous philosophers, mathematicians and historians have
approached advanced mathematics (in its three great
realms: classical, modern and contemporary) thereby
opening new perspectives for mathematical philosophy
that have been inexistent or 'effaced' from the point
of view of foundations or of elementary mathematics.
The pretension to exhaust the horizons of mathematical
philosophy with the 'fundamental' and the 'elementary',
and the unwillingness to see in modern and contemporary
mathematics an entire arsenal of problematics irreducible
to elementary examples or logical discussions (chapter 1),
has limited the reach of the traditional mathematical
philosophy inherited from analytic philosophy. Though
it has neglected the universe of advanced mathematics,
traditional mathematical philosophy has been able to pin­
point complex ontological and epistemological problems
(with respect to the notions of number, set and demon­
stration), which it has then treated with great precision.
A broad and current vision of traditional mathemati­
cal philosophy can be found in The Oiford Handbook ef
Philosophy ef Mathematics and Log;i.c.66 As we will see in
reviewing the text, the focus is clearly analytical, logical

66 Shapiro, Oiford Handbook. . .

101
 SYNTHETIC PHILOSOPHY

and Anglo-Saxon. Modern and contemporary mathemat­

ics, as we have defined these terms, and those who forged
modern and contemporary mathematics - Galois, Rie­
mann and Grothendieck, to name only the indispensable
figures - make minimal appearances or do not appear
at all.67 By contrast, another of the fundamental figures
of modern mathematics, Georg Cantor, is broadly stud­
ied throughout the volume, thus underscoring analytic
philosophers' interest in set theory. And so the range
of mathematics reflected upon in the volume is reduced
to a lattice of logics and classical set theory. This curi­
ous deformation of the mathematical spectrum, which
has been repeated for decades now in the Anglo-Saxon
world, should no longer be accepted. It would be another
matter if, with somewhat more humility, the volume in
question had been called The Oxford Handbook efAnalytic
Philosophy efLogic.
Starting with Shapiro's excellent general introduc­
tion (where he elaborates on certain remarks from his
earlier text, Thinking about Mathematics, mentioned in
our introduction), the compilation includes a review
of the philosophy of mathematics between Descartes
and Kant (Shabel); a chapter on empiricism and logical
67 The indexes (of both suhjects and proper names) at the end of the volume refer only
two pages (of the 833 in the volume) to Galois and Riemann; Grothendieck does not
even appear. Though the indexes are less than reliable (since, for example, in Steiner's
article, which concerns the problem of the applicability of mathematics, Riemann and
Galois are studied with greater patience), they are sufficiently indicative of the factual
situation.

102
 A BIBLI OGRAP H I CAL SURVEY

positivism (Skorupski); an introduction to Wittgenstein's

philosophy of logic and 'mathematics" (Floyd); three
chapters regarding versions of logicism (Demopoulos
and Clark, Hale and Wright, Rayo); one text on formal­
ism (D etlefsen); three chapters on forms of intuitionism
(Posy, McCarty, Cook); a text on Quine (Resnik); two
chapters on naturalism (Maddy, Weir); two chapters on
nominalism (Chihara, Rosen and Burgess); two chapters
on structuralism (Hellman, MacBride); one text on the
problem of the applicability of mathematics (Steiner), one
text on predicativity (Feferman); two chapters on logi­
cal consequence, models and constructibility (Shapiro,
Prawitz); two chapters on relevance logic (Tennant, Bur­
gess); and two chapters on higher-order logic (Shapiro,
Jane). All of the works demonstrate a high level of analysis,
extensive argumentative rigor and great professionalism.
Nevertheless, what seems to have been created here is an
extensive web of cross-references between the authors'
professional works and the stratum of logics linked to
those works: a secondary web that has been substituted for
the primary, underlying mathematics. Once this interest­
ing and complex web has been taken up - by means of
logical forms, problems associated with foundations,
detailed philosophical disquisitions and self-references
among specialists - very few of the authors included in
the handbook seem sufficiently self-critical to consider
that, perhaps, many other (possibly even more interesting

103
 OF CONTEMPORARY MATHEMATICS

and complex) forms o f mathematics have escaped their

attention. Of co.urse, we cannot (and should not) ask
the specialist to go beyond his field of knowledge, but
neither can we (nor should we) confuse the student or
professional interested in the topic, fooling him into think.
ing that the text covers the 'philosophy of mathematics
and logic' in its entirety. The disappearance efmathematics
and its supposed reducibility to logic make up the least
fortunate global perspective that Anglo-Saxon analytic
philosophy has (consciously or unconsciously) imposed
upon the philosophy of mathematics.
It seems surprising that, forty years after the publica­
tion of Benacerraf and P utnam's staple anthology Philoso·
phy efMathematics,68 the problems examined in Shapiro's
new compilation remain the same ones treated in the four
parts of the 1964 compilation: foundations, mathematical
objects, truth, and sets. The tools included in Shapiro's
compilation include a much broader and pluralistic web
of logics, as well as new unifying perspectives. But the
gigantic advances made by mathematics in the last fifty
years are dazzling in their absence. Again, it seems as if
mathematics has not evolved, as if the problems of the
philosophy of mathematics were fixed in time, leaving
room only for scholastic variations. We hope to show, in

68 P. Benacerraf & H. Putnam, eds., Philosophy efMathematics: Selected Readings, 2nd

edition (Cambridge: Cambridge University Press, 1984).

1 04
 A B I BLIOGRAPH ICAL SURVEY

the second and third parts of this essay, that the situation
we are dealing with here is unsustainable.
With respect to Benacerraf and Putnam's compilation,
Shapiro's opens perspectives onto two particular new
horizons for mathematical philosophy: naturalism and
structuralism. In her article, 'Three Forms of Naturalism',
Maddy explores the roots of naturalism in Quine, and
the later modifications of Quinean positions in Burgess
and in Maddy's own work. 69 Quine's self-referential natu­
ralist position, according to which the foundations of a
science and its fragments of certainty should be sought
in the science itself, and not in a first philosophy that
is external and alien to the science, provokes a robust
intramathematical perspective in Maddy, according to
which a naturalist philosopher of mathematics should
not slide into extramathematical metaphysical debates,
but must meticulously track the dynamics of concept
formation within her own discipline. Maddy has satisfied
this program with vigor and originality within set theory,
showing, in particular, that the supposedlyQuinean natu­
ralist position in favor of a reduced universe of sets (V=L)
receives no sympathy from the 'natural' arguments in
favor of large cardinals, conducted by the theory's chief
creators (Martin, Woodin and Shelah, among others).
Nevertheless, the 'mathematics' that the philosopher deals

69 Shapiro, Oxfrml Handbook... , 437-59.

105
 SYNTHETIC P H ILOSOPHY

with here is restricted, once again, to forms of logic and

set theory, without making any inroads into geometrical ,
algebraic or differential domains, and without coming
close to mentioning any of the Fields medalists (except
for Cohen, of course) who, presumably, have changed the
course of the discipline over the last fifty years.
In his 'Structuralism' article, Geoffrey Hellman pro­
poses four versions of a structural focus on mathemat­
ics (set-theoretical structuralism, generic structuralism,
categorical structuralism and modal structuralism), and
goes on to compare the advantages of each version with
respect to certain philosophical problems that arise within
the structural aspects themselves. These problems include
the following: the contrast between 'set' and 'structure'
and the choice of natural axiom-concepts; 2. the handling
of 'totalities'; the emergence of intractable 'ontologies­
epistemologies'; the handling of rigid and nonrigid struc­
tures from a philosophical perspective; the presence of
circularities in structures; the problems of theoretical
under-determination; the presence of primitive, undefined
conceptual substrata.7° Hellman's conclusions (carefully
delimited, in the style of all the authors of the handbook,
with sound lines of argumentation and with reference to
a minimum of mathematical cases) indicate that a mixture
efcategorical and modal structuralism could respond in the

1 06
 A BIBLIOGRAPHICAL SURVEY

best way possible to the problems confronted in the article.

We shall see, in part 3 of this book, how to construct and
significantly
extend that mixture, which is suggested by
Hellman and reclaimed by the extensive case studies that
we will undertake in the second part.
The analytic school of philosophy of mathemat­
ics, including, in particular, the great maj ority of its
Anglo-Saxon practitioners (with important exceptions,
of course71), could feel at home under the slogan, 'more

philosophy, less mathematics'. This has always been a per­

fectly valid option, but a restrictive one as well, no doubt.
The danger - which has always existed, and continues to
exist, and which Rota emphatically opposed - is that in
many academic environments, this option is the only one
available. Returning to behold, again, the complexity
of the mathematical world - as Lautman admirably suc­
ceeded in doing, along with many of the authors reviewed
in section 2.2, and which Corfield has again proposed as
an imperative - should reset the balance, and put forward
a new plan of greater equality: 'as much mathematics as
philosophy.' Part 2 of this work aims to cover the left side
of the balance; part 3, the right.

71 In addition to the authors mentioned in section 2.21 we could point out other Anglo�
Saxon philosophers and historians who try to cover a broad mathematical spectrum
(methodological, technical and creative), such as Jeremy Gray, Michael Hallett, Mark
Steiner and Jamie Tappenden, among others.

107
CHAPTE R 3

T OWARD A SYNT H E T I C P H I L O S O P HY O F

C O NT E M PO RARY MATHE MATI C S

In the introduction and preceding chapters we saw that a

contrasting (and often contradictory) multiplicity of points
ofview traverse the field of the philosophy of mathematics.
Also, we have delineated (as a first approximation, which
we will go on to refine throughout this work) at least five
characteristics that separate modem mathematics from
classical mathematics, and another five characteristics
that distinguish contemporary mathematics from modem
mathematics. In that attempt at a global conceptualiza­
tion of certain mathematical tendencies of well-defined
historical epochs, the immense variety of the technical
spectrum that had to be traversed was evident. Never­
theless, various reductionisms have sought to limit both
the philosophical multiplicity and mathematical variety
at stake. Far from one kind of omnivorous philosophical
wager, or one given reorganization of mathematics, which
we would then try to bring into a univalent correlation,
we seem to be fundamentally obliged to consider the
necessity of constructing multivalent correspondences
between philosophy and mathematics, or rather between
philosophies and mathematics in the plural.

109
 SYNTHETIC PHILOSOPHY

In a manner consistent with this situation, we will not

assume any a priori philosophical position until we have
carefully observed the contemporary mathematical land­
scape. We will, however, adopt a precise methodological
framework, which, we believe, will help us better observe
that landscape. Of course, that methodological schema­
tization wiU also influence our modes of knowing, but
we trust that the distortions can be controlled, since the
method of observation we will adopt and the spectrum we
presume to observe are sufficiently close to one another.
The philosophical and mathematical consciousness of
multiplicity at stake, in fact, requires a minimal instru­
mentarium that is particularly sensitive to the transit
of the multiple, that can adequately take stock of that
multiplicity, and that allows us to understand its processes
of translation and transformation. To those ends, we
will adopt certain minimal epistemological guidelines,
furnished in philosophy by Peirce's pragmatism, and in
mathematics, by category theory.
A vision moderately congruent with the multiformity
of the world should integrate at least three orders of
approximations: a diagrammatic level (schematic and
reticular) where the skeletons of the many correlations
between phenomena are sketched out; a modal level
(gradual and mixed) where the relational skeletons acquire
the various 'hues' of time, place and interpretation; and
a.frontier level (continuous) where webs and mixtures are

1 10
 TOWARD A SYNTHETIC PHILOSOPHY

progressively combined. In this 'architecture' ofvision, the

levels are never fixed or completely determined; various
contextual saturations (in Lautman's sense) articulate
themselves here (since something mixed and saturated
on a given level may be seen as skeletal and in the process
of saturation in another, more complex context) and a
dynamic frontier of knowledge reflects the undulating
frontier of the world. An adequate integration of dia­
grams, correlations, modalities, contexts and frontiers
between the world and its various interpretants is the
primordial object ofpragmatics. Far from being the mere
study of utilitarian correlations in practical contexts of
action-reaction (a degeneration of the term 'pragmat­
ics' that corresponds to the disparaging way in which
it gets used these days), pragmatics aims to reintegrate
the differential fibers of the world, explicitly inserting
the broad relational and modal spectrum of fibers into
the investigation as a whole. The technical attention to
contextualizations, modulations and frontiers affords
pragmatics - in the sense which Peirce, its founder, gave
it - a fine and peculiar methodological timbre. Just as
vision, like music, benefits from an integral modulation
through which one interlaces tones and tonalities so as to
create a texture, so pragmatics benefits from an attentive
examination of the contaminations and osmoses between
categories and frontiers of knowledge so as to articulate
the diversity coherently.

111
 O F CONTEMPO RARY MATHEMATICS

Various natural obstructions are encountered on the way

to any architectonic system of vision seeking to reintegrate
the Many and the One without losing the multivalent
richness of the differential. One obvious obstruction
is the impossibility of such a system's being stable and
definitive, since no given perspective can capture all the
rest. For, from a logical point of view, whenever a system
observes itself (a necessary operation if it seeks to capture
the 'whole' that includes it), it unleashes a self-referential
dynamic that ceaselessly hierarchizes the universe. As
such, a pragmatic architectonic of vision can only be
asymptotic, in a very specific sense interlacing evolution,
approximation and convergence, but without requiring
a possibly nonexistent limit. An 'internal' accumulation
of neighborhoods can indicate an orientation without
having to invoke an 'external' entity that would represent
a supposed 'end point' - it has thepower to orient ounelves
within the relative without needing to have recourse to the
absolute. This fact harbors enormous consequences, whose
full creative and pedagogical force is just beginning to
be appreciated in the contemporary world.
The maxim of pragmatism - or 'pragmaticism' (a
name 'ugly enough to escape the plagiarists'?•) as Peirce
would later name it in order to distinguish it from other
behaviorist, utilitarian and psychologistic interpretations

72 C. S. Peirce, Collected Papen (Harvard: Harvard University Press, 1931-1958),

vol. 5, 415.

112
 TOWARD A SYNTHETIC P H ILOSOPHY

- appears to have been formulated several times through­

out the intellectual development of the multifaceted North
American sage. The statement usually cited is from 1878,
but other more precise statements appear in 1903 and 1905:

Consider what effects that might conceivably have

practical bearings, we conceive the object of our
conception to have. Then our conception of those
effects is the whole of our conception of the object.
(1878)73

Pragmatism is the principle that every theoretical

judgment expressible in a sentence in the indicative
mood is a confused form of thought whose only mean­
ing, if it has any, lies in its tendency to enforce a corre­
sponding practical maxim expressible as a conditional
sentence having its apodosis in the imperative mood.
(1903)74

The entire intellectual purport of any symbol

consists in the total of all general modes of ratio­
nal conduct, which, conditionally upon all the
possible different circumstances and desires,

73 Peirce, 'How to Make Our Ideas Clear' (1878), in Collected Papers, vol. 5, 402.
74 Peirce, 'Harvard Lectures on Pragmatism' (1903), in Collected Papers, vol. 5, 18.

113
 SYNTHETIC PHILOSOPHY

would ensue upon the acceptance of the symbol.

(1905)75

What is emphasized in the i905 statement is that we come

to know symbols according to certain 'general modes',
and by traversing a spectrum of ' different possible circum­
stances'. This modalization of the maxim (underscored in
the awkward repetition of 'conception' in i878) introduces
into the Peircean system the problematic of the 'interlac­
ings' between the possible contexts of interpretation that
may obtain for a given symbol. In the i903 statement, on
the one hand, we see that every practical maxim should be
able to be expressed in the form of a conditional whose
necessary consequent should be adequately contrastable,
and on the other hand, that any indicative theoretical
judgment, in the actual, can be specified only through a
series of diverse practices associated with that judgment.
Expanding these precepts to the general field of semi­
otics, to know a given sign (the realm of the actual) we
must traverse the multiple contexts of interpretation
capable of interpreting that sign (the realm of the pos­
sible) and, in each context, study the practical imperative
consequences associated with each one of those interpre­
tations (the realm of the necessary) . Within that general
landscape, the incessant and concrete transit between the

75 Peirce, 'Issues of Pragmaticism' ( 1905), in Collected Papers, vol. 5, 438.

114
 TOWARD A SYNTHETIC PHILOSOPHY

p ossible, the actual and the necessary turns out to be

o ne of the specificities of mathematical thought, as we
will repeatedly underscore throughout this work. In that
transit, the relations between possible contexts (situated
in a global space) and the relations between the fragments
of necessary contradistinction (situated in a local space)
take on a primordial relevance - something that, of
course, finds itself in perfect tune with the conceptual
importance of the logic of relations that Peirce himself
systematized. In this way, the pragmaticist maxim indi­
cates that knowledge, seen as a logico-semiotic process,
is preeminently contextual (versus absolute), relational
(versus substantial), modal (versus determined), and
synthetic (versus analytic).
The maxim filters the world through three complex
webs that allow us to differentiate the One into the Many
and, inversely, integrate the Many into the One: the
aforementioned modal web, a representational web and a
relational web. In effect, besides opening onto the world
of possibilities, the signs of the world should, above all,
be representable in the languages (linguistic or diagram­
matic) utilized by communities of interpretants. The
problems of representation (fidelity, distance, reflexiv­
ity, partiality, etc.) are thus immediately bound up with
the differentiation of the One and the Many: the reading
of an identical fact, or an identical concept, dispersed
through many languages, through many 'general modes'

115
 O F C ONTEMPORARY MATHE MATICS

of utilizing information, and through many rules for the

organization and stratification of information.
One of the strengths of Peircean pragmatism, and in
particular of the fully modalized pragmaticist maxim, is
that it allows us to once again reintegrate the Multiple into
the One through the third web that it puts in play: the rela­
tional web. In fact, after decomposing a sign into subfrag­
ments in the various possible contexts of interpretation,
the correlations between fragments give rise to new forms
of knowledge that were buried in the first perception of
the sign. The pragmatic dimension emphasizes the coali­
tion of some possible correlations, discovering analogies
and transfusions between structural strata that, prior to
effecting that differentiation, had not been discovered.
In this way, though the maxim detects the fundamental
importance of local interpretations, it also insists on the
reconstruction of global approximations by means of
adequate gluings of the local. We shall later see how the
tools of the mathematical theory ofcategories endow these
first vague and general ideas with great technical preci­
sion. The pragmaticist maxim will then emerge as a sort
of abstract differential and integral calculus, which we will
be able to apply to the general theory of representation,
that is to say - to logic and semiotics in these sciences'
most generic sense, the sense foreseen by Peirce.
Overleaf we present a diagrammatic schematization
of the pragmaticist maxim, in which we synthetically

116
 TOWARD A SYNTHETIC PHILOSOPHY

condense the preceding remarks. This diagram (figure 4)

will be indispensable for naturally capturing the maxim's
structuration from the perspective of mathematical cat­
egory theory. Reading from left to right, the diagram
displays an actual sign, multiply represented (that is,
underdetermined) in possible contexts of interpretation,
and whose necessary actions-reactions in each context
yield its partial comprehensions. The terms 'pragmatic
differentials' and 'modulations' evoke the first process
of differentiation; the latter term reminds us of how a
single motif can be extensively altered over the course
of a musical composition's development. The process of
reintegration proper to Peircean pragmatics is evoked by
the terms 'pragmatic integral' and 'correlations', 'gluings',
'transferences', which remind us of the desire to return
that which has been fragmented to a state of unity. The
pragmatic dimension seeks the coalition of all possible
contexts and the integration of all the differential modu­
lations obtaining in each context, a synthetic effort that
has constituted the fundamental task of model theory
and category theory in contemporary logic.

117
 SYNTHETIC PHILOSOPHY

underdeterminations
of the sign

representation t
context i

sign
1 correlations

,[ gluings
l transferences

(ACTUAL) ' J
pragmatic

integral

Figure 4. Sketch efPeirce's pragmaticist maxim.

The pragmaticist maxim thus serves as a sophisticated

'sheaf of filters' for the decantation of reality. The crucial
role of the sheaf secures an amplified multiplicity of
perspectives which, for that matter, filters information
in more ways than one, thus establishing.from the outset a

118
 TOWARD A SYNTHETIC P H ILOSOPHY

certain plausibility for the claim that knowledge may be

sufficiently rich and multivalent. The Peircean pragmati­
cist maxim may come to play an extraordinarily useful
role in the philosophy of contemporary mathematics. Its
first upshot consists in notprivileging any point of view or
any fragment of language over another, thereby opening
the possibility of considering what Susan Haack called
'rival, incompatible truths', without supposing them to
be 'reducible to a privileged class of truths in a privileged
vocabulary' .76 In a pendular fashion, the second crucial
strength of the maxim consists in the power to compare
very diverse levels within the multiplicity of perspectives,
languages and contexts of truth that it has succeeded in
opening up. Indeed, the unwillingness to restrictively
assume any 'privileged 'foundation' does not oblige us
to adopt an extreme relativism without any hierarchies
of value. In questions of foundations, for example, not
reducing things to discussions referring to the supposedly
'absolute' base ZF or the dominant force of first-order
classical logic, but opening them instead to discussions
in other deductive .fragments ('reverse mathematics'),
semantic.fragments (abstract model theory) or structural
.fragments (category theory) ofZF, is a strategy that broad·
ens the contexts of contradistinction - and, therefore,

76 S. Haack, C. Bo, 'The Intellectual Journey of an Eminent Logician-Philosopher', in

C. de Waal, ed., Susan Haack: A Lady cfDistinctions (Amherst, NY: Prometheus Books,
2007), 27.

119
 O F CONTEMPORARY MATHEMATIC S

the mathematical richness at stake - without, fo r all that,

spilling into epistemological disorder. Our contention
is just the contrary: it is in virtue efbeing able to escape
various 'ultimate' foundations, and in virtue ef situating
ourselves in a relativefabric efcontradistinctwns, obstructions,
residues and gluings, that a genuine epistemological order
for mathematics asymptotically arises and evolves.
The Peircean pragmaticist maxim can be seen as a
sophisticated way of weaving between analysis/differ­
entiation and synthesis/integration. The contemporary
world requires new conceptual instruments of 'collation'
or 'gluing' (which respond with new arguments to the
'primordial' differentiation/integration dialecticn), and, to
a large extent, the Peircean pragmaticist maxim provides
one of those gluing instruments. As we understand it, if
there is any mathematical concept capable of serving as
a threshold between modern and contemporary math­
ematics, it is that of a mathematical sheaf, which is indis­
pensable for reintegrating adequate local compatibilities
into a global gluing.78 Correlatively, within the scope of
epistemology, we believe that the Peircean pragmaticist

77 A good presentation of the analysis/synthesis polarity and its subsumption into the
'greater' differentiation/integration polarity, can be found in Gerald Holton's article,
'Analisi/sintesi', for the Enciclopedia Einaudi (Torino: Einaudi, i977), vol. t, 49-522.

78 We will study, in detail, the multiple facets ofsheaves in topology, algebraic geometry
and logic in the second part of this work. The mobile plasticity of sheaves not only
lets us pass from the local to the global, but, in a natural fashion, allows for multiple
osmoses between very diverse subfields of mathematics. In a certain way, since their
very genesis, sheaves have acquired an incisive reflexive richness that has rendered them
extraordinarily malleable.

1 20
 TOWARD A SYNTHETIC PHILOSOPHY

maxim can serve as a remarkable methodological sheaf to

compare and then interlace the diverse. The Peircean web
of webs, in effect, opens onto the modal realms in their
entirety, and systematically attends to contrasting given
facts (within the phenomenological world) and necessary
behaviors (within well-defined contextual systems), so
as to then reintegrate them in an extended spectrum of
possible signs.
Differentiation and reintegration reach a high degree
of methodological precision in the mathematical theory
of categories. As a counterpart to the set-theoretical ana­
lytic championed by Cantor's heirs,79 category theory no
longer dissects objects from within and analyses them in
terms of their elements, but goes on to elaborate synthetic
approaches by which objects are studied through their
external behavior, in correlation with their ambient milieu.
Categorical objects cease to be treated analytically and are
conceived as 'black boxes' (with local Zermelo-Von Neu­
mann elements becoming invisible, while global elements
emerge) . Their movement through variable contexts is
observed by means of the significant accumulative effort
of synthetic characterizations. Within certain classes of
structures (logical, algebraic, ordered, topological, dif­
ferentiable, etc.), category theory detects general synthetic
79 Cantor himself is better situated in terms of a sort of general organicism (with
considerable and surprising hopes that his akphs would help us understand both the
living realm and the world around us), where analytic and synthetic considerations
relate to one another.

121
 SYNTHETIC PHI LOSOPHY

invariants and defines them by means of certain 'universal

properties'. Those properties , in the first instance, hold for
given universes of classes of structures (= 'concrete' cat­
egories), but can often be extended to more general fields
in which the minimal generic properties of those classes
are axiomatized (= 'abstract' categories). Between abstract
and concrete categories, multiple weaves of information
(= 'functors') are then established. An incessant process of
differentiation diversifies the universal constructions given
in abstract categories and, in contrasting forms, 'incar­
nates' them in multiple concrete categories. Inversely - in
a pendular fashion, we might say - an incessant process
of integration seeks out common constructors and roots,
at the level of abstract categories, for a great variety of the
special constructions showing up in concrete categories.
In this way, a quadruple synthetic strategy takes shape
in category theory. First of all, internally, in each concrete
category, we seek to characterize certain special construc­
tions in terms of their environmental properties in the
given class. Then, externally, in the general field of abstract
categories, we seek out certain universal constructions
that can account for the characterizations obtained in the
concrete categories. In the third stage, in a remarkable
weaving between concrete and abstract categories , we
go on to define adequate functors of differentiation and
reintegration. Finally, the same functors become the object
of investigation from a synthetic point of view, and their

1 22
 TOWARD A SYNTHETIC P H I LOSOPHY

osmoses and obstructions ( = 'natural transformations')

are studied systematically. Category theory, as we will see
in chapters 4-7, has acquired considerable mathematical
value in its own right, but for the moment we are inter­
ested only in accentuating its methodological interest for
a philosophy of mathematics open to incessant pendular
processes of differentiation and reintegration.
Indeed, ifthe philosophy ef mathematics could make use
ef the synthetic lessons on differentiation and reintegration
codified in both the Peircean prag;maticist maxim and in the
junctorial processes ef category theory, many of the funda­
mental problems in philosophy of mathematics might
acquire new glints and twists that, we believe, could enrich
philosophical dialogue. The objective of this essay's third
part will be ptecisely to discuss those problems, in light of
the contributions of contemporary mathematics, and in
light of a synthetic grafting of the Peircean pragmaticist
maxim onto the methodological lineaments of category
theory. However, in posing the same problematics from
the complementary perspectives of analysis and synthesis,
we can already indicate certain fundamental inversions
(see figure 5, overleaf) in the demands forced upon us by
analytic and synthetic perspectives.

123
 AN'' v 1 u· \/1.o,; 1 0 N SYN H l l·. l l C V I S I O N
(P l f l LO."JOJ'HY OF ( PRAGMATIC I S T
PROBl�f:MATIC
LANG UAca: + MAXIM +
Sf:T-T I I E O RE·1·1CAL FOUNDATIONS ) CATEGORY-TH EOR£TIC CONTEXTS )
rea!i1t ontology: mathmwtica! obja'fs exist in m u s t postulate the real existence of the must postulate the existence of a covering
a real world universe of sets, to which we are granted of the real by means of progressive
access by a reliable form of intuition hierarchies of structural contexts that
asymptotically approximate it
idealist ontology: mathematical objects are must postulate a dissociation between must postulate a dissociation between
linguistic subterfuges mathematical constructs and their classes of linguistic categories and classes
physical environments of categories from mathematical physics
realist epistemology: truth values reflect must postulate the existence of a set· must postulate the existence of categorical
objectiveforms efknowledge theoretic semantics as an adequate semantic adjunctions and invariant skeletons
transposition of semantic correlations in persisting through functorial weavings
the real world
idealist epistemology: truth values are must postulate a variability of must postulate the impossibility of
subjectivefonns efcontrol modalization of sets, and assume the 'archetypical' initial categories capable of
existence of stable transitions between generically classifying the truths of their
'compossible' worlds derived categories
realist metaphysics: must postulate the existence of a 'monstrous' must postulate the existence of multiple
to ti en einai ('the essential qfessence') exists model and reflexion schemas that would classifier toposes and additional inverse limits
mathematically accommodate every universe of sets where the classifiers can be 'glued together'
idealist metaphysics: must postulate the necessity of towers must postulate the necessity of functorial
to ti en einai does not exist mathematically of set-theoretical universes that can iterations ad infinitum, irreducible to
be controlled only through relative projections from a supposedly 'final'
consistencies classifier
 TOWARD A SYNTHETIC P H ILOSOPHY

Figure 5 (facing page). Complementary perspectives rm the 'pure' setting out ef

problematics in thephilosophy efmathematics.

As we shall see and discuss in part 3, some of the above

requirements seem too strong and go against the grain of
various advances achieved in contemporary mathematics.
For example, from a synthetic point of view - which is
better suited than an analytic one to mathematical practice
- an idealist ontology that dissociates linguistic categories
(a la Lambek) from categories of mathematical physics
(a la Lawvere) seems inviable, because it places itself in
immediate contradiction with advances in n-categories (a
la Baez) that allow us simultaneously to account for com­
plex torsors in linguistics and physics. Another example,
again with a synthetic focus, seems to show that an idealist
epistemology is similarly inviable, since it would conflict
with the (already actualized) possibility of constructing
classifier topoi and initial allegories (a la Freyd). In this
manner, it is thus easy to see how our alternative, double
strategy - to make use of 'synthetic' methodologicalfoci and
work closely with contemporary mathematics - can bear con­
siderable philosophical fruit. We hope to show, further
on, that directing our attention to more mathematics (and
not necessarily 'more philosophy') represents a reasonable
strategy, and one that opens attractive and unexpected
channels for philosophical dialogue.

125
 SYNTHETIC P H I LOSOPHY

The Peircean pragmaticist maxim and the methodological

lineaments of category. theory help to provide a vision of
mathematical practice that is faller and morefaitliful than
what an analytic vision offers. The reasons are varied and
have to do with the meanings ordinarily given to the terms
'full' and 'faithful' - provided we extend their scope in the
direction of the meanings they take on in the technical con·
text of category-theoretic functors. Observing that every
construction that is realized in a given mathematical envi­
ronment (topological, algebraic, geometrical, differential,
logical, etc.) is necessarily local in an adequate context, 80
we will call a context in which the construction can be
locally realized, but which does not, in addition, invoke
redundant global axioms, a minimal context efadequation.
When a vision of a determinate mathematical environment
allows us to associate a minimal context of adequation with
every mathematical construction in the environment, we
will say that the vision is.full. We will say that a vision of a
mathematical environment isfaitliful when it allows us to
reconstruct every mathematical construction of the envi­
ronment in a minimal context of adequation. The.fullness
of the vision ensures that the local richness of the theories
will not be diluted in a global magma; thefaithfulness of
80 The context of adequation can be very large: If the mathematical construction is, for
example, the cumulative universe of sets, a context rendering the cumulative hierarchy
local will have to reach some inaccessible cardinal or other. Nevertheless, the majority
of 'real' mathematical constructions (in Hardy's sense, taken up again by Corfield )
live in mathematical contexts that are under far greater control, with respect to both
cardinal and structural requirements.

126
 TOWARD A SYNTHETIC PHILOSOPHY

the vision ensures that the local richness i s really sufficient

for its full development. For example, the usual analytic
vision of mathematics - based on ZF set theory and its
underlying first-order classical logic - turns out to be
neitherfull norfaithful in this sense. Given the broad global
reach of the ZF axiomatic, the vision is not full, precisely
because the minimal contexts of adequation are forcibly
lost ( a situation of information loss - that is to say, a loss of
fullness - for which reverse mathematics proposes a pal­
liative) ; nor is it faithful, since most of the constructions
are realized by means of an uncontrolled invocation of the
full force of the axioms.
By contrast, mathematical practice turns out to be
much closer to a vision that genuinely and persistently
seeks to detect, between minimal contexts of adequation,
both transferences and obstructions alike. The notions of
obstruction and residue are fundamental here, since the
incessant survey of obstructions, and the reconstruction
of entire maps of mathematics on the basis of certain
residues attached to those obstructions, is part and parcel
of both mathematical inventiveness and its subsequent
demonstrative regulation.81 Now, the obstructions and

81 Riemann's C(s) function provides an exemplary case, here. From its very definition
(by analytic extension, surrounding its singularities in the line Re{s)=l), to its still
mysterious applicability in number theory (clustered around the proof that the zeros
of the Z·function lie on the line Re(s)=l/2), the C e.>:tendr its domains of invention and
-
proof in virtue efthe obstructions as much definitional as structural - on which mixed
constructions of great draught are dashed (here, the C function as a 'hinge' between
number theory, complex analysis and algebraic geometry).

127
 O F CONTEMPORARY MATHEMATICS

residues acquire meaning only locally, with respect to

certain contexts of adequation - something of which the
usual analytic vision often loses sight, and of which, by
contrast, a synthetic vision helps us take stock. As we saw
in the m ap of the Peircean pragmaticist maxim (figure
' '

4, p. 118), we are also dealing with a situation that is

particularly susceptible to being detected by the maxim,
insofar as the latter attends to local differentials and
contextual singularities, no less than to the subsequent
modal reintegration of local fractures.
A syntheticphilosophy efcontemporary mathematics mu s t
therefore seek to capture at least the following minimal
characteristics that naturally arise in a sort of generic 'dif­
ferential and integral methodology', in which mathemat­
ics, philosophy and history are interlaced:

l. a contextual and relational delimitation of the field

of contemporary mathematics with respect to the
fields of modern and classical mathematics;

2. a differentiation of the plural interlacings between

mathematics and philosophy, followed by a rein­
tegration of those distinctions in partial, unitary
perspectives;

3. a presentation of a full and faithful vision of math­

ematical practice, particularly sensitive to a pendular

128
 TOWARD A SYNTHETIC P H I LOSOPHY

weaving between transferences and obstructions, and

between smoothings and residues;

4-· a diagramming of the multivalences, ramifications

and twistings between spectra of mathematical theo·

rems and spectra of philosophical interpretations.

In what follows, we will take up precise case studies in

contemporary mathematics, by means of which we will
be able to repeatedly emphasize these four points, before
returning, in part 3, to additional 'skeletal'82 consider­
ations concerning the philosophy of mathematics.

82 In our strategy, one can observe an approximate analog to the practice in category
theory whereby, firstly, a category is delimited from other neighboring categories,
secondly, various concrete constructions of the category are studied in detail, and
thirdly, its skeleton and its free constructions are finally characterized. The three parts
of our work correspond - by an analogy that is not overly stretched - to the study of
the 'category' ofcontemporary mathematics and its various adjunctions with respect
to the various 'categories' of philosophical interpretations.

129
PART TWO

Case Studies
CHAPTER 4

GROTH E N D I E C K : F O R M S OF H I G H

MAT H EMATICAL C R EATIVITY

In this second part, we present briefcase studies from the

landscape of contemporary mathematics (1950-2000).
Our strategy will consist in providing mathematical infor­
mation (primarily conceptual and, to a lesser extent, tech­
nical information) that is usually taken to fall outside the
scope of philosophy - information whose philosophical
distillation and discussion will occupy us in part 3. Nev­
ertheless, although the primary objective of this second
part is to expand the concrete mathematical culture of the
reader, we will also go on to indicate and briefly discuss
a few generic lines of tension, both methodological and
creative, that a complete philosophical understanding
of mathematics will have to confront. Contemporary
mathematics has given rise to new forms of transit in
knowledge, which, in turn, generate new philosophical
problems, and new partial solutions of those problems.

133
 SYNTHETIC P HI LOSOPHY

4 . 1 GROTH E N D I E C K ' S L I F E AN D WORK: A BROA D

OUTLI N E

Alexander Grothendieck was born i n Berlin (1928),

where, during his early childhood (1933-39), he is edu­
cated under the care of a Lutheran minister (Heydorn) ,
while his parents actively dedicate themselves to political
agitation. His father, Alexander Shapiro, Russian anar­
chist, radical in Germany and France during the twen­
ties and thirties, Brigadista in the Spanish Civil War, is
murdered in Auschwitz. His mother, Hanka Grothendieck
- journalist for left-wing magazines, Shapiro's comrade
in France and Spain - is reunited with her son after the
defeat of the Spanish Republic. 83 Between 1940 and 1942,
Alexander and his mother are interned in the Rieucros
Concentration Camp, which he will later able to leave
for Le Chambon in order to be placed under the care of
another protestant pastor (Trocme) until the end of the
war. Reunited with his mother, Alexander completes his
degree in mathematics at the University of Montpellier,
where several of his teachers remark on his 'extraordinary
ability, unsettled by suffering'. It is then that the brilliant
young man, unhappy with the calculus being taught
to him at the university, proposes a complete theory of
83 '
A good overview of Grothendieck's life can be found in A. Jackson, Comme AppeU
-
du Neant As if Summoned from the Void: The Life of Alexander Grothendieck',
Notices ofthe AMS 51 No. 4, 10, 2004: 1037-56, n96-1212. A forthcoming biography of
Grothendieck, by Colin McLarty, should begin to fill in an inexcusable lacuna.

1 34
 GROTHE NDIECK

integration which, although he doesn't know it, turns out

to be equivalent to Lebesgue's theory.
From that day forward, it is the incessant making
of mathematics, rather than its study, that occupies
Grothendieck. 84 He is initiated into higher mathematics
while participating (in 1948) in the Cartan Seminar at the
Ecole Normale Superieure, completes his doctoral thesis85
under Dieudonne and Schwartz at Nancy between 1949
and 1953, and then visits America (Sao Paolo, 1953-54;
Kansas, 1955), where he becomes a well-known specialist
in topological vector spaces. 86 He then invents K-theory in
1957 and proposes a profound generalization of the Rie­
mann-Roch Theorem, with important consequences for
the mathematics of the late fifties and early sixties. (We will
expand on K-theory and the Riemann-Roch-Grothendieck

84 There is a famous anecdote about a visitor who had been to the Institut des Hautes
Etudes Scientifiques (!HES), (created for Grothendieck in the sixties) and had been
struck by the poverty of the library at such a Mecca of mathematics. Grothendieck
answered him, 'We don't read mathematics, here; we make mathematics.'

85 According to Dieudonne - an expert on analysis, if there ever was one - Grothendieck's

thesis could only be compared, in the field of topological vector spaces, with the works
of Banach.

86 Nuclear spaces, introduced by Grothendieck in his doctoral thesis, are topological

vector spaces defined by families of seminorm.s with a telescopic property (every
unit ball of a seminorm can be embedded in the balls of the remaining seminorms
by means of adequate multiplications). What is at stake here are spaces that capture,
in a natural way, important families of functions in complex and differential analysis
(entire holomorphic functions, smooth functions over compact differential varieties),
and that trace, in the infinite, certain good properties of finite-dimensional spaces. The
treatments of those properties by means of tensorial products begins to concretize
some of Grothendieck's later grand strategies: to study the properties of an object by
inserting it in a class (category) of similar objects; to construct transmitters of information
for the properties of an objectj to cmnpare similar behaviors in other categories, and
reutilize all of the pendular information accumulated in order to capture the initial
object in a new light.

135
 O F CONTEMPORARY MATHEMATIC S

theorem i n chapter 6, a s we approach Atiyah's work.) I n

1957 he also publishes his treatise-article, 'On a Few Points
of Homological Algebra' (on which we shall comment
in section 4.3), where he presents his program for the
renovation of algebraic geometry. 87
In the sixties, the IHES, with Grothendieck at the
helm, becomes the world's leading center for mathematical
inquiry. What ensues is a decade of creation, on the basis
of his central, driving ideas - schemes, topoi, motifs - with
the production of the two great series of writings that
would completely renovate the mathematics of the age:
the Elements efAlgebraic Geometry (EAG)88 and the Seminar
on Algebraic Geometry (SAG). 89 Grothendieck receives the
Fields Medal in i966, and in the panorama of subsequent
medalists, his spectrum of influence is enormous (figure 6).
Although he surprisingly retires from the mathemati­
cal world in 1970 (at 42 years of age!), after having left
behind a body of work that whole schools of mathemati­
cians would be hard-pressed to produce in a century, he
goes on to write great mathematical manuscripts9° and

87 A. Grothendieck, 'Sur quelques points d'algebre homologique', TolwkuMath.]ourna/

9, 1957: u9-221. The article is usually known as 'Tohoku', after the periodical in which
it was published.
88 A. Grothendieck (edited in collaboration with J. Dieudonne), llhnenJs de Geomitrie
Alg!brique, 4 vols., 8 parts (Paris: IHES, i96o-67).
89 A. Grothendieck et al., Seminaire de GeomitrieAlgtbriquedu Bois-Marie, 7 vols., 12 parts
(Berlin: Springer, 1970-3), original mimeographed fascicles, 1960--9.
90 La /ongue marche ti traverr la 1Morie de Galois (The Long March through Galois Theory],
1981, 1,8oo pages. Esquisse d'un programme �Sketch for a Program], 1983, a sort of
mathematical testament, 50 pages. us dirivateurr [Derivators], 1990, 2,000 pages.

136
 GROTHENDIECK

interminable (self-)critical reflections91 on the world (both

mathematical and theological). In sum, the body of work
left behind by Grothendieck is gigantic, both in terms
of its depth (the mathematics of the period 1970-2000,
particularly the Fields panorama, can to a large extent
be seen as a sort of 'commentary' on Grothendieck) and
its quantity (about ten thousand manuscript pages).
What we are dealing with here is a genuine gold mine
for commentators and philosophers, who have barely
begun to approach it.92

Gmi� (1966) __...... Deligne (1966) __...... Faltings (1966)

�
Atiyah (1966) - Connes (1982) �
Drinfeld (1990)
/ Kontsevich (1998)
Voevodsky (2002)
/
Perelman ( 2006)

Figure 6. Grothendieck's lines efir!fluence in thepanorama efFields medalists.

91 R£coltes et Semailles (Reaping and Sowing] , 1985-6, i,ooo pages. La clefdes songes [The
Key to Dreams] , 315 pages.
92 Several digitized fragments of Grothendieck's work, accompanied by a few studies,
can be found on the website http://www.grothendieckcircle.org, maintained by Leila
Schneps and Pierre Lochak.

137
 SYNTHETIC PHILOSOPHY

After the remarkable fifties (nuclear spaces, K-theory,

homological algebra) , Grothendieck's first great driving
_
idea during the golden age of the IRES would propel
a profound renovation of algebraic geometry. Situating
himself within what Thom would later call the 'found ­
ing aporia of mathematics'93 (that is to say, within the
irresolvably contradictory dialectic, discrete/continuous) ,
Grothendieck invents his schemes as a very powerful tool
in an attempt to resolve the Weil Conjectures {1949). A
precise grafting of the discrete and the continuous, the
conjectures seek out a way to measure the number of
points in certain algebraic varieties over finite fields, by
means of certain generating functions, such as the zeta
functions originating in Riemann's continuous, complex­
topological intuition.94 Dwork ( 1960) demonstrated the
rationality of the zeta functions, Grothendieck {1966)
the functional equation that governs them, and Deligne
(1974), Grothendieck's greatest student, the adequate
distribution of their zeros (which gives us combinato­
rial control of the points in a variety). Deligne's result,
which won him the Fields Medal, is a genuine technical
tour deforce.
Modern mathematics, in the first half of the twentieth
century, culminates in Weil's astonishing exploration.

93 R. Thom, 'L'aporia fondatrice delle matematiche', Enciclopedia Einaudi (Torino:

Einaudi, 1982), u33-46.

94 A. Weil, 'Numbers ofsolutions ofequations in finite fields',Bui.Am.Math.Soc. 55, 1949: 497-5o8·

138
 GROTH ENDIECK

Driven by a very subtle and concrete intuition, a s well

as an unusual capacity for uncovering analogies at the
crossroads between algebraic varieties and topology, Weil
succeeded in stating his conjectures with great precision.
Contemporary mathematics, in the second half of the

twentieth century, emerges with the work of Grothendieck,

giving rise to the entire apparatus of algebraic geometry

that will allow it, in turn, to resolve those conjectures.

While Zariski's topologies serve as mediations at the
algebraic varieties/topologies crossroads, and allow the
conjectures to be stated, the ('etale', l-adic) cohomologies
of Grothendieck and his school serve as mediations at the
schemes/topos crossroads, by means of which they can
now be resolved. Extending the algebraic varieties into
the field of schemes, the richness ofGrothendieck's generic
invention is not the outcome of gratuitous generaliza­
tions. At no point is generalization carried out without
adequate particularizations in mind, and what is really
at stake is a complex process of ascent and descent that,
as we shall see in greater detail in section 4.2, turns out
to be constantly governed by concrete consequences of
ever greater mathematical significance.
Indeed, with the creation of his schemes, Grothendi­
eck grafts together two of the major currents of modern
mathematics: the vision ef Riemann, which allows us to
understand a curve X by means of the ring M(X) of the
meromorphic functions over that curve, and the vision

139
 O F CONTEMPORARY MATHEMATICS

ef Galois-Dedekind, which allows us to understand an

algebraic variety V by-means of the spectrum Spec(V) of
its maximal ideals. In effect, Grothendieck generalizes
the situation so as to be able to envelope both visions at
once, and suggests that we might understand an arbitrary
(commutative, unitary) ring by means of a hierarchy of
three extremely mathematically rich objects: the spectrum
of its prime ideals, the Zariski topology over the spectrum
of primes, and the natural sheaves over the topologized
spectrum. Those sheaves, with certain additional condi­
tions on their fibers, tum out to be the 'schemes' (schemas)
of Grothendieck, who thus succeeds not only in unifying
some of modern mathematics' deepest intuitions (Galois,
Riemann), but in doing so, broadens the very conception ef
spa,ce, such that we are no longer concerned with points,
but with positions and movement (sections of the sheaf).
From its broad, general outlines to its most particular
technical concretizations (as we shall see in section 4.3),
Grothendieck's work yields a fundamental paradigm,
which we should like to call the practice efa relative math­
ematics. Grothendieck's strategies can indeed be under­
stood, in a conceptual sense, as close to the relativistic
modulations that Einstein introduced into physics. In
a technical manner, both Einstein and Grothendieck
manipulate the frame of the observer and the partial
dynamics of the agent in knowledge. In Grothendieck's
way of doing things, in particular, we can observe, firstly,

1 40
 GROTHENDIECK

the introduction of a web of incessant transfers, transcrip­

tions, translations of concepts and objects between appar­
ently distant regions of mathematics, and, secondly, an
equally incessant search for invariants, protoconcepts and
proto-objects behind that web of movements. In their
technical definitions, sheaves and schemes incarnate both
flux and repose. Beyond sheaves as singular objects, the
'protogeometry' that underlies certain classes of sheaves
then gives rise to the Grothendieck topoi.
Grothendieck topoi (1962) are categories of sheaves
issuing from certain 'natural' abstract topologies.95 Topoi,
which are something like parallel universes for the develop­
ment of mathematics, are categorical environments suf­
ficiently vast for the development of an entire sophisticated
technology of the relative to be possible. Generalizing
the action of certain groupoids on the fibers of a sheaf,
Grothendieck seeks to movethetopoi (environments that are
no longer just set-theoretical but topological, algebraic,
differential, combinatorial, etc.) and study, in a generic
fashion, the actions of various functors on enormous

95 In categories with certain good properties of compositionality and covering, an

abstract topology (Grothendieck topology) can be defined by means of (sub)collections
of morphisms that are 'well matched' with one another. The categories of presheaves
(categories of functors to values in the category of sets) verify those good properties
of compositionality and covering, and abstract topologies can be defined there.
Grothendieck topoi issue from categories of presheaves that are 'situated' around
a given abstract topology. (Those categorical environments are also called sites.)
A simplification of Grothendieck topoi is provided by Lawvere's elementary topoi
(1970), where the abstract topologies (by means of Yoneda's lemma) can be easily
described thanks to a single endomorphism of the subobject classifier, which registers
the algebraic properties of a closure operator.

141
 SYNTHETIC P H I LOSOPHY

classes of topoi. The results come without delay, and it is in

the genericgeometrical realm eftopoi that certain cohomological
obstructions disappear: it is there that Grothendieck and his
school could develop the etale cohomology that would
allow Deligne to resolve the Weil conjectures. In top oi,
objects are no longer 'fixed' but 'unfold through time ' .
We are dealing here with variable sets, whose progressive
parametric adjustments allow us to resolve a multitude of
obstructions that seemed irresolvable from within a 'punc­
tual', classical or static mathematics. From this alone, one
can intuit the enormous philosophical impact that such
a relative mathematics might have - a mathematics atten­
tive to the phenomenon of shifting, but with the capacity
to detect invariants behind the flux, a mathematics that
goes against the grain of supposedly ultimate founda­
tions, absolute truths, unshakable stabilities, but which is
nevertheless capable of stabilizing asymptotic webs of truth.
In Grothendieck's work, objects tend to be situated
over certain 'bases' (the sheaf over its underlying topo­
logical space, the scheme over its spectrum), and many
important problems arise when base changes are carried out.
'Relative' mathematics then acquires a great mathematical
incisiveness, in inquiring as to which properties are trans­
ferred in the effectuation of base changes (descent theory:
the search for the conditions under which one is able to
carry out transfers; and its counterpart, the detection of
the conditions of obstruction in base changes). Conditions

1 42
 GROTHENDIECK

of coherence and abstract gluing arise quite naturally from

these processes of transference/translation - conditions
that, it seems, can be easily defined only in the context of
Grothendieck topoi. In particular, Zariski's 'site', which
allows us to articulate the Weil conjectures, is replaced
by Grothendieck's 'etale site' ,96 in which is constructed
- following a general procedure that we will come back
to in the Tohoku, according to which certain categories
of sheaves give rise to natural cohomology groups - the
etale cohomology that Deligne will later need in order
to resolve the conjectures.97
In reality, the conceptual dynamics of topoi far sur­
pass the theory's first technical objectives, as brilliant as
they were. Indeed, surfacing behind the Grothendieck
topoi Lawvere's elementary topoi, where we see that the
number-theoretic, algebraic, topological and geometrical
considerations advanced by Grothendieck also possess
surprising log;i.cal counterparts.98 As we will see in chapters
5 and 7, as we approach Lawvere and Freyd, the categories

96 Etale: smooth, without protuberances (the term comes from a poem of Victor Hugo's,
about an 'iftale' sea). Grothendieck's metaphorical use of 'etale' condenses the idea of
the nonramifi.ed, where Grothendieck combines, once again, some of the central ideas
of Galois and Riemann's: extensions of nonramified fields (Galois's separability) and
nonramified Riemann surfaces, enveloped in a generic unifying concept.
97 See P. Deligne, 'Quelques idCes maitresses de l'oeuvre de A. Grothendieck', in
Matbiaux pour l'histoire des matMmatiques au xx" siecle, Seminaires et Congrl:s 3,
Societe Mathematique de France, 1998, H-19.
98 Oddly, Grothendieck, who explored almost every field ofmathematics with rremendous
penetration, hardly bothered with mathematical logic. That disquieting logico­
mathematical separation - by one of the two or three major mathematicians of the
twentieth century - should give logic-centered philosophers of mathematics much to
think about.

1 43
 OF C O NTEMPORARY MATHEMATICS

and allegories that situate themselves between cartesia n

categories and topoi encode an entire legion of intermedi­
ary logics, whose relative web reflects a good part of the
greater mathematical movements that are based on them.
Base changes in underlying logics thus give rise to a complex
landscape - we could call it relative logic that allows us -

to return to the historical origins of mathematical logic

(Peirce's 'logic of relatives') and reinterpret, in a new
light, many of the problematics concerning foundations,
which were taken up in a conventional way by analytic
philosophy.
Tue Grothendieckian attention to the movement of
mathematical concepts and objects is accompanied by an
oscillating search for archetypes for mathematical reason
and imagination. Between the One (the 'form') and the
Many (the structures: schemes, topoi, etc.), Grothendieck
discovers and invents99 suitable invariants of form: the
cohomologies. Although the homology and cohomology
groups for algebraic topology tend to satisfy certain con­
ditions of univocity, the possibilities for cohomological
invariants multiply as we move through algebraic geom­
etry (Hodge, De Rham, crystalline, etale, /-adic, etc.). It
is for this reason that Grothendieck proposes his motifs,
99 We will deepen, in section 4.2, the dialectic of discovery and invention in Grothendieck,
a dialectic that canMt be reduced to either of its two poles; and we will study with
greater care, in part 3, the fact that a position as much realist ('discovery') as idealist
('invention'), is indispensable in advanced mathematics, once the latter goes beyond
a certain threshold '![complexity for structures, languages and the transits/obstructions
at stake.

144
 GROTHENDIECK

deep generic structures underlying distinct cohomologies.

Reading Grothendieck's own words· is well worth the
trouble, since we will be taking up several ideas from the
following quotation throughout this essay:

This theme [that of motifs] is like the heart or the soul,

the most hidden part, the most concealed from view,
of the schematic theme, which is itself at the heart
of a new vision. [ .. ] Contrary to what happens in
.

ordinary topology, [in algebraic geometry] we find

ourselves faced with a disconcerting abundance of
different cohomological theories. One gets the very
clear impression that, in a sense that at the beginning
remains somewhat vague, all of these theories should
'turn out to be the same', that they 'give the same
results'. It is to be able to express this intuition of
'kinship' between different cohomological theories
that I have extracted the notion of a 'motif associ­
ated with an algebraic variety. By this term I mean to
suggest that what is at stake is the 'common motif
(or the 'common reason') underlying this multitude
of different cohomological invariants associated
with the variety, thanks to the entire multitude of
cohomological theories that are possible a priori.
These different cohomological theories would be
something like different thematic developments,
each in its own 'tempo', 'key' and 'mode' ('major'

1 45
 SYNTHETIC P H I LOSOPHY

or 'minor'), of the same 'basic motif' (called ' motivic

cohomological theory'), which would be at the same
time the most fundamental, or the most 'fine', of all
these different thematic 'incarnations' (that is to say,
of all these possible cohomological theories) . And so,
the motif associated with an algebraic variety would
constitute the 'ultimate' cohomological invariant, the
cohomological invariant 'par excellence', from which
all the others [ . . . ] would be deduced, as so many
different musical 'incarnations' or 'realizations'. All
the essential properties of 'the cohomology' of the
variety would already be 'read off' (or 'heard in') the
corresponding motif, so that the properties and struc­
tures familiar to the particularized cohomological
invariants (l-adic or crystalline, for example), would
simply be the faithful reflection of the properties and
structures internal to the motif.100

Both homologies (mathematical constructions that help

us resolve 'the discrete/continuous aporia' [Thom] and
that consist of chains of abelian groups with which one
captures ample information about the topological object

too Grothendieck, 'Prelude', in Recoltes et Semailks 45-6 (qU<>tation marks and italics are the
author's). The (conceptual, mathematical, stylistic, methodological, phenomenological)
richness of this paragraph will give rise to many reflections in our work. For the
moment, it is enough to underscore the movement between the One and the Many,
the. tension between the 'ultimate' and the differences, the problematic of fidelity and
variation, the dialectic between the internal and the external, the modal spectrum of
possibilities and realizations, the interlacing ofvagueness and precision, the grafting
of corazan and rtu:6n (heart and reason), the aesthetic equilibrium.

146
 GROTHENDIECK

under investigation) and cohomologies (dual construc­

tions involving more familiar set-theoretical limits [prod­
ucts, pullbacks, etc.]) become, thanks to Grothendieck,
some of the 'most powerful [mathematical] instruments
of the century' . 101 At the end of his research at the IHES,
after his work on schemes and topoi, Grothendieck
envisioned the difficult and ambitious motivic prof!7am.
Retiring from the world of mathematics and ceasing to
publish, the major lines of development ofGrothendieck's
program went on to circulate as manuscripts, and many
of his suggestions were considered excessively vague. 102
Nevertheless, Voevodsky introduced motivic cohomol­
ogy (1990-2000), a contribution that, in part, answered
Grothendieck's expectations, and that won him the Fields
Medal ( 2002). Instead of working, as in algebraic topol­
ogy, with algebraic surgeries on space (singular cohomol­
ogy, ring of cohomology groups), Voevodsky proposed a
more delicate collection of surgeries on an algebraic variety,
introducing new forms of topology for algebraic objects
(fine Grothendieck topologies over sites of schemes),

101 Grothendieck, Ricoltes et Semailles, 43.

102 The same could be said of another very influential 'vision' of Grothendieck's, the
'moderate program' that he sketched out in his 1983 Esquisse d'une programme.
Grothendieck sought new focms of topology, which would turn out to be natural and
would smooth over the singularities that a set-theoretical topology must endure (replete
with artificial examples coming from analysis). Grothendieck had the intuition that
'
a son of deconstruction ( dtuissage', Esquisse, 25) of stratified collections of structures
would be tied to the discovery of a 'moderate topology'. Amid the developments of the
'moderate program' one can find tame model theory or a-minimality in contemporary
model theory - another unsuspected resonance of Grothendicck's ideas with logic.

147
 O F CONTEMPORARY MATHEMATIC S

and defining a sophisticated concrete category for theH(V)

homologies functorially associated with varieties V. A
central tntnk ef cohomologies had then begun to 'surface' ,
concordant with Grothendieck's extraordinary mathemati­
cal intuition.

4 . 2 M ETAP H O R S , M ETHODS, STYLE

In this section we analyze some of the major metaphors

that Grothendieck himself used in explicating his modes
of creation and work methods, and we observe some of
the resonances we encounter between Grothendieck's math­
ematical production (in its imaginative phase no less than
its definitional and theorematic phase), the mathemati­
cian's reflection on that production, and the formal expres­
sion of that reflection. All of these resonances constitute
what could be called the peculiar style of Grothendieck.
The metaphors of the 'hammer' and of the 'rising tide'
preside over much of Grothendieck's conceptual vision.1"3
For Grothendieck, a problem can be imagined as a sort
of 'nut', whose hard shell has to be penetrated in order
to get to its 'soft flesh'. In Grothendieck's conception,
there are two essentially distinct strategies for opening
the shell: hitting it with a hammer and chisel - sometimes
slipping and sometimes smashing the inside to pieces

io3 Grothendicck, Ricoltes et Semaille.r, 552-3.

1 48
 GROTHENDIECK

along with the shell - and immersing i t i n a liquid ('the

tide') in such a way that, after weeks or months, its exte­
rio r softens and opens up 'with a squeeze of the hands
[ . . J like a ripe avocado'. The first strategy (yang) aims
.

to resolve the problem; the second strategy (yin) aims to

dissolve it. Through an adequate immersion in a natural,
ambient medium, the solution should emerge within a generic
landscape that outstrips the particular irregularities of
the shell . The metaphors capture a precise mathematical
methodology that Grothendieck had constantly put into
practice over at least thirty years: immersing a problem
in an appropriate general category (K), performing a
profound labor of conceptual and definitionalprescission104
inside that category, decomposing examples and objects
inside that general frame, and proceeding finally to the
study of the correlations, transits and osmoses within the
category. After an incessant abstract (de )construction
('divissage'), the problem can be resolved with the great­
est possible seftness ('a ripe avocado'), without blows and
without artificial ruses, as the direct testimony of Deligne
indicates.105
Going further still, Grothendieck's strategy of the 'ris­
ing tide' goes on to place questions, notions and points

104 Prescissian,
in Peirce's sense, at once cuts andspecifies the boundaries of the entity under
analysis.
105 Deligne, 'Qµelques idies maitresses .. . ", 12.

149
 SYNTHETIC PHILOSOPHY

of view at the center of mathematical attention, above

and beyond the resolutions themselves:

More than anything, it's really through the discovery

of new questions, and likewise new rwtions, and even

new points ef view - new 'worlds', in fact - that my

mathematical work has turned out to be fruitful,

rather than through the ' solutions' that I have con­

tributed to questions already posed. This very strong

drive, which has carried me toward the discovery of

good questions, rather than toward answers, and

toward the discovery of good notions and statements,

much more than towards proofs, is another strong

'yin' trait in my approach to mathematics.106

Behind a problem, Grothendieck always seeks out the

wellsprings ( the sources) of natural questions associated
with the problem. What is at issue, therefore, is a vision

106 Grothendieck, Rico/tes et Semai//es, 554. Of course, such a paragraph can only be
appreciated from a great height, seeking to clarify the most salient movements of the
topography. We must not forget the (literally) thousands of pages that Grothendicck
devoted to 'answers' and 'demonstrations' in his major fields of production: 'tensorial
prod.ucts and nuclear spaces, cohomologies of sheaves as derived functors, K-theory
and the Grothendieck-Riemann-Roch Theorem, emphasis on work relative. to a base,
definition and construction of geometrical objects via the functors to that which
must represent them, fibered categories and descent, stacks, Grothendieck topologies
(sites) and topoi, derived categories, formalisms of local and global duality (the "six
operations"), Ctale cohomology and cohomological interpretation of L�functions,
crystalline cohomology, "standard conjectures", motifs and the "yoga of weights",
tensorial categories and motivic Galois groups' (following a 'brier list of contributions,
in P. Cartier et al., 'Ihe Grothendieck Festschriji [ Basel: Birkhauser, 1990] , vol. I, viii).
As Dieudonne points out (ibid., i4), 'there are few examples in mathematics of so
monumental and fruitful a theory, built up in such a short time, and essentially due
to the work of a single man'.

150
 GROTHENDIECK

of the foundations o f mathematics that differs radically

from the one proposed by set theory. Grothendieck's
'reading' is a transversal one, in which an ultimate base
is of no importance . What is under investigation, instead,
is the base's movement (its shifiing),10? and what matters,
more than an accumulative resolution of knowledge, is
the mobile interlacing of natural questions underlying
the solutions.108
In fact, it is not even a question of a 'reading' in
Grothendieck, but rather a listening. An articulation
between images, intuition and ear, as opposed to other
merely formal manipulations of language, seems to be
fundamental for him. In addition to the metaphor of
the nut and the rising tide, another of Grothendieck's
central metaphors is, in fact, the image of the creative
mathematician attending to 'the voice of things ' . 109

107 For Merleau-Ponty, the 'height of reason' consists infeeling thegroundslip away, detecting
the movement of our beliefs and our supposed knowledge: �every creation changes,
alters, elucidates, deepens, confirms, exalts, recreates or precreates all the rest' (M.
Merleau-Ponty,Noter der coumie College de France [1958-59, 1960-61] [Paris: Gallimard,
1996], 92). In L'oeil et /'esprit (Paris: Gallimard, 1964), Merlcau-Ponty describes the
body as operating in the domains of knowledge as a 'sheaf of functions interlacing
vision and movement'. Through incessant levels of self-reference, the sheafpermits the
conjugation of inner and outer, essence and existence, reality and imagination. And,
moreover, it is in the murky and antinomic frontiers of such apparent contradiction that
the sheaf gives rise to invention and creation. We shall return, in part 3 (chapter 10) to
certain connections between Grothendieck, Merleau-Ponty and Rota, as regards the
apparently fundamental opposition between invention and discovery in mathematical
philosophy.
108 Recall, here, the similar position of Lautman, who pointed to an ·urgency of problems,
behind the discovery of their solutions' (pp66- 8, above).
109 Grothendieck, Recoltes et Semailles, 27.

151
 OF CONTEMPORARY MATHE MATICS

Th e 'hidden beauty o f things'"0 appears t o be the hidden

beauty of mathematical structures, an intrinsic beauty
that the mathematician discovers by means ef the extrinsic
invention of sufficiently expressive languages. And so,
in Grothendieck's perspective, mathematical structures
appear in the phenomenological spectrum of the world,
and so they are discovered - but these are discoveries that
can only be made by inventing, in an almost synchronic
dialectic, adequate representations of the structures in
question. The (musical, cohomological) metaphor of
the motif itself shores up the idea that there exist hidden
germs efstrncturation, which a good 'ear' should be able to
detect . And so Grothendieck's motifs appear to be already
present in the dynamic structure of forms, independent
of their future discoverers (Voevosdky, Levine, Morel,
etc.) , whose work would consist essentially in creating the
adequate languages, the theoretico-practical frameworks,
and the sound boxes required to register their vibrations .
Again, i t i s instructive t o listen to Grothendieck himself:

The structure of a thing is not in any way something

that we can 'invent'. We can only patiently, humbly
put it in play - making it known, 'discovering it'. If
there is inventiveness in this work, and if we happen
to perform something like the work of the blacksmith

uo Ibid., 28.

152
 GROTHENDIECK

or the tireless construction worker, this i s not a t all to

'fashion' or 'build' structures. They do not wait for
us in order to be, and to be exactly what they are!
It is rather to express, as faithfully as we can, these
things that we are in the midst of discovering and
sounding out - that reticent structure toward which
we try to grope our way with a perhaps still-babbling
language. And so we are lead to constantly 'invent'
the language that can express, ever more finely, the
intimate structure of the mathematical thing, and
to 'construct', with the help of that language, thor­
oughly and step by step, the 'theories' charged with
accounting for what has been apprehended and seen.
There is a continual and uninterrupted back-and-forth
movement here, between the apprehension of things
and the expression of what has been apprehended, by
way of a language that has been refined and recreated
as the work unwinds, under the constant pressure of
immediate needs. m

The incessant recreation and invention through the disci­

pline's unwinding environs, the step-by-step construction,
and the groping expression all evince Grothendieck's
eminently dynamical and dialectical perception. What we
are dealing with here, in effect, is a 'continual weaving'

111 Ibid., 27. The quotation marks and italics are Grothendieck's.

153
 SYNTHETIC PHILOSOPHY

in mathematical thought, back and forth with a probing

instrumentarium, whose ontological and epistemological
categories cannot be stabilized in advance, independently
of the (practical, historical) action of the discipline. I n
that ubiquitous transit in which mathematics returns to
itself, in that always-moving and often enigmatic sea,
Grothendieck's manner nevertheless provides a profound
orientatwn and a surprising relative anchorage.
In effect, Grothendieck relies on multiple mathemati­
cal methods in order to maintain an orientation within
the variable scenery that looms up ahead. Above all, a
persistent ascent and descent allows him to overcome the
obstructions that lay in wait in local and excessively par­
ticularized labyrinths. The ascent to the general is never,
in Grothendieck, a gratuitous operation, but is controlled
by certain crucial modes of mathematical practice: the
insertion of a particular local situation (object, property,
example) into a universal, global environment (category)
- with subsequent osmoses between manifestations of
the singular and forms of the continuous; the plural
construction of webs and hierarchies so as to collate the
particular within wider relational universes; the discovery
of proximities in a topography with clearly defined eleva­
tions and projections of various types. Generalization is
then a weapon of contrast, a method for elevating vision,
that helps us to orient ourselves in a complex terrain.

1 54
 GROTH ENDIECK

On the other hand, Grothendieck's 'manner' i s entirely

governed by a ubiquitous (conceptual, linguistic, techni­
cal) dialectic. From the most vague (the yin and the yang)
to th e most precise (functorial adjunctions), passing
through incessant tensions between polar regions of
mathematics, Grothendieck's thought comes and goes
without the slightest rest. Many of his great technical
constructions - nuclear spaces, K-theory and generalized
Riemann-Roch, cohomologies, schemes, topoi, motifs
- straddle apparently distant mathematical nuclei, cul­
minating in i983 in the dessins d'enfants (children's draw­
ings), which propose strange combinatorial invariants
for number theory, by way of surprising mediations on
analysis (Riemann surfaces) and algebra (Galois groups).
The dialectic functions on multiple levels, from the vague
and imaginary to the technically restrained, in a 'vast
counterpoint - in a harmony that conjugates them ' . 112
Restricting the dialectic to the subdefinition of transits
and obstructions within mathematical activity, the genericity
of mathematical concepts and objects (a la Grothendieck)
gives rise to other original concretizations in the spectrum
of the arbitrary - arbitrariness being understood as a
simultaneous topos of mediation ('arbiter', transit, continu­
ity) and opposition ('arbitrary', imposition, discretion) .113

112 Ibid., 23.

n3 I owe this beautiful dialectical and etymological readiog ofthe 'arbitrary' to Roberto Perry
and Lorena Ham, in whose work it gives rise to a complex 'therapeutic of the arbitrary'.

155
 OF CONTEMPORARY MATHEMATICS

In effect, a generic entity combines its implicit definabili ty

within a horizon of possibility with its explicit concretiza -
tion within a horizon of stratification (Desanti),114 so as
to project its abstract capacity for transit (amid possibilia)
onto the concrete panorama of impositions that it encoun­
ters (in the hierarchies of the actual). Whence a third
method emerges, one that is very much present in all of
Grothendieck's works, in the 'art' of making mathematics.
What we are dealing with here is, indeed, a new ars combi­
natoria, which proposes to explicate, in four well-defined
steps, mathematics' unity in multiplicity, which is to say,
'the very life and breath'"5 of the discipline: the incessant
stratification of mathematical activity; the ramification of
ambient categories of interpretation; the recursive decon­
struction ('divissage') of the concepts at stake, throughout
the many available categorical hierarchies; the framing
of relational interlacings (diagrams of transferences and
obstructions) between realized deconstructions.
Grothendieck's manner116 - a mixture of vertical
114 J.-T. Desanti, Les idialitis mathtimatiques (Paris: Seuil: 1968).
115 Grothendieck, R/coltes et Semail/es, i6.

n6 In the artistic theory ofthe sixteenth and seventeenth centuries, maniera appears at the
nucleus of critical discussions on great painter's ·ways of doing things' (of inventing,
creating). With the degeneration of maniera into mannerism, the notion of style later
emerged as a conceptual substitute for capturing the major categories of the history
of art (baroque, classical, romantic, etc.). In part 3 of this work, we will take up the
problematic of how we may try to define - intrinsically, and not just diachronically, as
we have done so far - some of the great demarcations of mathematical styles: classical,
modem, romantic� contemporary. The maniera of Grothendieck opens important
channels for attempting to approach such intrinsic demarcations. It is something
that we have already begun in our chapter I - with conditions i-5 in terms of which
Lautman investigated modem mathematics, and with conditions 6-10, closely tied

156
 GROTHENDIECK

(ascent/descent), horizontal (dialectics/polarities) and

diagonal (reflections between stratified hierarchies) weav­
ings gives rise to a style that allows us to naturally express
-

these ways of making mathematics. By 'style', here, we

will understand the superposition of webs of'inscriptions'
(recalling the stylus with which the Babylonian tablets
were inscribed) and the intermeshing 'gears' between the
webs, in three fundamental mathematical registers: (i) the
initially vague invention, (ii) the subsequent delimitation
of that vagueness and consequent demonstrative expres­
siveness, (iii) the critical reflection on the demonstrative
body whose elaboration has been made possible. "7 From
this perspective, Grothendieck's style is of the great­
est interest because it amply extends through the three
registers,"8 and, in the epistemological sense of the word
'style', it effects a genuine 'saying-thinking' conjugation
(lexis)"9 in each field. Grothendieck, in fact, describes his
'particular genius'120 as a capacity for introducing great
new themes and unifyingpoints efview amid diversity; and
to Grothendieck's maniera, regarding which we have approached the 'contemporary'
- which we shall study more carefully, however, in Chapter u.
117 These three registers correspond to forms of the three Peirccan categories: (i) firstness
and abduction; (ii) secondness and contradistinction; (iii) thirdness and mediation.
The log;icefscientificinquiry, extensively studied by Peirce, links togetherprecise modes
of transforming information between these three categories. We will return to these
questions in chapter IO.
118 (i) : Correspondence; (ii): EGA, SGA, articles; (iii) Rlcoltes et Semailks. In each register,
the studious reader can count on hundreds of pages for developing her observations.
ug See the entry on 'Style' in B. Cassin, ed., Vocabulaire europeen desphilosophU!s (Paris:
Seuil, 2004), 1,226.
120 Recoltes et Semoilles, 15.

1 57
 SYNTHETIC P H ILOSOPHY

his skill i n 'saying' - that is, the rich traces efhis style - is
a vital instrument. for his unitary capacity for 'thought'.
The correspondence with Serre reveals Grothendieck's
indomitable energy, his potent mathematical inven tive­
ness, his staggering capacity for abstraction and con­
centration, but also his doubts and errors, his desire to
'cultivate himself' with the help of his correspondent's
enormous breadth of mathematical knowledge, as well as
the melancholy twilight of his great critical brilliance (in
the final letters of i985-1987, around the time of Recoltes
et Semailles).121 The enormous technical complexity of the
correspondence122 does not prevent us from being able
to detect many moments in which Grothendieck's ideas
continually emerge over the course ef days. From the point
of view of style, let us underscore, among other things,
the 'cohomology deluge' (1956) that gave rise to the
article 'submitted to Tannaka for the Tohoku', the writ­
ing of which is contrasted with 'Weil's truly intimidating
demonstrations';123 Grothendieck's incessant preoccupa­
tion with defining 'dearly natural' concepts and ideas,
which distinguish his manner of making mathematics from
other artificial practices;124 the presence of a 'plausible

121 P. Colmez, J.-P. Serre, eds., Correspondance GrothendiecJ.:-Serre ( Paris: Societi'

Mathematique de France, 2001).

1.22 Extramathematical rem.arks make no appearance in the correspondence, with the

exception of a brief Grothendieck-Serre-Cartan exchange ( 1961) about the problems
that military service is causing for young mathematicians. Ibid., 121-8.

123 Ibid., 38, 49·

124 Ibid., UL

158
 G ROTHENDIECK

yogical reason' that helps to specify and orient cer­

tain general speculations;125 the explosion of gigantic
' river letters [lettresjleuve]' in 1964, at the moment of
the invention of the Grothendieckian ideas about l-adic
cohomology (which ten years later would lead to the
resolution of the Weil conjectures). The nourishing cor­
respondence - which would be accompanied by hours ( ! )
of telephone conversations between Grothendieck and
Serre, an entirely original way (and one incomprehensible
to mere mortals) of sharing high mathematics - allows us
to glimpse a gestating mathematics: mobile, imbued with
a remarkable vigor, replete with approximations between
distant concepts, in a process of refinement, with a per­
petual correction of (natural) errors between one letter
and the next. Being able to spy on the blacksmith at his
forge, we can now see a real mathematics, so far from that
which the predominant currents of analytic philosophy
have taught us to appreciate.
As a counterpart to his rare talent for 'discovery' and
his facility for training the ear to listen to the 'voice of
things', Grothendieck also had a rare sensitivity for the
use of language, which he succeeded in concretizing in
terminologi.cal'inventions' as explosive as his mathematical
inventiveness itself. An integral . part of Grothendieck's

125 Ibid. The comical neologism 'yogical [yogique]' invokes Grothendicckian yoga (the
'vague' interlacings of yin-yang dialectics) and is contrasted with a supposedly formal
and precise logic.

159
 OF CONTEMPO RARY MATHEMATICS

mathematical style, the terminology, i n fact, formed a n

entire universe i n i�self. With the same fascinatio n t hat
Proust expresses in 'Place Names: The Name',126 G rothend­
ieck declares that 'one of my passions has constantly bee n
to name the things that discover themselves to me , as a
first means of apprehending them ' , 121 and points out th at

from a quantitative point of view, my work during

those years of intense productivity has principally
concretized itself in some twelve thousand pages of
publications, in the form of articles, monographs
and seminars, and by means of hundreds, if not
thousands, of new notions that have entered into
the common patrimony with the same names that I
gave them when I delivered them . In the history of
mathematics, I believe I am the one who has intro­
duced the greatest number of new notions into our
science, and the one who has been led, by that very
fact, to invent the greatest number of new names in
order to express those notions with delicacy, and in
as suggestive a fashion as I could . 128

126 'Du cote de chez Swann (Swann's Way]', in M. Proust, A la recherche du temps perdu
(Paris: Gallimard, i997). The latest edition, by J.-Y. Tadic, collates all the variants
of the original manuscript and thereby thoroughly explores the place of Proustian
inventiveness. [Tr. C. K. Scott Moncrieff & T. Kilmartin, In Search efLost Time (New
York: Modem Library, i9!)&).J
127 Rico/Jes et Semailles, 24.
128 Ibid., 19.

1 60
 GROTHENDIECK

'Jhe 'delicacy' of the name, the 'suggestiveness' of meta­

phors, the profusion of names that go on to modify a
'common patrimony' are not contemplated in the treatises
of mathematical philosophy. Even in schools ensnared
by a fascination with language, no study of the language
of real mathematics has been undertaken, even though
lengthy disquisitions have been elaborated on a language
supposedly capable of supplanting mathematics itself.
It is not our intention to go further into mathematical
language, but it must be pointed out that Grothendieck's
terminological inventiveness ought to be explored, else­
where, with the care that it deserves.

4.3 T H R E E PARAD I G MATIC TEXTS: TOHOKU, EGA,

CARTAN S E M I NAR

In this final section, we will review three specific texts by

Grothendieck, aiming to unveil some of the great lines of
tension in Grothendieck's production, as underscored in
the first section, as well as some of the methodological,
metaphorical and stylistic procedures indicated in the
second. We will look at three paradigmatic texts:
I. the Tohoku (published in i957, but which emerges from

i955 on, as we see mentioned in the correspondence with

Serre);
2. the first chapters of the great treatise, Elements efAlge­

braic Geometry (published in 1960, but already in gestation

161
 SYNTHETIC PHILOSOPHY

a few years earlier, as we can also deduce from the cor­

respondence with Serre) ;
3. the extraordinary series of expositions in the Cartan
Seminar of 1960-61, concerning themes in 'analytic geom­
etry' (understood, in Cartan and Dieudonne's manner,
as the geometry of analytic functions) .
As Grothendieck points out at the beginning of the
article, the Tohoku aims to explicate a 'common frame­
work' that would allow us to nourish the 'formal analogy'
between the cohomology of a space with coefficients
in a sheaf and the series of derived functors of a func­
tor on a category of modules.129 .The work was written
at the very moment when the notion of sheaf came to
constitute an indispensable component of mathematical
investigation,13° an investigation during the development
of which Grothendieck made a point of mentioning his
'conversations' with Cartan, Godement and Serre.131 Now,
from the very beginning of the Tohoku, vague 'analogies'
and dynamic forms of mathematical creativity appear in

129 Grothendieck, 'Sur quelques points d'algi:bre homologique', n9.

130 The concept of mathematical sheaf emerges in the work ofJean Leray, in a course 'on
algebraic topology in the O!lag XVII-A* (1943-5), a series of notes in the Comptes &ndm
de l'Acadimiedes Sciences (1946), courses on spectral successions at the Col!ege de France
(1947-50), and reaches its definitive development in the Henri Cartan Seminar at l'tcolr
NomwleSupirieure (1948-51). (O!lag XVII-A was an unusual German prisoner-Of-war
camp in which officers were imprisoned during WWII; Jean Leray directed the camp's
university.) A sheaf is a type of mathematical object that allows for the global gluing of
whatever proves to he coherently transferable in the local. Certain mathematical objects
can then be better understood, thanks to a logic of neighborhoods and mediations over
a continuous space (getting away from yes·no binarisms), and to the natural actions
of groupoids in the fibers of the sheaf.

131 Ibid., 120-1. Grothendieck also spoke of the 'interest' of his colleagues.

162
 GROTH ENDIECK

the text, and Grothendieck immediately proceeds to invent

the language (additive and abelian categories) and discover
the richness of the mathematical structures (sheaves,
inj ective objects via products and infinite sums, actions
of a group) that explicate the initial 'formal analogy'. In
particular, the actions of the groups at stake are stratified
in a precise hierarchy oflevels - action on a spaceX (first),
action on a sheaf 0 of rings over X (second), action on a
sheaf of modules over 0 (third) - thereby concretizing
one of the typical forms of the Grothendieckian process.
This strategy gives rise to the article's various parts:
(I) abelian categories (language); (II) homological algebra
in abelian categories and (III) cohomology with coeffi­
cients in a sheaf (structures); (IV) Ext calculi for sheaves of
modules and (V) cohomologies with spaces of operators
(transferences and actions).
The 'grand vision' of category theory proposed by
Grothendieck in section (I) remains extraordinarily fresh
after fifty years.13• It is dealt with in the lengthy introduc­
tory section of the text, where Grothendieck establishes
the three clear levels of categorical thought - morphisms,
functors, natural transformations (called 'functorial

132 We could consider the present essay (originally published in 2007), to a large extent,
as an homage to the Tohoku, fifty years after its publication. It would pcrhap.s be too
much of an exaggeration to describe contemporary mathematics as a series of footnotes
to the work of Grothendieck, but the exaggeration would have an unquestionable
grain of truth to it.

163
 O F CONTEMPORARY MATHEMATICS

morphisms' in the text).'33 In the same section, he also

introduces his additive and abelian categories, compares
existence axioms for infinite products, establishes the
existence of sufficient injectives (via diagrams, generators
and products) and develops the quotient categories. '34
In the process, Grothendieck produces some remarkable
examples, placing them like little crystalline gems in the
weave of ascent and descent between the universal and the
concrete. As instances ofadditive non·abelian categories,'35
for example, we are shown separated topological modules,
filtered abelian groups, and holomorphic fiber spaces over
a I-dimensional manifold. In this way, the topographical
richness of Grothendieck's mathematical thought is never
gratuitous, its ascent is never driven by an artificial impulse
to generalize, and it always contemplates a vivid landscape
of specific hills and valleys. This is something that receives
further corroboration with the comparison of various axi­
oms for infinite products in terms of the classes ofconcrete

i33 Ibid., i24.

134 Grothendieck gives the name 'Serre's module language C' to the idea of variation
over the base (ibid., 137). Given that what is at issue here is one of the central ideas
governing a good deal of 'relative mathematics' and the Grothendieckian 'tide', it is
veryintcrestingto observe how, mtheverymuvemmt eftlieidea's emergence, Grothendieck
sees Serre as the 'creator' of ' modulation through C'. This is one more proof of the
incessant contamination of mathematics, with every sort of residue in a web of mixtures
and impurities, that falls outside of analytic philosophy's modes of observation.

135 Ibid., 127.

164
 GROTHENDIECK

categories that they distinguish,136 with a 'fun example'137

having to do with cohomologies over an irreducible space,
with a profound example dealing with sheaves of germs
and differential forms over holomorphic manifolds,138
and with further examples of functorial manipulations
that allow us to reconstruct earlier arguments in which
Grothendieck utilizes the machinery of Cech coverings.139
The 'common framework' emerging from the Tohoku
- constructed in order to allow for the study of natural
interlacings between algebraic geometry, topology, com­
plex analysis and cohomological calculi - went on to
modify the mathematical landscape. Focusing his efforts
on a pivotal mathematical concept/object ( the mathemati­
cal sheaf), defining the general environments in which
sheaves can be studied in their unity/multiplicity (abelian
categories) , and putting this entire instrumentarium at
the service of comprehending the deep forms of structures
(the cohomologies) , Grothendieck brings about not only a

136 Ibid., 129. The category of abclian groups, its dual category of compact topological
abelian groups, and the category of sheaves of abclian groups over a given topological
space all make appearances. The instruments of transfer (the Pontrjagin duality) and of
integra.1 leaps from one level to another (sheaves) are put in the service or a differential
understanding between categories - gemis of a very abstract, contemporary, differential
and integral calculus.
137 'Un exemple arnusant', ibid., 160. 'Fun' is not well represented in 'formal' mathematics,
but is, without a doubt, among the important motors of the creative mathematician,
something which, of course, appears to be indiscernible in both 'normal' mathematics
(the series or texts published in the community) and in philosophical discussions of
this normal tradition.
138 Ibid., 165-6.
139 fbid., 16!, 213. The interlacing between functorial descriptions and Cech-style coverings
can probably be seen as the very origin of the Grothendieck topologies.

165
 SYNTHETIC PH ILOSOPHY

'Copernican turn' i n mathematics, but a genuine 'Einstei­

nian tum', if we allow ourselves to stretch the metaphor a
little.'4° Grothendieck's vision even comes to transcend the
framework that he himself elaborates, when, in a brilliant
premonition, he observes that 'it would be a good idea
to provide a treatment of "noncommutative homologi­
cal algebra"' in a context of functors and categories that
envelopes the theory of fibrations and the extensions of
Lie groups'4' - a startling anticipation of fragments of
Connes's program of noncommutative geometry, which
we will review in chapter 6.
Like the Tohoku, the Elements de Geometric Algebrique
(EGA) incorporate the most visible characteristics of
Grothendieck's procedure. Although the metaphorical,

140 For his part, Peirce brings about what could be called an 'Einsteinian tum' in philosophy.
Of course, although Peirce preceded Einstein and the label is therefore paradoxical,
the universal Peircean semiosis and its associated construction of relational invariants
is precisely fitted to the 'revolution' in modern physics that Einstein would bring about
just one decade later. In Peircean semiosis, subject and object are considered not a�
monadic predicates but as relational webs of various signs, inserted in scaffoldings of
reference subject to a perpetual dynamism ('unlimited semiosis'). In that dynamism of
relative movements, even the observation of an object can undergo modification. Peirce
therefore tries to find invariants in that complex relational flux: the 'Einsteinian tum'
of his philosophy seeks (and finds) what we could call the philosophical invariants of
a general logic of relations and higher-order logic. The relativity of perspective, the
unlimited dynamism of interpretation and the modification of interpret.ants are some
of the great conquests of the Peircean system - conquests that the twentieth century
repeatedly corroborates in the most diverse guises. Nevertheless, with his system's
permanent processes of reintegration and gluing, Peirce overcomes the extreme relativism
into which certain rehabilitations of the ephemeral and the local in the last stages
of the twentieth century can be seen to have led. We will take these ideas regarding
a supposed 'Einsteinian tum' in philosophy (Peirce) and an 'Einsteinian turn' in
mathematics (Grothendieck) further in part 3. If those approximations are more or
less correct, the philosophy of mathematics should, in turn, undergo a comiderable
tum.
141 Ibid., 213.

166
 GROTHENDIECK

analogical and stylistic levels are reduced to a minimum in

the EGA (an asepsis owing, to a large extent, to Dieudon­
ne's steely coauthorship ), they incorporate a 'grand vision'
(a mathematics in action that would provide the bases from
which to attack the Weil conjectures, 'a labor scarcely
undertaken'142 in i960), an open way of propelling and

presenting the discipline,143 a clear global landscape with

some well-delineated local techniques,144 a general way of
doing mathematics within contexts (categories) sensitive
to the transit/obstruction ofinformation (functors, natural
transformations) , 145 and a persistent search for the natural
notions governing those osmoses . 146

14" EGA, I, 9·

143 'All ofthe chapters are considered to be open', ibid., 6. Observe, indeed, how chapter
o ends with the phrase 'A suivre [to be continued]' (ibid., 78), something rather
uncommon in the mathematical literature, where texts are usually presented as
'finished'. Mathematics in gestation always ends by emerging (more obscurely than it
sometimes seems) in Grothendieck's works.
144 Given two algebraic varieties X,Y (or, more generally, given two schemes), the study
of the properties of a problem P in a neighborhood of y E Y is approached by way
of its transformations/obstructions through a (proper) morphism between X and Y,
following precise steps in the analysis of the problem: introducing the study of an
adequate local ring A over y; reducing that study to the artinian case (with which
one moves to a 'greater understanding of the problem, which on this level is of an
'infinitesimal' nature' [ibid., 8]); effecting adequate passages by means of the general
theory of schemes; permitting the discovery of algebraic extensions ofA (the primordial
task of algebraic geometry} by means of adequate multiform sections of the schemes.
145 Several sections ofchapter I ("The Language of Schemes') answer precisely to the study
of schemes from the categorical point of view of the preschemes that envelop them:
products (§3); subobjects (§4}; separability conditions (§5); and finitude conditions
(§6). As in the Tolwku, sophisticated examples are introduced (in polynomial rings,
ibid., 139) in order to distinguish, by means of suitable models, the various conditions
of separability.
i46 "Ihe usual constructions suggested by geometrical intuition can be transcribed,
essentially in a single reasonable mo.nner, into this language [of schemes]' (ibid., 9, our
italics). Grothendieck once again expresses one of his deep convictions: behind the
plurality of structures and signs, the construction (invention) of an adequate language
should allow for the naturalness of certain 'One'-structures (discovery) from which

167
 OF CONTEMPORARY MATHE MATICS

A s we saw i n section 4.1, schemes let u s elaborate an

important unification of the visions of Galois and Rie­
mann, a global fact that gets countersigned locally when
Grothendieck and Dieudonne explain that the 'lag in con­
ceptual clarification' in the theory of schemes may have
been largely due to the identification between the points
of a proper scheme X over a discrete valuation ring A and
the points of the tensorial X®A K over the.field of fractions
K of A.'47 In fact, the usual analytic perspective - busy­
ing itself with the observation of points and far from the
synthetic reading that would attend to the sections in a
sheaf (which do not, in general, proceed from points) - is
one that impedes scheme theory's emergence, and obstructs
the natural change of category that should have led us to
work on rings rather than fields. We thus encounter ways
of seeing that are apparently vague and general (analysis
versus synthesis) , but that take on an enormous richness
and technical importance that profoundly ojfects the
development of precise and well-defined theories. Amid
many other examples that we will later review, this concrete
fact within universal tensions, this material residue within
ideal dialectics, shows the importance - delineated, instru­
mental, traceable - of a fundamental analytic/synthetic

one should be able to project the remaining 'Many'-structures that are at stake. What
we are de.a.ling with here is neither more nor less than a surprising rebirth of a sort of
mathematical metaphysics that seeks (and finds) new archetypes behind the relative. We
will have a chance to discuss this situation at length in part 3.

t47 Ibid., ng.

168
 GROTHENDIECK

opposition that certain currents o f philosophy, following

Quine, would rather see disappear.
In a footnote in the EGA, Grothendieck and Dieudonne
point out that algebraic geometry, extended to the uni­
verse of schemes, should be able to serve 'as a sort of
formal model' for analytic geometry (that is to say, for
the theory of analytic or holomorphic spaces) . 148 Shortly
thereafter, Grothendieck will develop his 'Construction
Techniques in Analytical Geometry', a remarkable series
of notes from the 'seminars', written up in the heat of the
moment and surfacing in a matter of days.149 Grothendi­
eck's objective is made explicit at the very beginning of
his expositions: to extract ( 'degager') a general functorial
mechanism for the manipulation of modules, applied in
particular to the case of complex variables; to extract a
'good formulation' of the problems of modules within
the framework of analytic spaces; to interlace projectivity
properties with existence theorems within that frame­
work (Teichmiiller space); to bring the framework of
schemes (algebraic geometry) and the framework ofholo­
morphic manifolds (analytical geometry) closer to one
another, in both cases making particular use of the crucial
properties of certain nilpotent elements in local rings. 15°

148 Ibid., 7.

149 A. Grothendieck, 'Techniques de construction en geometrie analytique I-X', Shninaire

Henri Cartun, vol. 13 (Paris: Secretariat Mathematique, 196o-1 ). They cover about two
hundred pages in total.

150 Ibid., exp. 7, 1.

1 69
 SYNTHETIC P H I LOSOPHY

Once again, Grothendieck's generic methods are incar­

nated in the concrete: the processes of ascent (general
functoriality, problematics in abstraction, global frame­
works) and descent (projectivity, complex modules, local
rings, nilpotence); the dialectic of the One and the Many
(the drawing together of frameworks, the specific func­
toriality of modules, interlacings between projectivity
and existence); and, in sum, the hierarchical structuration
efmathematical knowledge in circulating levels, in which are
combined the rich conceptual multiplicity of each object
(or morphism), the collection of functors that allows us
to 'measure' the differential multiplicity of each level,
and the collection of (natural) transformations that let
us reintegrate the differential frameworks encountered.
Though we cannot go into excessive technical details
here, we may note that Grothendieck's first paper in his
'Techniques of Construction in Analytic Geometry' on its
own encapsulates the entire richness of his mathemati­
cal thought. The objective is to construct an analytical
space of universal representation (Teichmiiller space)
that classifies aU other algebraic curves over analytical
spaces;1s1 Grothendieck goes on to axiomatically describe
the functorial properties1s2 that this space should satisfy,
and theorematically derives its (global) existence from

151 Ibid., exp. 7, 1.

152 Ibid., exp. 7, 6. Grothenclieck first describes these properties 'vaguely', and only later
gives them precise technical descriptions.

1 70
 GROTHENDIECK

the existence of a (local) echeloned hierarchy of adequate

fibering functors ('Jacobi functors') 'that let us control
the number of automorphisms of the structures in play
(rigid functors).153 These technical conditions are, in tum,
explicated through coverings, group actions, base changes
and free operators over the same rigid functors. '54 We thus
encounter a genuine mathematics in motion, a relative
mathematics that, nevertheless, allows us to encounter
certain universal invariants ('Teichmiiller space') in virtue
of the sheer variation of the local mathematical objects,
through the subtle hierarchy of mediations that give rise
to the global mathematical object.
Grothendieck's legacy in the landscape of contempo­
rary mathematics is becoming increasingly evident as the
discipline unfolds and technical advances corroborate
many of French mathematician's major intuitions.'55 In
the following chapters, we will try to amplify that con­
temporary spectrum, as we review the works of other
mathematicians of the first rank, works that partly con­
tinue and partly complement Grothendieck's own, so
that later, in the third part of this essay, we may try to
reflect - in ontological, epistemological, methodological
153 Ibid., exp. 7, 32.
154 Ibid., exp. 7, 18-21 (coverings and groups), 2ff. (bases), 27ff. (functors).

155 This is particularly conspicuous when it comes to his 'Sketch of a Program' , with the
connections inaugurated there between combinatorics, number theory and functional
analysis, which have been extended, to the great surprise of the scientific community, to
theoretical physics and cosmology. We will return to this in chapter 6, as we approach
Connes and Kontsevich.

171
 OF CONTEMPORARY MATHE MATI C S

and cultural realms - on that extensive spectrum, which is

usually ignored in the tracts of mathematical philosophy.

1 72
C HAPT E R 5

E I DAL MAT H E MATI C S : S E RR E , LANGLAN D S ,

LAWVERE, S H E LAH

In the next three chapters we will review other examples

of high mathematics in action, in the conviction that
advanced mathematics - and, in this case, contemporary
mathematics - provides philosophy with new problematics
and instruments, as we shall see in part 3 of this work.
But firstly, in order to present the mathematical landscape
more commodiously, we will sort a few of its striking
creative contributions into three complementary spectra:
eidal, quiddital and archeal mathematics. Of course, these
mathematical realms (and the neologisms that denote
them) do not exist in a well-defi,ned way as such, and must
be understood only as expository subterfuges for easing
the presentation and to help us get our bearings in a
complex terrain.
In fact, behind the central question of phenomenol­
ogy - How do we transit between the Many and the
One? - along with its polar subquestions: How do we
unify phenomena by means of categories, and how do we
multiply the universal in the diverse? - lie crucial modes ef
transformation of knowledge and the natural world. Media­
tions, hierarchizations, concatenations, polarizations,
inversions, correlations, and triadifications, for instance,

173
 SYNTHETIC P HILOSOPHY

are series of transformations leading to the partial exp li­

cation of certain universal categories,156 in knowl edge
(reorganizing the Kantian inheritance as regards the
transit between the noumenon and the phenomenon) an d
the physical world alike. 157 Within that universal trans­
fannism present since the very beginnings of Greek
-

philosophy and, in the field of mathematics, now codifi ed

in the mathematical theory of categories - it has always
been possible to detect a double movement in perpetual
readjustment: an oscillating series of ascents and descents
in the understanding; and a search for invariants behind
those natural oscillations. We will call movements of
ascent eidal (from eidos [idea]), movements of descent
quiddital (from quidditas [what there is]), and the search
for conceptual invariants in the various forms of transit
archeal (from arkhe (principal]).
In its very etymology, eidos involves an interlacing of
seeing (idein) and knowing (oUia). In 'raising' herself from
the world toward the ideas, the observer contemplates an
open landscape from a higher perspective and can 'see'

156 This tramformalional operation can be seen ingreat detail, for example, in the emergence
of the three Peircean categories, as shown in the doctoral thesis of Andre de Ttenne,
L 'analytique de rqmfsentatian cha. Peirr:e. La genese de la thiorie des categorier (Brusse1 s:
Publications des Facultes universitaires Saint-Louis, 1996).

157 lhis is particularly visible in contemporary physics and biology, which are becoming ever­
more imbued with dynamical considerations, linked to the description-comprehension
of 'diagrams of transit'. As we have indicated throughout this work, mathematics -
straddling the 'pure' understanding and the physical world - incorporates to an even
greater extent, in a visible and conspicuous manner, the study ofgeneralandparticular
problematics eftransit.

1 74
 E IDAL MATHEMATICS

further. Extended vision thus implies a greater breadth

knowledge. In mathematics, the interest in great 'ideas'
is no different: they open an immense field of action,

bv which
work programs can be organized, horizons
u� cluttered, and subspecialists oriented. The ideas, in
turn, combine with images (eidola) , and so often comprise
su rprising
transfusions efJorm. In what follows, we shall
see how
certain incisive contemporary contributions in
mathematics respond, in a technical manner, to sophis­
ticated distillations of form in the conceptual world of
mathematical ideas.

Figure 7. The Eidal Realm.

175
 OF CONTEMPORARY MATHE MATICS

The great interlocutor and proponent of Grothendieck's

work, Jean-Pierre Serre (France, b. 1926) - a sort of
sparring partner for his friend, as we have seen in the
previous chapter - may be considered one of the greatest
mathematicians of the twentieth century in his own right.
Awarded the Fields Medal (1954.) and the Abel Prize
(2003),158 Serre has received the mathematical commu­
nity's highest distinctions, in recognition of his brillian t
investigative beginnings,'59 as well as his mature work as
an exemplary creator and scientist. Serre's main works
cover a very broad mathematical spectrum: the study
of the homotopy groups of hyperspheres by means of
pathspace fibration (with spectacular calculations, such as
the determination of the twelve continuous applications
between a 6-dimensional and a 3-dimensional sphere) ;
foundational works at the intersection of algebraic geom­
etry and analytic geometry, based on the emerging theory
of sheaves, FAC100 and GAGA ; 161 the Galois representa­
tions associated with formal groups, abelian varieties and
modular forms (which, among other conjectures, would
interlace Fermat's Theorem with key advances in algebraic

158 Tue Abel Prize citation honors him 'for playing a central role in shaping the modem form
of several parts of mathematics, including topology, algebraic geometry and number
theory'. Note the importance ofjorm in the citation. The casual use of the adjective
modern, however, turns out to be inadequate according to our own delineations.
i59 Serre remains one of the youngest Fields medalists in history.

160 J.-P. Serre, 'FaisceauxalgCbriquescoherents' (FAC),AmwlsefMmhematicsfil, 1955: 197-278.

161 J.-P. Serre, 'Geometric algebrique et geometric analytique' (GAGA), Annales de/'institut
Fourier 6, 1956: 1-42.

1 76
 E IDAL MATHEMATI C S

geometry and number theory) . Serre indicated how his

works, despite their apparent eclecticism, responded to
the same way of observing and transforming problemat­
thanks to the use of transversal instruments and the
ics,
emergence of mixtures (recall Lautman) in which the One
and the Many are technically conjugated:

I work in several apparently different topics, but in

fact they are all related to each other. I do not feel
that I am rea1ly changing. For instance, in number
theory, group theory and algebraic geometry, I use
ideas from topology, such as cohomology, sheaves
and obstructions.
From that point of view, I especially enjoyed working
on l-adic representations and modular forms: one
needs number theory, algebraic geometry, Lie groups
(both real and l-adic), q-expansions (combinatorial
style) . . A wonderful mdange . 162
.

162 M. Raussen, C. Skau, 'Interview with Jean-Pierre Serre', NotU:es '![the American
MathematU:al Society 51, 2004: 211. Regarding the rrwde ef creation underlying the
emergence of that 'wonderful melange', it is interesting to point out that Serre speaks
of how 'you have some ideas in mind, which you feel should be useful, but you don't
know exactly for what they are useful', and of working 'at night (in half sleep)', which
'gives to the mind a much greater concentration, and makes changing topics easier'
(in C.T. Chong, Y.K. Leong, 'An Interview with Jean-Pierre Serre', Mathematical
Medley 13, 1985, http://sms.math.nus.edu.sg;lsmsmedley/smsmcdley.aspx#Vol-13). The
combination ofthefaay buundary and the great potentialfor exactitude is a fascinating
theme for mathematical philosophy, which an analytic approach is unable to take
up. In chapter 10, we will see how a broader mathematical phenomenology (which
incorporates the instruments of Peirce, Merleau-Ponty and Rota, among others) can
help us better understand those transits of creation.

177
 SYNTHETIC P HI LOSOPHY

Various concrete cases of the 'transfusion of forms' emerge

in Serre's work. When, for example, in GAGA, Serre estab­
lishes an equivalence between algebraic (coherent)163
sheaves over a projective variety and (coherent) analyti c
sheaves over the analytic space associated with that vari­
ety, an equivalence whereby cohomological groups appear
as invariants,'64 he is precisely pinpointing the transit
between algebraic and analytic forms, while carefully con­
trolling the osmoses and obstructions in play. When, at the
end of the fifties, Serre tests various structures in an effort
to produce a 'good' cohomology for varieties defined over
finite fields (trying to make some headway on the Weil
conjectures in this way), and when the cohomology group
with values in a sheaf of Witt vectors shows up165 (some of
the transpositions and obstructions of which will become
the very inspiration for Grothendieck's l-adic, crystal­
line and etale cohomologies), Serre once again brings
a tremendous labor of precision to bear on the manipu­
lation and transference of forms. In the environment

163 Coherenee codifies a finite type of property in sheaves. Coherent sheaves come from
both analytic geometry (the sheaf of germs of holomorphic functions), and from
algebraic geometry (the structural sheaf of a Noetherian scheme). A cmnmon idea/form
thus hides behind coherence. This has to do with a technical stratum of unification that
allows for a still-greater unity on the higher level of cohomology groups.

164 We are drawing here (and in what follows) on the excellent overview, P. Bayer, 'Jean­
Pierre Serre, Medalla Fields', la Gaceta de la Real Sociedad Matemfilica Espanola 4' 2001:
211-47. This is perhaps the best survey of Serres's work in any language.

165 Witt vectors (1936) are infinite successions ofelements in a ring, by means ofwhich sums
and products ofp-adic numbers can be represented in a natural manner. 'They therefore
have to do with underlyingform.r, hidden behind certain incomplete representations.
Here, one can sec how it is that� though only certain hierarchies of eidal adequation
are under consideration, new mathematics are already emerging.

178
 E I DAL MATHEMATI C S

of eidal mathematics, one can thus catch a glimpse of

a complex dialectic that delineates both the movement
of concepts/objects (the functorial transit between the
algebraic, the geometrical, and the topological) and the
relative invariants of form (cohomologies). At stake is a
profound mathematical richness - a richness that vanishes
and collapses if one restricts oneself to thinking in terms
of elementary mathematics.
Serre underscores a deep continuity in his creative career,
beyond certain apparent cuts or ruptures:166 an interlacing
of homotopy groups and C-theory (see p. 164, n. 134);
a natural osmosis between certain structures from both
complex analysis (sheaf cohomologies in the range of
functions of various complex variables or complex pro­
jective varieties) and algebra (cohomology-sheaves in
the range of rational functions or algebraic varieties); a
study of algebraic geometry over arbitrary fields (from
algebraic closures to finite fields, by way of generaliza­
tions of the theory of class fields) in terms of groups
and Lie algebras as 'mother' structures, where we find
an intersection of the extant contextual information.
Indeed, looking over certain contiguities/continuities
between the Riemann hypothesis, certain calculations over
modular forms and certain calculations of characteristics
of discrete subgroups of the linear group, Serre exclaims,

166 Chong, Leong, 'An Interview with Jean-Pierre Serre'.

179
 O F C O NTEMPORARY MATHEMATICS

'such problems are not group theory, nor topology, nor

number theory: they are just mathematics'.167 Advanced
mathematics thus contemplates a series of sophisticated
technical transits over a continuous conceptual ground,
something that, again, is lost from view when we restri ct
our perspective to the (essentially discrete) fragment of
elementary mathematics.
This continuity of mathematical knowledge has been
energetically emphasized in the work of Robert Lang­
lands (Canada, b. i936). The Langlands program, in fact,
consists of an extensive web of conjectures by which
number theory, algebra, and analysis are interrelated in a
precise manner, eliminating the official divisions between
the subdisciplines. The program first emerges in a long
letter (1967) from the young (and unknown) Langlands
to the eminent master of the epoch, Weil. The letter turns
out to be full of admirable conjectures that approach the
world of the complex variable and the world of algebraic
extensionsjunctorially by way ofgroup actions. Langlands
,

arrives at these surprising approaches by following the

precise elevations of group actions in the ei,dal realm.1 68
In effect, Langlands's intuition emerges through a

167 Ibid.

168 The extremely meticulous Weil is somewhat irritated by the young Langlands's
flights of fancy, and he immediately sends the reckless young man's interminable,
handwritten letter to be typeset. The letter to Weil, along with a great deal of additional
material by (and on) Langlands, can be found online: http:/jwww.sunsite.ubc.ca/
Digita!MathArchive/Langlands/intro.html.

180
 E IDAL MATHEMATICS

correlative contemplation oftwo ascending paths: the transit

from modular forms (analytic functions of complex vari­
ables respecting certain actions of the group SL2(R)) to
automorphic forms (analytic functions respecting actions
of Lie g;roups), passing through the intermediate actions
of the Fuchsian group on Poincare's modular forms and
those of the symplectic group on Siegel's modular forms;
the transit in the hierarchy of L-representations169 of the
Galois group Gal(K*:K) . 170
The correspondences between theforms of that transit
come to express the celebrated Langlands correspondence:
the automorphic forms associated with the linear group
Gal(n:K) correspond (functorially) to the n-dimensional
£-representations of the Galois group Gal(K*:K) . The
series of conjectures thus achieved concretize the continu­
ity of mathematical thought in a remarkable manner.111

169 The L·series (Dirichlet) appear as analytic objects for representing Riemann's �
function. The L-functions (Artin, Hecke) are analytic continuations of the L-series that
serve to measure the ramification of prime ideals in algebraic extensions. The abstract
construction (eidal elevation) of the concept of L-function therefore integrates analytic,
algebraic and arithmetical considerations.

170 K is an arbitrary field and (K*:K) need not be commutative. The commutative case had
already been resolved, before Langlands, in classfield theory (Hilbert, Takagi, Hasse,
Herbrand), and, in fact, considerations pertaining to class fields guided a great deal
of the Canadian mathematician's intuition.

171 It is worth the trouble to reproduce the beginning of the letter to Weil here: 'While
trying to formulate clearly the questions I was asking you before Chem's talk, I was led
to two more general questions. Your opinion of these questions would be appreciated.
I have not had a chance to think over these questions seriously and I would not ask
them except as the continuation of a casual conversation. I hope that you will treat
them with the tolerance they require at this stage.' The combination of informality
(supposed lack of seriousness, casual conversation, tolerance) and profundity (as the
seventeen-page letter is far from lighthearted or frivolous) should remind usof Serre's
observations regarding mathematical thought 'in the middle of the night'.

181
 SYNTHETIC P H I LO SOPHY

Just as Grothendieck sensed the existence of motifs

beneath the various manifestations of cohomologie s,
Langlands senses the existence ofstructuralforms oftransit
beneath the various natural manifestations of group
actions in complex analysis and number theory. In con­
temporary mathematics, as we shall see in chapte r 7,
this leads to the recognition of a series of archetypes, a
series of structures/concepts lying in the depths of the
mathematical continuum, from which many other partial
forms, deriving from the archetype, are extracted through
representational 'cuts'.
From the point ofview of the global concepts in play,
the Langlands correspondence proposes an unexpected
equivalence between certain dijferentiable structures associ·

ated with an extended modularity (the automorphic forms

associated with the linear group) and certain arithmetical
structures associated with analytic continuations (the
L-representations of the Galois group). The profound
proximity of the differentiable and the arithmetical, in the
restricted context of modular action and analytic continu­
ation, constitutes a truly major discovery for contemporary
mathematics. In fact, the Langlands program has given
impetus to many highly technical results, including vari­
ous proofs of the correspondence, for every n and for
specific cases of the field K under consideration: for fields
of formal series over finite fields (Laumon, Rapoport,
Stubler 1993); for p-adic fields (Harris, Taylor 1998); for

182
 E IDAL MATHEMATICS

fields of rational functions over curves defined over finite

fields (Lafforgue 2000, work which won him the 2002
Fields Medal).
Nevertheless, the program is currently encountering
a fo rmi dable obstruction in tackling the 'natural' fields of
characteristic zero (Q, R, C), for whose study, it seems, the
in disp ensable instrumentarium is yet to be constructed.
Here, technique encounters one of those paradigmatic
conceptual fissures that may be of the greatest interest
for the philosophy of mathematics. Now, the leap to the
analytic (L-functions) gives us a better understanding of
certain fragments of number theory, but then, once this
leap has been made, we encounter major obstructions in
returning to what should be the natural structures of the
analytic (fields of characteristic zero). And so we find
ourselves faced with new obstructions in transit ( obstruc­
tions that are carefully maintained, in the scope of the
Langlands program, through a sophisticated theory of
the structural transfusions of form), revealing once again
the incessant presence of what, according to Thom, is the
'founding aporia of mathematics': that inherent contradic­
tion between the discrete and the continuous that drives
the discipline. Fundamentally, nothing could therefore be
further from an understanding of mathematical inven­
tion than a philosophical posture that tries to mimic the
set-theoretical analytic, and presumes to indulge in such
'antiseptic' procedures as the elimination of the inevitable

183
 OF CONTEMPORARY MATHEMATICS

contradictions of doing mathematics o r the reduction of the

continuous/discrete dialectic. In approaching the works of
a few of the great contemporary mathematicians, we see
how the labor of the 'real' mathematician (in Corfiel d's
sense) points in the exact opposite direction: toward the
multiplication of the dialectic, and the contamination of the
spaces of mathematical knowledge. At stake is a tremen­
dous leveraging of knowledge, which the philosophy o f
mathematics should begin to take stock of.
Given an algebraic group112 G and an L-function, one

can construct a new group La (the Langlands group) that

combines the absolute Galois group over the field underly·
ing G and a complex Lie group associated with L. What is
at issue here is a mixture (in the full Lautmanian sense) - a
mixture that helps us control the theory of representations
of G. Within this framework, some of the underlying tran.
sits in the Langlands program correspond to thefunctorial
fact (plausible, correct in particular cases but generally
undemonstrated) that every morphism LG - LG issues
from a morphism between the associated automorphic
forms, which is well behaved in every p-adic stratum of
representation.for almost every p. '73 We are faced here with

172 An algebraic group is an algebraic variety that has, in addition, the structure of a
group. Examples of algebraic groups include finite groups, linear groups GL(n), and
elliptical curves.

173 See R. Langlands, 'Where Stands Functoriality Today', in T.N. Bailey & A.W. Knapp,
Representation Theory andAutomurphit: Forms,
Proceedings of the Symposium on Pure
Mathematics 61, (Providence: American Mathematical Society, 1997), 457-71.

1 84
 E I DA L MATHEMATICS

a recu rring situation in advanced mathematics, one that

cannot be contemplated in mathematical contexts of a lower
level of complexity - a situation in which the general forms
ofknowledge (universality, functoriality) and the restricted
calculations underlying them174 (particularity, diophantine
objects) depend on a complex intermediary hierarchy that
Jordhly structures both the generic transits (going up) and
the specific obstructions (going down).
In an interview, Langlands sums up his perception
of mathematics:

I love great theories, especially in mathematics and

its neighboring domains. I fell in love with them,
but without really grasping their significance, when
I was still a student. [ ... ] What I love is the romantic
side of mathematics. There are problems, even big
problems, that nobody knows how to tackle. And
so we try to find afootpath that leads to the summit, or
that lets us approach it. [ . J I love having the impres­
..

sion of standing before a virgin continent. I love

the problems whose solutions require unpublished

174 We cannot evoke, here, the enormouscancreterichnessof the calculations, but it is enough
to imagine that they include the entire, enormous tradition of nineteenth.-.century and
early twentieth-century Gennan arithmetic Oacobi, Dirichlet, Eisenstein, Kummer,
Hilbert, Hecke, Artin, Hasse, etc.).

185
 SYNTHETIC P H I LOSOPHY

and unsuspected theories. In other words, I love the

mathematics that make us dream .11s

The ascent to the peaks of the eidal thus grants us privi

leged visions, like the one which Langlands had at age
thirty-one, and which he looks back upon with admira­
tion: 'That I had nevertheless arrived somewhere always
seemed to me a miracle [. . . ] that I had seen so much in
a single blow, I do not think that this will ever ce as e to
astonish me.'176 His astonishment is sincere and, to a cer­
tain extent, tied to the marvels of his discovery (a 'virgi n
continent') , but one can already observe ( though we will
make this more precise in chapters io and n) that, in con·
temporary mathematics, many great creators are able to
climb to the summits precisely in virtue of their capacity
to correlatively transit the world of the eidal, approaching
polarities, hierarchies, relative invariants and intermediary
structural correspondences in a systematic manner. What
we are dealing with here is a dynamics of the mathemati­
cal world that is completely different from what we can
find in elementary mathematics, whose low threshold
of complexity does not allow for those aforementioned
transformations to emerge. It is also completely different

175 S. Durand, 'Robert Langlands. Un explorateur de l'abstrait', Qµlbec Science 2000.

http://www.crm.umontreal.ca/math2000-1/pub/langlands.html (italics ours). We
will come back to the indispensable romanticside of mathematics in our finalchapters.

t76 Response to the Gold Medal ofthe French Academy '![Sciences (2000), http://sunsite.ubc.
ca/Digita!MathArchivejLanglands/misc/gror.ps .

186
 E I DAL MATHE MATICS

from what is ordinarily discussed in the analytic phi­

losophy of mathematics, whose dismemberment of math­
ematical objects, in order to procure a static and ultimate
foundation that would sustain them, does not allow us to
take stock of their incessant torsions - and so impedes
any approach to mathematics' transitory specificity (on three
levels: phenomenological, ontological and epistemologi­

cal, as we shall see later on) .

Langlands's revealing commentary on the difference
between the Taniyama-Shimura Theorem•?? and Fermat's
Theorem, shows the importance of a mathematical think­
ing that is attentive to eidal generic structures:

Fermat's Theorem is an unexpected consequence

of another theorem (Taniyama-Shimura-Weil). The
latter has to do with a coherent framework, whose
beauty in my eyes comes from the fact that it cor­
responds to an order I'm used to seeing in number
theory. On the other hand, according to my intuition
or imagination, Fermat's Theorem could have turned
out to be false without that order being disturbed' . 118

177 The Taniyama-Shimura conjecture (1955) suggests the equivalence (modulo L-series)
of elliptic curves with modular forms. Frey ( 1985) conjectured that a nontrivial Fermat­
style solutionx•+y•=x• would yield a nonmodular elliptic curve (a 'Frey curve'). Ribet
demonstrated (1986) the Frey conjecture, thus establishing the implication Not (Fermat)
= Not (Taniyama-Shimura), or, equivalently, Taniyama�Shiumura � Fermat. Wiles

(1993-4) demonstrated the Taniyama-Shimura conjecture for semistable elliptic curves

(among which appears Fermat's curve), thereby proving Fermat's Theorem. The full
proof ofTaniyama-Shimura, for all elliptic curves, was finally obtained in 1999 (Breuil,
Conrad, Diamond, Taylor).
178 Durand, 'Robert Langlands .. .' (italics ours).

187
 O F CONTEMPORARY MATHEMATIC S

General counterweights, pendular oscillations, coun­

terpoints in an abstract order, and hidden aesthetic har­
monies therefore guide the theoretical vision, whereas the
particularity of the concrete case may come to distract
it. It is in the equilibrium between the broadest abstract
universality and the most restricted concrete particular­
ity - that is to say, in the broad register of mediations
- that mathematical inventiveness emerges with all its
force, haloed with 'order and beauty, luxury, peace and
voluptuousness'. '79
William Lawvere (USA, b. i937) has been one of the
main proponents of a broad hierarchy of mediations within
the general horizon of category theory. Since his doctoral
dissertation,iBo Lawvere has consistently insisted on think.
ing differently, and has succeeded in situating himself in a
conceptual undergroundwhere the synthetic and the global
take precedence, setting himself at some distance from the
usual analytic and local approaches. Lawvere's thought
combines various strategies that aim to fully capture the
movement of mathematical concepts, thereby following the
lines of Grothendieck's work. Whether through an 'inver­
sion of the old theoretical program of modeling variation
within eternal constancy', or by breaking the 'irresolvable

179 C.Baudelaire, 'L'invitation au voyage, ' Fleursdu ma/ (1857).

i8o F. W. Lawvere, FwictorialSemantics efAlgebrau Theories (New York: Columbia University
Press, 1963). A summary is given in 'IM Proceedings eftheNatWnal Academy efSciences
50, 1963: 869-72. Lawvere was a student of Eilenberg and Mac Lane, the creators of
category theory (Eilenberg, Mac Lane, 'General Theory of Natural Equivalences')

188
 E I DAL MATHEMATICS

contradiction' that is the 'metaphysical opposition

between points and neighborhoods', or by constructing
a topos theory that allows for a back-and-forth between
constant and variable sets,181 Lawvere seeks to build a
dynamic conception of mathematics, capable of captur­
ing both the physical world's continuous becoming and
the continuous enfolding of the relational canopy spread
over it.182 As we shall see, Lawvere's procedure consists in
elevating, into the eidal, a sophisticated web of weavings
and oppositions between concepts, calculi and models,
extending them both vertically (hierarchically) and hori­
zontally (antithetically). The synthetic comprehension of
that web, 'under the guidance of that form of objective
dialectics known as category theory',183 is a fundamental
contribution to contemporary thought - one that is still
in the process of transcending the strictly mathematical
horizon of its origin.
One of the few great contemporary mathematicians
who has dared to interpolate his writings with vague and

181 F. W. Lawvere, 'Continuously Variable Sets; Algebraic Geometry = Geometric Logic',

in Proceedings efthe Logic Colloquium 1973 (Amsterdam: North·Holland Publishing Co.,
1975), 135-56.

182 directed
Lawvere was, first, a student of Truesdell in continuous mechanics; although he later
himselftoward the categorical foundations ofmathematics and logic, he has
always given preference to the categorical instruments dealing with transit and flux
(variable sets and sheaves, beyond the static and the punctual) as being better suited
for an understanding of the physical world.

183 F. W. Lawvere, 'Introduction', in Proceedings efthe Halifax Conference on Toposes,Algebraic

Geometry and Logic, (New York: Springer, '972), 1.

189
 SYNTHETIC PH ILOSOPHY

nondiscip!inary184 references as a source of subsequen1

technical refinements, Lawvere has been able to construct
an uneasy equilibrium between the seer, who peers into
abysses ('gauging to some extent which direction s of
research are likely to be relevant'18S), and the climber,
who secures every step in his ascent (confronting some of
the major technical challenges of the age). In a remark­
able article on the 'future' of category theory,'86 Lawvere
characterizes the theory of categories as

the first to capture in reproducible form an incessant

contradiction in mathematical practice: we must,
more than in any other science, hold a given object
quite precisely in order to construct, calculate, and
deduce; yet we must also constantly transform it into
other objects.'87

Category theory's capacity to axiomatize, with great

precision,188 the fundamental weaving between static con­
siderations (states, points, objects) and dynamic ones
(processes, neighborhoods, morphisms) is one of the dee p

184 Including up-front and provocative mentions of Engels, Lenin, or Mao.

185 F. W. Lawvere, 'Adjointness in Foundations' in Dialectica 23, 1969: 281-96: 281).

i86 F. W. Lawvere, 'Some Thoughts on the Future of Category Theory', in Proceedings of
the Como Meeting on Category Theory, (New York: Springer, i991), i-13.
187 Ibid., I.

188 Lawvere speaks of the 'crystalline philosophical discoveries that still give impetus w
our field ofstudy', ibid.

1 90
 E IDAL MATHE MATICS

reasons for its success. The theory presents a permanent

back-and-forth between the three basic dimensions of
the semiotic, emphasizing translations and pragmatic
correlations (functorial comparisons, adjunctions'89) over
both semantic aspects (canonical classes of models) and
syntactic ones (orderings of types). In Lawvere's vision,
we find opposed - and, in fact, springing from that very
opposition - two classes of categories corresponding to
'Being' and 'Becoming', between which is set to vibrat­
ing a 'unity and identity of opposites'•go that gives rise to
remarkable mathematical conjectures in an intermediate
terrain between the ascent to the general ('from below,
from real space') and the descent to the particular ('from
above, by classifying abstract algebra').191 A 'descending'
functor sends a given category to a smaller one and, in
two linked movements of oscillation (two adjunctions,
see figure 8), two skeletal counterpoints appear (in 'posi­
tive' and 'negative'192 ) as extreme images of Becoming.
The category is set.free by this back-and-forth: the skeleton
surfaces as a static filtration (in Being) efier its immersion

189 A categorical adjunction generalizes a residuation in an ordered set (which is to say,

a pair of morphisms/,g, such thatf:r S y iff x S gy). Adjunction consists in a pair of
functorsF, G, such thatMar(FX,Y) "'Mar(X,GY) (natural isomorphism). Residuations
run through all of algebra, and, in particular, the algebra of logic, giving rise to
implication and the existential quantifier. In category theory, adjunctions, linked to
free objects, appear even more ubiquitously.
190 Ibid., 2.
191 Ibid., 12 .
192 Ibid., 7.

191
 OF CONTEMPORARY MATHEMATICS

in a dynamic fluid (in Becoming). The skeletons (positive

and negative), together with the descending functor, form
a 'unity and identity of opposites', in which what appears
as contradictorially fused on one level can be separated and
distinguished on another. The descent to the abyss hap­
pens to be perfectly controlled by a hierarchical strategy
- levels and contexts, which is to say, functors and catego­
ries - in a way that is quite close to Peircean pragmatism.

{�"?���>i&\ , '
· · positive
g s kelet�� ·� ; , ,
higher
' ., category
· (large)
.
.

' ,. ' .·· ·.

level

t I
I
descent
I
left I adjoint

unity and identity of opposites

lower
category
(small)

Figure 8. 'Unity and identity efopposites': descents, levels, andskeletons, according

toLawvere.

1 92
 E I DAL MATHE MATICS

Lawvere's examples cover a multitude of modem and

contemporary mathematical fields: general and algebraic
top ology; algebraic and differential geometry; abstract
algebraic structures - categories, lattices, rings - nonclas­
sical logics; functional analysis; mathematical physics; etc .
Within the framework of that collection of concrete situ­

ations, Lawvere nevertheless performs an incessant eidal

exercise: knowledge emerges in the wake of a meticulous
strategy of descent and ascent, combined with a dialectic
ofcontrapositions. In Lawvere's words, 'This explicit use
of the unity and cohesiveness of mathematics sparks the
many particular processes whereby ignorance becomes
knowledge' .193 The processes of weaving, the progressive
freeing of objects,194 the contraposition of opposite skel­
etons allows us to stand at a distance, decant the objects,
and look on with other eyes.
The union between the static and the dynamic antici­
pated by Novalis is realized with the greatest originality
in the theory of categories . A natural combinatorics of
levels allows a single object to be represented as simultane­
ously fixed (in a given category) and variable (through

193 Ibid., 2.

194 Afree object in a category has a precise technical definition, but can he seen as a universal
object - disincarnate and skeletal - with astonishing projective ductility; it has the
capacity to project, in a unique manner, its entire formal structure onto any other object
in the category having no more than a similar basis. Free objects allow for a remarkable
formal passage from the part to the whole, in mathematical situations involving rich
possibilities of transit (which are not always given, since not every category possesses
free objects). Within this framework, an adjunction can be seen as a generic) uniform
and cohesive process of constructing free objects.

1 93
 SYNTHETIC PHILOSOPHY

its functorial transformations) . Schlege!'s aphorism that

links universality and transformation - 'the life of the Uni­
versal Spirit is an unbroken chain of inner revolutions' ,'95
yet another form of the 'Great Chain of Being' and the
infinite 'Tree of Knowledge' studied by Lovejoy - is, in
category theory, incarnated with precise sophistication.
The universals of the theory are always dynamic univer­
sals, never rigidified in a fixed Absolute. The entire set
of categorical instruments - composition, morphisms,
natural transformations, sketches, limits, adjunctions,
sheaves, schemes, etc. - is converted into an enormously
powerful technical arsenal that unexpectedly revitalizes
the romantic dialectic between Being and Becoming. We
are not far here from the 'romantic side' of mathematics
divined by Langlands.
Within this very general dynamic framework, Lawvere's
work manifests concrete distillations offorms in multiple man­
ners. An admirably simple fixed-point argument in closed
cartesian categories lets us interlace Cantor's Theorem
card(X) < card( f.JX) , Godel's Incompleteness Theorem,
Tarski's satisfiability theory and the occurrence of fixed
points in complete lattices.196 A subtle understanding of
certain exactness properties in topoi allows us to interpret

195 F. Schlegel, PhilosophUa/Fmgments, tr. P. Firchow (Minneapolis: University of Minnesota

Press, 1991), 93 (fragment 451).
196 F. W. Lawvcre, 'Diagonal Arguments and Cartesian Closed Categories', in LectureNotes
in Mathematics 92 (New York: Springer, 1969), 134-45.

1 94
 E I DAL MATHE MATICS

quantifiers as adjoints, describe the logical behavior of

sheaves, and launch a program for the geometricization
of logic in which, to everyone's surprise, intuitionism
takes a central place. '97 An elementary axiomatization
of topoi (with regard to the subobject classifier) allows
us to reinterpret topologies as modal operators, abstract
the properties of Grothendieck topologies by means
of closure operators f over the subobject classifier, and
define an abstract notion of sheaf with respect to those
'topologies' f.198 These are all examples of how the various
subfields of mathematics tend to contaminate one another,
and how certain formal subspecializations transit simul­
taneously through logic, geometry and algebra. Indeed,
we find ourselves before a genuine ontic/epistemic break,
emerging in modern mathematics and fully defined in
contemporary mathematics, in which the old subdivisions
of the discipline tend to disappear.

197 F. W. Lawvere, 'Quantifiers and Sheaves', in Actes du Congres international des

maJhimaJiciens 1970, (Paris: Gauthier·Villars, 1971) 329-34. Lawvere explicitly indicates
that 'in a sense, logic is a special case of geometry' (329). Th.is is the announcement
of a growing geometricization of logic, some of the manifestations of which we will
review in chapter 7 (Freyd, Zilber), and which constitutes a renewed geometrici.wtion
efcuntemporary mathematics, as we have indicated in chapter 2. It may be noted that
the Nice Congress, where Lawvere proposed his program, is also where Grothendieck
announced his premature abandonment of the mathematical world. 'Though it could
not be clearly seen at the time, the Frenchman's ideas were to undergo a dramatic
reincarnation at the hands of the North American.

198 F. W. Lawvere, 'Introduction', in LectureNot,,, in Mathematics 274, (New York: Springer,

1972), 1-12. Lawvere and Tierney axiomatized the topoi (coming from the Grothendieck
school) in elementary terms, which is to say, without requiring conditions of infinity
or choice. A subobject classifier generalizes the idea of power set, and allows the
subconstructions of ordinary set-theoretical mathematics to be governed through the
universal behavior of the morphisms in play.

195
 O F CONTEM PORARY MATHEMATICS

T h e work o f Saharon Shelah (Israel, b. 1945) offers

new and profound technical arguments for understand­
ing a powerfully stratified mathematics, imbued with
multiple transversal tensions, and attends to the study
of that stratification's structural limitations, both in its
horizontal levels and along its vertical skeleton. Lying
far beyond the Godel's Incompleteness Theorem (regard·
ing the deductive limitations of theories whose threshold
of complexity meets or exceeds that of Peano arithme­
tic), Shelah's nonstructure theorems (1980-90) uncover
the semantic limitations of natural classes of models in
advanced mathematics.'99 Shelah's results reveal an unex·
pected polarization in the study of classes of models for a
classical theory T. His Main Gap Theorem shows that the
number of nonisomorphic models (of a given size) for
T faces an abrupt alternative: either it literally explodes,
reaching the maximum possible number of models, or
else it turns out to be perfectly controllable. There is no
place for semicontrol or semi-explosion: either the class
of models for T cannot count on having any structure
at all (all of the possible models are given, everything
that can possibly exist also exists in actuality); or else the
class can be fully structured (all that actually exists also
exists in a 'coordinated' form, according to subtle scales

199 Sec the works collected in the monumental (73g-page) text, S. Shelah, Classification
7heory and the ){umber ofNtmisomorphic M�ls, (Amsterdam: North-Holland, i990) .

1 96
 E I DAL MATHEMATICS

of invariants). Indeed, the impetus behind Shelah's most

original ideas is ultimately the acquisition of a general
theory ef dimension (subtle invariants for the structured
case). His 'theory of excellence', in particular (the but­
tress of the difficult part of the Main Gap proof, which
took ten years to complete) requires a series of 'algebraic'
interactions in arbitrarily high finite dimensions, which
by far transcends the usual 2-dimensional interactions
that show up in the independence proofs of traditional
algebraic geometry. 200 We thus find ourselves faced with
yet another situation in advanced mathematics where a
leap in complexity gives rise to new mathematics, which
are not at all reflected in the lower strata.
After detecting the generic presence of the gap in the
set-theoretical universe, the tremendous labor of Shelah
and his team201 was devoted to describing many observ­
able concrete conditions, in order to be able to detect
whether an arbitrary theory can be classified as structure

200 Thanks to Andres Villaveces for these clarifications. According to Vtllaveces (personal
communication), 'There is a quantity of structures "waiting to be discovered", in
geometry and in algebra, that demand that the algebraic interaction accounts for all
those high-dimensional diagrams. [ . .. ] Already in group theory they're beginning to
make 3-dimensional amalgams. They are very difficult and correspond to group-theoretic
properties that are truly more profound than most of the traditional ones. [ . .. ] It is sort
of as if, in geometry, we had been working, up to now, with a .. 2-dimensional projection"
of phenomena that would have seemed more narural if we had contemplated them in
their true dimension.'
201 Shelah's charisma has given rise to a genuine logic workshop, distributedover numerous
countries. The work of Shelah and his collaborators is nearing a thousand articles,
almost altogether implausible in the mathematical world. In virtue of the profundity
of his general ideas, his technical virtuosity, the tenacity of his daily work, and his
influence in the community, Shelah can easily be seen as the greatest logician of the
twenty-first cenrury.

197
 SYNTHETIC PHILOSOPHY

or nonstructure. The classes of models for such a theory

range, in principle, between two extremes: proximity to
a Morley-style categoricity theorem, where the mod.els
turn out to be isomorphic by strata, 202 or else freedom
from any structural restriction. A first dichotomy for
classifying classes of models distinguishes between stable
and unstable theories. The deep mathematical meaning of
stable theories comes from the structure of the complex
numbers C, with addition and multiplication, and from the
algebraic geometry that can be undertaken within it. The
current notions of dimension and algebraicity extend to
stable theories and can be used as natural logico-algebraic
invariants for 'coordinating' the models of those theories.
By the same token, unstable theories are theories in which
certain generic orders can be defined; 2"3 in that case the
classes of models tend to disintegrate and their diversity
explodes. One example that is now being thoroughly

202 The Morley Theorem (1965) asserts that a denumerable first-order theory that is
categorical in a nondenumerable cardinal lC (that is to say, such that all of its models
of size K are isomorphic) is likewise categorical in every cardinal greater than K. This
is the strongest possible result of the 'collapse' of the infinite for first-order theories
(collapse by strata), since, by contrast, from one stratum to another, far from collapsing,
the models are multiplied, due to the properties of first-order logic (compactness,
LOwenheim-Skolem).

203 We face a situation contrary to that ofC, in which we cannot define an order congruent
with the operations. The fact that complex numbers are not an ordered field, for
many decades seen as an important limitation in the architectonic construction of
sets of numbers, has today come to he seen as a strength (as a reason for stability).
The pendulariJy of the mathematical understanding is evident here. No description
of the ontic richness of C should be allowed to neglect this fundamental oscillation.
Nevertheless, not only has this pendular movement gone unstudied in the analytic
tradition; its existence has not even been registered! This is one example, among many
others in advanced mathematics, thatJon;es us to change our philosophical perspective,
if we are really to be in a position to accept the advances of the discipline.

198
 E I DAL MATHEMATICS

studied (and that we will come back to i n chapter 7,

when dealing with Zilber) is the structure of the complex
numbers, with addition, multiplication and exponentiation
added on; the complex exponential in fact introduces a
sophisticated hierarchy of analytic submodels that remains

beyond the control of first-order logic, and the theory

turns out to be profoundly unstable. 2°4
Beyond the stable/unstable dichotomy, Shelah's pro­
gram broaches the problematic of describing and studying
many other dividing lines by which the Main Gap may
be refined, with important mathematical (and not just
logical) content on each side of the division. One robust
dividing line is the dichotomy: superstable + non-DOP +
non-OTOP / not superstable or DOP or OTOP, with pow­
erful structure theorems on the superstable + non-DOP
+ non-OTOP side, where the models can be analyzed by
means of trees of countable models. 2°5 We then see that a
general eidalpolarity can come to be incarnated in multiple
concrete polarities.

204 For complements and clarifications, see the existent overview, A. Villaveces, 'La tensiOn
entre teorfa de modeios y anilisis matematico: estabilidad y la exponencial compleja',
Boletin de Matemiiticas Nueva Serie XI, 2 004: 95-108.
205 Omitting types order property (OTOP) indicates that a certain order that is notdefinable
in first-order logic becomes so through the omission of types (in the logic Lwiw of
the greatest expressive power). Dimensional order property (DOP), is another form
under which an order that is hidden from the eyes of first-order logic may be expressed.
Thanks to Andres Villaveces for this information.

1 99
 O F CONTE M PORARY MATHEMATICS

In a prospectus on the future of set theory, 206 Shelah

reflects that the main sources of interest in the theory's
development have their roots in its beauty (earning it n ine
points), its generality (six points), its concrete proofs (five
points), and its wealth of internal developments (fou r
points) .2°7 He shores up this polemical vision by adding,
'My feeling, in an overstated form, is that beauty is for
eternity, while philosophical value follows fashion'. 00s
For Shelah, beauty is rooted in 'a structure in which
definitions, theorems and proofs each play their part
in the harmony'. 209 Even though many of Shelah's major
theorems exhibit, analyze and synthesize the inharmo­
nious and nonstructured behavior of certain classes of
models, it should be observed that, in its entire concep­
tion, there nevertheless exists a pendular counterpoint
between the structured and the lack of structure, and
that oscillation is in itselfprofoundly harmonious. Once
again, transfusions and obstructions of forms, with global
equilibria and incessant local tensions, govern the main
outlines of a decisive body of work in contemporary

206 S. Shelah, "The Future ofSetTheory' (2002), http://arxiv.org/pdf/math.L0/02u39].p<lf.

207 Shclah completes the list with other minor sourees of interest (in descending order):
applications, history, 'sport', foundations, philosophy.

208 Ibid., 2. Note that the major exponent ofset theory neglects its supposed philosophical
value. 1his fact should cause philosophers who see the philosophy of mathematics
only in the philosophy of set theory to seriously question their perspective.

209 Ibid. As examples of beauty, Shelah proposes Galois theory (and 'more exactly whar is
in the book of Birkhoff-Mac Lane') and the Morley Theorem {with its proof). Observe
how a great mathematician insists on theform of proofs and on their exposition; once
again the fundamental pertinence ofstyle in mathematics enters the scene.

200
 E I DAL MATHEMATICS

mathematics. As often happens with great creative math­

ematicians, advances in one direction of thought find
a counterpoint in unexpected advances in the opposite
direction. After becoming a specialist in relative consistency
results in set theory,2 10 and above all after his very difficult
demonstration of the independence of the Whitehead
11
problem, 2 Shelah turns toward a new understanding
of approximations in the infinite, with an ambitious
program of cardinal arithmetic in which he proposes to
redesig;n adequate invariants for the operations.212 Given
that cardinal exponentiation effectively turns out to be
a great obstruction to cardinal arithmetic - due to the
'wild' behavior of 21( for an infinite cardinal K (Cohen's
1963 independence results) - Shelah proposes to seek
out a robust alternative skeleton for infinitary operations.
Shelah finds the supports of that skeleton in his PCF
theory (an acronym for 'possible co.finalities'), in which

210 Given a statement <(J and a subtheory T of the ZF set theory, cp is relatively consistent
with T ifCon (T) � Con ( T+q>), where Con(.!:) means that the throry is consistent, which
is to say, that one cannot deduce a contradiction from L. rp is independent from T if
both qJ and •'P are relatively consistent with T. Godel (1938) began this line of study,
with the relative consistency of the continuum hypothesis with respect to ZF. By other
routes, relative strategies later came to be converted into one another, as we have seen,
in one of Grothendieck's major predominant tendencies.

21! The Whitehead problem (1950) aimed to characterize a.freeabelian group A by means
of a condition on its contextual behavior (residual A condition: for every morphism
g over the group A, with nucleus Z, there exists a section s such that gs=idA.)· Shelah
demonstrated that, for abelian groups, the conjecture IJ (A is residual � A is free) is
independentof'lF set theory, since, on the one hand, V= L implies IJ (hence Con (ZF+IJ)),
and, on the other hand, MA + -,HG implies -,IJ (hence Con(ZF+-,IJ )).

212 S. Shelah,ProperandlmproperForcing, (Amsterdam: North-Holland, 1992); S. Shelah,

Cardinal Arithmetic (Oxford: Oxford University Press, 1994)·

201
 SYNTHETIC PHILOSOPHY

he introduces a web of tame algebraic controls013 for cardinal

cofinalities,214 and discovers that, beneath the erratic and
chaotic behavior of exponentiation, lies a regular behavior
of certain reduced products by means of which the upper
reaches of the cardinals can be approximated. What we
encounter here, again, is the construction of an intermedi­
ate hierarchy that allows for the relative adequation of
transit and the discovery of its proper invariants.
A sort of correlation ( PCF/cardinals = algebraic topol­
ogy/topology) issues from Shelah's works. In fact, the
search for a robust skeleton (co finalities) and a tame
algebraic calculus (reduced products) in the theory of
singular cardinals corresponds to the idea of seeking
natural algebraic invariants (homotopies, homologies) for
topology. This closes one of the circles we entered in this
chapter, beginning with Serre's works on the homotopy of
spheres. 215 If perpetual transfusions of form have been the
213 The strategy of a 'tame' mathematics, which neglects certain singularities (such as those
artificial counterexamples ofgeneral topology, based on the axiom of choice, or such
as cardinal exponentiation), goes back to Grothendieck (see p. 147 n. 102).

214 The cqfinality co(1<) ofa cardinal K is defined as the minimal cardinal of the cofinal subsets
in (the order of) K. A cardinal is regular if it is equal to its cofinality (example: K0+1) and
singular otheiwise (example: Kw}· PCF theory helps to control the subsets of a singular
cardinal through cofinalities, something which cannot be done through exponentials.

215 The circle is closed even tighter ifwe observe (Villaveces, personal communication) the
parallel PCF/cardinals := schemes/varieties. In effect, cardinal arithmetic is localited in
PCF, by controlling cofinalities around fixed cardinals and then 'gluing' the infonnation
together - in a manner similar to that by which schemes help to localize the arithmetic
of varieties, by controlling local rings over prim.es and then 'gluing' the information
together. In the third part of this essay, we will come back to the crucial impurtance
efsheaves - underlying processes of localization and gluing, or, more generaHy, of
differentiation and reintegration - in order to try to capture the passage from modern
to contemporary mathematics in an intrinric fashion (and not merely diachronically1
in the environment 1940-50) .

202
 E I DAL MATHE MATICS

fundamental motifof this chapter, we have also been able

to contemplate a rich multiplicity of concrete modulations
in which this motif is diversely incarnated. The majority
of these transits have taken place in the world of the eidal,
in the vast space of the mathematical imaginary.
In the next chapter we will see how, in tum, those
ascents of mathematical inventiveness succeed in descending
once again into the physical world, in the most unsettling
ways possible. In fact, in the last thirty years, mathematical
physics has been impregnated with an extraordinary host
of abstract methods from high mathematics (to a large
extent through a rejuvenated perspective on Grothendi­
eck's work, thanks to the Russian school), the technical
consequences of which we are just now beginning to
glimpse, and the philosophical consequences of which
may tum out to be utterly explosive.

203
CHAPT E R 6

Q.U I D DI TAL M AT H EMATI C S : ATIYA H , LAX,

CONNE S , KONTSEVI C H

Since its very beginnings, mathematics has been very

close to physics. The observation of natural phenomena
has sought to avail itself of the mathematical apparatus
at every moment of history. Mathematics, which has
always been described as the universal language of the
sciences, had, at the beginnings of modern mathematics
(thanks to Riemann's spectacular revolution) come to
understand itself as a sort of structural machinery for the
sciences. In Riemann's vision, far from being reduced to
a mere language, an expedient that would serve only to
display what other sciences had discovered, mathematics
would in fact be the discipline that allows us to codify
the deep structures underlying the natural world. The
situation was complicated even further in the last half of
the twentieth century, with some of the most formidable
advances in contemporary mathematics. As we shall
see, certain arithmetico-combinatorial structures (the
Grothendieck-Teichmiiller group) may come to govern
certain correlations between the universal constants of
physics (the speed of light, the Planck constant, the
gravitational constant), while, conversely, certain math·
ematical theories originating in quantum mechanics

205
 SYNTHETIC PHILOSOPHY

(noncommutative geometry) may help to resolve difficul t

problems i n arithmetic (the Riemann hypothesis) . We are
dealing here with absolutely unanticipated results, which
bring together the most abstract mathematical inventions
and the most concrete physical universe. The problems
that these new observations pose for the ontology o f
mathematical objects are enormous: Where do these
objects 'live' - in arithmetic or in the physical world? Can
this alternative really be contemplated? Can we instead
elaborate a transitory, nonbipolar ontology? We will be
sure to tackle them in part 2 of this work.
In this chapter, we will use the neologism quiddita l
(from quidditas [what there is]) to designate the process of
descent from the highly abstract constructions of contem­
porary mathematics and their application to the physical
world (what there is). This Being is subdivided by a tense
contraposition between essence (ousia) and existence
(huparxis), and by a counterpointing'' 6 oftranifusions ofthe
real that should remind us of the mathematical dialectic
between essence and existence studied by Lautman.

216 A fundamental neologism introduced by Fernando Ortiz (1940), in his understanding

of Latin America. See his Contrapunteo cubano del tabaco y el atUcar (Havana: Jesus
Montero, 1940) [Tr. H. de Oniz as Cuban Caunte7paint: Tobacco and Sugar (New York:
Alfred A. Knopf, i947)].

206
 Q,UIDDITAL MATHE MATICS

111 11 --
11 111
--1111 -----Ill"•
"""
1::1 1•ll""'-------11---
·,111·111·

quiddita.s

\ t
'that which is'

I I

' '
o>""�'�""�"'°'"""'a'�"'.m,,.,,,j,•»ci.l\>o;�;!

Figure 9. The Qpiddital Realm.

Tue works of Sir Michael Atiyah (England, b. 1929) have

given a decisive impulse to the possible application of
contemporary mathematics' sophisticated instruments
to the understanding of associated physical phenomena.
Awarded the Fields Medal (1966)217 and the Abel Prize
(2004), Atiyah is best known for his famous Index Theo­
rem (in collaboration with Singer, 1963),218 a very deep
result that can be considered one of the major theorems of
217 The 1966 Fields medalists were Atiyah, Cohen, Grothendieek and Smale: an impressive
mathematical revolution in full bloom!

218 M. Atiyah & I. Singer, 'The Index of Elliptic Operators on Compact Manifolds',
Bulletin eftheAmerican Mathematical Society 69, 1963: 322-433. Further developed in
M. Atiyah & I. Singer, 'The Index of Elliptic Operators 1-V', Annals efMathematics 87,
1968: 484-604 and 93, 1971: n9-49.

207
 O F CONTEMPORARY MATHEMATICS

the twentieth century. The theorem effectively combin es:

a statement of great simplicity and universality (a pre­
cise interlacing of transferences and obstructions in the
domain of elliptic equations); a very diverse collection
of proofs, drawn from apparently contrasting realms of
mathematics (K-theory, Riemann-Roch theory, cobord­
ism, heat equation, etc.); a remarkable radiation into a
very broad mathematical spectrum (differential equatio ns,
mathematical physics, functional analysis, topology, com­
plex analysis, algebraic geometry, etc.).
In rough, conceptual terms, the Index Theorem states
that the balance between transits and obstructions in
certain changes in nature is completely characterized by
the geometry of the environment where this change takes
place. In more precise terms, given an elliptic differential
operator ( 'change'), the index ('balance') - defined as
the number of solutions ('transits') minus the number of
restrictions ('obstructions') for the operator - is completely
determined by adequate topological invariants ('geometry
of the environment'). I n still more precise terms, if we
give ourselves an elliptical differential operator219 D, and
if we define an analytic index for D with indarwlyt(D)=dim
Ker(D) - dim coKer(D) (kernel Ker(D) : 'solutions' - that is,

219 Elliptic operators (whose coefficients in higher-order partial derivatives satisfy a suitable
condition of positivily) appear ubiquitously in mathematics: the Laplace operator
{y�+ ... y�) associated with the heat equation; the Toeplitz operator (given a continuous
f, take tbe holomorphic part in tbe multiplication off by a holomorphic), associated
with the Cauchy-Riemann equations; the Fredholm operator (derivation in the tangent
bundles of a marrifold), associated with ellipticity equations in fibers.

208
 Q.UI D DITAL MATHEMATICS

harmonic functions; cokemel coKer(D): 'obstructions' -

that is, restrictions in nonhomogeneous equations of the
type Df=g), the Index Theorem asserts that the analytic

index can be characterized by means of a topologi,cal index

ind1"P(D) linked to purely coho mo logical invariants220 of

the operator's geometrical environment. A fact of the

greatest interest that follows from this theorem is that we
can observe how the solutions and obstructions, which
are separately completely unstable (owing to the great local
variation in the differential equations) , nevertheless tum
out to be stable in their difference (global unification of the
indices). We will come back to the profound philosophical
interest in results like this, which, again, do not exist in
elementary mathematics. But we can already sense the
richness of a philosophical approach that really takes
seriously the dialectic of contrapositions of mathematical
flows, demonstratively incarnated in the Index Theorem.
The Index Theorem affords a striking quiddital transit
between analysis and topology, with an sorts of appli­
cations, since elliptic equations serve to model many
situations in mathematical physics. But even more aston­
ishing is that transit's eidal support, rooted in the hidden

220 This is a sophisticated invariant that involves, among other constructions, the Thom
isomorphism between the homology groups of a manifold and their cotangent space
(modulo boundary), the Chern characters coming from K-theory, and the Todd
classes of a manifold. Without being able to enter into technical details, we can see
how eminently abstract concepts thus appear, typical of the transfusions efform that
we were able to contemplate in the previous chapter, and which are here applied to
the transfusions of the real.

209
 SYNTHETIC PHILOSOPHY

algebraic-geometricalfoundations underlying the theorem's

proof. The genesis of the Index Theorem is revealing in
this sense. The Riemann-Roch Theorem (1850) proposes
to algebraically (dimensionally) control the space of
meromorphic functions associated with a curve through a
topological control (the genus) of the Riemann surfaces
associated with the curve. It thus presents a first great trans·
lation of mathematical concepts, which yields the general
problematic of how the solutions to certain linear systems
of algebraic parameters can be controlled through appro­
priate topological invariants. The generalizations of the
Riemann-Roch Theorem would later turn out to be legion,
and they are situated at the (diachronic) frontier between
modern and contemporary mathematics: Schmidt (1929,
for algebraic curves) ; Cartan-Serre-Hirzebruch ( 1950-56,
for one system of sheaves); Grothendieck (1957, for all
systems of sheaves with algebraic parameters: K-theory) ; 221
Hirzebruch-Atiyah ( 1958, for all sheaves with continuous
parameters: topological K-theory) . As a consequence of
these advances in the genericproblematic of transit between

221 Grothendieck's K·theory (K for Klassen: the study of classes in their totality) aims to
study algebraic classes of sheaves, and shows how to pass from natural structures of
monoids of sheaves to certain rings of sheaves (by way of formal inversions in the
sheaf's vectorial fibers). Out of K-theory emerges the famous conjecture of Serre
(1959) - every finitely generated projective module over K(X) is free - which was
settled in the affirmative in 1976, by Q.uillen and Suslin. (Quillen's 1978 Fields Medal
was, in part, due to that proof.) Compare the fortune of Serre's conjecture with that
of Shelah's work on the undecidability of the Whitehead conjecture: we should be
struck by the complexity of a universe in which such apparently similar statements
nevertheless find themselves so profoundly demarcated. It is a (romantic) abyss into
which the philosophy of mathematics must enter.

210
 Q,UIDDITAL MATHEMATICS

the algeb raic and the topological o n the basis o f complex

analys is (Riemann-Roch), Gelfand, in i960, proposes a
generic statement concerning the homotopy invariance
of the index. The rupture finally emerges in i963, when,
working with elliptic differential equations instead of
linear systems with parameters, and considering algebraic
functions as holomorphic functions satisfying ellipticity
(Cauchy-Riemann equations), Atiyah and Singer intro­
duce the radical shift in perspective that leads them
to combine the statement of the Index Theorem (a la
Gelfand) with the entire instrumentarium of concepts
inherited from Riemann-Roch (a la Grothendieck) .
In Cartan's presentation on Atiyah's work,222 which
he gave when the latter was awarded the Fields Medal,
and in Atiyah's later retrospective, 223 both mathematicians
underscore the importance of the analytic index inda1Wlyt
being stable underperturbations and how it is that we can,
thus, reasonably hope for a topological formula ind1,ap in
purely geometrical terms. Since K-theory (whether alge­
braic or topological) provides the precise mathematical
instrumentarium for capturing extensions of morphisms
between (algebraic or topological) structures linked to
perturbations, Atiyah remarks that 'K-theory is just the
right tool to study the general index problem' and that 'in

222 H. Cartan, 'L'oeuvre de Michael F. Atiyah' ( 1966), in M. Atiyah, D. Iagolnitzer, eds.,

Fields Medallists' Lectures (NewJersey: World Scientific, 2003), u3-18.
223 Atiyah, 'The Index of Elliptic Operators' (1973), ibid., 123-5.

211
 O F CONTEMPORARY MATHE MATI C S

fact, the deeper one digs, the more one finds that K-theory
and index theory are one and the same subject! '024 The
horizon we are facing here is similar to the one we passed
through in Grothendieck's work: incorporating a transit
between objects (variations, perturbations) so as to then
proceed to determine certain partial stabilities (invariants)
beneath the transit. Note that this general strategy, in
the most diverse subfields of mathematics, gives rise to
remarkable concreteforms oforigi,nal knowledge, forms that
go completely unnoticed in static realms where move·
ment is lacking (for example, in the realms of elementary
mathematics, where levels of complexity are low).
Atiyah remarks that 'any good theorem should have
several proofs, the more the better'. 225 Such is the case with
the Index Theorem, which, in virtue of its very centrality,
benefits from techniques from many domains; each proof
and each point of view goes on to amplify the 'freedom',
'variety' and 'flexibility' of the mathematician. 226 As Atiyah
describes it, mathematics, in such various decantations, 'is
always continuous, linked to its history, to the past', and
it 'has a unity', which nevertheless should not become ' a
straitjacket. Tue center of gravity may change with time.
It is not necessarily a fixed rigid object in that sense;

224 Ibid., 133, 134.

225 M. Raussen and C. Skau, 'Interview with Michad Atiyah and Isadore Singer', Notices
qfthe American MathematicalSociety 52, 2005: 225.
226 Ibid., 226, 227.

212
 Q.UI D DITAL MATHE MATICS

I think it should develop and grow'. 227 As we have come

to see, in advanced mathematics (whether modern or

contemporary) the transit of objects is vital, and the varia­

tion of the discipline's centers of gravity seems to be no
less inevitable. In the dynamics of mathematical activity,
Atiyah recalls that 'a theorem is never arrived at the way
logical thought would have you believe', that everything
is 'much more accidental', and that 'discoveries never
happen as neatly' as posterity thinks. 228 M oreover, no
reduction of mathematics to a series of logical analyses
can be anything but a (philosophically unacceptable)
impoverishment of the discipline.
The richness of a quiddital mathematics, in profound
proximity to reality, 229 is reinforced by the works of Peter
Lax (Hungary/USA, b. 1926) in two privileged realms
where the variability of the real is modeled with precision:
differential equations and the theory of computation. The
approach to the quidditas in Lax consists in a sort ofpen­
dular oscillation, the inverse of the movement that we have
observed in Atiyah. The latter produces a descent.from the
eidal into the quiddital from Atiyah's technical mastery in
-

algebraic topology we pass to its subsequent application

to the I ndex Theorem. Inversely, in Lax, a very concrete

227 Ibid., 225, 230.

228 Ibid., 225.

229 According to Atiyah, 'almost all ofmathematics originally arose from external reality'.
Ibid., 228.

213
 SYNTHETIC PHILOSOPHY

origjnary pragmatics in the quidditas leads to an ascent into

the eidal, so as to be able to then project itself onto the
fragment of reality with which it began. If, indeed, the
mathematical 'heart' of the effort to capture the physical
world is to be found in partial differential equations, and if
an adequate 'tomography' of that heart is to be found in
the computational calculations efthose equations' solutions,
then the knowledge lying at the intersection of those two
fields - precisely Lax's speciality - will help to describe
something of a 'real ground' of mathematics.
The citation of the 2005 Abel Prize awarded to Lax
underscores 'his groundbreaking contributions to the
theory and application of differential equations and to the
computation of their solutions'. 23° In Lax's own words,
these contributions can be sorted into four essential
subareas:231 integrable nonlinear hyperbolic systems ; 232
shock waves;•33 the Lax-Phillips semigroup in scattering

230 Portal of the 2005 Abel Prize, http://www.abelprisen.no/en/prisvinnere/2005/index

hUnl.

231 C. Dreifus, 'From Budapest to Los Alamos. A Life in Mathematics', 'IMNew rork Times.
March 29, 2005, http://www.nytimes.com/2005fo3f29/science/29conv.html.

232 Integrable systems are systems of differenti2' ...quations for which there exists a well­
defined collection of'conserved quantities' ( Jified in the spectrum of the differential
operator)� by means of which one can achieve a complete knowledge of the system 's
solutions. What we are dealing with here is yet another incarnation of one of the cenffal
problematics of modern and contemporary mathematics: the study of transits (here,
partial differential equations) and invariants (here, conserved quantities) in the realms
of exact thought.

233 Shock waves are perturbations that propagate energy in a given (usually fluid or
electromagnetic) medium, and that are characterized by an abrupt discontinuity in
their initial conditions. Supersonic waves constitute a paradigmatic case.

214
 Q,UI DDITAL MATHE MATICS

theory; 234 solutions of dispersive systems when dispersion

approaches zero. We can reinforce the pendularity between
Atiyah and Lax by observing how both mathematicians
cover complementary fragments in the universe of differ­
ential equations: elliptic operators (Atiyah; paradigm:
heat equation) and hyperbolic operators (Lax; paradigm:
wave equation). A hyperbolic system can be associated,
in a natural manner, with 'conservation laws', by means
of suitably integrating the flows inhering in the system's
evolution. In this realm, Lax works according to various
alternative strategies: a general theory and structure (trans­
formations that preserve the spectrum of a hyperbolic
differential operator) ; particular theories and structures
(entropy conditions in the case of shock waves, the 'Lax
pair' for comprehending solitons in the KdV equation), 235

234 Scattering studies forms of deviation in radiation that are due to certain deficiencies
of uniformity in the medium in which the radiation is propagated. Many forms of
dispersion in the physics of elementary particles (including X-rays) are paradigms of
scattering. Scattering techniques, in turn, allow radar photos to be interpreted.

235 The KdV equation (after Korteweg-De Vries, 1895: ui+u=x+6uu,=O) is one of the
best-known examples of a nonlinear hyperbolic equation_ The equation models the
behavior of waves in a fluid (a liquid surface, for example) and has been particularly
helpful for applications in naval architecture and for the study of tides. The KdV
equation gives rise to a completely integrable system, and its solutions (solitons) behave
well, since they can be described as solitary waves that uniformly displace themselves
in the medium, repeating the same pattern of propagation (Kruskal & Zabusky, 1965).
Knowledge of those solitons can be lineariud by means ofmethods that are inverse to
those of scattering, and the 'Lax pair' allows us to discover that inversion by means of
suitable noncommutative linear operators. We see here how a very concrete quiddital
situation (waves in water, the KdV equation) gives rise to an entire, subsequent eidal
architecture (integrable system, solitons, the Lax pair), which then subsequently
returns in the quidditas. But the richness of mathematical transits is not restricted
to just one direction. Indeed, Kontsevich (1992) has succeeded in demonstrating a
conjecture ofWitten's, according to which the generative function of the intersection
numbers of spaces over algebraic curves (moduli) satisfies the KdV equation. In this
manner, some of the most abstract constructions in mathematics are governed by an

215
 O F CONTEMPORARY MATHE MATICS

specific computational calculations (the 'proximity' of a

general system to an integrable system236) .
Lax emphasizes the importance o f 'looking at prob­
lems in the large and in the small', of 'combin[ing] bot h
aspects' and then benefiting from their combinatorial
'strength'.2s1 What we are dealing with here is an oscil­
lating conceptual equilibrium that Lax himself take s
his style to reflect, a style that seeks a certain elegance,
understood as revelation, simplicity and equilibrium
between the abstract and the concrete - an elegance
that ought to be reflected in the diversity of proofs an
important mathematical theorem deserves. In this man­
ner, the richness of mathematical thought, according to
a practitioner of the first rank of the discipline, is rooted
in the multiple transits of proof, and not in the.fixed state­
ment being demonstrated. We therefore, once again, see
how any logical reduction (or tautological reduction, in
the style of the early Wittgenstein) of a statement with
a high threshold of complexity would eliminate all of its
equation that is ubiquitous in mathematical physics! When we ]ater come to approach
the works of Connes and Kontsevich, we will see how the richness of the junctures
between abstract mathematics and physics is astonishing beyond measure.

236 Lax evokes 'the Kolmogorov-Arnold-MoserTheorem which says that a system near a
completely integrable system behaves as if it were completely integrable. Now, what
near means is one thing when you prove theorems, another when you do experimenls.
It's another aspect of numericalexperimentationreuealingthings.' (M. Raussen & C. Skau,
'Interview with Peter D. Lax', Notices qfthe American Mathematical Society 53, 2006:
223-9: 224; emphasis ours.) And so, extensive calculations in the quidditas hel p s lO
uncover structure in the eidos: a position quite close to Grothendieck's commentarie-s
concerning certain concrete cohomological calculations that would provoke the
structural emergence of motifs.

237 Ibid., 224.

216
 QUIDDITAL MATHEMATI C S

genuine mathematical content, encoded i n diverse and

contrasting structural proofs - in diverse and contrasting

calculative experiments.
The back-and-forth between calculation and structure,

between the physical world and mathematical abstrac­

tion, between the quiddital and the eidal, is, for Lax, an
indispensable process - one that explains the tremendous
vigor of mathematics:

My friend Joe Keller, a most distinguished appl ied

mathematician, was once asked to define applied
mathematics and he came up with this: 'Pure math­
ematics is a branch of applied mathematics.' Which
is true if you think a bit about it. Mathematics origi­
nally, say after Newton, was designed to solve very
concrete problems that arose in physics. Later on,
these subjects developed on their own and became
branches of pure mathematics, but they all came from
an applied background. As Von Neumann pointed
out, after a while these pure branches that develop
on their own need invigoration by new empirical

material, like some scientific questions, experimental

facts, and, in particular, some numerical evidence.
[ .] I
.. do believe that mathematics has a mysterious

217
 SYNTHETIC PH ILOSOPHY

unity which really connects seemingly distinct parts,

which is one·of the glories of mathematics. 238

The robust mathematical abstractions and large-s ca le

computations that Lax practicesfold back into one another.
The 'unity' - and the consequent 'glory' - of such transits
constitutes one of the specificities of advanced mathemat­
ics. Moreover, when the transit between the 'pure' and the
'applied' - moving openly and without any privileged
direction between both poles - breaks with the reasonable
expectations of the mathematical community, the 'glory'
and the 'honor of the human spirit' are intensified. This is
the case, as we shall now see, with the 'applications ' of the
Lax-Phillips semigroup in number theory, and, even more
surprisingly, as we shall see later, with Connes's strategy
in his (ongoing) effort to find a proof for the Riemann
hypothesis by means of instruments drawn from physics,
and with his work dealing with the 'appearance' of the
Grothendieck-Teichmiiller group in cosmology.
In their works concerning the spectrum of an opera tor
over a hyperbolic variety, Lax and Phillips introduce a
formal semigroup - a collection of operators Z(t) associ­
ated with orthogonal projections of waves over suitable
subspaces of the Hilbert spaceL2(R) - in order to control
238 Ibid., 225 (emphasis ours) . Von Neumann, the young Hungarian mathematician's
mentor in Los Alamos, is, for Lax, the exemplar to follow in mathematics: a powerful
vision and great calculative capacity, always breaking the supposed barriers between
pure and applied mathematics.

218
 Q,UIDDITAL MATHE MATICS

the scattering associated with the propagations of waves

( that is to say, the asymptotic behavior of waves in the

remote past or future) . 239 Fadeev and Pavlov later observe

(igri) that revealing connections with the harmonic

a nalysis of certain automorphic functions emerge when

we apply Lax and Phillips's theory to the non-euclidean
wave equ ation.240 Refounding scattering theory on non­
euclidean bases, Lax and Phillips go on to character­
ize the meromorphic properties of Eisenstein series,241
pro ducing explicit formulas and exhibiting proofs that
are b oth concise and general ( 'elegant', that is, in Lax's
aforementioned sense) , and that thereby uncover a totally

unexpected transit between the differential and the arith­

me tical, through the properties of the semigroup Z(t) .242
Nevertheless, in a subsequent revision of the theory,•43
Lax and Phillips explain how the connection between the
natural non-euclidean geometry modeled on the Poincare

239 P. Lax & R. Phillips, &attering'lheory (New York: Academic Press, 1967).
240 Given a wave equation u.n=c2V2u. (with the Laplacian V2=I:a2/an, the non-euclidean
wave equation is obtained by means of the perturbation un=c2V1u. +u./ 4.

241 Given the Poincare plane z (that is, z EC with Im ( z) >O) and given k � 2, the associated
Eisenstein series is defined by l:-m,n# (m + nz)-2k. This is a holomorphlc function that
converges absoluteJy on the PoinrarC plane, which turns out to be invariant under the
modular group SL2(Z) and which extends into a meromorphic function over C. The
remarkable Ramanujan identities, which the ingenious Indian mathematician proposed
with respect to the coefficients of the Eist:nstein series, are well known, and correspond
to sophisticated differential identities between the series.

242 P. Lax & R. Philips, 'Scattering Theory for Automorphic Functions', in Annals ef
Mathematics Studies 87 (Princeton: Princeton University Press, 1976).

243 P. Lax & R. Phillips, 'Scattering Theory for Automorphic Functions', Bulletin '![the
American Mathemati.cal Society, New Series 2, 1980: 261-95.

219
 O F CONTE M PORARY MATHEMATICS

plane (the group of rational transformations w --+ (aw+b) /

(cw+d) with a, b, c, d E R, and ab-bc==I and the various types
of natural invariants associated with that geometry L 2 ,
Dirichlet, Laplace-Beltrami)•44 turns out to be the deep
connection that lets us unfold the 'intrinsic meaning''4s
hidden in differential equations like the non-euclidean
wave equation, a meaning that can be glimpsed pre­
cisely in virtue of the semigroup Z(t). In this manner, we
observe how a complete mixture in Lautman's sense (the
Lax- Phillips semigroup) allows us to naturally mediate
between the (apparently distant) realms of the differentia\
and the arithmetical, thanks to the discovery of a single
natural model that pertains to both realms. That model is
the Poincare plane, seen as a non-euclidean model, with its
differential Riemannian geometry and analytic invariants,
on the one hand, and the same plane, seen as a complex
model, with its theory of automorphic functions and
arithmetical invariants, on the other. In situations of this
sort, we confront a sophisticated web of transits between
the quiddital and the eidal, with multiple contrasting
supports in the web: physical motivations (scattering,
waves); concrete models (Poincare plane, non-euclidean
geometry, modular forms), generic structures (geometries,
invariants, semigroups).

244 Ibid., 262.

245 Ibid.

220
 Q,UIDDITAL MATHEMATICS

A sophisticated web of motivations issuing from phys­

ics, a very broad spectrum of examples from functorial
analysis, geometry and algebra, and a powerful abstract
theorematic machinery are combined in the work of
Alain Connes (France, b. 1947): the concretization of a
rnathematics profoundly oriented toward the quiddital,
but which is also reflected in the (pendular and inevitable)
eidal transit - of higher mathematics. A Fields medalist
(1982) for his works on the classification of operator alge­
bras in Von Neumann algebras, and for his applications
of the theory of C*-algebras246 to differential geometry,
Connes has, from early on (since his doctoral thesis in
1973), worked on the unification of various, apparently
distant, abstract conceptual instruments (modular and
ergodic operators, projectivity and injectivity properties),
and on their manifold uses•47 in functional analysis and the

246 A C*-algebra is a Banach algebra (an associative algebra with a complete normed
tOPology) with an involution operator ( )* that behaves multiplicatively with respect
to the norm. The original examples of C*-algebras are matrix algebras (linked to
Heisenberg's matrix mechanics) and linear operator algebras over Hilbert spaces
(linked to quantum mechanics, following Von Neumann). Von Neumann algebras are
C*-algebras of operators that are closed under certain weak topDlogies. C*-algebras are
mixed mathematical objects in Lautman's sense, in which the linear and the continuous
are interlaced, through a hierarchy of intermediary properties that have to do with
convexness, order, identities and quotients. For a presentation of Connes'searlyworks.,
see H. Araki, 'The Work of Alain Connes', in Atiyah, Iagolnitzer, Fields Medallists'
Lectures, 337-4+
247 The weaving between the One and the Multiple isfully bipolar in Connes. Indeed, he
makes use of the abstract instruments of mathematics for applications in physics (the
use of C*-algebras for understanding quantum mechanics, sharpening Von Neumann's
program), but also, as we will later see, he makes use of the concrete instruments of
physics for 'applications' in mathematics (the use of spectroscopy for understanding
the Riemann hYPothesis) · As regards the evanesccnt.ftvntier between the pure and
the applied, recall Keller s paradoxical definition of pure mathematics as a branch of
applied mathematics, as evoked by Lax.

221
 SYNTHETIC PH ILOSOPHY

underlying mathematical physics. Connes subsequently

obtained an index theorem for foliations248 (1981), an d ,
in the eighties, began to develop his noncommuta tive
geometry. In the wake of such great unifying works as
those of Von Neumann, Grothendieck and Atiyah, Connes
opens mathematics onto research programs of tremendous
breadth. 249
The emergence of the noncommutative paradigm in
Connes rests upon three basic pillars: 2s0 the real (qui d ­
dital) ubiquity of spaces whose coordinate algebras are
non-commutative; the technical power of abstract (eidal)
instruments that can be extended to noncommutative situ­
ations (cyclic cohomology, K-homology, spectral theory,

248 A foliation is a differential manifold that is locally decomposed into parallel

affine submanifolds (the 'leaves' of the foliation). Foliations appear everywhere in
mathematics: given a immersion/:M ...... N between varieties with dim(M) ?dim (Nl
we obtain an n-foliation overM, the leaves ofwhich are the components of/1(x)
given a Lie group G acting on a manifold M in a locally free fashion (which is to say,
such that for all x in M, {gEG: gx=x} is discrete), the orbits of G make up the leaves
of a foliation over M; given a nonsingular system of differential equations, the farnil)
of solutions to the equation make up a foliation, and the global determination of the
solutions determines the behavior of the foliation. An index theoremfarfoliatiom, such
as the one obtained byConnes, therefore interlaces certain general constructions from
differential geometry with the underlying techniques of algebraic geometry in the
Index Theorem. Advanced mathematics thereby continues to soar over the incessant
transits between its subdomains.
249 Fittingly, Connes is (in Grothendieck's footsteps) a permanent professor at the HIES
(since •979), and (in Serre's footsteps) a permanent professor at the College de France
(since •984).
250 Connes is a magnificent expositor and defender of his ideas. See, for example, A.
Connes, Noncommutalive Geometry (San Diego: Academic Press, t994); A. Connc;
& M. Marcolli, A Walk in the Noncommutalive Garden, preprint, http://arxiv.org/abs/
math/06o1054. Connes'swebsite (www.alainoonnes.org) has a complete bibliographY,
available as an archive of pelf files. The descriptions appearing in our text come from
Connes's writings, with certain emphases added by us.

222
 Q.UIDDITAL MATHEMATICS

the 'thermodynamics' o f operators); the harmonic richness

of certain very general motivations:
Euclidean Geometry Commutative Terrestrial Physics

Non-euclidean Geometry Noncommutative Cosmic Physics

In fact, we see noncommutativity appear in a natural man­

ner in essential areas of physics (phase spaces of quantum

mechanics, cosmological models of space-time), geometry

(duals of discrete non-abelian groups, non-abelian tori,
foliation spaces), and algebra (adele spaces, modular
algebras, Q-lattices) . Here we discover a certain ubiquity
of the noncommutative in actual nature,251 which goes
hand in hand with an extension efthe notion efspace from
the point of view of its conceptual nature: the passage
from infinitesimal manifolds (Riemann) to C*-algebras
of compact operators (Hilbert, Von Neumann); the pas­
sage from dual K-homology (Atiyah, Brown, Douglas,
Filmore) to noncommutative C*-algebras (Connes); the
passage from the Index Theorem (Atiyah, Singer) to the
handling of noncommutative convolutions in groupoids
(Connes); the passage from the groups and algebras of
modern differential geometry (Lie) to quantum groups
and Hopf algebras;•s• the passage from set-theoretic

251 It would not be out of place here to recall Lautman's remarkable study of symmetry
and dissymmetry in mathematics and physics, as a sort of prelude to subsequent
noncommutative studies (see p. 59, above) .

252 Hopf algebras are structures that show up in the proofs of representation theorems
for algebraic groups (combinations of groups and algebraic varieties, such as linear
groups, finite groups, elliptic groups, etc.). Vladimir Drinfeld (1990 Fields medalist)
introduced quantic groups (1986) as nonrigid deformations of Hopf algebras, and

223
 OF CONTEMPORARY MATHEMATICS

punctuality to the actions of noncommutative mono ids

k .,
in Grothend"1ec top01 etc.
Constrained as we are to make a selection from the
many an d various results and subprograms of inves ti-
tion brought forward by Connes in his approach to
:e quiddital, let us emphasize two of them here : t he
of a 'cosmic' Galois group that is close to
emergence
the 'absolu Galois group in number theory, a form of
te'
een a well-known eidal configuration (abso­
transit betw
and an unexplored quiddital one (cosm ic
lute group)
e utilization of spectroscopic techniques in a n
group); th
dem onstrate the Riemann hypothesis, an inverse
effort to
ansit between the quiddital and the eidal. In a
form of tr
ticle, Pierre Cartier, one of Grothendieck's mai n
famous ar
ad conjectured that 'there are many reasons
disciples, h
ing in a "cosmic Galois group" acting on the
for believ
sh0wed
a
that they appear naturally i n the Yang-Baxter equation, a pivotal equ tion for
statistical mechanics. In turn, string theory in contemporary physics - the
ins of
� ian ut
opia of a harmony between the infinitely small (quantum mechanics)
e ;nfinitely
large (general relativity) - stands in need of a sophisticate d
andththmatical theory of knots, which can be adequately handled only by mean s of
ma
�
uan c g
roup s and n-categories (categories in which one climhs the scale qftrans1_ts -
q between morphisms [functors], and morphisms between functors
d morphisms
t>eyonal transformations], one studies morphisms between natural transfonnations1
n tu
�
[ � en
morphisms between those lower-level morphisms, and so on). Drinfrld·s
an rks (written at the age of twenty!) resolved the Langlands conjecture for
earI wo
the�L(Z:k) case, with k being a global field of finite characteristic. As we will see,
also proposed a combinatorial description of the Grothendieck-Teichmuller
Drtn�eId
grou
with surprising applications in physics. Beginning with Drinfeld, and perhaps
.p�cingin Kontsevich, the Russian school has generated an extraordinary profusion
cul
ical results in physics, making use of both the categorical abstractions in
t� ret
of :
Gro end
ieck's work, and the functorial conjectures of the Langlands program. A
and categorical inter/aci1Z{f of arithmetic-algebra-geometry would thus seem to
"'
be
�
; en in the
continuous mysteries of physics. The philosophical consequences of
uation are of enormous relevance, but are nevertheless invisible in the usual
sue a sit
ts of the philosophy of mathematics.
treatmen

224
 Q.UIDDITAL MATHEMATICS

fu ndamental constants o f physical theories. This group

should be closely related to the Grothendieck·Teichmiiller
group.'253
One of those reasons consists in a result of Connes
)
and Kreimer's (1999 , in which it is demonstrated that
the Lie algebra of the Grothendieck·Teichmiiller group25-4

acts naturally on the algebra corresponding to Feynman

diagrams. Shortly thereafter, Connes and Marcolli ( 2004)
succeeded in demonstrating that the cosmic Galois group
can be described as the universal symmetry group U of
renormalizable physical theories, and that, in effect, its
Lie algebra can be extended to the Lie algebra of the
Grothendieck·Teichmiiller group. 255

253 P. Cartier, 'A Mad Day's Work: From Grothendieck to Connes and Kontsevich. The
Evolution of Concepts of Space and Symmetry', Bulletin oftheAmerican Mathematical
Society (New Series) 38, 2001: 407. Cartier's article was originally written in French in
1998; in 2000, he added a postscript, from which we have taken our quotation._The
ahsolut'!_Galois group is the Galois group of the (infinite) algebraic extension Gal ( Q:Q)
where Q is the algebraic closure of the rationals. The Grothendieck-Teichmiiller group
(GT) offers a combinatorialdescription of the absolute Galois group. It remains an open
conjecture �hether or not the algebraic and combinatorial descriptions are equivalent
(GT "' Gal ( Q:Q)). The Grothendieck-Teichmii!ler group appears in a natural manner
in Grothendieck's dessins d'enfants (i983): finitaryobjects aimed at characterizing the
behavior of number fields through certain associated Riemann surfaces. Yet to be fully
understood, dessins d'enfants - forms of the combinatorial understanding of algebra
by way of analysis - constitute a typical Grothendieckian transit.

254 The Lie algebra of GT can be described as a free algebra over the Euler numbers l;(3),
l;(5),l;(7), . . . (where l;(k)=l:,,1n-•). The Euler numbers appear in many comers of
number theory, but are still almost entirely unknown objects: only the irrationality
of l;(3)has been proven (Apery 1979, a tour de force that remained isolated for many
years), and, recently, the irrationality of infinitely many l;(k)when k is odd (Rivoal
2000). See Cartier, 'A mad day's work .. .', 405-6.

255 The Lie algebra ofUis the free alg!:braover l;(l), 1;(2),1;(3), . . .. The proximities between
the absolute Galois group Gal ( Q:Q) and the cosmic Galois group U, through the
mediation of GT, thus allow us to pinpoint a totally unexpected actWn ofan arithmetical
group on the universal constants efphysics (the Planck constant, the speed of light, the
gravitational constant, etc.).

225
 SYNTHETIC P H I LOSOPHY

'Cartier's dream', as his conjecture was known for some

years, had therefore 'come true', thanks to the results
of Connes and his team, and it thus represents a sort of
intensified, in.finitely re.fined Pythagoreanism, harking back
to the first, rough and original hypotheses regarding the
existence of harmonic correspondences between math­
ematika (the study of quantity) and kosmos (order) . o;r;
In Connes's work, this arithmetico-geometrico-physical
refinement extends to deeper analogies between physical
divergences in field theory and arithmetical mixtures in
Tate motifs,257 thus approaching what Connes calls the
very 'heart' of mathematics: 'modular forms, L-functions,
arithmetic, prime numbers, all sorts of things linked to
that'. 258 From the dream to the heart, we therefore achieve a
progressive revelation in the order of discovery, which we
have already carefully observed in Grothendieck's work,
and which Connes takes up for his art as well: 'There
are several phases leading to the "finding" of new math,
and while the "checking" phase is scary and involves just
256 I t would not b e out of place here to make a fresh return to Plato's Trmaeus. Independently
ofthe calculations contemplated there, which have obviously been outstripped., Plato's
underlyingrelationalstrategy does not tum out to be too far from the relational search for
correspondences between arithmetico-geometrical forms and cosmological structurrs
that is now being contemplated by Cartier, Connes and Kontsevich. We will return to
these questions in this work's third part.

257 Tate's mixed motifs (1965) show up in the representation of the homology classes of a
variety by means of linear combinations of subvarieties (algebraic cycles) and in the
connections between that representation and /-adic cohomology. Tate motifs serve as
a concrete guide to Grothendieck's general conjectures concerning motifs (standard
conjectures).

258 C. Goldstein & G. Skandalis, 'An Interview with Alain Connes', EMS Ner11sletter 63,
2007: 25-30: 27.

226
 QUI D DITAL MATHE MATICS

rationality and concentration, the "creative" phase is of

a totally different nature'. The emergence of simple ideas
after very lengthy experimentations and the transit through
'mental objects which represent intermediate steps and
results at an idealized level' underpin the specificity of
doing mathematics. 259
Connes's inventiveness reaches a still-higher pitch in
his program to demonstrate the Riemann hypothesis by
means of strategies and techniques drawn, essentially,
from physics. 260 The passage from the quiddital to the
eidal that surrounds the Riemann hypothesis is supremely
original. 261 On the one hand, Connes points out that a
quantum-theoretical study of the absorption spectrum ef

259 Alain Connes, 'Advice to the Beginner', http:/jwww.alainconnes.org;ldocs,K:ompanion.ps, 2.

260 Connes shows himself to be a devoted lover of formulas and calculations. See Goldstein
& Skandalis, 'An Interview with Alain Connes' 28.

261 The Riemann hypothesis encodes certain arithmetical properties in terms of analytic
properties. Riemann's zeta function is a function of a complex variable that is initially
defined by means of an absolutely convergent series �(s)=Ln2:ln--s for the case where
R.e(s)> 1 (over natural numbers greater than l, this therefore coincides with the Euler
numbers) that subsequently, through an analytic extension, gives rise to a meromorphic
function over C, with a simple pole ins= 1 (residual 1 ). A functional equation obtained
by Riemann for the zeta function shows that it possesses 'trivial' zeros (roots) in the
negative even integers. Riemann conjectured ( 1859) that allthe other uros of the zeta
function lie in the complex lineRe(z)=l/2 (the so-called Riemann hypothesis). Various
mediations serve to link Riemann's zeta function to arithmetic: the Euler formula
l:ne?:I n--s=flpprime 1/(1-p -s); other 'mixed' functions of a complex variable determined
by the zeta function; intermediary functional equations between these; and the subtle
asymptotic behavior of the functions. Riemann's strategy inaugurates a profound
understanding of the discrete through underlying continuous instruments7 which
would go on to be furthered by the German school of abstract algebra (Artin, Hecke),
and which would give rise to the Weil conjectures and to Grothendieck's grand
cohomological machinery. The consequences of the Riemann hypothesis in number
theory are quite extensive, and the hypothesis may perhaps be considered, today. as
the greatest open problem in mathematics. For a description of the situation, see E.
Bombieri, 'The Riemann Hypothesis', in J. Carlson et al., 1he Millennium Pria Problems
(Providence: The Clay Mathematics Institute, 2oo6), 107-24.

227
 O F CONTEMPORARY MATHEMATICS

light, using the instruments ofnoncommutative geometry,

allows us to recalculate, with all the desired precision, all
of the constants that appear in the limited developments
of Riemann's zeta function. 262 A fundamental critical turn
in that approach consists in calibrating the appearance
of a negative sign (which Coones qualifies as 'cohomo­
logical') in the approximations of the zeros of the zeta
function through absorptions (and not emissions) of a
spectrum. On the other hand, Connes offers a sweeping
construction of analof;i.es in an attempt to tranifer to the
case offinite extensions ofQ ('number fields') Weil's i942
demonstration of the generalized Riemann hypothesis for
global fields of characteristic p > 0. 2 63 Connes's strategy
(announced in 2005 with Consani and Marcolli, and with
an eye to the near future) here consists in progressively
eliminating the obstructions in the transit by means of
an elucidation of concepts, definitions and techniques
in noncommutative geometry that correspond to Weil's
successful undertakings in algebraic geometry. 2 64 In this
262 A. Conncs, "Trace Formula in Noncommutative Geometry and the Zeros of the Riemann
Zeta Function', SelectaMathemo.tica (New Series) 5, •999: 29-1o6. The idea ofusing the
spectrum and the trace of an operator in a suitable Hilbert space to capture the zeros
of the zeta function comes from Hilbert and Polya, as Connes himself indicates. H JS
originality consists in combining the natural instruments of noncommutative geometry
connected to Hilbert spaces with the univmal physical situations underlying tho,.
instruments.

263 For details see Connes & Marcolli, A Walk in theNuncummutative Garden, 84-99.
264 Connes's program clearly shows how certain webs of invention and discovery proceed in
higher mathematics. The analogies - or harmonious conjectures - correspond to precise
(but not theorematic) translations between algebraic geometry and noncommutativc
geometry, with transferences and technical redefinitions ofconcepts in either context
The refined structural organization of each realm allows for an intuition of synrhetic

228
 Q,UI D D ITAL MATHE MATICS

way, we may gain access t o characteristic zero a s the 'limit'

(in noncommutative geometry) of those that are well

behaved in characteristic p. In each of these processes,

a back-andforth between the quiddital and the eidal, without
any privileged direction being.fixed in advance, allows for
the emergence of mathematical results of great depth - a

depth that is not only conceptual and technical but also

philosophical.
Maxim Kontsevich ( Russia, b. 1964) is another

remarkable contemporary mathematical author who has

been able to unite high speculative abstraction and the
concrete richness of physical phenomena. Kontsevich
was awarded the Fields Medal in 1998. In his acceptance

speech, he said, '65

For myself, as a mathematician, it is very interest­

ing to decipher the rules of the game of theoretical
physics, where one doesn't see structures so much
as the symmetry, locality and linearity of observ­
able quantities. It is very surprising that those weak

correspondences, which are then analytically delimited and contrasted with the
many examples available, thereby producing a sort of dictionary between algebraic
geometry and noncommutative geometry. A series of 'analogies' can be found in
Connes & Marcolli, A Walk in theNoncommtaative Garden, 9. As the authors indicate,
the fluctuations implicit in the analogies are what driues the subsequent development of
the mathematics. To eliminate this indispensable, initial vagueness from mathematics - as
a century of analytic philosophy presumed to do - therefore impedes understanding
of the complex creative forms of the discipline.

265 For a technical description of Kontsevich's work prior to the Fields Medal, see C.
H. Taubes, 'The Work of Maxim Kontsevich', in Atiyah, Iagolnitzer, Fields Medalists'
Lectures, 703-10.

229
 SYNTHETIC PHILOSOPHY

conditions ultimately lead t o such rich and compli­

cated structures. 266

This is the case with Feynman diagrams in theoretical

physics,267 whose formal use in mathematics was intro­
duced by Kontsevich in order to resolve some formidable
problems: Witten's conjecture regarding moduli spaces
of algebraic curves; the quantization of Poisson varieties;
and the construction of invariants of knots.
In the arithmetic of moduli spaces of algebraic curves,
certain cohomological invariants appear ('intersection
numbers'), which may in turn be represented as complex
combinatorial coefficients of a formal series F(t0t). For­
mally manipulating two quantum field theories, Witten
conjectures that the formal series U=a2F/at02 should
satisfy the KdV equation, 268 which would give rise to
numerous interrelations between the intersection numbers
in the arithmetic of algebraic curves. Making use of his
great talent for combinatorial calculations, Kontsevich
266 M. Kontsevich, reception speech at the Acadbnie des Sciences (2003), http://www.
academie--sciences.fr/membresjK/Kontsevich_Maxim_discours.htm, 2.
267 Feynman diagrams are graphs by which one may represent the perturbations ofparticles
in quantum field theory. Certain combinatorial (dis)symmetries and (dis)equilibria
in the diagrams no only suffice to simplify the calculations but predict new physical
situations, which subsequent mathematical calculations confirm. The heuristic use of
the diagrams in theoretical physics has been very successful, opening an important
gateway onto visuality in theoretical knowledge. A mathematica/farmalkalion of the
diagrams appears in A. Joyal & R. Street, Tue Geometry of Calculus l', Advances in
Mathematics 88, i99i: 55-112 (using the instruments of category theory). On the other
hand, the mathematical use of the diagrams in order to resolve deep mathematical
problems is due to Kontsevich.
268 Seep. 215, n. 235.

230
 Q,UIDDITAL MATH EMATICS

proved Witten's conjecture and succeeded in exhibiting

those interrelations in an explicit fashion, s tarting with
constructive models for moduli spaces, based on Riemann
surfaces of diagrams with metrics. 269 This is an extraordi­
nary example of the inventive richness of contemporary
mathematics, whereby arithmetic and physics are woven
together in ways that are rwt predetermined in advance:
the conjecture is arithmetico-differential, motivated by a
comparison with physics, and the proof interlaces com­
binatorial, arithmetical and continuous fr agments on
the basis of certain physical images (diagrams, graphs,
surfaces, metrics). The bipolar transit between physics
and mathematics is thus truly the generator of a new
knowledge. What matters here is not a supposedly origi,nary
base (the physical world, the world of mediations or the
world of ideas) that would firmly support the edifice of
knowledge, but a tight correlational warp that supports the
transit of knowledge (see chapters 8 and 9).
Phenomena of quantization - the deformation of
observable quantities by means of new parameters and the
asymptotic study of the deformations as the parameters
tend toward zero - show up on many levels in physics,
and, in particular, in the s tudy of the infinitely large
and the infinitely small. On the one hand, in general
relativity, it has been observed how the Poincare group

269 M. Kontsevich, 'Intersection Theory on the Moduli Space of Curves and the Matrix
Airy Function', Communications in Mathematical Physics, 147, 1992: I-.23.

231
 OF CONTEM PO RARY MATHE MATICS

(isometries of Minkowski space-time) tends toward the

Galileo group (isometries of euclidean space-time) when
the parameter bound to the speed ef light tends toward
zero. On the other hand, in quantum mechanics, it has
been observed how the 'natural structures' of quantum
mechanics tend toward the 'natural structures' of classical
mechanics when the parameter tied to the Planck constant
tends toward zero. The structures of classical mechanics
are well known, and correspond to the Poisson manifolds, 2;0
in which one can naturally formalize the Hamiltonian
as an operator for the measurement of the energy (well­
determined energy and momentum) of classical physical
systems. Although, from a mathematical point ofview, the
quantizations of an algebra have, since the 1950s, been
understood as quotients of formal series over that algebra
(Kodaira), the quantii:.ations ef Poisson manifolds (which
later showed up in quantum mechanics) had not been
rigorously studied before Kontsevich. Here, Kontsevich
discovers that such a quantization is linked to a new kind
ef string theory, where the use of Feynman diagrams is
significant; and he exhibits, again in an explicit manner,
that the deformation is linked to certain perturbations
of quantum fields and to extremely subtle calculation s
270 A Poisson algebra is an associative algebra with a Lie bracket that acts as a derivation of
the algebraic operation (the law[x,yzJ=[x,yJ z + [x, z J y, is to be read as an analogue of
'Leibniz's Law': a,(yz)= (a,y)z+(a,,z)y). A Poisson manifold is a differential marnfold
with the structure of a Poisson algebra. The paradigm of a Poisson manifold i s the
algebra of smooth functions over a symplectic manifol.d (a generalization of a manifold
with a Hamiltonian) .

232
 Q.UI DDITAL MATHEMATICS

2
of certain terms of asymptotic expansions. 71 Even more
22
astonishingly, 7 what emerges in Kontsevich's calculations is
an action of the Grothendieck-Teichmiiller group on the

space of possible universal formulas of physics, a group

that can also be seen as the symmetry group of the pos­
sible quantizations of the original Poisson manifold. And
so we find, in a totally unexpected way, a corroboration
ofConnes's simultaneous discovery regarding the action
of the absolute Galois group on the universal constants
of physics.
A third unexpected place where Feynman diagrams
show up, to help resolve highly sophisticated mathematical
problematics, is in the construction of universal invariants
in mathematical knot theory. 273 In his work, Kontsevich
introduces an entire series of novel constructions over
which the invariants are layered: complex differentials of

271 M. Kontsevich, 'Deformation Quantization of Poisson Manifolds I', !HES preprints

M/97/72 (1997).
272 Kontsevich describes the 'surprise' caused by the emergence of new strings in his
quantization calculations (Kontsevich, speech, Acadimie des Sciences, l) . The surprise
caused by the action of the Grothendieck-Tcichmiiller group was in no doubt greater
still.
273 A mathematical knot corresponds to the intuitive image ofa string tied into a knot, with
its extremities identified (formally, a knot is therefore an immersion of the circle s1 in
R3). A general and complete classification of knots is still an open problem. Poincarl:,
Reidemesteir and Alexander, in the first half of the twentieth century, proposed some
initial instruments for such a classification. But it was above all in the last decades of
the twentieth century, with the work ofJones and Witten (1990 Fields medalists), that
knottheory made its theoretical breakthrough. Vassiliev proposed a series of topological
invariants connected to the Jones polynomial, which Kontsevlch had reconstructed
by means of abstract integrals over suitable algebraic structures, with strong universal
properties. See M . Kontsevich, 'Feynman Diagrams and Low-Dimensional Topology',
Firrt European Congress efMathematics (Paris i992) (Boston: Birkhauser, 1994), 97-121,
and M. Kontsevich, 'Vassiliev's Knot Invariants', Advances in So'Oiet Mathematics, 16/2,
1993: 137-50.

233
 SYNTHETIC P H I LOSOPHY

graphs; cohomology groups of those complex differen­

tials, differential forms over those groups; integrability of
those forms through a generalized Stokes argument; etc.
And so we find ourselves in the company of a mathemati­
cian extraordinarily skilled in combinatorial manipula­
tions, and endowed with tremendous plasticity in the most
diverse forms of exact transit: the technical and calculative
passage between nearby subbranches of mathematics; the
analogical and structural passage between more distant
realms of mathematics; and the visual and conceptual
passage between mathematics and physics.
We encounter another example of Kontsevich's trans­
gressive power in his ideas for homologically formal -
izing the phenomena of mirror symmetry in theoretical
physics.274 Witten described a topological unfolding in
supersymmetry phenomena, which corresponds to a
sort of specular reflection between strings (A-branes
and B-branes, sophisticated models that incorporate
Riemann surfaces and holomorphic manifolds). Kontse­
vich conjectured that the mirror symmetry between two
manifolds X, Y should correspond to an equivalence eftwo
triangulated categories,•15 one coming from the algebraic

274 M. Kontsevich, 'Gromov-Witten Classes, Quantum Cohomology and Enumerative

Geometry', Communication.r in Mathematical Physics, 164, 1994: 525-62; M. Kontsevid
& Y. Soibelman, 'Homologic.al Mirror Symmetry and Torus Fibrations', in K. Fukaya
et al., Symplectic Geometry and Mirror Symmetry (Singapore: World Scientific, 2001),
203-63.
275 Triangulated categories supply axioms (the naturalness of which is still under
discussion) that can be put to use in trying to universally capture the properties

234
 Q,UIDDITAL MATHEMATICS

geometry of X, and the other from the symplectic geometry

of Y. In this manner, complex symmetry phenomena in
the physics of the infinitely small correspond to transfers
of structure between the discrete (algebraic varieties)
and the continuous (symplectic manifolds), thus giving
rise to another new, totally surprising and unexpected
connection between physics and mathematics. Kontsev­
ich's hypothesis was later mathematically demonstrated in
many cases - for elliptic curves (Kontsevich, Polischuk,
Zaslow), for tori (Kontsevich, Soibelman), for quartics
(Seidel) - and physically confirmed with the discovery of
new strings (D-branes) anticipated by the theory.
Kontsevich's other works explore very deep connec­
tions between motifs, 'operads',•16 the cohomology of Lie
algebras and the topology of varieties, seeking to provide
the foundations for a ubiquitous quantum cohomology,
which would reveal the presence of certain universal

of the derived category of an abelian category. Given an abelian category A (which

generalizes the properties of the category of abelian groups), Com(A) is the category
of its simplicial complexes with chain morphisms, and Der(A) is the derived category
whose objects are homotopy classes of the objects of Com(A), and whose morphisms
are localizations (modulo quasi-isomorphism) ofthe morphisms of Com(A). Com(A) and
Der(A) are triangulated categories. The notion comes from Grothendieck and Verdier
(at the beginning of the sixties, and with Verdier's •967 thesis - not published until
i996!), where it is used to express certain properties of duality in a general fashion.
276 Operadr can be understood in terms of the analogy: algebras/operads = representations/
groups. Operads are collections of operations that compose with one another well and
that realize a sort of minimal (dirincamate) compositional combinatorics, underlying
the higher algebras in which the operads are incarnated in a concrete fashion. Operads
can thus be seen as one more example of generic - or archetypical, we do not fear the
word - constructs, like many other mathematical objects that we have been parading
through these pages (universal objects in categories, cohomologies, motifs, possible
cofinalities, etc.). In the next chapter we will supply additional mathematical facts
concerning this emergence of 'archetypes,, and in chapters 8 and 9 we will study their
ontic and epistemic status.

235
 O F CONTEMPORARY MATHEMATICS

algebraic 'archetypes' behind the many continuous func­

tions of physics. •n At stake here is a situation one may
find at the very 'center' of mathematics, and which would
answer, in a novel manner, to what Thom called the
'founding aporia' of mathematics. Indeed, Kontsevich
has explicitly pointed to a possible enlarging of the heart
of mathematics (recalling Connes), in which the radical
importance of the current connection between physics
and mathematics is emphasized:

The impact that the new discoveries in physics have

had on mathematics is enormous. One could say
that, before, in mathematics, there existed a principal
center of mysteries, the group of all the conjectures
interlacing number theory, the motifs of algebraic
varieties, the L-functions (generalizations of Rie­
mann's zeta function) and automorphic forms, which
is to say the harmonic analysis of a locally homog­
enous space. Now, however, the theory of quantum
fields and string theory constitute a second center of
mysteries, and offer a new depth and new perspectives
to mathematics. 218

277 M. Kontsevich, 'Operads and Motives in Deformation Quantization', Letters in

Mathematical Physics, 48, 1999: 35-72; M. Kontsevich, 'Deformation Quantization of
Algebraic Varieties', Letters in Mathematical Physics, 56, 2001 : 271-94.

278 Kontsevich, speech at the AcadCmie des sciences, i.

236
 Q.UIDDITAL MATHE MATIC S

In this manner, the quidditas imposes its massive imprint

on the eidal signs aimed at helping us understand the
world. Amid all of these tensions, we see how the 'world'
consists in a series of data/structures (Peircean fi.rstness),
registers/models (Peircean secondness) and transits/
functors (Peircean thirdness), whose prog;ressive interlac­
ing into a web not only allows us to better understand the
world, but constitutes it in its very emergence. In the next
chapter ,we will see how contemporary mathematics is
finding new, stable supports in that web (invariants,
'archetypes'), thus solidifying the relational and synthetic glu­
ingefboth phenomena and concepts, without any need of an
analytic foundation to ensure the security of the transit.

237
C HA P T E R 7

ARC H EA L M AT H E MAT I C S : F REYD, S I M P S O N ,

Z I LB E R, G RO M OV

A metaphor for understanding the complex transits that

take place in contemporary mathematics can be found
in the image of an articulated pendulum. Unlike a simple
pendulum, which, as it sweeps its course, determines a.fixed
frontier, equidistant between its extremes, an articulated
pendulum - built by linking together two pcndula oscil­
lating in opposite directions - defines an altogether extraor­
dinary dynamic curvature, unimaginable if one werejust to
consider the two pcndula separately. In fact, in a chrono­
photograph of an articulated pendulum by Marey (1894
- see figure 10, overleaf), we can see how an entire extensive
spectrum efthe intermediateemerges in the reticulated undu­
lation to the left, thereby opening itself to the curvatures
characteristic of the living and the organic. The contrast
between an articulated and a simple pendulum serves as a
simple, metaphorical contrast between advanced and ele­
mentary mathematics. Indeed, on the one hand, advanced
mathematics - and especially contemporary mathematics,
as we have shown in the second part of this work - harbors
an entire series of dialectical concretizations arising from a
sophisticated articulation between webs and scales of con­
cepts and models, on multiple eidal and quiddital levels.

239
 SYNT H E T I C P H I LO S O P H Y

On the other hand, the low levels of complexity i n the

techniques of elementary mathematics inherently simplify
the underlying conceptual movement; they therefore have
no need of truly subtle (infinitely discerning) articulations
or hierarchies in order to erect their edifice. We thus have
a metaphorical counterpoint between the articulated and
the simple, which is further supported by a counterpoint
between a 'relative mathematics' , a mathematics in motion,
in the style of Grothendieck (contemporary mathematics,
articulated pendulum), and an 'absolute mathematics',
a mathematics at rest, in the style of Russell (elementary
mathematics, analytic foundation, simple pendulum).

Figure 10. Articulatedpendulum, after a chronophotograph by Marey.

240
 A R C H E A L M A T H E M AT I C S

Tue (dynamic, organic, living) curvature we see o n the

left side of figure 10 may seem to transcend the composite
and the projective, while the situation can actually be
reconstructed on the basis of prototypical underlying ten­
sions (the contrary impulses between the greater and the
lesser pendula) by anyone who knows how the articulated
pendulum works. Likewise, behind the processes of ascent
and descent that we have been describing in the preceding
chapters, behind the pendular oscillations between frag­
ments of ideality and reality, behind what we have called
the bipolar dialectic between the eidal and the quiddital
- which is to say, behind the incessant transit in both direc­
tions between concepts and data, between languages and
structures, between mathematics and physics, between
imagination and reason - contemporary mathematics has
gone on to produce deep archetypes, by which the transit
can be stabilized, the opposing polarities mediated, and
the pendular movements balanced.
With the neologism archeal (from arche [principle or
origin, see figure n, overleaf]) we will designate, in this
chapter, the search for (and discovery of) remarkable
invariants in contemporary mathematics, by which the
transits can be soundly controlled, without any need efbeing
anchored in an absolute ground. Those invariants will serve
as relative origins ( arkhO), commanding (arkhen) movement
on a particular level (that is, in specific concrete catego­
ries) . We will therefore take up a revolutionary conception

241
 O F C O NT E M P O RA R Y M AT H E M A T I C S

which has surfaced i n contemporary mathematics i n a

theorematic manner: the register of universals capable ef
unmooring themselvesfrom any 'primordial' absolute, relative
universals regulating theflow of knowledge. I n this chapter
we will discover certain technical constructions in that
register of 'decantations of the universal', and in chapter 9
we will inquire as to how we may contain the apparent
contradiction in terms 'relative universal', which will give
rise to a newfounding synthetic aporia of mathematics (one
that is nonanalytic - that is, nonfoundational).

Figure n. 7he Archeal Realm.

The works of Peter Freyd (USA, b. i935 ) in category the­

ory forcefully bring to light the emergence of archetypes

242
 A R C H E A L M AT H E M AT I C S

in the structuration o f mathematical thought. As we have

already seen with Grothendieck, the dialectic of the One
and the Many finds one of its happiest expressions in
categorical thought, where one object defined by means
of universal properties in abstract categories will in turn
appear or turn out to be many, over the plurality of con­
crete categories where it ' incarnates'. The One and the
Universal enter into perfect counterpoint and dialogue
with the many and the contextual. Going one step further,
Freyd's allegories279 arc abstract categories of relations,
axiomatized through a generic relational combinatorics,
beyond the very restrictions of functionality. Here, full
attention to the categorical diagrams of relational com­
position serves to reveal the precise mechanisms of one/
many adaptation that mediate between logical theories
and their various representations - mechanisms that a
functional, set-theoretic reading would fail to detect.
Indeed, Freyd's categorical and relational machinery
furnishes axiomatizations for hitherto unnoticed inter­
mediate categories - categories of lesser representational
power than Lawvere's topoi - and shows that they con­
stitute classes of natural models for intermediate logics
between certain 'minimal' logics and intuitionistic logic.
With the discovery of a remarkable, ubiquitous process
in categorical logic, which we will turn to in a moment,

279 P. Freyd & A. Scedrov, Categories, Allegories (Amsterdam: North-Holland, 1990).

243
 SYNT H E T I C P H I LO S O P H Y

Freyd shows how, starting from pure type theories with

certain structural properties (regularity, coherence, first­
order, higher-order) , one can uniformly construct, by
means of a controlled architectonic hierarchy,Jree categories
that reflect the given structural properties in an origin
(regular categories, 2 80 prelogoi,281 logoi,282 and topoi283) .
I n reaching the free categories, we obtain the most 'dis­
incarnate' categories possible, categories that can be
p rojected into any other category with similar properties:
Freyd thereby succeeds in constructing something like
initial archetypes of mathematical theorization. His dis­
covery is doubly significant, because it not only describes
the invariants of logico-relational transit, but seizes them
in a universal manner, beyond the particular fluctuations
of each logical fragment. As often happens with great
turns in mathematics, Freyd's results will only be fully
understood in the future, but it is already easy to predict
their extraordinary importance .

.280 Regular categories are categories with the exactness properties (cartesianity, existence
of images, preservation of coverings under pul1backs) that are necessary and sufficient
for achieving an adequate composition of relations.
281 Pre1ogoi are regular categories for which the subobject functor takes values in the
category of lattices (and not only in sets) . A preorder P, understood as a category,
will turn out to be a prclogos if and only if P is a distributive lattice with a maximal
clement.
282 Logoi are prelogoi for which the subobject functor (with respect to its values in the
category of lattices) possesses a right adjoinc. A preorder P, considered as a category,
will turn out to be a logos if and only if P is a Heyting algebra.
283 "Ve have already witnessed the appearance of topoi in Grothendieck's work and its
subsequent elementary axiomatization due to Lawvere. The categoryP associated with
a preorder will turn out to be a topos if and only if P reduces to a point.

244
 ARC H EAL MATHE MATI CS

Freyd's procedure begins by taking a given logical theory,

and then goes on to capture, by way ef an intermediate
free category ef relations, the free terminal category of
morphisms that faithfully represents the properties
of the initial theory. The structure of the process 2 84 is
T ( theory) __, A T ( allegory) MapSplitCor(A T) (category) ,
__,

which yields a free result when one starts with a pure type
theory, and which shows, in each of its stages - relation­
ality, subsumption in identity (Car), partial invertibility
(Split), functionality (Map ) 285 - how a determinate math­
ematical conglomerate goes on to be 'filtered'. Out of this
'filtration' come two observations of great interest, both
mathematically andphilosophically: L the analytic process
of decomposing the transit is linked to the exhibition
of a universal synthetic environment that emerges in the
process ( the allegory A T), once again stressing the exis­
tence of an indispensable2 86 analytic-synthetic dialectic

284 Ibid., 2 77-

285 Corref!.exivity generalizes (in the axiomatic environment of allegories) the property of
a relation being contained in a diagonal (basic example: partial equivalence relations,
'PERS', which are increasingly used in computability theory). Cor functorially captures
correflexivity. Partial invertibilil)' generalizes the invertibility property of morphisms
to the right (basic examples: regular elements in a semigroup, sections of a sheaf).
Split functorially captures partial invertibility. Functionalif)1 generalizes (as always,
in the allegorical environment) the usual set-theoretic restriction of functionality on
relations. Map functorially captures functionality.
286 That dialectical indispensability turns out to be necessary in the context of Freyd's
representation theorems. The fact that delicate philosophical problems depend upon
partial theorematic reflections in contemporary mathematics is one of the great strengths
of those advanced mathematics, as compared to elementary mathematics, in which,for
lack efcomplexity, such reflections do not appear. The low threshold of complexity in
elementary mathematics therefore turns out to be a true obstruction from a philosophical
point of view. \Ve will come back to these questions in Part Three.

245
 O F C O N T E M P O RARY M AT H E M A T I C S

i n mathematics; 2 . beyond the terminal bipolar objects in

a one-to-one correspondence with one another (theories
and regular categories, coherent theories and prelogoi,
first-order theories and logoi, higher-order theories and
topoi), the richness of Freyd's procedure consists in its
progressive adaptation efsynthetic, intermediary machinery (its
explication of the actions of the Cor, Split, Map functors) .
A natural searchfor archetypes is essential to categorical
thought. The various levels of categorical information
(morphisms, functors, natural transformations, n-mor­
phisms, etc.) , on every level, allow for sub-definitions
of.free objects - universal projective objects (archetypes:
'commanding origins', etymologically speaking) - with
far more general universal constructions emerging over
each level's projections. This, according to Freyd, is what
happens with the process T ->A T -> MapSplitCor(A T) , as well
as with Yoneda's famous lemma"87, which allows us to
embed any small category into a category of presheaves:

287 Freyd recalls that the lemma does not actually appear in Yoneda's original article ('Note
on Products in Ext', Proceedings oftheAmerican Mathematical Society, g, 1958: 873-75), but
in 'a talk that Mac Lane gave on Yoneda's treatment of higher EXt functors' (see http://
www.tac.mta.ca/tac/reprints/articles/)/foreword.pdf, p. 5). The lemma's immense
philosophical content (to which we will return in the book's third part) corroborates
the Lautmanian transit between 'notions and ideas' and 'effective mathematics'. It is
interesting to point out that the lemma - so close to the structural grounds of Laurman's
thought - in fact arose through a vivid discussion between Yoneda and Mac Lane in
the Gare du Nord in Paris (see http://www.mta.ca/·cat·dist/catlist/1999/yoneda), so
close to the French philosopher's own physical neighborhood.

2 46
 A R C H E A L M AT H E M ATI C S

category of

category C A presheaves
ideal
over C
objects
'copy ' of c

discrete context continuous context

Figure 12. Yoneda 's lemma.

The representable functors hA capture the 'coronas' of mor­

phisms surrounding A (hA =Morc (A.--) , A being an object
in C), but, in the Yoneda embedding (A --> hA) , many other
nonrepresentable functors ('ideal' presheaves) show up to
complete the landscape. In fact, the category of presheaves
over C can be seen as a context of continuity into which
the initial discrete category is inserted (and by which it
is completed) , since, on the one hand, the category of
presheaves possesses all the (categorical) limits, and, on
the other, representables preserve limits. The universality
of the Yoneda embedding (which obtains for every small
category) brings two facts of enormous significance to
light: mathematically, it demonstrates how forms of an
archetypical continuum lie beneath many discrete situa­
tions, forms composed of ideal objects that allow us to
reinterpret the initial discrete context (something Hilbert
had already predicted in his brilliant text, On the Infinite, 288
and which we have found to be carefully articulated,

288 D. Hilbert, 'On the Infinite' ( 1925), in]. van Heijenhoort, FromFrege to Godel: A Source Book
in Mathematical Logi£ 1879-1931 (Cambridge: Harvard University Press, 1967), 367-92.

247
 SYNT H ET I C P H I LO S O P H Y

for example, i n Langlands's program) ; philosophically, i t

shows how, in the philosophy of advanced mathematics,
the imbrication of fragments of both realism and ideal­
ism is not only possible but necessary ( the archetypical
completion of 'real' representable presheaves by means
of non-representable, 'ideal' ones, and the need for the
latter in order to fully understand the former) .
In revealing the archetypical nuclei of mathematical
thought, the reverse mathematics program of Stephen
Simpson (U SA, b. 1945 ) takes on a special significance.
Initiated in collaboration with Harvey Friedman, 289 the
program tries ( successfully) to locate minimal and natural
subsystems of second-order arithmetic that are equivalent to
the ordinary theorems of mathematical practice derived
from those axioms. The equivalence is complete since, in
the eyes of adequate underlying theories, the theorems
in turn entail the conglomerate of axioms by which they
are proved, and so an entire dialectic of deduction and
retroduction ( whence the name: ' reverse mathematics' )
exists in demonstrative transit. There are no fixed founda­
tions in this back-and-forth ( it is not a question of an
absolute foundation from which everything else would,
presumably, follow) but multiple relative proefs ef equid­
eduction between fragments of mathematical practice.

289 Srephen Simpson, 'Friedman's Research on Subsystems of Second Order Arithmetic',

in L. Harrington et al., eds., Haroey Friedman 's Reuarch in the Foundatiom efMathematics,
(Amsterdam: ::-iorth·Holland, 1985), 137-159.

248
 A R C H E A L M AT H E M AT I C S

I n that movement o f proofs, Simpson detects certain sub­

sys tems of second-order arithmetic that are 'canonicaI' or
'archetypicaI' enough to organize collections of theorems,
and with them obtains a natural stratification of ordinary
mathematics, where, for example, statements of type (a.0 )
the Heine-Borel Lemma, (a1) the Balzano-Weierstrass
Theorem, or of type (b0) the existence of prime ideals
in rings, ( b 1 ) the existence of maximal ideals in rings,
can be classified in a precise hierarchy of equideductive
complexity ( in this case, a;<='> bi i = l .2, in the eyes of a
minimal constructive system) . 09°
In the language of second-order arithmetic,291 Simp­
son defines certain canonical subsystems of arithmetic
in the following manner: RCA0 incorporates the basic
axioms of arithmetical terms, the axiom of induction
cp ( O ) i\ Vx(<p (x) -> cp (x+l ) ) -> Vx cp (x ) restricted to I:6 for­
mulas and the comprehension axiom :JXV x ( cp ( x ) <-> x E X)
restricted to !<. b formulas ( formulas in the Kleene-Mostowski

290 In what follows, we will define that minimal base, which we will denote RCAo. The
important point that should be emphasized here, before getting into the details, is
that of equideduction in the eyes ofa weak axiomatic base. Of course, from the absolute,
analytic point of view of ZF set theory, we also have ZFf- a 1 � b 1 , simply because both
Slatements are tautological from the absolute point of view of the axioms of set theory;
but, in this case, the equideduction is trivialized and loses the logical richness of an
intermediate derivation, without strong premises tha[ distort it.

291 The language of second-order ari[hmetic extends that of the first order with variables
of two types ('set' or second-order variables, in addition to 'numerical' or first-order
variables), with an additional relational symbol ( E), with new formulas of the type t EX
(where t is a numerical term, andX a set variable), and wilh additional quantification
over set variables (whence the 'second' order). A formula in the second-order language
is called arithmetical if it does not quantify over any set variables.

249
 O F C O NT E M P O RA RY M AT H E M AT I C S

hierarchy) ;292 WKL0 consists o f RCA0 + a weak version of

' Konig's lemma' (every infinite subtree of a binary tree
possesses an infinite branch); ACA0 incorporates the basis
axioms for arithmetical terms, the axiom of induction
restricted to arithmeticalformulas and the comprehension
axiom, likewise restricted to arithmetical formulas. Other
major subsystems of greater expressive power allow us
to codify more sophisticated infinitary combinatorial
manipulations, but we will not mention them here. 293
With those nuclei ('archetypes') of deductive power,
Simpson thus obtains a real stratification of the theorems
of ordinary mathematics, demonstrating deep, logical
equiconsistencies such as the following: in the eyes
of WKL0, a full equivalence (deduction and retroduc­
tion) between the system WKLu and many significant
statements that can be demonstrated within WKL0 (the
Heine-Borel lemma, the completeness of first-order logic,
the existence of prime ideals in countable commutative
rings, the Riemann-integrability of continuous func­
tions, local existence theorems for differential equations,
the H ahn-Banach Theorem for separable spaces, etc.);
in the eyes of RCA0, a full equivalence between the system

292 I� includes (first-order) formulas with recursive matrices, which can be put into a
normal form with n alternations of quantifiers and with 3 as its outermost quantifier.
IT� is defined likewise, for normal forms with v as their outermost quantifier. We then
define 6. � formulas as those that can be put into either L ·� '"', n � form.
293 See Simpson, Subsystems ofSecond Order Arithmetic, chapters 5 and 6, 167-241 ( transfinite
recursion for arithmetical formulas and comprehension for fl: formulas ) .

250
 A R C H E A L M AT H E M AT I C S

ACAr, and many o f its theorems (the Bolzano-Weierstrass

Theorem, the existence of maximal ideals in countable
commutative rings, the existence of bases for vectorial
spaces over countable fields, the existence of divisible
closures for countable abelian groups, etc.) . The usual
deductive transit, which goes from the global (the system
in its entirety) to the local (a particular theorem of the
system) , is here reversed in a striking manner, allowingfor
an unexpected passage.from the local to the global, thanks to
the theorem's equiconsistency with the entire system.
It is instructive to observe how that reverse mathemat­
ics can emerge only after one makes explicit the natural
candidates to act as archetypical deductive nuclei in second­
order arithmetic.
Simpson describes one part of reverse mathemat­
ics as a restructuring of Hilbert's program: 294 in show­
ing what the minimal axiomatic systems for proving the
theorems of ordinary mathematics are, we can measure
the latter's complexity in a precise fashion, and, in many
cases, reduce that complexity to strictly finitary argu­
ments. I n this m anner, although the absolute version
of Hilbert's program fell with Godel's Incompleteness
Theorem, the program can nevertheless be relativized, and
go on to secure certain partial and intermediary results.
This, for example, is the case with II5 sentences provable

294 Ibid., 381-2.

251
 SYNTH E T I C P H I LO S O P H Y

i n WKL0: Simpson has proven a conservativity result•95

for WKL0 over Peano arithmetic for I: /1 formulas, which is
in tum conservative over recursive arithmetic, and so the
I H sentences provable in WKL0 can be proved with purely
finitary arguments. In particular, this yields finitary proofs
for many sophisticated mathematical results (expressible
by II3 sentences provable in WKL0) : the existence of
functional extensions (Hahn-B anach), of solutions to
differential equations, of prime ideals in commutative
rings, of algebraic closures, etc.
Certain archetypes in the spectrum of proofs in second­
order arithmetic thus acquire a special relevance, since
they end up being fully reflected in avatars of ordinary
mathematics (abstract algebra, point topology, functional
analysis, differential equations, etc . ) . As Simpson has
pointed out, the archetypical nuclei of proofs emerge
through a 'series ef case studies' leading to the discovery
of 'the appropriate axioms', 296 showing how it is always
mathematical experimentation that helps us declutter the
landscape and find (when they exist) suitable invariants

295 Given a theory T in a language L, and another theory Ti in a language L 1 � L, T1 i s a

conservative extension of T if, for every L-scntence <p, T1 - <p if and only ifT I <p. Simpson
demonstrates that ACA.0 is a conservative extension of Peano arithmetic (first-order
PA), and that both RCA.0 and \VKLu are conservative extensions of PA restricted to r.J
formulas (in other words, L l -PA is a first-order fragmen t of RCA0and \VKL0, while
PA is a first-order fragment of ACAn). The conscrvativity proofs exhibited by Simpson
employ existential techniques from model theory, and their effective (finitary) character
has been brought into question. Harvey Friedman has nevertheless stated that, i n
reverse mathematics, existential proofs of conservativitycan indeed b e converted into
effective proofs .

296 Ibid., vii (xiii in 2nd edition).

252
 A R C H E A L M AT H E M AT I C S

lying behind the movement. The processes o f mathematics

(and no longerj ust its objects) are thereby forged in multi­
ple webs of contradistinction, whether on the level of their
logical representations, tied to well-defined archetypes of
proof, or on the level of their structural correlations, tied
to great spectra of regularity/singularity in domains of
transit/obstruction. We once again find ourselves before
an abstract, relative differential and integral calculus, like the
one we have encountered in Grothendieck's work, and
which, in the realm of reverse mathematics, consists in
subtly differentiating the scales of demonstration, so as to
afte rward reintegrate them around canonical nuclei of proof.
The works of Boris Zilber (Russia, b. 1949 ) take the
search for canonical structures for logic, algebra and
geometry to a far greater depth. 297 His works install
themselves in a new paradigm, progressively emerging in
model theory, which shifts, in the middle of the twentieth
century, from being understood as 'logic + universal alge­
bra' (Tarski, Birkhoff; a paradigm normalized by Chang
and Keisler), 298 to being considered, by the century's end,
as 'algebraic geometry - fields' (Shelah, H rushovski,

297 Vie are grateful to Andres Villaveces for his lessons on Zilber (through an article we
have referred to, conversations and conferences). An excellent overview of Zilber and
his epoch can be found in B. Poizat, 'Autour du thi:oreme de Morley' (the section
'1980-90: les annees-Zilber', in particular), inj .-P. Pier, ed.) Development ofMathematics
1950-2000 (Boston: Birkhauser, '.woo), 879-96.
298 H.J. Keisler & C.C. Chang, Mode/ Theory (Amsterdam: North·Holland, 1973) .

253
 O F C O NT E M P O RARY M A T H E M A T I C S

Zilber; a paradigm standardized by Hodges).299 At issue

here is an important change of perspective that brings
model theory together with Grothendieck's visions (logi­
cal invariants of dimension that approach geometrico­
algebraic invariants, o-minimal structures that approach
' tame' structures) . Above all, this perspective situates
logic in a series efprogressive geometrical decantations of the
objects and processes of mathematics. Along these lines,
a revealing work3°0 of Zilber's shows how certain strongly
minimal theories3°' can be intrinsically associated with
combinatorial geometries: the models of those theories are
obtained as 'limits' of suitable finite structures, and the
intrinsic geometries of those models, which are perfectly
controlled, are either trivial (the algebraics do not expand
to the model) or else affine or projective geometries over
a finite field, something which allows optimal 'coordinate
systems' for the models to be found. We see here how,
behind general logical properties (strong minimality,

299 W. Hodges, Model 7heory (Cambridge: Cambridge University Press, 1993). A

contemporary version of model theory, including Zilbcr's contributions, appears in
B. Poizat, A Course in Model Theory: An Introduction to Contemporary Logic (New York:
Springer, 2000 ) .
300 B . Zilber, 'Totally categorical theories, structural properties, and the non-finite
axiomatizability', in L. Pacholski et al., Proceedings ofthe Conference on Applications of
Logic to Algebra andArithmetic (Karpacz, 1979) (Berlin: Springer, 1980 ) , 380-4rn.
301 A theory is strongly minimal if the subsets definable in the models of the theory are
finite or cofinite (for example, the theory of vector spaces, the theory of algebraically
closed fields of characteristic p) . This situation gives rise to natural logical notions
analogous to those of dimension and algebraic closure, thereby opening the way to a

reading of aspects of logic as fragments of a generalized algebraic geometry. The first

result of Zilber's mentioned here refers to w-categorical, strongly minimal theories
(countable isomorphic models).

254
 A R C H E A L MATH E MATI C S

co - categoricity), lie deep geometrical nuclei. I f, following

Grothendieck, we take up the terminological-conceptual
tension between discovery and creation, we are thus con­
fronted with archetypical synthetico-geometrical structures
that are 'discovered' through the ' invention' of analytico­
logical languages.
Another fundamental work3°2 of Zilber's aims to
classify the underlying geometries in strongly minimal
theories. The Zilber trichotomy conjectures that strongly
minimal theories can be split into three classes, according
to their associated geometry: (i) theories with 'dismem­
bered' models, which therefore possess a set-theoretic
notion of dimension (dim (Xu Y) =dim (X) +dim (Y) ) , and
which cannot be interpreted in the theory of an infinite
group; (ii) theories with basically linear models, whose
geometry is modular, and therefore possesses a vecto­
rial notion of dimension (dim (Xu Y) =dim (X) +dim (Y)­
dim (X n Y) ) , with finite fragments of the model being open
to interpretation through abelian groups; (iii) theories
with models that are bi-interpretable with the algebra­
ically closed field of complex numbers (a particularly
rich group), in which case the geometry is algebraic, and
therefore has a natural notion of algebraic dimension.
As Poizat has pointed out, among the great riches of the

302 R. Zilber1 'The structure of models of uncountably categorical theories', in Proceedings

ofthe International Congress efMathematiciam (Va7'.Sovia 1983) (Varsovia: Polish Scientific
Publications, 1984), 359�68.

255
 S Y NT H E T I C P H I L O S O P H Y

trichotomy, a s envisioned b y Zilber, i s the emergence o f

'groups everywhere' - invisible a t first, b u t lying i n the
depths ('archetypes') .3°3 In a sort of renaissance of Klein's
Erlangen Program (18r2),3°4 groups - and their associ­
ated geometries - thus help to classify the deep forms of
logic, and we may once again reflect on how logic can in
no way precede mathematics, as is sometimes presumed
from analytic perspectives.
Zilber's conjectural trichotomy aims to elucidate certain
geometrical nuclei behind logical descriptions. About ten
years after the conjecture's appearance, H rushovski suc­
ceeded in demonstrating that there is, at least, a fourth case
that it does not cover:3°5 by means of a sophisticated amal­
gamation in the limits of models, H rushovski constructed
a strongly minimal structure whose geometry is neither
trivial, nor modular, nor 'algebraic' .3°6 Nevertheless, Zilber
and Hrushovski conjectured and demonstrated that the
trichotomy is indeed valid3°7 for theories whose intrinsic
geometries are Zariski geometries.3°8 Afterward, by means

303 Poizat, 'Au tour du thCortmc de Morley', 890.

304 For an illustrated modern edition of Klein's text, with a preface by Dieudonne, see F.
Klein, Le programme d'Erlangen ( Paris: Gauthier-Villars, 1974).

305 E. Hrushovski, 'A new strongly minimal set', Annals ofPure and Applied Logic 62, 1993:
147-66.

306 lhat is, not bi·imerpretable with an algebraically closed field, in the sense of algebraic
geometry.

307 E. Hrushovski & B. Zilber, 'Zariski Geometries' ,Journal ofthe American Mathematical
Society 9, 1996: 1-56.

308 A Zariski geometry is a sort of variable topological structure (Xn: n2::: 1 ) , with no ethericity
conditions and coherence between the X �· Zariski geometries can be seen as Lautmanian

256
 A R C H E A L M A T H E MA T I C S

of a very subtle analysis o f Hrushovski's counterexample,

Zilber was led to conjecture3°9 a new extended alternative:
the intrinsic geometry of a strongly minimal theory must
be (i) trivial, (ii) modular, (iii) algebraic (bi-interpretable
with (C. + . . . 0. 1 ) , or (iv) 'pseudo-analytic' (bi-interpretable
with an expansion of (C. +. . . 0. 1 ) with an analytic function
of the type exp) .
The fundamental point here is the surprising appear­
ance (again hidden and profound) of the complex exponen­
tial, with which Zilber conj ectured that the classification
could be closed. A first pathway in that direction con­
sists in exploring model-theoretic notions of pseudo­
exponentials m infinitary logics whose nuclei can be
located,310 so as to then construct limits of structures
with pseudo-exponentials that allow us to cover case (iv) .
Another, completely unexpected, path of exploration lies

mixtures between model theory, algebraic topology and algebraic geometry. Hrushovski.
drawing on techniques emerging in Zariski geometrics, succeeded in demonstrating a
Mordcll·Lang conjecture regarding the number of rational points over curves in fields
of functions (1996), which generalized the famous Mordell conjecture for curves over
Q (proven by Faltings, winning him the Fields Medal in 1986). This is, perhaps, the
most famous example of techniques comingjrom logic that help to resolve a problem
in the 'heart' of mathematics. (Remember Cannes.)
309 B. Zilber, 'Pseudo-exponentiation on algebraically closed fields of characteristic zero',
Annals Pure andAppliedLogi,e 132, 2004: 67-95.

310 The nucleus of the complex exponential up(2inx1 contains Z, where, with sum and
multiplication, Peano arithmetic can be reconstructed, which gives rise to the many
phenomena of incompleteness, instability, profusion of nonstandard models, etc.
".'./everthcless, using a logic with countable disjunctions Lrniwi the nucleus ofthe exponential
can beforced to be standard, by means of the sentence 3a 'ix ( e..rp ( .r ) =O ___, \/m>=Z .r=am)
(a= 2 i 1T in the classical case), for example. The pseudo-exponentials generalize various
properties of the complex exponential to arbitrary classes of models in Lell w• including
the standard forcing of the nucleus,

257
 O F C O NT E M P O RARY M AT H E M A T I C S

i n the new connections311 that Zilber has found between

pseudo-analyticity, foliations, and noncommutative geom­
etry. Specifically, Zilber has 'discovered' that, on the one
hand, the counterexamples to the trichotomy correspond
to deformations ef Zariski curves by means ef noncommuta­
tive groups, such as 'quantum tori' groups, and that, on
the other hand, other model-theoretic counterexamples
are linked to certain foliations studied by Connes.312
B ut, plunging even further into the depths,313 into the
abyssal model theory/noncommutative geometry hia­
tus, Zilber seems to be sensing the presence of the com­
mon structures efphysics,314 whose logical and geometrical

3u B . Zilber, 'Koncommutativc geometry and new �table structures' (preprint 2005,

available at http ://www 2 . maths.ox.ac. uk/-zilbcr/publ.html).

312 Ibid., 2-3.

313 \Visdom lies sunken in the deep, in the infinite abysses, as Melville suggests in Moby
Dick, in recounting Pip's second fall from the whaling vc!>scl and his immersion in
the ocean's lower strata: 'The sea had jeeringly kept his finite body up, but drowned
the infinite of his soul. l\�ot drowned entirely, though. Rather carried down alive to
wondrous depths, where strange shapes ef the unwarped primal world glided to and fro
before his passive eyes; and the miser-merman, V/isdom, revealed his hoarded heaps',
Herman :"v!elville, Moby Dick (1849-5 1 ) , H. Hayford, JI. Parker & G.T. Tanselle), eds.
(Evanston and Chicago: Northwestern Cniversity Press and The Newberry Library,
1 9 88), 414. Archeal mathematics active1y explores those 'strange shapes of the unwarped
primal world' that escape the frightened Pip.

314 Regarding the junction that Zilber envisions between the objects of model theory and
the deep structures of physics, Andres Villaveces remarks, 'The structures that are most
linked to noncommutative geometry and to the structures of physics arc the nonclassical
<;ariskigeometries. These form part of the "positive" side of the trichotomy, and appeared
in the article by Ilrushovski and Zilber. But only recently, only in the last tvm or three
,
years, has Zilber begun to see that certain cases of "finite coverings. that should have
been understood in terms of algebraic curves cannot be reduced to the lalter. This
dramatically changes two things: the junction with physics (more closely linked to a
refined analysis of Zariski geometries - of finite but not unitary coverings of varieties
that can only be undersrnod in terms of actions of noncommutative groups), and the
still more open role of pseudo-analytic structures' (personal communication, 2007).

258
 A R C H E A L M AT H E MA T I C S

representations would only be different facets o f deep,

unitary phenomena.315
The work of Mikhael Gromov ( Russia, b. i943 ) con­
firms, by way of other, utterly different routes, the richness
of certain 'archetypical' connections revealed in contem­
porary mathematics. Considered perhaps the greatest
geometer of recent decades, Gromov has completely
revolutionized the various fields of investigation he has
entered: geometry, where he introduced new perspec­
tives on 'smoothing' and 'globalization' that are tied
to the notion of metrics everywhere; partial differential
equations, where he introduced homotopic calculations
through partial differential relations ( PDE via PDR) ; sym­
plectic varieties, where he introduced into real analysis
pseudoholomorphic curves, and, with them, new techniques
of complex analysis; groups, where he introduced notions
of polynomial growth, asymptotic behavior and hyperbolic­
ity. I n all of these fields, Gromov's works combine many of

3 1 5 The great Russian school - a s w e have seen with Drinfeld, Kontsevich and Zilber, and
as we will soon see \vi th Gromov � consistently tends to reveal deep unitary structures
behind multiple mathematical and physical phenomena. This is also the case with the
works of Vladimir Voevodsky (2002 Fields medalist), who has succeeded in providing
a technical support for Grothendieck's motifs, as a central trunk ef cohomologi,es. By
introducing new Grorhendieck topologies for algebraic objects, Voevodsky managed
to construct subtle forms of 'surgery' for algebraic varieties - analogous to 'surgeries'
for topological spaces, but having to overcome far more delicate obstructions - and
succeeded in defining homotopic theories for algebraic varieties and schemes ( 2000).
At the intersection between algebraic geometry and algebraic topology, Voevodsky's
eidal ascent toward Grothendieck topologies would later allow him to bring about
a quiddital descent toward the singular cohomology ('enriching it' in Voevodsky's
sense) and pinpoint, in the last instance, the archeal motifs sought by Grothendieck
For technical introduction to Voevodsky's works, see Christophe Soule, 'The work of
Vladimir Voevodsky'. in Atiyah, Iago1nitzner, Fields Medalists' Lectures, 769-72.

259
 S Y NT H E T I C P H I L O S O P H Y

the qualities that Tao enumerates for what 'good math­

ematics' should be : 316 programmatic breadth of vision,
conceptual inventiveness, technical mastery, abstract
treatment, calculative skill, breadth of the spectrum of
examples, deep interlacing of the global/abstract with
the local/calculative, usefulness and applicability. The
influence of the Russian schoo[3'7 is particularly palpable
in this fantastic junction of geometrical vision, analytic
virtuosity, and physical applicability.
To a large extent, Gromov's ideas surface in a com­
plex counterpoint between refined webs of inequalities
and series of suitable invariants in those webs. This is
how it is with the triangular,3'8 isoperimetric,3'9 and

316 T. Tao, 'What is good mathematics?', preprint, arXiv:math.H0/0702396v1 February

i3, 2007. Gromov's works attain a level of excellence in the majority of the qualifications
that Tao specifies for 'good' mathematical work: the resolution of problems, technique,
theory, perspicacity, discovery, application, exposition, vision, good taste, beauty,
elegance, creativity, usefulness, power, depth, and intuition. Tao emphatically presents
his list 'in no particular order' (ibid., i) and, above all, makes an effort to illustrate the
correlationality of certain of those qualities in concrete works of higher mathematics.
And so, for the mathematician (Tao, for example), the synthetic configuration ofgood
qualities, in their coherent agglutination, matters more than placing those qualities on a
well-ordered, well-founded analytic scale.
317 Gromov completed his doctorate in 1969, at the University of Leningrad, under Rochlin.
Concerning the influence of his soviet colleagues, see R. Langevin, 'Interview: Mikhael
Gromov', in Pier, Development ofMathematics, 1,213-27 (1,221 in particular). Between
1974 and 1981, Gromov was a professor at Stony Brook; since 1981, he has been a
permanent professor at the IHES. Profiting from Cannes, Gromov and Kontsevich
(among others), the !HES has been able to perpetuate the great tradition of higher
mathematics opened by Grothendieck.
318 M. Berger, 'Renconcres avec un gComCtre' (1998), inJ.-M. Kantor, ed., OU en sont les
mathimatiques (Paris: Veuibert/SociCtC MathCmatiquc de France, 2002)1 399-440.
Berger's text emphasizes the inequalities mentioned here (400 ) .
319 M. Gromov, Metric Structures for Riemannian and Non·Riemannian Spaces (Boston:
Birkhauser, 1999). Notes from a course given by Gromov at Paris VII, 1979-80. First
published in French as M. Gromov, J. Lafontaine, P. Pansu, Structures mftri.ques pour

260
 A R C H E A L M AT H E M AT I C S

systoJic320 inequalities, and how i t i s with quite diverse

archeal constructs such as simplicial321 and minimal322
volumes, L2-invariants, homotopic invariants linked to
the geometry of partial differential relations, the Gromov­
Witten invariants, etc. From a philosophical point of view,
the emergence of these last few invariants is of particular
interest. Over a given symplectic manifold, we can define
many quasi-complex structures323 that do not necessarily
correspond to a complex manifold; in an effort to nev­
ertheless study the symplectic/real with techniques from
complex analysis, Gromov manages to overcome the
obstruction by introducing a new notion ofpseudoholomorphic
les variitis riemanniennes(Paris: Cedic-Nathan, i981); the English edition contains
extensive complements and appendices.) The sixth chapter, 'Isoperimctric inequalities
and amenability' (321-49), presents a detailed study of various forms of isoperimetric
inequality, by which the volume of a compact subspace is determined by means of
the volume of its border. According to Berger ('Rencontres ... ' , 415), isoperimetrics
in infinite dimensions should be seen as a form of geometric surgery.
320 Sections 4E+ ('Unstable systolic inequalities and filling') and 4F+ ('Finer inequalities
and systoles of universal spaces') (264-272) take up this theme directly. 'Systoles' are
minimal volumes of nonhomologous cycles in the Riemannian manifold; a particular
case consists in minimal noncontractible curves.
321 Given a compact manifold, its simplicial volume is defined as the infimum (greatest
lower bound) of the sums of (real) coefficients such that the fundamental class of the
manifold is covered by the sums of those coefficients multiplied by simplicial sets. The
simplicial volume turns out to be an invariant linked to the geometry of the manifold
in the in.finite, and it is therefore useful to study the varieties' asymptotic properties
(Berger, 'Rencontres ... ', 412).
322 Given a compact manifold, its minimal volume is defined in the class of all the Riemannian
structures linked to the manifold, through the metric that least raises the local behavior
of protuberances in the manifold (Berger, 'Rencontres . . . ' , 413).
323 Given a manifold M, a quasi-complex structure over M is a section J of the fibration
End ( TM) (TM being a tangent space) such thatJ2� -Id. If the manifold M is a complex
manifold, then multiplication by i defines such a structure. For details, see G. Elek, 'The
Mathematics of Misha Gromov', Acta MathematicaHungarica n3, 2006: 171-85. Elek's
article - prepared for the occasion of the awarding of the Bolyai Prize to Gromov in
2005 (a prize previously received only by PoincarC, Hilbert and Shelah!) - constitutes
an excellent technical presentation of Gromov's work.

261
 O F C O N T E M P O RARY M A T H E M A T I C S

curve, which behaves magnificently well in the n-dimen­

sional complex plane (any two points whatsoever can be
connected by means of an appropriate pseudo-holomor­
phic curve) ; moving on to search for invariants of those
curves, Gromov shows that the spaces modulo the curves
are compact, and that it is therefore possible to work
out a natural theory of homology, which leads to the
Gromov-Witten invariants; in the last instance, the new
invariants allow us, on the one hand, to distinguish an
entire series of hitherto unclassifiable symplectic varieties,
and, on the other, help to model unexpected aspects of
string theory.324
In this manner, a direction of transit (real-complex),
an obstruction in that transit (the multiplicity of the
pseudo-complex), a partial saturation of the obstruction
(pseudo-holomorphic curves) , and an archeal deepening
behind the new saturating concept (the Gromov-Witten
invariants), show that mathematics - far from striving
toward an analytical 'flattening' of phenomena's contra­
dictory oscillations needs that deeplyfissured topog;raphy
-

for its full development. I n fact, in a brilliant analysis of

situations of this sort, Gromov has pointed out3°5 that

324 I n the A model of string theory, 6 temporal dimensions are united in a 3-dimensional
symplectic manifold, and the 'leaves of the universe' are parameterized as
pseudoholomorphic curves over that manifold. The Gromov-\r\littcn invariants are
thus linked to deep physical problems. The interlacing of higher mathematics and
cosmology is once again underwritten in unexpected ways.
325 Langevin, 'Interview: Mikhael Gromov', u13-15.

262
 A R C H E A L M AT H E M AT I C S

'Hilbert's tree' (the ensemble o f mathematics' branches) ,

far from being simply planar and deductive, is crisscrossed
by multidimensional geometrical objects: exponential nodes
(sites in the tree bearing great, amplified oscillations:
number, space, symmetry, infinity, etc.) , clouds ('guides',
or coherent gluings, such a s geometrical nuclei a la Zilber,
inside a tree that is a priori disconnected for reasons of
complexity or undecidability), local wells (sites where
mathematical information 'sinks' and is lost) , etc.
Gromov's geometrical style recalls (implicitly) two of
Grothendieck's synthetic strategies - a global vision of
classes of structures and observation of properties on
a large scale - and supplements them with an incisive,
comparative analytic, through a doublejrag;mentation and
reintegration of the webs of inequalities under investiga­
tion. Gromov, in fact, offers a new understanding of
Riemannian geometry by contemplating the class of all
Riemannian manifolds and working with multiple metrics
within that class - thereby setting the manifolds into
motion and finding in that movement the appropriate
relative invariants. Likewise, his works in the realm of
partial differential relations inscribe themselves within
a double matrix that allows surprising gluings to be per­
formed along two primordial axes: synthetic/analytic
and global/local. The h-principJe326 (h for homotopy)

326 M. Gromov, Partial Differential Relations (New York: Springer, i986).

263
 O F C O N T E M P O RA RY M AT H E M AT I C S

effectively postulates, in certain geometrical realms, the

existence of certain homotopic deformations between
continuous sections of a sheaf (linked to local differential
correlations that codify the local conditions in a partial
differential equation) and the sheaf's holonomic sections
(tied to global solutions, through global differentials) . The
monumental work undertaken by Gromov in his Partial
Differential Relations succeeds in exhibiting the ubiquity
of the h-principle in the most remote areas of geometry
(the synthetic richness of the principle) and constructing
a multitude of local methods and practices for verifying
the h-principle in particular conditions (analytic richness)-
The group structure, rej uvenated by the likes of
Connes and Zilber, seems to enjoy infinite lives in Gro­
mov's hands. His program in geometrical group theory can
be described as an intent to characterize finitely generated
groups, modulo quasi-isometries, which is to say, modulo
'infinitesimal' deformations of Lipschitz-type distances.327
I n that program, Gromov has demonstrated that many
properties of groups turn out to be quasi-isometric invari­
ants; in particular, the (word-)hyperbolicity of a group328 is
one such invariant, by which we can characterize the linear

327 For technical details, see Elek, 'The :Mathematics of Misha Gromov', 181-2.
328 The basic example of a hyperbolic gro11p, in Gromov's sense, is the fundamental group
of an arbitrary manifold of negative curvature. The generalization of certain properties
of ·rhin' triangles in the universal covering of that variety leads to abstract definitions
ofhyperbolicity (ibid., 183).

264
 A R C H E A L M A T H E M AT I C S

complexity o f a group's associated word problem.329 O n the

other hand, using the definition of a metric in a Cayley
graph of a finitely generated group, Gromov is able to
define the 'polynomial growth' of a group and study that
asymptotic growth's correlations with classical properties:
solub ility, nilpotence, Lie sub-representations, etc .33°
We thus find ourselves before a quite typical situa­
tion in contemporary mathematics, where certain classical
nuclei are seen as the limits ofdeformations (be they logical,
algebraic, topological or quantic) in very broad classes of
spaces. In virtue of these great synthetic processes ( in the
style of Zilber's 'groups everywhere' , or Gromov's 'metrics
everywhere' ) , the classical invariants are recuperated, but
many new invariants (archeal in our terminology) are
discovered as well, invariants that a restricted vision -
whether local, analytic or classical - cannot catch sight of.

329 Given a recursively presented group G, the wordproblem associated with (; consists in
decid ing if two finite products of the generators of G (that is to sa;\ words in the free
group) coincide or not. Some groups for which the word probl em is soluhlP include
finite groups and simp le ) finitely generated groups. It can be demonstrated that a
uniform solution of the problem for all groups docs not exist, and so the measure of
the problem's complexity for certain classes of groups turns out to be a result of great
interest. Gromov has demonstrated that the complexity of the word problem for a
given group is linear if and only if the group is hy perb olic. By means of a metric in
the class of fin itely generated groups, it has been demonstrated that the closure of the
subclass of hyperbolic groups contains the 'Tarski "M on s ters ' . (The latter are infinite
groups whose nontrivial subgroups are cyclical groups of order p, for a fixed prime p;
the existence of such groups was proven by Olshan ski i [1980], with p> 10 �-. - a result
that should make philosophers dream I )

330 Ibid., 183-4.

265
 PART TH R E E

Synthetic S ketches
C HAPTE R 8

FRAG M E N T S OF A T RAN S I T O RY O N T O L O GY

In this, the third part of the work, we will reflect (philo­

sophically, methodologically and culturally) on the case
studies that we have presented in the second part. Accord­
ingly, when we refer to 'mathematics' (and its derivative adjec­
tives) in whatfollows, we mean 'contemporary mathematics' ,
unless we explicitly state otherwise. Now, we should note
right away that this essay cannot cover all the forms of
doing mathematics, and, in particular, will not dwell
on the practices peculiar to elementary mathematics.
We therefore do not aim to produce anything like an
all-encompassing philosophy of mathematics, only to
call attention to a very broad mathematical spectrum that has
rarely been accountedfor in philosophical discussions, and
which should no longer be neglected. In the final chapter
we will try to provide an intrinsic characterization of the
interval between i950-2000 (open on both extremes)
with regard to 'contemporary mathematics'; but for the
moment, we we will merely ground ourselves on the con­
crete cases of mathematical practice reviewed in part 2 .
We will provide a n extensive number o f cross references
to those case studies; to that end, we will systematically use
references between square parentheses, such as [x] , [x-y] and

269
 S Y N T H E T I C P H I LO S O P H Y

[x,y, ... ] in the body efthe text, which will direct the reader
to the pages x, x-y, or x, y, . . . of this book.
The case studies of the second part should have made
it clear that contemporary mathematics is incessantly
occupied with processes eftransit in exact thought, involved
in multiple webs ef contradistinction, both internal and
external. From this it immediately follows that the ques­
tions concerning the content and place of mathematical
objects - the ontological 'what' and 'where' - through
which we hope to describe and situate those objects,
cannot be given absolute answers, and cannot be fixed
in advance. The relativity of the 'what' and the 'where'
are indispensable to contemporary mathematics, where
everything tends toward transformation and flux. I n
this sense, the great paradigm of Grothendieck's work,
with its profound conception of a relative mathematics
[140-141] interspersed with changes of base of every
sort in very general topoi [141-142], should be fully
understood as an 'Einsteinian turn' in mathematics. As
we have seen, we are dealing with a vision that ramifies
through all the mathematics of the epoch, and which is
also capable of giving rise to a genuine Einsteinian turn
in the philosophy efmathematics.
Now, the point of Einstein's theory of relativity, once
we assume the movement of the observers, consists in
finding suitable invariants (no longer euclidean or Gali­
lean) behind that movement. Likewise, the point of a

270
 F RA G M E N T S O F A T RA N S I T O RY O NT O L O GY

relative mathematics a la Grothendieck, once we assume

the transit of mathematical objects, consists in finding
suitable invariants (no longer elementary or classical) behind
that transit. This is the case with many of the archeal
situations in mathematics that we have been looking at:
motifs [144-146] , PCF theory [201-202] , intermedi­
ate allegories [245-246], Zilber's extended alternative
[256], the h-principle [263-4], etc. A skeptical relativ­
ism , which leads to disorientation and allows for an isotropy
of values, in the style of certain postmodern subrelativisms
or the infamous pensiero debole, is thus very far from being
the same as the Einsteinian or Grothendieckian projects,
where, though there be neither absolute foundations nor
fixed objects, not everything turns out to be comparable
or equivalent, and where we can calculate correlative archeal
structures - that is, invariants with respect to a given context
and a given series of correlations - which, precisely, allow
differences to be detected and reintegrated.
The first important point in specifying 'what' math­
ematical objects are consists in really taking relativity and
transit in contemporary mathematics seriously. Objects
in this realm cease to be fixed, stable, classical and well
founded - in sum, they cease to be 'ones' . I nstead they
tend toward the mobile, the unstable, the nonclassical,
and the merely contextually founded - in short, they
approach 'the many'. Multiplicity everywhere underlies
contemporary transit, and the objects of mathematics

271
 O F C O N T E M P O RA RY M AT H E M AT I C S

basically become webs andprocesses. Determinate 'entities' ,

firmly situated in one absolute, hard and fast universe , do
not exist; instead we have complex sig;nic webs interlaced
with one another in various relative, plastic and fluid
universes. The levels contemplated by these 'complex
signic webs', where mathematical objects are constituted,
are multitudinous, and nofixed level exhausts the richness
ofthe object (web) .
This is obvious, for instance, with the mathematical
'object', group; we have seen how that object appears and
captures disparate information (under the most diverse
representation theorems) in the most distant realms
of mathematics: homology and cohomology groups
[142-148, 178-1 79] , G a l o i s gro u p s [150, 155,
2 25] , group actions ( 162-163, 180-181), abelian
groups [165), homotopy groups [176], algebraic groups
[ 184] , the Grothendieck-Teichmiiller gro up [ 225,
233] , Lie groups [223] , quantum groups [223] , Zil­
ber groups [255-256] , hyperbolic groups (264] , etc.
It is not that we are confronted here, ontologically, with a
universal structure, that takes on supplementary proper­
ties at each supposed level of reading (logical, algebraic,
topological, differential, etc.); what happens is rather that
the diverse webs of mathematical information codified
under the structure of group overlap ('presynthesis') and
compose ('synthesis') so as to transmit information in a coherent
fashion. It is not that there exists 'one' fixed mathematical

272
 F RA G M E N T S O F A T R A N S I TO RY O N T O L O GY

object that could be brought to life independently of the

others, in a supposedly primordial universe; what we find
instead is the plural existence of webs incessantly evolving as
they connect with new universes of mathematical interpre­
tation. This is particularly evident in the inequality webs
[260] studied by Gromov, or in the webs of equideduc­
ible theorems [ 248-249] that Simpson has displayed;
the progressive leaps and bounds taking place in the webs
go on to configure the global landscape, and this modifies
in turn the entities that the global environment locally
internalizes.
Given that the objects of mathematics 'are' not stable
sums but reintegrations efrelative differentials, the question
concerning their situation ('where they live') takes on an
asp ect that is almost orthogonal to the way in which this
question is posed from an analytic perspective (grounded
in set theory). For if mathematics finds itself in perpetual
transit and evolution, the situation of an object cannot
be anything but relative, with respect to a certain realm
('geography') and to a moment of that realm's evolution
('history') . This just goes to reinforce the position of
Cavailles, who understood mathematics as gesture - a
position that has echoed throughout the century, all the
way to Gromov, who pointed out how we should come
to 'admit the influence of historical and sociological fac­
tors'w in the evolution of Hilbert's tree [262] .
331 Langevin, 'Interview: Mikhael Gromov', 1,214.

273
 S Y NT H E T I C P H I L O S O P H Y

Needless to say, this reading o f mathematics a s a histori­

cal science goes against the grain of the readings offered
by the analytic philosophy of mathematics - readings
according to which fragments of the edifice arise a tempo­
rally against absolute backgrounds, codified in the various
analytic 'isms' [102-103], and within which each com­
mentator plays at undermining contrary positions, and
proposing his version as the most 'adequate', which is to
say, as the one most capable of resolving the problematics
at stake. Curiously, however, the supposed local reconstruc­
tions of 'mathematics' - studied a hundred times over in
the analytic texts - go clearly against what mathematical
logic has uncovered in the period from 1950 to 2000. In
fact, as we have seen, following the style of Tarski (logics
as fragments of algebras and topologies) and Lindstrom
(logics as coordinated systems of classes of models), the
most eminent mathematical logicians of the last decades
of the twentieth century (Shelah, Zilber, Hrushovski)
have emphasized the emergence of deep and hidden
geometrical kernels [195, 253] lying behind logical
manipulations. Just as Jean Petitot, in his mixed studies332
on neurogeometry, Riemannian varieties and sheaf logic
(Petitot declares himself a great admirer of Lautman's
332 See]. Petitot, 'Vers une neuro·geometrie. Fibrations corticales, structures de contact et
contours subjectifs modaux', Mathbnatiques, Infonnatique et Sciences Humaines 145, 1999:
5-101, or 'The ncurogeometryof pinwheels as a subriemannian contact structure'Journal
efPhpiology 97, 2003: 265-309. His remarkable doctoral thesis (Pourun schimatisme de
la structure [EHESS 1982, 4 vols. J, part of which is contained in Morphogenise du sens
[Paris: PUF, 1985]) has still not been made use of much in mathematical philosophy.

274
 F RA G M E N T S O F A T RA N S I TO RY O N T O L O GY

'mixtures'), has begun to defend the idea that geometrical

proto-objects underlie neuronal activity, and, moreover,
that a proto-geometry should take heuristic precedence over a
language, contemporary mathematical logic has likewise
come to demonstrate how a proto-geometry necessarily pre­
cedes a logic. What is at stake, then, is a situation that leads
us to completely overturn - again, almost orthogonally
- the usual approaches of analytic philosophy.
Within contemporary mathematics, and following
the 'double orthogonality' that we have pointed out,
an object is not something that 'is', but something that
is in the process of being, and these occurrences are not
situated in a logical warp, but in an initial spectrum of
proto-geometries. The consequences for an ontology of
mathematics are immense and radically innovative. On
the one hand, from an internalist point of view, the 'what'
involves webs and evolving transits, while the 'where'
involves proto-geometrical structures anterior to logic
itself. On the other hand, from an externalist point of view,
the 'what' takes us back to the unexpected presence of
those proto-geometries in the physical world (the actions
of the Grothendieck-Teichmiiller group on the universal
constants of physics [224-225,!l33] , the Atiyah index
[208-209] , the Lax-Phillips semigroup [218-219] ,
the interlacing of Zilber's pseudo-analyticity with the phys­
ical models of noncommutative geometry [257-259],
etc.), while the 'where' conceals a profound dialectic of

275
 O F C O N T E M P O RARY M AT H E M AT I C S

relative correlations between concrete phenomena and

their theoretical representations. In fact, what comes to
light in these readings is that the questions concerning
an absolute 'what' or 'where' - whose answers would
supposedly describe or situate mathematical objects once
and for all (whether in a world of 'ideas' or in a 'real'
physical world, for example) - are poorly posed questions.
The richness of mathematics in general (and of con­
temporary mathematics in particular) consists precisely
in liberating and not restricting the experience and influ­
ence of its objects. In a certain sense, a base from which
we may better understand mathematics' indispensable
transitability could be furnished by the mixture of (i)
structuralism, (ii) categories and (iii) modalization that Hell­
man proposes [ 106-107] , but in reality the situation
is more complex, as comes to light in the multiple oscil­
lations, hierarchizations and ramifications cast in relief in
this book's second part. In fact, if a categorical reading
(ii) seems indispensable in contemporary mathemat·
ics (something we have emphasized with multiple case
studies beneath which, implicitly or explicitly, lies the
Grothendieckian categorical and relativistic program) and
if, beyond a certain threshold of complexity, mathemati­
cal modalization (iii) seems equally indispensable (the
multiform transit between hypotheses, models, calculations
and contradistinctions, carefully hidden in the classical
formalization of set theory, and carefully ignored by so

276
 F R AG M E N T S O F A T RA N S I T O RY O N T O L O G Y

many of the 'hard' currents of the analytic philosophy of

mathematics), contemporary mathematics nevertheless
underscores the importance of processes over structures
(i) , since the latter emerge on a second level as the invari­
ants of suitable processes. The relative ontologization of
objects in contemporary mathematics thus forces us to
further dynamize Hellman's base (which appears fixed in
a quadrant of Shapiro's square [ 10] ) , so that we can
better take stock of the contemporary processes.
Ultimately, when faced with contemporary mathemat­
ics we cannot escape a certain 'transitory ontology' that,
at first, terminologically speaking, seems self-contradictory.
Nevertheless, though the Greek ontotetes sends us, through
Latin translations, to a supposedly atemporal 'entity' or
an 'essence' that 'ontology' would study,333 there is no
reason (besides tradition) to believe that those entities or
essences should be absolute and not asymptotic, governed
by partial gluings in a correlative bimodal evolution between
the world and knowledge. The philosophical bases of such
a 'transitory ontology' can be found in Merleau-Ponty's
theory of shifiing334 (in the general realm of knowledge
and the particular realm of visuality), and in the specific

333 J. -F. Courtine1 'Essence', in Cassin, Vocabulaire europien des philosophies, 400-14.
334 See M. Merleau-Ponty, Notes des cours du College de France (1958-59, 1960-61) (Paris:
Gallimard, 1996), and L 'oeil et !'esprit (Paris: Callimard, 1964) . The latter is the last
text he published in his lifetime, in 1961, and is a magnificent way to be introduced to
Merleau-Ponty's work See also his two posthumous works, La prose du monde (Paris:
Gallimard, 1969) and Le visible et !'invisible (Paris: Gallimard, 1964). The hiatus between
the visible and the invisible can only be understood while shifting within it.

277
 SYNTH E T I C P H I L O S O P H Y

transitory ontology proposed by Badiou335 (in the field of

mathematics). For Merleau-Ponty, the 'height ofreason'336
consists in feeling the shifting ef the soil, in detecting the
movement of our beliefs and supposed claims of knowing:
'each creation changes, alters, clarifies, deepens, confirms,
exalts, recreates or creates by anticipation all the others' ,337
A complex and mobile tissue of creation surges into view,
full of 'detours, transgressions, slow encroachments and
sudden drives', and in the contradictory coats of sediment
emerges the force of creation entire. This is exactly what we
have detected in contemporary mathematics, as we have
presented it in part 2 of this work. For Badiou, mathemat­
ics is a 'a pseudo-being's quasi-thought', distributed in
'quasi-objects'338 that reflect strata of knowledge and world,
and whose simple correlations (harmonic and aesthetic)
gmw over time. Those quasi objects not only escape fixed
-

and determinate identities: they proceed to evolve and

distribute themselves between warps of ideality and real­
ity. Through the dozens of situations surveyed in part 2 ,
we have indeed been able to make evident the distribu­
tion, the grafting and gluing of those quasi-objects, not only
between fragments internal to mathematics, but between

335 Badiou, Brir;ftngs on Existence.

336 Mcrleau-Ponty, L'Oeil et /'esprit., 92 .
337 Badiou, Brif!fings on Existence, 4 7 .

338 Ibid., 42-3.

278
 F R A G M E N T S O F A T RA N S I T O RY O NT O L O GY

theoretical mathematics (the realm of the eidal) and the

concrete physical world (the realm of the quiddital) .
With new perspectives and with new force, Badiou's
transitory ontology allows us to bring an untrivialized Plato
back to the landscape of mathematical philosophy. In the
style of Lautman's dynamic Platonism attentive to the -

composition, hierarchization and evolution of the mix­

tures in the Philebus339 Badiou, too, insists on a Platonism
-

essentially open to a 'cobelonging of the known and the

knowing mind', through which one derives an 'ontological
commensurability' 34o that incorporates movement and
transit. We thereby immerse ourselves in a Platonism that
is not static, not fixed to a supposedly transcendent world
of Ideas, and very far from the image most often used to
sum up the doctrine.w Badiou's dynamic Platonic reading
opens onto quite different philosophical perspectives - in
particular, onto an understanding of mathematics as evolv­
ing thought342 and a study of the problems of saturation,
maximization, and invariance in the movements of

339 Lautman, Essai sur l'uniti. . . , 143-7, 203, 227-8, 303. See N.-1. Boussolas, L'Etre et la
composition des mixtes dans le Phitebe de Platon (Paris: PUF, 1952).

340 Badiou, Briefings on Existence, 90.

341 This is what we find in the 'trivialization' presented by Benacerraf: 'Platonists will be
those who consider mathematics to be the discovery' of truths about structures that
or
exist independently of the activity thought of mathematicians'. In Benacerraf &
Putnam, Philosophy ofA1athematics, 18.
342 Badiou, Briefings on Existence, 39-54, 102. Mathematics as 'thought' studies the exact
transitions of being and approaches a transformative/productive gesturalil)' (CavaillCs,
of
Ch<ltelet) nontrivial information. This position is quite far from an understanding of
mathematics as a mere language game, or as a set of tautological deductive transitions.

279
 O F C O NT E M P O RARY M AT H E M A T I C S

thought - which, following Merleau-Ponty, falls within

the asymptotic order ef 'shijting', and which covers some of
the deeply mathematical problematics to which we have
called attention in the case studies contained in part 2 .
Th e process ef being, commensurably i n between the webs
ef cognition and the webs efphenomena, is a fundamental
characteristic of 'transitory ontology' that contemporary
mathematics requires. What we are dealing with here is an
apparently innocuous change efgrammatical category in pos­
ing ontological questions, which nevertheless affords them
a whole new dimensionality and enriches the problematics
at stake: the 'what' and the 'where' - which were initially
taken up in an absolute and actual present, leading us to
ask poorly posed questions - come to be understood in a
modal and relative present perfect. The major consequence
of this opening up to transit is to immediately shatter the
dualisms and pigeonholes of Shapiro's square [IO] .
As we have seen, the objects of mathematics - that is,
in reality, its processes and quasi-objects - ceaselessly
transit between eidal, quiddital and archeal webs, whether in
the world interior to mathematics, or in their perpetual
contradistinction with the physical exterior. What in one
context turns out to be eidal (Grothendieck-Teichmuller
(G T ) , linked to the Euler numbers [225], for example),
in another appears as quiddital (GT acting on the uni­
versal formulas of physics [233]) , and in yet another
appears as archeal (GT in the dessins d'enfant (225]) .

280
 F R AG M E NT S O F A T RA N S I T O RY O N T O L O GY

The classical dissociations and exclusions ( of the either-or

variety, in the style of Benacerraf's dilemma [n ]) have
caused excessive and unnecessary damage to the philoso­
phy of mathematics, and the time to overcome them is at
hand. To that end, we can already count on at least three
great tendencies in contemporary mathematics - where
binary positionings are cast aside and where perspectives
of continuity between diverse webs are opened up - that
philosophical reflection should begin to take seriously:
the understanding of the 'positive' ( classical, commuta­
tive, linear, elementary, structured, etc. ) as a limit of
'negative' mediations ( intuitionism, noncommutativity,
nonlinearity, nonelementarity, quantization, etc. ) ; the
theory of sheaves, with its continuous handling of coher­
ence and gluing between the local and the global; the
mathematical theory of categories, with its handling of
differentiation and reintegration between the particular
and the universal. In what follows, we will clarify these
three deep technical tendencies from a conceptual and
philosophical perspective, and explain how they allow us
to reinforce a transitory ontology of mathematics.
Among the characteristics specific to mathematics
between i950 and 2000, we have indicated the act of
working with the jluxion and deformation of structures'
typical boundaries [42] . This is obvious, for example,
in Grothendieck's K-theory [135-136, 210-212] ,
where we study the perturbations of morphisms over

281
 S Y N T H E T I C P H I L O S O P HY

classical fibers; in Shelah's abstract elementary classes

[196-199] , where we study limits of algebraic invari­
ants beyond first-order classical logic; in the Lax-Phillips
semigroup [218-219] , where non-euclidean geom­
etry allows us to understand the scattering of waves;
in Connes 's noncommutative geometry [ 222-224),
where the dispersion and fluxion of quantum mechan­
ics and thermodynamics are unveiled; in Kontsevich's
quantizations [229-233] , where classical structures
obtain as limits of the quantized; in Freyd's allegories
[243-246] , where topoi are seen as limits of interme­
diate categories; in Zilber's extended alternative [ 256),
where counterexamples to the trichotomy arise as defor­
mations of noncommutative groups; or in Gromov's
theory of large-scale groups [264-265] , where we study
the nonlinear asymptotic behavior of finitely generated
groups. These are all examples of high mathematics, with
great conceptual and concrete content (and are thus far from
being reducible to mere 'language games'). In them, a
new understanding of mathematical ( quasi-)objects is
demonstrated, by means of fluxions, deformations and
limits, and by passing through intermediate strata that are
nonpositive ( nonclassical, noncommutative, nonlinear,
nonelementary, and nonquantized).

282
 F R A G M E N T S O F A T RA N S I T O RY O N T O L O GY

A result due to Caicedo343 shows that classical logi,c in a

'generic' fiber of a sheaf effirst-order structures is no more
than an adequate limit efintuitionistic logi,c in the 'real'fibers
of the sheaf This remarkable situation shows that the
construction of the classical and the positive as 'limit
idealizations', as seen in the aforementioned mathematical
examples, is reflected in the realm of logi,c as well, and
in exactly the same fashion. What comes to the surface,
once again, is the continuity efmathematical knowledge, for
which watertight compartments are worthless. The rich­
ness of Caicedo's sheaf logic is precisely due to its being
elaborated in an intermediate zane between Kripke mod­
els344 and Grothendieck topoi [ 141-144] , benefiting
from the many concrete examples of the former and the
abstract general concepts of the latter. In a total crossing
of algebraic, geometrical, topologi,cal and logi,cal techniques,
Caicedo constructs an instrumentarium to incite transit

343 X. Caicedo, 'L6gica de 1os haces de estructuras', Revista dela Academia Colombiana de
Ciencias Exactas, F£sicasy Natura/es XIX, 7 4, i995: 569-85. Caicedo provides a framework
of great depth and breadth - mathematical, logical, conceptual, and philosophical
- which is unfortunately still unknown by the international community. He has
announced (in 2012 ) a forthcoming publication in English. For a partial overview in
Italian, see F. Zalamea, 'Ostruzioni e passaggi nclla dialettica continuo/discreto: ii
caso <lei grafi esistenzia1i c della logica dei fasci', Dedalus. Riuista di Filoscifia, Scien.w e
Cultura - Uniuersitii di Milano 2, 2007: 20-5.
344 Kripkemodels are 'trees' that can be used to represent a branchingtemporal evolution;
from a mathematical point of view, they are simply presheaues over an ordered set (seen
as a category). Kripke models furnish a complete semantics for intuitionistic logic. Other
complete semantics for intuitionism are provided by the class of topological spaces,
or by the class of elementary topoi [111]. Intuitionism and continuity are interlaced
over an archeal ground in this manner, and Thom's aporia emerges in a new form:
classical-discrete versus intuitionistic-continuous. This new appearance of intuitionism,
untethered from its original constructivist aspect, has not been sufficiently exploited
in the philosophy of mathematics.

283
 O F C O NT E M P O RARY M A T H E MA T I C S

and transference. The result i s his 'fundamental' theorem

of model theory, where pivotal statements in logic such
as the Loz theorem for ultraproducts, the completeness
theorem for first-order logic, forcing constructions in sets,
and theorems of type omissions in fragments of infinitary
logic, can all be seen, uniformly, as constructions of
generic structures in appropriate sheaves.
From a philosophical point of view, a striking con­
sequence follows from all of this. Through fluxions and
transit we effectively see that the classical perspectives are
no more than idealities, which can be reconstructed as limits ef
nonclassical perspectives that areJar more real. A particular
case of this situation is a new synthetic understanding of
the point-neighborhood dialectic, where - contrary to
the analytic and set-theoretic perspective, according to
which neighborhoods are constructed from points - the
classical, ideal 'points', which are never seen in nature, are
constructed as limits of real, nonclassical neighborhoods,
which, by contrast, are connected to visible, physical
deformations. From this perspective, the ontology of
the ( quasi-)objects at stake once again undergoes a radi­
cal turn: an 'analytic' ontology, linked to the study of
set-theoretic classes of points, can be no more than an
idealized.fiction, which forgets an underlying 'synthetic' /
'transitory' ontology that is far more real, and linked to
the study of physically discernible, continuous neighbor­
hoods. The transitory and continuous ( quasi-)objects of

284
 F R AG M E N T S O F A T RA N S I T O RY O N T O L O GY

nonclassical mathematics are thus interlaced with transitory

and continuous phenomena in nature, through webs of
informative correlation that are gradual, nonbinary and
nondichotomous.
This situation is corroborated by the principal para­
digms of sheaf theory.345 The ancient philosophical prob­
lematic, 'how do we get from the Many to the One?'
(corresponding to a phenomenological transit) becomes
the mathematical problematic, 'how do we get from
the local to the global?' ( technical transit) , which, in
turn, can be subdivided into the questions (i) 'how do
we differentially register the local?' and (ii) 'how do we
globally integrate those registers?' The natural mathemati­
cal concepts of neighborhood, covering, coherence and
gluing emerge in order to analytically tackle question (i),

345 Sheaves emerge in the work of Jean Leray, in the study of indices and coverings for
differential equations (with his works in the Oflag XVII, and articles from i946-50;
the term 'sheaf' first appears in 1946). In the general realm of the study of a manifold
through its projections into manifolds of lower dimensions (Picard, Lefschetz,
Steenrod), there arises the problematic of studying the topology of the initial manifold
by means of the coherent information provided by the fibers in the projection, and
sheaves are precisely what help to capture (and glue together) the continuous variation
in the fibers. The Cartan Seminar of the Ecole Normale SupCrieure (1948-51) served
ro distill Leray's ideas and present the sheaf concept as it is known today: as a fibered
space, or «:talC space' in Lazard's terminology (not to be confused with Grothendieck's
notion of etale: distinguished by a mere accent aigu, these two concepts are almost
diametrically opposed - the first is ramified, the second is non ramified), and as the
sheaf of germs of sections. Godement unified the concepts and terminology in his
Topologiquealge'briqueet tMoriedesfasceaux (1958), in parallel with Grothendieck's Tohoku
[162-166]. For a detailed history of sheaf theory, see ]. Gray, 'Fragments of the
history of sheaftheory', in LectureNotes in Mathematics 753 (New York: Springer, 1979)1
1-79, and C. Houzcl, 'Histoire de la thi:orie des fasceaux', in La giomitrie algibrique
(Paris: Albert Blanchard, 2002), 293-304. In our final chapter we will situate the
appearance of sheaf theory as the concretization of a deep conce,ptual break between
modern and contemporary mathematics, and we will try to show that the diachronic
break around 1950 is not a mere historical accident.

285
 S Y N T H E T I C P H I L O S O P HY

while the natural mathematical concepts of restrictions,

projections, preservations and sections emerge in order
to synthetically tackle question (ii) . Pres heaves (a term
we owe to Grothendieck) cover the combinatorics of the
discrete interlacings between neighborhood/restriction
and covering/projection, while sheaves cover the continu­
ous combinatorics tied to the couplings of coherence­
preservation and gluing-section (figure i3) . In this way, the
general concept efsheafis capable of integrating a profound
web of correlations in which aspects both analytic and
synthetic, both local and global, and both discrete and
continuous are all incorporated.
correlations -- (prescissions, in Peirce's sense) +- (conditions)

seclions +:::!: preservations :;:::t projections +:t restricrions SYNTHESIS

SHEAF

gluing � coherence � covering � neighborhood ANALYS IS

SHEAF
PRESHEAF

Figure 13. The sheafconcept: general transitori.city efanalysis/synthesis, local/global,

and discrete/continuous.

I n prescinding346 the fundamental notion of gluing

in the sheaf, the notions of coherence, covering and
346 Prescisrion, in Peirce's sense, allows for the passage from the panicular lo the general,
descry·ing the most abstract in the most concrete. It is a process thal is ubi quicous in
mathematics, and one that also takes place in the passage from the Many to the One,
which is to say, in the phenomenological search for universal categories. Peirce's three
cer1ofl)•thagorean categories ( fi.rstness: immediacy; secondness: action-reaction; thirdness:
mediation) surface in meticulous dialecrics of prescission. For a detailed study of those

286
 F RA G M E N T S O F A T R A N S I T O RY O N T O L O GY

neighborhood emerge progressively and necessarily. These

las t two notions are of the greatest importance in onto­
logical discussions. On the one hand, if it turns out that
the 'real' cannot, as we have seen, be anchored in the
absolute, and if the 'real' can therefore only be under­
stood by way of asymptotic conditions, then strategies for
covering fragments of the real take on a pivotal ontologi­
cal importance. It is in this sense that we should recall
Rota's exhortations that, with respect to mathematical
objects, we must escape the 'comedy of existence' staged
by analytic philosophy, and instead attend to a 'primacy
of identity'347 that is bound to the grafting of the real's
modal coverings, and which will allow us to classify the
possible identities between ideas and physical objects,
identities that evolve - and in the most interesting cases
remain invariant - over large-scale periods (here we
should remember Gromov) . On the other hand, a turn
toward a real logic efneighborhoods, as a counterpoint to a
logic that is ideal and punctual - a turn to sheaflogic as a
counterpoint to classical logic - likewise leads to a radi­
cal ontological turn. Many of the exclusive disjunctions
presupposed by analytic and classical thought are under­
mined in synthetic environments (whether physical or

dialectics - linked to 'bipolar tensions' similar to those that have been discovered in
the adjunctions of mathematical category theory - see De Tienne, L 'anal;•tique de la
ripresentation . . .
34i Rota, Indiscrete 7houghts, i84-6.

287
 O F C O N T E M P O RARY M AT H E M AT I C S

cognitive) where mediations are the rule. I n such syn­

thetic/transitory/continuous realms, in particular, there
is no reason why the disjunctive question concerning
the 'ideality' or 'reality' of mathematical (quasi-)objects
should be answered in an exclusive fashion [6-14) ,
as is suggested by the watertight compartments of Shap­
iro's table [10] . In concrete terms, sheaves, those (quasi-)
objects indispensable to contemporary mathematics, acquire
all their richness in virtue of their double status as idealj
real, analytic/synthetic, local/global, discrete/continu­
ous; the mutual inclusion (and not the exclusion)348 of
opposites - an incessant exercise of mediation - secu res
their technical, conceptual and philosophical force.
Category theory axiomatizes rqjons of mathematical
practice, according to the structural similarities of the
objects in play and the modes of transmission of infor­
mation between those objects, in harmony with Peircean
pragmaticism [1n-121] , which is equally sensitive
to problems ef traniference. In an inversion of set theory,
where objects are internally analyzed as a conglomerate
of elements, category theory studies objects through their
external, synthetic behavior, in virtue of the object's rela­
tions with its environment. A morphism is universal with

348 Rota clearly expresses the urgency of not adopting exclusions in an a priori fashion:
'Mathematical items can be viewed either as analytic statements derived within an
axiomatic system or as facts about the natural world, on a par with the facts of any
other science. Both claims are equally valid. l · · · l The contextual standing of an item
as analytic or synthetic is not fixed'. Ibid., 168.

288
 F RA G M E NT S O F A T R A N S I T O RY O NT O L O GY

respect to a given property if its behavior with respect to

other similar morphisms in the category possesses certain
characteristics of unicity that distinguish it within the
categorical environment. The basic notions of category
theory associated with universality - the notions ofjree
object and adjunction - respond to problematics bound up
with the search for relative archetypes and relative dialectics.
In the multiplicity, in the wide and variable spectrum of
mathematics' regions, category theory manages to find
certain patterns of universality that allow it to overtake the
splintering of the local and the surmounting of concrete
particulars. In a category, for example, a free object can
be projected into any object of the category's sufficiently
large subclasses: it is therefore something of a primordial
sign, incarnated in all of the correlated contexts of inter­
pretation. Beyond relative localizations, certain relative
universals thus arise, giving an entirely new technical
impulse to the classical notions of universality. Though
we can no longer situate ourselves in a supposed absolute,
nor believe in concepts that are universally stable in space
and time, the notions efuniversality have been redimensioned
by category theory, and coupled with a series of transfer­
ences relative to the universal-free-generic, facilitating a
transit behind which are revealed remarkable invariants.
In this manner, category theory explores the structure
of certain generic entities ( 'generalities') by a route akin
to the 'scholastic realism' of the later Peirce. Categorical

289
 O F C O N T E M P O RARY M AT H E M AT I C S

the diagram first places, o n the left, a sign in a n abstract

category, and, on the right, the same sign partially incar­
nated in various concrete categories. The various 'modula­
:ions' and 'pragmatic differentials' let us pinpoint a sign
:hat is one, abstract and general, and convert it into the
multiple, the concrete and the particular. It is a task that,
in category theory, is achieved by means of the various
functors at stake, which, depending on the axiomatic
richness of each of the categorical environments on the
right, incarnate the general concepts in more or less rich
mathematical objects.
This first process of specializing to the particular,
concretizing the general, and differentiating the one,
can thus be understood as an abstract differential calcu·
!us, in the most natural possible sense: to study a sign,
one first introduces its differential variations in adequate
contexts of interpretation. But, from the point of view
of the pragmaticist maxim, and from the point of view
of category theory, this is only the first step in a pendular
and dialectical process. For once the variations of the
sign/concept/object are known, the pragmaticist maxim
urges us to reintegrate those various pieces of information
into a whole that constitutes knowledge of the sign itself.
Category theory, likewise, tends to show that behind the
concrete knowledge of certain mathematical objects, and
between these objects, exist strong functorial correlations
(adj unctions, in particular) , which are what actually and

292
 F R AG M E N T S O F A T RA N S I TO RY O N T O L O G Y

profoundly inform us about the concepts at stake. In either

of these two approaches, we are urged to complete our
forms of knowing, following the lines of an abstract inte­
gral calculus, the pendular counterpart of differentiation,
which allows us to find certain approximations between
various concrete particulars - particulars that may appear
disparate, but which respond to natural proximities on
an initially imperceptible, archeal ground.
The differential/integral back-and-forth, present in
bo the pragmaticist maxim and the theory of categories,
th
is situated not only on the epistemological level mentioned
above, but continuously extends to the 'what' and the 'where'
of the ( quasi-)objects at stake. The vertical interlacings to
the right of figure 4 [n8] - denoted by 'correlations,
gluings, transferences', and situated under the general
sign of the 'pragmatic integral' (f ) - codify some of the
most original contributions coming both from a broad
modal pragmatism and from category theory, something
we have already corroborated with sheaf theory. Both
(mathematical) ( quasi-)objects and (Peircean) signs live
as vibrant and evolving webs in those environments of
differentiation and integration. Sheaves, categories and
pragmaticism therefore seem to answer to a complex
regime of the prefix TRANS, on both the ontological and
the epistemological level. We will now make our way into
that second dimension.

293
C H AP T E R 9

C O M PARATIVE E P I S T E M O L O GY A N D

S H EA F I F I CATI O N

We have described the ( quasi-)objects o f contemporary

mathematics as webs of structured information and as (in
Leibniz's sense) 'compossible' deformations within those
webs, open to relative and asymptotic composition in vari­
able contexts. The dynamism of those webs and deforma­
tions straddles the eidal and the quiddital, in an iterated
weaving of conceptual and material approaches. These are
bimodal objects in Petitot's sense, situated in both physical
and morphological-structural space, acting and reacting
along spectra of formal and structural correlations with
the many mobile environments where their partial transit
takes place. In this way, the ontological 'what' and 'where'
are blurred, and their frontiers become murky. We thus
confront an ontologi,calfluctuation that may provoke a
predictable horror vacui in certain analytic approaches
to the philosophy of mathematics, which seek to delimit
and pinpoint their perspectives in the clearest possible
way, fleeing from smears and ambiguities, and situating
those tidy delimitations over fragments of the absolute.
Nevertheless, for a contrapuntal, synthetic approach,
open to relative transit, as heralded by contemporary
mathematics, it becomes imperative to consider the mobile

295
 SYNTH E T I C P H I LO S O P H Y

frontiers between the conceptual and the material..149 A

transitory ontology, as described in the previous chapter,
thus gives rise to a fluctuating ontology, one that is not
easily pigeonholed into the watertight compartments of
Shapiro's square [10] : a natural variation of the 'what'
and the 'where' gives rise to an associated variation of
the ' how'. Once the correlative spectrum of disparate
epistemological perspectives is opened up - in what we
could call a comparative epistemology - we will, in this
chapter, go on to reintegrate several of those perspectives
in a sort of epistemological sheaf, sensitive to the inevitable
complementary dialectic of variety and unity that con­
temporary mathematics demands.
In part 2, we repeatedly saw how, behind the differenti­
ated, many of mathematics' most pivotal constructions
bring pendular processes of differentiation and reintegra­
tion into play, beneath which emerge invariant, archeal
structures. Recall, for example, Grothendieck's motifs
beneath the variations of cohomologies [144-148] ,
Freyd's classifier topoi beneath the variations of rela­
tive categories [245-246] , Simpson's arithmetical
349 For Petitot, 'there i s a rational solidarity between the conceptual, the mathematical,
and the experiential, which challenges the positivist conception of the sciences and
leads to a rehabilitation of critique on new bases' (protogeometries, morphological­
structural order, local/global dialectics, phenomenological invariants of the world
and not only of language, etc.). See Petitot 's contributions to the Enciclopedia Einaudi,
and in particular the entry, 'Locale/globale', Enciclopedia Einaudi (Torino: Einaudi,
1979), vol. 8, 429-90, and 'Uniti delle matematiche', Enciclopedia Einaudi (Torino:
Einaudi, 1982, vol. 15), 1,034-85 (p. 1,084 cited above). In the latter, Petitot places
Lamman at the center of his argument ('La filosofia matcmatica di Albert Lautman',
ibid, 1034-41) : this was the first deep presentation of Lautman's work outside France.

296
C O M P A RA T I V E E P I S T E M O LO GY A N D S H E A F I F I C AT I O N

nuclei beneath the theorematic variations of 'ordinary'

mathematics [ 248-250] , Zilber's pro to-geometric
nuclei beneath the variations of strongly minimal theo­
ries [255-257] , the Lax-Phillips semigroup beneath
the non-euclidean variations of the wave equations
[218-220] , the Langlands group beneath the varia­
tions of the theory of representations [184-185] , the
Grothendieck-Teichmuller group beneath arithmetical,
combinatorial and cosmological variations [225, 233] ,
Gromov's h-principle beneath the variations of partial
differentials, etc. In all these examples, through webs
and deformations, the knowledge of mathematical processes
advances by means efseries ef iterations in correlative triadic
realms: differentiation-integration-invariance, eidos-quidditas­
arkhe, abduction-induction-deduction, possibility-actuality­
necessity, locality-globality-mediation.
Advanced mathematics invokes this incessant 'tri­
angulation', and perhaps finds its most ( technically
and conceptually) striking reflection in the notion of
sheaf [285-288] . A fact of tremendous importance
in contemporary mathematics is the necessity efsituating
oneselfin ajulljledged thirdness without reducing it (without
'degenerating it', Peirce would say) to secondnesses or
firstnesses. If thirdness disappears from the analytic modes
of understanding, this is because existing (classical) forms
of analysis are basically dual, as either-or exclusions, in the
style of Benacerraf [11] . This means that the analytic

297
 S Y N T H E T I C P H I L O S O P HY

the two approaches: without the web of analytic inequali­

ties, the synthetic invariants do not emerge [260-262];
and without the invariants the web of inequalities drifts
about pointlessly (whether this be from a technical or a
conceptual point of view) . The knowledge of Rieman­
nian geometry thus incorporates both analytic elements
and synthetic configurations, situating itself in both
'idealist' (set of all metrics) and 'realist' (local physical
deformations) perspectives simultaneously. In the case
in question, the sheaf of the different local metrics gives
rise to continuous sections tied to the invariants at stake.
Going beyond this one case, however, the same Gromov
points out [ 262] how the contrast between the expo­
nentials, clouds, and wells in Hilbert's tree governs the
multidimensionality of mathematical knowledge, which is
never reducible to just one of its dimensions.
As we have seen in the generic analysis of the notion
of sheaf [283-288] , a sheafification in comparative
epistemology requires us firstly to prescind the notions of
neighborhood, covering, and coherence, before moving
on to possible gluings. The notion of a neighborhood between
epistemological perspectives requires us to postulate, first
of all, the mutual commensurability of the latter; here, we
explicitly position ourselves against the supposed incom­
mensurabilities of 'paradigms' (Kuhn), a phenomenon

300
C O M PARAT I V E E P I S T E M O L O GY A N D S H E A F I F I CAT I O N

that rarely occurs, at least i n mathematics.351 I n this context,

an epistemologjcal neighborhood should bring together dif­
ferent perspectives, either in their methods (for example,
the 'synthetic' method applied to a realist or idealist
stance: pigeonholes 23 and 21 in figure 5 [125] regard­
ing the 'pure' posing of problematics in the philosophy
of mathematics , where i1 denotes the ;"th pigeonhole in
column i), in their objectives (in their approach to the
real, for example: pigeonholes 1:3 and 23 in figure 5 ) , or in
their mediations through the warps at stake (for example,
asymptotic mediation: pigeonholes 16 and 25 in figure 5 ) .
This third option is the most rich and innovative from a
conceptual point of view, since it can help us eliminate
dualistic exclusions by working with epistemic webs ef
approximation.
Once a certain neighborhood has been established
(pursuing the cases above, a neighborhood of 'synthetic
methodology', 'realist finality' or 'asymptotic mediation',
for example) , we pose the question of how that neighbor­
hood may be covered. The neighborhoods could be covered
in a binary fashion (by means of the pairwise-indicated
pigeonholes, in the cases above, for example), but the
most interesting cases might correspond to nonbinary
351 From an epistemological point of view, not only do the various paradigms ('isms') in
mathematics (logicism, intuitionism, formalism, structuralism, etc.) not cancel each
other out, in fact, they benefit from mixedfusions (through their reinterpretation i n
terms o f category theory, for example). From a logical point of view, the amalgams i n
model theory are legion, a n d many interesting works in the area correspond t o the
crossing of apparently incommensurable entities.

301
 SYNTH E T I C P H I LO S O P H Y

'geography' can, i n turn, b e founded i n a sort o f squared

sheafthat traces out all of the richness, both extrinsic and
intrinsic, of the concept.
In accordance with the multidimensionality of math­
ematical vision, with the depth of 'Hilbert's tree' , with
relativity of the webs of categorical perspectives, and with
those webs' sheafification, we can sense the existence of
the complex protogeometry that underlies a comparative
mathematical epistemology. Let us make explicit here,
behind these considerations of ours, the hypothesis of a
continuity35° between the world of phenomena, the world
of mathematical ( quasi-)objects associated with those
phenomena, and the world of the knowledge of those
objects - which is to say, the hypothesis efa continuity between
thephenomenal, the antic and the epistemic. The mathematical
constructions (and discoveries) that we have extensively
reviewed in this work's second part show that contem­
porary mathematics offers new support for the possible
soundness of this continuity hypothesis. Advanced math­
ematics, here, offers a finer illustration than elementary

352 This continuity is one expression of Peircean synechism, which postulates an even stronger
continuity hypothesis, by supposing the existence of a complete�v operative continuity
in nature (in which the human species appears, according to Peirce, both materially
and semiotically). Another expression of this synechism is constituted by the three
cenopythagorean, universal categories, which1 according to the Peircean hypothesis,
continually traverse boch the world of phenomena and the forms by which those
phenomena are known. For a description of synechism, of the generic (nonclassical�
nonCantorian, nonextensional) concept of the continuum according to Peirce, and of
certain partial mathematical models for this nonstandard continuum, see F. Zalamea,
F.I continua peirceano (Bogota: Universidad Nacional, 2001 ) . An English translation is
available ar http://unal.academia.edu/FcrnandoZalameaj\Vebsites.

304
C O M PA RATI V E E P I ST E M O L O GY A N D S H E A F I F I C AT I O N

mathematics, since the latter's low thresholds of com­

plexity impede the emergence of the continuous/discrete
dialectic that ceaselessly traverses the contemporary space
of mathematics. From an epistemological point of view,
the distinct perspectives are nothing other than breaks in
continuity. In those breaks (as in Peircean abduction) ,
new forms of knowledge are generated,353 and - in an
epistemology open to transit - those forms of knowledge,
when they are coherent, can be subsequently reintegrated
in an adequate fashion.
The protogeometry underlying a comparative epistemol­
ogy of mathematics exhibits several peculiar characteris­
tics, tied to a combinatorics efcoherent couplings between the
webs in play (deep, multidimensional, iterative webs). On
the one hand, in fact, the inverse bipolar tensions between
prescission and deduction [ 286] show that, in many
mathematical cases past a high threshold of complexity
(among which we may place the case of sheaves, in any of
their geometrical, algebraic or logical expressions) , there
emerges a ('horizonal') hierarchy of partial couplings,
whose striated resolution yields important forms of math­
ematical knowledge. This is the case, for example, with
Gromov's h-principle [263] , where an 'inverse bipolar
tension' between local and holonomic sections yields an
353 There may exist, here, a profound analogy between the processes of symmetry brealdng
in the physics of the first instants of the universe, and the processes of continuity
breaking in the continuous archeal groups by which those forms of symmetry can be
represented.

305
 O F C O NT E M P O RA RY M AT H E M AT I C S

entire series of homotopic mediations, with calculative

fragments of enormous practical interest in the resolution
of partial differential equations. On the other hand, the
self-referential exponential nodes in Hilbert's tree [2fo1] yield
anotherhierarchy (now 'vertical') of partial couplings whose
stratified resolution likewise results in remarkable advances.
This, for example, is what happens with the fundamental
notion of space, whose breadth expands as it transits
between sets, topological spaces, Grothendieck topoi
[140] , and elementary topoi [195] , on each stratum
generating distinct (though mutually coherent) conglomerates
of new mathematical results. But there also exists at least a
third hierarchy (now 'diagonal') of partial couplings between
mathematical webs, directly influenced by the transgressive
spirit of Grothendieck. Beyond the horizontal or vertical
displacement, and beyond their simple combination, there
do indeed exist diagonal mediations with archeal traits
('nondegenerate' Peircean thirdnesses), between remark­
ably far-flung realms of mathematical knowledge (dessins
d'enfants between combinatorics and complex analysis
[ 225], the Grothendieck-Teichmiiller group between
arithmetic and cosmology [225], noncommutative groups
between logic and physics [258] , etc.).
The protogeometry of those coupli ngs thus
incorporates an entire complex interconnection of
multidimensional elements, in concordance with the 'intu­
itive' multidimensionality of mathematical knowledge.

306
C O M PARAT I V E E P I S T E M O L O GY A N D S H E A F I F I CA T I O N

The image o f that mathematical knowledge i s far removed

from its logical foundation,354 and a new preeminence of
geometry appears on the map ( item 7 in the distinctive
tendencies of contemporary mathematics [41] ) . The
beginning of the twenty-first century may indeed be a
good time to begin to seriously consider ageometricization
ef epistemology such as we are proposing,355 that would
help us to overcome ( or, at least, to complement) the
logi,cization ef epistemology undertaken throughout the
twentieth century. The influence of analytic philosophy,
whose logical support boils down to first order classical
logi,c, should be countered by a synthetic philosophy, one
much closer to the emerging logic ef sheaves. What is at
issue here is an important paradigm shift in mathematics
and logic ( as in Caicedo's fundamental results [283]),
which philosophy ( ontology, epistemology, phenomenol­
ogy) too should begin to reflect. Indeed, the changes in

354- This is something that impresses itself upon even such a studious partisan offoundations
as Feferman: 'lhc logical picture of mathematics bears little relation to the logical
structure of mathematics as it works out in practice'. S. Feferman, 'For Philosophy of
:Y1athematics: 5 Questions', 13. Course material, http://math.stanford.edu/-fefrrman/
papers/philmathfive.pdf. Fcfcrman nevertheless repeats the usual prejudices against
an 'ingenuous' or 'trivial' Platonism: 'According to the Platonist philosophy. the
objects of mathematics such as numbers, sets, functions and spaces are supposed to
exist independently of human thoughts and constructions, and statements concerning
these abstract entities are supposed to have a truth value independent of our ability
to determine them' (ibid., u). Compare this (caricatural) description with the more
complex Platonism of a Lautman [52-60] or a Badiou [277-280].
355 Petitot 's program for the naturalization ofphenomenology covers similar bases, and offers
a great deal of room to geometry. See]. Petitot et al., Naturalizing Phenomenology: Issues
in Contemporary Phmomenology and Cognitive Science (Palo Alto: Stanford University
)
Press, 2000 . Although Petitot makes use of techniques in neuroscience, which we du
not mention here, his invocation of Riemannian geometry and sheaf logic anticipates
our own perspectives.

307
 SYNTH E T I C P H I L O S O P H Y

the logical base register the deformability of mathematical

( quasi-)objects through relative transits, and allow us to
overcome the classical 'rigidity' an object exhibits in a
supposedly absolute universe.
In a detailed approach to Grassmann's work, Chatelet
lucidly describes one of the fundamental problematics of
mathematical knowledge:

Like the contrast continuous/discrete, the equal

and the diverse are the result of a polarization; it
is thus that algebraic forms 'becoming through the
equal' and combinatory forms 'becoming through the

diverse' can be distinguished. It is a matter offinding

the articulation that makes it possible to pass continuously

from the equal to the diverse.356

C hatelet's mathematico-philosophico-metaphorical
instruments in this search for a continuous articulation
include 'dialectical balances', 'diagrammatic cuts', 'screw­
drivers', 'torsions', and 'articulating incisions of the suc­
cessive and the lateral', which is to say, an entire series
of gestures attentive to movement and which 'inaugurate
dynasties ofproblems'3s7 and correspond to a certainfluid

356 G. Chatelet, Les enjeux du mobile, 167 (our emphasis).

357 Ibid., 37, 33, 38, 218, 32.

308
C O M PARAT I V E E P I S T E M O L O GY A N D S H E A F I F I C AT I O N

electrodynamics358 of knowing. Here, we are once again

facing the multidimensionality of mathematical (quasi-)
objects, a multiplicity that can only be apprehended by
way of 'gestures', that is, by way of articulations in motion,
which allow for partial overlaps between the 'what' and
the 'how'. Mathematical epistemology - in so far as it
wishes to be able to incorporate the objects of contem­
porary mathematics (and in many cases, those of modern
mathematics) into its spectrum - should therefore be
essentially mobile, liable to torsion, capable of reinte­
grating cuts and discontinuities, and sensitive to fleeting
articulations - in sum, genuinely in tune with the objects it
aspires to catch sight of. No fixed position, determined a
priori, will be sufficient for understanding the transjorm­
ability of the mathematical world, with its elastic transits,
its unstoppable weavings between diverse forms, and its
zigzagging pathways between modal realms.
In the case studies in part 2, we can concretely detect
various properties of the epistemic protogeometry that we
have just been discussing. Both in the ( quasi-)objects
at stake, and in the forms by which they are known,
we observe common protogeometrical features, among
which we must emphasize: (a) multidimensional cuts
358 Chitelet interlaces Grassmann's 'fluidity' (ibid., 166) with Maxwell's electromagnetism
in order to study an electrophilosophy in proximity with Faraday's electrogeometrical
space (ibid., chapter 5 ) . A geometrical perspective on Maxwell and Faraday's
'
allusive operators' (ibid., 219; emphasis ours) breaks with punctual or instantaneous
interpretations, and explores the asymptotic deformations of entities within given
neighborhoods ('the pedagogy of lines of force', ibid., 238-48).

309
 O F C O NT E M P O RARY M AT H E MAT I C S

and reintegrations, ( b) triadic iterations, (c) deformations

of objects and perspectives, (d) processes of passage to
the limit by means of nonclassical approximations, (e)
asymptotic interlacings and couplings, and (f) fragments
of sheafification. In contemporary mathematics, there is a
continuous overlapping of the 'what' and the 'how', whose
very protogeometry tends to emerge from a common
and unitary base. Grothendieck's schemes [139-140]
combine many of the above properties: 'ontically', the
schemes are constructed by (a) introducing structural
representations of rings, (c) viewing points as prime ideals
and topologizing the spectrum, (f) locally defining fibers
as structural spaces of rings and gluing them together
in an adequate fashion; 'epistemically', schemes involve
(a) a process of generalization and specification between
multidimensional objects, (b) an iterated threeness359
between a base, its widening, and their mediation (pro­
jection) , (c) an comprehension of objects through their
relative positions, etc. Similar registers can also be speci­
fied for Grothendieck's other great constructions: topoi
[141-144] and motifs [144-148] . In the same way,
the Langlands correspondence [181-186] includes,
'ontically', ( a) representations of groups, (b) iterated
359 The.first level of iteration corresponds to the (algebraic) step from the initial, unitary,
commutative ring to the set of its prime ideals; the second level, to the (topological)
step from the combinatorial spectrum to the spectrum with the Zariski topology; the
third level, to the (categorical) step from neighborhoods to fibers; thefourth level, to the
ultimate shcafification of the whole. It is interesting to observe how those iterations
are not absolute, and can be realized in a different, or better still, mixed order.

310
C O M PARATIVE E P I ST E M O LO G Y A N D S H EA F I F I C AT I O N

transits between the modular, the automorphic and the

L-representable, (c) mixed differential and arithmetical
structures, and (d) manipulations of noncommutative
groups; 'epistemically', the strategy of the program rests
on (a) an understanding of arithmetical objects as 'cuts'
(projections) of more complex geometrical objects, (b)
a systematic search for geometrical mediations between
(discrete) arithmetic and (continuous) complex analy­
sis, and (c) - (d) an observation of the objects' analytical
deformations, etc. Another similar situation occurs with
the general theory of structure/nonstructure according
to Shelah [ 196-197] , where, 'ontically', there arise
(a) amalgams in high finite dimensions, (d) cardinal
accumulations by means of nonexponential objects (PcF
theory) , and (e) monstrous models and asymptotic warps
of submodels; 'epistemically', Shelah's vision integrates
(a) a celebration of mathematics' multidimensionality,
(b) an incessant moderation between the structured and
the nonstructured, (d)-(e) a deep understanding of the
objects at high levels of the set-theoretic hierarchy as
limits of 'moderate' and 'wild' fragments, etc. In this way,
certain common protogeometric characteristics naturally
lace together the ontic webs and the epistemic processes in
play.360 Throughout part 2 of this work, we have implicitly
360 In the last instance, the transformations leading from a fixed ontology to a transitory
ontology, and from a fixed epistemology to a sheqjified, comparative epistemology, cause the
'entities' under investigation in each of these approaches ( 'what', 'where', 'how' ) to draw
closer to one another, and cause their mobile frontiers to become much less excJusive.

311
 SYNT H ET I C P H I LO S O P H Y

carried out other analyses, and additional examples could

be explicated here (the works of Connes, Kontsevich,
Zilber, and Gromov, especially, lend themselves to this) ,
but the aforementioned cases may provide sufficient
illustration.
The principal limitation that seems to encumber any
'analytic' epistemology, as opposed to the 'synthetic', com­
parative epistemology that we are hinting at here, pivots
on the analytic difficulty in confronting certain inherently
vague environments, certain penumbra! zanes, certain 'out­
posts of the obscure', as Chatelet calls them,361 certain elas­
tic places of 'spatial negativity', certain 'hinge-horizons'
where complex mixtures emerge that resist every sort of
strict decomposition. We will study the problematic of the
penumbra with greater care in the next chapter, where we
take up the (sinuous, nonlinear) dynamics of mathemati­
cal creativity, but in what remains of this chapter we will
approach the dialectic of the obscure and the luminous in
conjunction with the three ubiquitous polarities, analysis/

'lhc ontic 'webs' and epistemic 'processes' are therefore only relative specifications
( in ontological and epistemic contexts) of a single and common kind of 'proto·actions'
(something which coincides with certain tendencies of Pcirccan universal semiotics). In
such a 'blurring' of frontiers between the ontic and the epistemic, it is worth pointing
out how Badiou, on the one hand, sees mathematics as being, basically, ontology,
while Petitot, on the ocher hand, considers it to be, basically, epistemology. On an
analytic reading, such blurrings are taken to be improper, but, as we have observed,
from a synthetic reading - once transits, osmoses, and contaminations are accepted - new
anal_-rses of the processes of transference can be carried out, without having to spill
over into an extreme relativism or into ingenuous forms of skepticism. The forms
of the decomposition (analysis ) of transit (synthesis) can no longer be forgotten in
mathematical philosophy.
361 Chitelet, Les enjeux du mobile, 2 2, 37.

312
C O M PARATIVE E P I ST E M O L O GY A N D S H E A F I F I C AT I O N

synthesis, idealism/realism and intensionality/extensionality,

which appear in any epistemological approach. Our objec­
tive consists in mediating (or 'moderating': Grothendieck,
Shelah) between them, and proposing reasonable cou­
plings from the 'outposts' of contemporary mathematics.

CLASSICAL LOGIC

C O · I NTUITIONISTIC LOGIC INTUITIONISTIC LOGIC

LINEAR LOGIC

CLASSICAL LOGIC f- A V -iA

preserves LEM and LNG h (/I A -iA )
INTUITIONISTIC LOGIC fL A V -iA
preserves LNG, rejects LEM h (A A -iA )
CO· INTUITIONISTIC LOGIC f-A V -iA
preserves LEM, rejects LNG fL -i ( A A -iA )
LINEAR LOGIC f- A 'lhA
all ifthe above, h (A ® -iA )
but in differentforms fL A EB -iA
fL -i (A & -i A )

Figure i4. Logic: classical, intuitionistic, co-intuitionistic, and linear.

A very interesting case of this search for mediations in

penumbra! zanes corresponds to the dualization of the
central Clas - Int logical axis (the continuum of superin­
tuitionistic logics between intuitionistic logic and classical
logic) into its counterpart Cofnt - Linear (logics between

313
 OF C O NT E M P O RA RY M AT H E M AT I C S

linear logic and paraconsistent, co-intuitionistic logic).

The Class Int axis can be seen as a line along which the
-

law of the excluded middle progressively gives way, by

continuous degrees. Now, the law of the excluded middle
has a dual, which is the principle efnoncontradiction. In the
eyes of classical logic, these two laws are equivalent, but
the Class Int axis progressively breaks this symmetry,
-

dissolving the excluded middle while maintaining non­

contradiction. What this suggests, then, is the possibility
of a dual axis, that would evaporate noncontradiction
while maintaining the excluded middle, terminating a
logical system that is the dual of intuitionistic logic: a
'paraconsistent' system called co-intuitionistic log;i.c.
What these two logics sacrifice in symmetry, they
gain in unearthing a coherent and profound computa­
tional dynamism, allowing for a mapping of proofs onto
functional prog;rams, and the dynamic of cut-elimination
(the procedure by which their proofs can be reduced to a
unique normal form, a 'direct proof' where no lemmas are
used) onto the process of program execution. Now, as the
French proof-theorist Jean-Yves Girard has observed, the
transformation of classical logic into intuitionistic logic
can be seen as the result of an asymmetrical restriction of
the 'structural rules' that govern the former - rules which
allow formulas to be reused and forgotten as needed in the
course of a proof, and so encode their measure of ideal­
ity. A similar observation can be made for the shift from

314
C O M PARAT I V E E P I S T E M O L OGY A N D S H E A F I F I CA T I O N

classical to co-intuitionistic logic. As Girard observes, it is

therefore possible to recover both the coherent dynamics
of intuitionistic (and co-intuitionistic) logic, and the pro­
togeometrical symmetry of classical logic, by symmetrically
suspending the structural rules - which gives us linear logic,
a logic where formulas work more like transient actions
(and their negations as reactions) than like atemporal and
ideal propositions. The forms of conjunction and disjunc­
tion, here, are each split in two - into 'multiplicative'
and 'additive' forms - which allows this synthetic logic to
partially preserve and partially reject both the law of the
excluded middle and its dual, the law of noncontradic­
tion. The power of the structural rules, alone capable of
reuniting these forms, is then reintroduced in the form of
local modal operators, which, in fact, constitute some of the
most stubborn obstructions to geometrical and dynamical
analyses of linear proofs (Girard calls them the 'opaque
modal kernels' of 'essentialism').360
The diamond of the four logics363 complete the pri­
mordial Class - Int axis in a natural manner, and Luke
Fraser has posed the fundamental question as to what
concepts would correspond to the well-known landscape

362 J-Y. Girard, 1he Blind Spot: Lectures on Logic (Zurich: European Mathematical Society,
2011), 11.
363 The 'diamond' is outlined and briefly discussed in Samuel Troniron's outstanding
genealogy of the Girardian research program in S. Tron�on. Dynamique des dimonstrations
et thiori.e de /'interaction, PhD thesis, University of Aix·Marseil1e I, Marseille, April
2006, 278.

315
 SYNT H E T I C P H I L O S O P H Y

Class (analysis, discreteness, extensionality, sets) - Int (syn­

thesis, continuity, intensionality, categories) if we attempt
to dualize them toward the Colnt - Linear side. Indeed, the
penumbra! ::.one is all the more accentuated given that, in
mathematics' current state of development, we do not
seem to be able to even glimpse natural completions of the
dyads of analysis/synthesis, discrete/continuous, exten­
sionality/intensionality, and sets/categories. If we think
of Thom's 'founding aporia' as describing the conceptual
galaxy that turns around just one of these axes (Class Int),
-

this leaves, in the penumbra, three axes around which

may revolve entire galaxies, conceptual spaces of which
we have not yet learned to dream.
Lawvere has introduced limit co-operators and co­
Heyting algebras in topoi, the Brazilian school has worked
on categorical models for paraconsistent (co·intuitionistic)
logics and, recently, the Dutch school has given impetus
to the study of co-algebraic techniques in computational
modelings. The analyses carried out, nevertheless, have
turned out to be very localized and correspond to the
resolution of minor technical questions. A.fundamental
philosophical study of the galaxies associated with the
penumbra! zone Colnt Linear, by contrast, could, in the
-

coming years, produce a genuine conceptual revolution - yet

another example of how contemporary mathematics can
serve to tune the strings of philosophy and make them
vibrate. The triadic Peircean phenomenology could be of

316
C O M PARAT I V E E P I S T E M O L O GY A N D S H E A F I F I C AT I O N

great assistance here, a t least i n relaxing the dyads s o as

to yield a third opening (with Giovanni Maddalena we
have, for example, proposed an extension of analysis/syn­
thesis into analysis/synthesis/horosis,364 through the concept
of horos, border, frontier), but the labor of conceptual
completion through the 'Fraser Side', Co!nt - Linear still
entirely remains be to done.365
The analysis/synthesis polarity has always been a
source of equivocations and advances, in philosophy366 no
less than in mathematics.367 In a dialectical fashion (of the
thesis-antithesis type, e 8) the polarity brings into oppo­
sition knowledge by decomposition and knowledge by
composition, but in a relativefashion, since 'there exist no
absolute criteria, either for processes of analysis or for pro­
cesses of synthesis' .368 Indeed, nature always presents itself
by way of mixtures that the understanding decomposes
364 G. Maddelena & F. Zalamea, 'A New Analytic/Synthetic/Horotic Paradigm. From
Mathematical Gesture to SyntheticjHorotic Reasoning,' (preprint) submitted to Kurt
Godel Research Prize Fellowship 2010).
365 We are grateful to our translator, Zachary Luke Fraser, for having presented this
problematic to us, and for helping us ro develop our discussion of the 'logical diamond'.
366 See G. Holton, 'Analisi/sintesi', Enciclopedia Einaudi (Torino: Einaudi, 1977), vol. 1, 491-
522. On many occasions we have tried to (technically, conceptually and analogically)
use the Enciclopedia Einaudi, whose eminently transductive project quite naturally
brings together modern and contemporary mathematics. The last two volumes (vol.
15, Sistematica and vol. 16, lndici) are of great service, as well as the fascinating maps,
tables, and reading diagrams developed by Renato Betti and his collaborators, where
the many osmoses and shiftings of contemporary thought are made manifest. (Recall
Merleau-Ponty's philosophy of the 'shifting' [277].) Petitot's remarkable contributions
to the Enciclopedia introduced us to the local/global, centered/decentered, and One/
Many problematics that, as we have seen, reemerge in contemporary mathematics.
367 See the anthology edited by M. Otte and M. Panza,Anarysis andSynthesis in Mathematics,
History and Philosophy (Dordrecht: Kluwer, 1997).

368 Holton, 'Analisi/sintesi', 502.

317
 O F C O N T E M P O RA R Y M AT H E M AT I C S

and recomposes i n iterated oscillations. Analytic differen­

tiation (typical in the mathematical theory of sets: objects
known by their 'elements') can lead to a better resolu­
tion of certain problematics, though subdivision in itself
(Descartes) does not assure us of any result (Leibniz) .369
Synthetic integration (typical of the mathematical theory
of categories: objects known by their relations with the
environment), for its part, helps to recompose a unity
that is closer to the mixtures of nature, but, as with its
counterpart, inevitable obstructions arise in its use. One
should therefore try to bring about a 'continuous effort ef
balancing analysis and synthesis' (Bohr),37° something that
meets with peculiar success in Peircean pragmaticism, and
particularly with its pragmaticist maxim [111-uu] .
An epistemological orientation can then emerge, leading
to an abstract differential/integral (analytic/synthetic)
'calculus' , situated over a relativefabric efcontradistinctions,
obstructions, residues and gluings [118]. The case studies
that we have carried out in contemporary mathematics
show that such an orientation is never absolute, but
only asymptotic, that it depends on multiple relativiza­
tions - eidal ascents and quiddital descents, 'first, reason
climbs up with analysis, and then down with synthesis'37'
369 Ibid., 503.
370 Ibid., 505.
371 Ibid. The oscil1ations of thought have been s tud ie d with geometrical instruments at
least since Ramon Llull, Libra de/ ascenso y decenso de! entendimiento (1304) (Madrid:
Orbis, 1985).

318
C O M PARAT I V E E P I S T E M O L O GY A N D S H E A F I F I C AT I O N

- but also requires us to have enough points of reference

(archeal invariants) to calibrate the relative movements.
Lying still further beyond the pendular analysis/
synthesis equilibrium invoked by Bohr, contemporary
mathematics hints at how that equilibrium can be pro­
duced. We have seen how the notion of sheaf, in a very
subtle manner, combines the analytic and the synthetic,
the local and the global, the discrete and the continuous,
the differential and the integral [285-288] . In this way,
the sheafification of the analysis/synthesis polarity generates
a new web of epistemological perspectives, following the
directions of contemporary mathematics. If, in fact, we
look at the diagrammatization of the general concept of
sheaf that appears in figure i3 [286] , we can see how,
behind the double opposition, analysis/synthesis (the
vertical arrangement in the diagram) and local/global
(horizontal arrangement), lies a very interesting diagonal
mediation that is rarely made manifest. Taking the pivotal
analytic concept of ' covering' and modulating it by means
of a synthetic 'section-preservation-projection-restriction'
hierarchy (figure i3) , the natural notion of a transversal
traniform if the covering emerges, whereby the open and
closed372 strata of the various representations (or ' covers')

3i2 The notion of 'covering' comes from the Latin cooperire ( operire [close]; cooperire [close
or cover completely], eleventh century). In counterpoint with aperire (open), the notion
of covering thus includes, by way of its etymological roots, a conception of the transit
bettveen open and closed environments (a translt reflected in other derived frontiers:
operculum [overlay]; aperitivus [aperture]) .

319
 S Y N T H E T I C P H I LO S O P H Y

involved are superposed on one another. The transform

- tied to a sort ofprotogeometry of position ('pretopos')
- combines the analytic capacity to cover through decom-
positions (whether with elements, neighborhoods , or
asymptotic approximations) , and the synthetic capacity
to recombine the fragmented through variable contexts
(whether with partial couplings or more stable structural
gluings). We will call this type of sweeping mediator
between the analytic and the synthetic a Grothendieck
traniform,373 a sweeping reticulation particularly attentive
to the relative mixes between the concepts in play.
By its very definition, the Grothendieck transform
incorporates peculiar modes of knowing. Through the
transverse, it introduces a reticular warp over what contra­
distinctions, couplings and asymptotes stabilize. Through
the covering, it introduces a fluid dynamics (Merleau-Pon­
ty's 'shifting [glissement]'), incarnated in the gerund itself:
the ' to be covering' over a situation ('geography') and
a duration ('history'). The Grothendieck transform thus
changes our epistemological perspectives by the simple
fact of smoothing them as a whole (Gromov [259, 263]):
integrating them into an evolving tissue, it desingularizes
373 Looking back at chapter 4, it seems clear that this transformational process of
mathematical concepts has always been pre.sent in the global, conceptual inventiveness
of Grothendieck. From a local and technical point of view, moreover, Grothendieck
topologies [138) help to incarnate, in a restricted manner, what we refer to here
as a 'Grochendieck transform'. Indeed, for any arbitrary site (any category with a
Grothendieck topology), the category ofpresheaves over that site gives rise to a category
of sheaves, through a general process of sheafification that corresponds, precisely, to
bringing about the mediations contemplated in the 'Grothendieck transform'.

320
S O M PARAT ! V E E P I S T E M O L O G Y A N D S H E A F I F I C AT I O N

punctual perspectives, i n favor o f a comparability that

emphasizes the regularities, mediations and mixtures
between them. This is evinced in the mathematicalpractice
that we have reviewed in the second part of this essay.
The transversality and 'wonderful melange' found in
Serre's methods [177], the transversal contamination in
Langlands's program [184] , the omnipresent dialectic
of adjunctions and the evolution of objects in Lawvere
[193] , the covering moderation in Shelah's PCF theory
[201-202] , Atiyah's I ndex Theorem with its medley
of eidal transversality (equilibrium between transits and
obstructions) and quiddital coverage (from algebraic
geometry to physics) [208-209], the harmonic analysis
(precise technical form of covering transversality) applied
to the non-euclidean wave equation in Lax [219] , the
noncommutativity that runs transversally over quantum
mechanics in Cannes [223] , the quantizations whose
transversal recoverings allow us to reconstruct classical
structures in Kontsevich [232] , the intermediate cat­
egories between regular categories and topoi, produced
as transversal cuts (Map, Split, Car) over free allegories in
Freyd [246], the logical nuclei covering second order
arithmetic in Simpson [250] , the protogeometrical ker­
nels covering strongly minimal theories in Zilber [255],
and the transversality of the h-principle in Gromov [263]
are all very subtle and concrete examples behind which
lurk the mediations and smoothings of Grothendieck's

321
 O F C O NT E M P O RARY M AT H E M A T I C S

'transversal covering transform'. In all o f these construc­

tions, the epistemological vision of mathematics is one
that is flexible enough to cover a hierarchical stratification ef
ideality and reality, dynamically interchang;i,ng the contexts
of interpretation or adequation of the ( quasi-)objects
whenever the technical environment or creative impulse
demands.
These considerations help us to better understand
how - in contemporary mathematics and, more broadly,
in advanced mathematics [27] the poles of idealism -

and realism should not be considered as separate (discrete

exclusion, through an analytic approach) , but infull inter­
relation (continuous conjunction, through a synthetic
approach) . Indeed, a good understanding of the ideal/
real dialectic in advanced mathematics can be achieved
through the notion of an epistemological back-and­
forth.374 The back-and-forth postulates not only a pendular
weaving between strata of ideality and reality, but, above
all, a coherent covering of various partial approximations.
374 The back-and-forth emerges in Cantor's proof concerning the isomorphism between
two dense, countable linear orders without limit points. The Cantorian technique
(formalized in a modern fashion by Hausdorff in 1914) uses approximations to the
isomorphism by way of a collection of partial homomorphisms that cover the sets
little by little (surjectively), while continuing to preserve the orders, and whose well­
behaved limit furnishes the desired isomorphism. Observe, here, the gerundial, the
asymptotic and the limit, which anticipate several of the themes taken up in this chapter.
The back-and-forth was later used by Fralsse (1954) to characterize the elementary
equivalence between abstract structures (with arbitrary relations, beyond orders),
and by Lindstrom ( 1969) in his surprising characterization theorems for classical,
first-order logic (maximal with respect to the compactness and LOwenheim-Skolem
properties). Ever since, the back-and-forth has become an indispensable technique in
abstract model theory. See J. Barwise, S. Feferman, eds., Model Theoretic Logics [New
York: Springer, 1985].

322
C O M PARAT I VE E P I ST E M O LO GY A N D S H E A F I F I CAT I O N

A direct example of this situation i s found i n Lindstrom­

style back-and-forth technique used in abstract model
theory: the semantic ( stratum of reality, constituted by a
collection of models with certain structural properties)
is understood through a series of syntactic invariants
( strata of ideality, constituted by languages with other
partial reflection properties) , and more specifically, the
elementary ( real) equivalence is reconstructed through
( ideal) combinatory coherences in a collection of partial
homomorphisms that are articulated in the back-and­
forth. Other indirect examples appear in the case studies
that we have carried out: the progressive back-and-forth
in the elucidation of the Teichmiiller space's functorial
properties according to Grothendieck [170] , the struc­
tural amalgams by strata according to Shelah [196], the
approximations of hyperbolic groups and the polynomial
growth of groups according to Gromov [259], etc.
In these processes, mathematical ( quasi-) objects'
modes of creation, modes of existence, and the modes
by which they are known are interlaced and reflected in one
another (general transitoriness between phenomenology, ontol­
ogy and epistemology) . The relative ( partial, hierarchized,
distributed) knowledge of those transits therefore becomes
an indispensable task for mathematical epistemology.
Beyond trying to define the ideal or the real in an absolute
manner ( a definition that, from our perspective, reflects
a poorly posed problem), the crucial task of mathematical

323
 S Y NT H E T I C P H I L O S O P HY

epistemology should instead consist in describing, pin­

pointing, hierarchizing, decomposing and recomposing
the diverse forms of transit between the many strata
of ideality and reality of mathematical ( quasi-)objects.
Through the samejorces that impel the internal develop­
ment of the mathematical world, we have, for example,
explicated some contemporary forms of transit with great
expressive and cognitive power: protogeometrization,
nonclassical approximation, sheafification, Grothendieck
transformation, back-and-forth modulation.
Another important epistemological opening is found
in mediating the usual 'extensional versus intensional'
dichotomy, which grosso modo corresponds to the 'sets ver­
sus categories' dichotomy. One of the credos of Cantorian
set-theoretic mathematics - and, for that matter, of tra­
ditional analytic philosophy - has been the symmetry
of Frege's abstraction principle, introduced locally by
Zermelo with his axiom of separation: given a set A and
a formula cp (x) , there exists a subset B= {a E A : cp (a) }, and
so an equivalence obtains (locally, within the restricted
universe A) between cp (a) (intensionality) and a E B ( exten­
sionality) . But, beyond the credo and its indisputable
technical convenience, there is no philosophical or math­
ematical reason to prevent an asymmetrization of Frege's
principle of abstraction.375 A non-Cantorian continuum, for
375 Various mathematical texts consider that a breaking ifthe symmetry between extension
and intension may turn out to be beneficial. From our perspective, this symmetry

324
C O M PARAT I V E E P I ST E M O L O GY A N D S H EA F I F I C AT I O N

example, seems t o involve profound intensional charac­

teristics that are impossible to achieve with an extensional
modeling.376 In the same manner, many of the intensional
characterizations of mathematical objects and processes
obtained in category theory offer new perspectives ('rela­
tive universals') that do not coincide377 with extensional
set-theoretical descriptions.
Though the extensional/analytic/set-theoretic influ­
ence has, until now, been preponderant in mathematics,
its intensional/synthetic/categorical counterpart every
day becomes more relevant, and a new 'synthesis of the
analysis/synthesis duality' is the order of the day. Of course,
the self-reference just mentioned in quotation marks is not
breaking could correspond to a defonnation of the local symmetries codified in
Zermclo's separation axiom (symmetries that obtain fiber by fiber, but that should
collapse as a consequence of slight deformations of those fibers). See, for example,].
Bfnabou, 'Rapports cntre le fini et le continu', in j . M . Salanskis, H. Sinaceur, eds.,
Le labyrinth de continu (Paris: Springer-Verlag, 1992), 178-89; E. Nelson, 'Mathematical
Mythologies', ibid., 155-67; R. Thom, 'L'anteriorite ontologique du continu sur le
discret', ibid., 137-43.
376 From the point ofview of the axiomatic bases required to capture a generic continuum
such as Peirce's (non-Cantorian) continuum, Zermelo's local separation axiom is an
excessively demanding postulate. By contrast, giving precedence to the intensional
offers, from the outset, an important support for the inextensibility of the Peircean
continuum, which is to say, the impossibility of defining it through an accumulation
of points. In effect, when we asymmetri.ze the axiom of separation, only certain formulas
can give rise to classes, and the a priori 'existence' of points can be eliminated: singleton
sets {x} do not always exist, and only in certain, specific (constructible) cases can
they come to be actualized. At the same time, by permitting the manipulation of
contradictory intensional domains (in the potential) without having to confront the
associated contradictory extensional classes (in the actual) that would trivialize the
system, we achieve a greater flexibility in our generic approach (freed from the tethers
to the actual) to the continuum. See Zalamea, El continua peirceano, 84-6.
377 \i\That we are dealing with here is a crucial, mathematical noncoincidence, although
logically' the terms may be equivalent. The mathematical richness of category theory,
as we have seen, does not reduce to a logical counterpart in the fashion of 'topos theory'
� 'restricted set theory', but it is directed, rather, toward the discovery of symmetries
and synthetic equilibria, unobservable in light of analytic decompositions.

325
 O F C O N T E M P O RA RY M AT H E M AT I C S

only noncontradictory, but exerts a multiplicative force in

a hierarchy of knowledge. Throughout chapters 8 and 9,
we have tried to forge some mediating perspectives within
the 'synthesis efthe analysis/synthesis duality' . By expanding
our spectrum to more general cultural horizons, we will,
in the two final chapters, see how the incessant mediations
of contemporary mathematics are grafted, on the one
hand, to the ( local ) 'creative spirit' of the mathematician,
and, on the other, to the complex and oscillating (global)
'spirit of the age' in which we find ourselves immersed.

326
C H APT E R 10

P H E N O M E N O L O GY O F

MAT H E MAT I CA L C REATIVITY

Traditional philosophy of mathematics tends to neglect

the ways in which mathematical thought emerges. Sev­
eral texts dealing explicitly with mathematical invention
have come to us from such practitioners of the disci­
pline as Poincare, Hadamard, Grothendieck and Rota,
but, curiously, the professional philosopher neglects
the phenomenology efmathematical creativity as something
foreign to his reflection. Nevertheless, in science, and,
more generally, in every area of knowledge, the way in
which knowledge emerges is (at least) as important as the
knowledge itself. As Valery reminds us, 'the interest of
science lies in the art of doing science':378 the art of inven­
tion and the practices associated with creativity constitute
the true interest of science. This is all the more obvious
in the realm of mathematics, the specificity of which is
rooted in the incessant transit (ars) of concepts, proofs
and examples between the possible (abduction), the
necessary (deduction) and the actual (induction) . Valery,
a true connoisseur of mathematics and an extraordinary
investigator of creative modulations in the twenty-seven
378 J. Prevost, P. Valery, Marginalia, Rhumbs et autres (Paris: Editions Uo Scheer, 2006),
229 (Valery's emphasis).

327
 S Y NT H E T I C P H I L O S O P H Y

thousand pages of his Cahiers,379 used to point out, for

his part, that 'the origin of reason, or of its notion, is,
perhaps, the transaction. One must transact with logic,
on one side, with impulse, on the other, and, on yet
another, with facts' .380 The phenomenology of math­
ematical creativity should confront those transactions,
those contaminations, those impurities, which are what
ultimately afford us the entire richness of mathematics.
The reduction of the philosophy of mathematics to the
philosophy of 'logic' (according to the usual tendencies
of analytic philosophy, as we saw when we opened the
Oxford Handbook ef Philosophy ef Mathematics and Logi,c
[101-107] , from which ' Mathematics' had vanished),
or the very reduction of the philosophy of mathematics
to questions of 'logic' and of 'facts' (according to the
slightly broader tendencies that take stock of mathemat­
ics' interlacing with physics), are approaches that leave
behind the indispensable creative 'impulse' to which
Valery refers. In this chapter we will try to elucidate
that (apparently vague and undefinable) impulse, which
responds, nevertheless, to a complex phenomenological
web of catalysts and graftings of inventiveness that can
be made explicit.

379 Facsimile edition: P. Valery, Cahiers (Paris: Editions du CNRS) (29 vols.), i957-61.
Critical edition: P. Valery, Cahiers 1894-1914 (Paris: Gallimard) (9 vols. at present),
i987-2003. Thematic anthology: P. Valery, Cahiers (Paris: Gallimard/Pleiade) (2 vols.),
i973-4.
380 Prevost & Valery, Marginalia . . . , 225 (Va!ery's emphasis).

328
 P H E N O M E N O L O G Y O F M AT H E M AT I C A L C R EA T I V I TY

The nondualist instruments that afford us an adequate

perception of creative transit in mathematics, at bottom,
have been under our nose all the while, even if often
poorly interpreted: namely, the work of Plato, understood
as a study of the mobility of concepts, as a metamorpho­
sis of knowledge, as a description of connections and
interlacings, as a subtle analysis of the inter- and the
trans- has always been there, under our nose, though it
has often been poorly interpreted. Beyond certain trivial
readings, there nevertheless stands the dynamic Plato
of Natorp,38' thanks to whom it has become impossible
to not see 'the genuine, dynamic sense of the idea' that
renders 'untenable the interpretation of ideas as things' .382
The processual, nonstatic Plato, a Plato not fixed to a
reification of the idea, a Plato whom Natorp recuper­
ated at the beginning of the twentieth century, and to
whom Lautman [56] and Badiou [279] would later
return, seems to constitute the nondual, mobile base that
mathematics requires: an apparent contradiction in terms
- for the approaches customary to analytic philosophy of
mathematics, the 'base' should not turn out to be mobile.
Nevertheless, in the rereading of Plato that Natorp pro­
poses, we see how 'the logoi do not have to be governed by

3 81 ]. Servais, Paul Natorp et la thiorie platonicienne des /dies (Villeneuve d'Ascq: Presses
Universitaircs du Septentrion, 2004). An exccllem and brief introduction to Plato
by Natorp can be found in F. Brentano & P. Natorp, Plat6n y Arist6teles (Buenos Aires:
Quadrata, 2004).

3 82 Ibid . . 120, 92 (in the order of the texts cited).

329
 O F C O NT E M P O RARY M A T H E M AT I C S

the onta', how 'the definitions o f mathematics, i n reality,

define methods; in no way do they define existent things
or simple properties inherent in such things', how 'kinesis '

the movement, the transformation, or the gait, let us say,

of concepts' governs all of mathematics, how the Platonic
theory of ideas refers to the 'method of its positions', to
its 'becoming, mutation or peregrination', and how the
thought of Plato 'installs itself in the relative position' and
opens itself to the study of the 'correlative [ . . . ], of change
[ . . J and transition' 383 Mathematics - bound up with the
. .

study of logoi, methods, partial representations, and rela­

tive positions, as we have abundantly corroborated in the
contemporary realm, a realm that oscillates between the
eidal, the quiddital and the archeal - can thus assume at
the outset, as its mobile base, this Platonic thought that
is so alert to the transformability of concepts/proofs/
examples. The mobility ofthe base, indispensable for under­
standing Grothendieck's work [14Q], underlies Platonic
philosophy from the beginning.
Later in this chapter, we will draw from the works of
Merleau-Ponty, Blumenberg and Rota to further refine
the 'mobile base' just indicated. But already - simply by
way of the possibility and plausibility of the hierarchy of
modulations and 'transactions' afforded by a dynamic
Platonic philosophy - we can better understand the modes

330
P H E N O M E N O L O GY O F M AT H E M AT I C A L C R E A T I V I TY

by which mathematical creativity emerges. Indeed, we can

already begin to situate the tension between discovery and

invention under 'reasonable' presuppositions. Taking the
word 'reasonability [razonabilidad]' in the sense given to
it by Vaz Ferreira - as a gluing together of 'reasonable'
and 'sensibility'384 - we can say that some reasonible Pla­
tonic presuppositions underlying the invention/discovery
polarity are: that the polarity is not antagonistic or dual,
but, rather, entwined in a web; that through this web transit
various types of quasi-objects, modalities, and images;
that the positions of those quasi-objects, modalities, and
images are not absolute, but relative; that a progressive
gradation determines, depending on the context in ques­
tion, the mathematical quasi-objects' proximity to each of
the polar extremes; that in that gradation, observations of
structuration tend to approach (in spiraling or asymptotic
turns) the processes of discovery, while constructions of
language tend to approach (after perhaps one more 'turn
of the screw') processes of invention.
In the emergence of mathematical thought, contamina­
tions are legion. Recall Grothendieck's magnificent text
on motifs [145-146] . In it, mathematical creativity is
distributed across a great variety of registers: the initial
'listening' to the motif (Peircean firstness), its incarnation
384 C. Vaz Ferreira, L6gica viva (Caracas: Biblioreca Ayac:ucho, 1979) . On Vaz's 'reason­
ability', see A. Ardao, L6g;i.ca dela raz6n y l6gi,ca dela inteligencia (Montevideo: Marcha,
2000). The works ofVaz Ferreira (Uruguay, 1872-1958) open up very interesting (and
untapped) narura] osmoses between 'pure' and 'human' sciences.

331
 S YN T H E T I C P H I L O S O P H Y

in a 'multitude of cohomological invariants' ( secondness),

and its pragmatic interlacing through modulations of
the 'basic motif' ( thirdness) . But the process does not
stop there; it is not static, cannot be isolated, and iterates
itselfrecursively: given a cohomology ( first) , we study the
topological spaces ( second) captured by that cohomol­
ogy, and then determine the transits ( third) between the
spaces that the given cohomology codifies. And so, in
succession: a position relative to a given space, a position
relative to a given transit, and so on. Mathematics thus
proceeds by means of maximal connections of information
( 'saturations', as Lautman would say) in evolving strata
of knowledge. Creativity emerges through that variable
multiplicity: in virtue of singular 'impulses' and hypoth­
eses, examples that let one visualize how hypotheses are
grafted to concepts, inventive forms of demonstration
that let us ascertain the scaffolding's soundness. Indeed,
the Peircean methodology of scientific investigation - a
cycle between abduction ( first) , deduction ( third) and
induction ( second) - is paradigmatically incarnated in
mathematics, if we extend the planar 'cycle' to compre­
hend a 3-dimensional spiral, both recursive and amplifying.
Pursuing certain images in Grothendieck's work will
help us further comprehend the amplifying spiral of
mathematical creativity. From the 'cohomology deluge'
[158] expressed in his correspondence with Serre (1956),
to his musical vision of motivic cohomology [145-146]

332
P H E N O M E N O LO G Y O F M AT H E M AT I C A L C R E A T I V I TY

in Recoltes et semailles (1986) , passing through the techni­

cal construction of the great cohomologies ( 1964) that
would lead to the resolution of the Weil conjectures
[142, 159], Grothendieck sets out from a vague image
(firstness, abduction: deluge), which he submits to the
complex and extensive filter of definability and deduc­
tion (thirdness, deduction: etale cohomology), which
he then contrasts with other invariants (secondness,
induction: other cohomologies) , which provokes a new
vision (firstness, abduction: motifs, musicality) . Note
that mathematical inventiveness is not uniquely restricted
to the realm of abduction or firstness, where the creative
hypothesis obviously takes precedence, but also takes
place in the realm of demonstration and the contrasting
of examples. Indeed, as the Platonic mobile base suggests,
neither invention nor discovery are absolute; they are
always correlative to a given flow of information, be it
formal, natural or cultural. It is in an antecedent transit,
for instance, that Grothendieck discovers motifs, though
a manner of representing them would later be invented

by Voevodsky [147' 259] . In a similar manner, Zilber

discovers the emergence of 'groups everywhere' [256] ,
hidden in the theory of models, though only later, together
with Hrushovski, would he invent the Zariski geometries
[256] by which the ubiquity of groups could be rep­
resented. There are a profusion of other examples, all
of which seem to be governed by an initial, elementary

333
 O F C O NT E M P O RA RY M AT H E M AT I C S

typology: a perception/vision/imagi,nation efa generic situ­

ation, associated with a broad spectrum of applicability
(the realm of discovery), which is interlaced (without
determinations of priority regarding the direction of the
interlacing) with a constructioniframework/realization ef
manyparticular concretizations in the adopted spectrum of
applicability (the realm of invention) . In this way, we once
again find the One (discovery) interlaced with the Many
(invention), through an abstract integral and differential
calculus that is both mobile and complex.
The recursive modal transit between the possible, the
actual and the necessary is one of mathematical creativity's
greatest strengths. The conjunction of the three emphasized
terms ('transitoriness', 'recursivity' , 'modality') to some
extent explicates the specificity of mathematical thought.
On the one hand, we have seen how, throughout the twen­
tieth century - with the works of Godel, Grothendieck,
Lawvere, Shelah, Zilber or Gromov, among many oth­
ers - mathematics has opened the vital floodgates to the
relative, while always searching behind that movement for
the proper invariants: the transitoriness of ( quasi-)objects
and processes is recognized, but certain ubiquitous modes
are sought in the flux (an epistemological leap from the
'what ? ' to the 'how?' ). On the other hand, we have also
seen how the hierarchization of mathematics involves
incessant processes of self-reference, which yield knowl­
edge by recursion of the ( quasi-)objects and processes

334
 P H E N O M E N O L O G Y O F MATH E M A T I C AL C RE A T I V I TY

at stake: knowledge distributes itself in layers and strata

(mathematics as architectonics) , and the interrelation of
local information hints at a global vision of mathematical
'entities'. Finally, we have observed howfree combinations
(firsts) , in the abstract and the possible, contrast with
facts (seconds) of the physical world, and help flesh out
a comprehension (third) of the cosmos as a whole: the

weavings between mathematics and physics have been

perennial, and have once again reached startling heights
(Arnold, Atiyah, Lax, Witten, Conncs, Kontsevich) . I t
i s between the inventive freedom o f concepts and the
inductive and deductive restrictions of calculation that
mathematics has situated itself.
The work of Jean-Yves Girard is emblematic in this
sense.385 His monumental Locus Solum386 breaks with our
received ideas and completely upends our understanding
of logic. Beyond syntax and semantics, Girard pierces
through to the geometry ef logic's rules and extensively
explores logic's locative transformations. As a substitute
for proofs and models, Girard introduces the concept
of 'design' as the 'locative structure of a proof in the
sequent calculus', and elaborates a subtle and complex

385 We are grateful to Zachary Luke Fraser for having indicated the pertlnence of Girard's
works to us in the context of ideas of 'flow' and 'obstruction', in both a logico­
gcometrical sense (Locus Solum) and in a methodological sense (Du pourquoi au
comment).

386 J.-Y. Girard, 'Locus Solum: From the rules of logic to the logic of rules', Mathematical
Stmctures in Computer Science 11, 2001: 301-506.

335
 S Y NT H E T I C P H I L O S O P H Y

instrumentarium for the combinatorial and geometrical

transformations of designs (processes of normalization,
theorems for addition, multiplication and quantification,
soundness and completeness for nets) . The de-ontologi­
zation of the enterprise - or, better, its new collocation in
a transitory ontology - is remarkable. 'Designs' require no
material substrates, whether syntactic or semantic, to be
understood. Their pragmatic behavior, the abstract dis­
engagement of their flows and obstructions, and their cat­
egorical manipulation are entirely sufficient.387 In tune with
the search for an initialprotogeometry - having this in com­
mon, as we have seen, with the works of Zilber, Freyd and
Petitot - Girard situates his methodological reflection in
a leap from the 'why? ' to the 'how?' 388 The French logician
situates the lambda calculus, denotational semantics and
his own linear logic in relation to this 'how'. The underly­
ing substrate of the 'what' is not only unnecessary, but may
even become noxious. The liberation of objects, the power
to look at rules categorically and handle them in terms of
a pragmatico-geometrical 'how', does indeed constitute,

387 Girard indicates three steps in the development of logic: 1900-1930, 'the time of
illusions'; 1930-1970, 'the time of codings'; 1970-2000, 'the time of categories'.
Girard's own work can be seen as attempting to cross the threshold of this third era,
and hears the deep influences of category theory's mathematical development in an
effort to exceed its reach. While category theory reveals thefunctional, structural, and
universal dimensions of logical proofs (what Girard calls their 'spiritual' aspect),
Girard's 'ludics' and 'geometry of interaction' aim to uncover their interactive, dynamic
and singular side( what he calls their 'locative' aspect).
388 J.·Y. Girard, 'Du pourquoi au comment: la thCorie de la demonstration de 1950 a nos
jours,' in Pier, Development ofMathematic, 515-45

336
 P H E N O M E N O L O G Y O F M AT H E MA T I C AL C R E A T I V I T Y

as we have seen throughout this book, one of the highest

points of contemporary mathematical creativity.
In that environment of flows and obstructions, a
general thinking of residues and sedimentations comes to
be of great assistance for the better understanding of the
processes of genesis at work in mathematics. Merleau­
Ponty proposes that a science of 'sedimentations' be
founded,389 whereby the circle of man and nature would
be closed through an operative body that interlaces the
visible and the invisible - a body, that is to say, that
conjoins the horizon of a general world (the visible) with
the horizon of an underworld (what is seen by the seer) ,
that is inserted in turn into the first horizon. Knowledge,
rooted in a body, but mixed with a web of world horizons,
refuses the Cartesian caesura of mind and world, and
connects knowing and nature in a continuous fashion.
Nothing happens 'outside the world', the senses, or vision
in particular. Cultural horizons, the interpreter's contex­
tualizations, and sedimentations are vital for knowledge,
and for mathematics in particular, which turns out to be
profoundly human. Phenomenology interlaces the human
eye, the general horizons of the world into which vision
is inserted, and the particular subhorizons where 'things'
are reborn through the body of the observer. And so, on
perceptual grounds, the sediments of culture, knowledge,

389 :;vferleau-Ponty, Notes des cours... , 44.

337
 O F C O NT E M P O RA RY M A T H E M AT I C S

and social life continue to accumulate, and 'things' are

modalized through multiple horizons, whereby we detect
them in their various registers (structure,fibration,junc­
tion, sensation, etc.). Figure and ground likewise outstrip
their dualization, as they interlace with each other in an
incisive and visible continuum, not only in the modern
manifestations of art (Mallarme, Proust, and Cezanne, as
reread by Merleau-Ponty), but also, as we have seen, in
the contemporary manifestations of mathematics.
In Eye and Mind, Merleau-Ponty describes the body
operative in the domains of knowing as a 'sheaf of func­
tions interlacing vision and movement' [151] . As we
have been indicating, that sheaf serves as an interchange
(a la S erres) between the real and the imaginary, between
discovery and invention, and allows us to capture the
continuous traniformation efan image into its obverse, through
the various visions of interpreters. Two of the major
theses of the late Merleau-Ponty combine the necessity
of both thinking the recto/verso dialectic and thinking in
a continuous fashion:

What is proper to the visible is to possess a fold of

I.

invisibility, in the strict sense.

2. To unfold the world without separating thought is,

precisely, modern ontology.39°

3 90 Ibid., 85, 2 2 .

338
P H E N O M E N O L O GY O F M AT H E M A T I C A L C RE A T I V I T Y

As we have seen, many constructions specific to contem­

porary mathematics allow us to corroborate, with all
the desired technical support, both of Merleau-Ponty's
theses. The 'fold of invisibility' is particularly striking in
what we have called the 'structural impurity of arithmetic'
[41], where the most important signposts leading
to the resolution of Fermat's Theorem [187] are liter­
ally invisible from the discrete perspective of the natural
numbers, until we pass through to the obverse side of
the complex plane. Likewise, to 'unfold' mathematics,
without separating its subregions, is one of contemporary
mathematics' major strengths, and underlies, in particular,
the exceptional richness of Grothendieck's thought, the
constant transgressor of artificial barriers and explorer of
natural continuous connections between apparently dispa­
rate images, concepts, techniques, examples, definitions
and theorems.
The 'manner' (156] , or style, through which the
great mathematicians produce their works is another
problematic that analytic philosophy of mathematics
intrinsically neglects. The works of Javier de Lorenzo
(76-78] opened up an important seam at this point,
several decades ago, but one which has nevertheless not
been sufficiently mined. In this work's second part, we have
described a few concrete registers of local ways of doing
mathematics, ways we have not analyzed (and cannot
do so: it is work for another essay) from the viewpoint

339
 SYNT H E T I C P H I LO S O P H Y

of the style's global configuration. Nevertheless, from the

p henomenological perspective on creativity adopted
here, we can specify certain forms of overlapping and
sedimentation, which, counterpointing [206] forms of
break and rupture, help us delineate the stylistic spectrum
of the creators of mathematics. A basic observation, in
fact, shows us that the mathematical creator proceeds by
way of gradual exercises of weaving (or 'back-and-forth')
between generic (powerful and vague) images and various
successive restrictions (definitions, theorems, examples)
that allow her to go on sharpening the original 'impulses'
or intuitions. The counterpoint between the deductive
overlapping and the imaginal rupture is thus necessary in
the first steps of creation, although, later, sedimentation
tends to overwhelm the break.
That counterpointed textile involves peculiar modes
of enlacing and correlating. Perhaps close to music in this
sense, mathematics discovers both symmetries/harmonies
and ruptures of symmetry/harmony, which it must then
distill by way of many developments, variations, and
modulations, in well-defined languages that allow it to go
on to 'embody' the great harmonies or ruptures, whether
seen or at first 'hidden' (remember Grothendieck, and his
listening to ' the voice of things' [151, 159] ) . Reason­
ability ( reason[a][sensi]bility, in Vaz Ferreira's sense) is
=

vital here, and we should literally glue together a free, dia­

grammatic, imaginal sensibility (the realm of the aesthetic),

340
P H E N O M E N O L O G Y O F M AT H E M AT I C A L C R E A T I V ITY

and a normative, ordering, structuring reason (the realm

of the ethical), whether in lengthy disquisitions on the
pentagram or long series of mathematical deductions.
But the very polar coherence of mathematical creativity
(which, by contrast, does not occur in musical creation)
obliges us to observe the situation's underside, counterpoint­
ing it with an inventive and liberating 'aesthetic reason',
and with, above all, a contrastive and communal 'ethical
sensibility'. Mathematics thus succeeds in transcending
the imagination, unreined by the isolated individual, and
becomes the greatest imaginary construction of which an
entire community is capable.
The accelerations and decelerations in the processes
of mathematically 'gluing' together conjectural images,
partial hypotheses, imaginary residues, real examples and
theorematic sedimentations cover the most diverse situa­
tions possible. In many eidal ascents deductive stratifica­
tions and generic visions may take precedence, over and
above secondary contras tings (residues, examples): this is
the case, for example, in Grothendieck-Dieudonne's EGA
[166-169] , in Lawvere's conception of a set theory
without local Von Neumann elements [195], in the
beginnings of Langlands's functorial program [ 180] ,
or in Shelah's first glimpses of the nonstructure theorems
[196] . In turn, in what we have called quiddital descents,
the richness of residues, obstructions and examples ( dif­
ferential equations a la Atiyah or Lax, noncommutative

341
 O F C O NT E M P O RARY M AT H E M AT I C S

or quantum groups a l a Connes or Kontsevich), tends to

take precedence over 'primordial' images, generic instru­
mentaria, or deductive machinery. In the archeal findings
that we have pointed out (such as the 'geometrical nuclei'
in Zilber or Gromov, or the 'logical nuclei' in Freyd or
Simpson), we do not observe a particular directionality
or predominance in the combination of images, hypoth­
eses, residues, examples, definitions and theorems that
develop the mathematical archetypes which correlatively
'dominate' certain classes of structures. In the latter, in
fact, particularly in the works of Zilber and Gromov, the
incessant weaving between the most concrete and most
abstract is not only noncircumventable, but constitutes
something like a genuinely original, systematically oscil­
latory, manner, one we could perhaps consider proper to
the Russian 'school' .
Going back t o the work of Nicholas of Cusa, Hans
Blumenberg recalls that the world divides into 'visibilia,
invisibilium imagines [visible "things", images of the invis­
ible]', and how the universe of 'mathematicalia [math­
ematical "things"]' allows us to refine our vision from a
clear 'methodological vantage point: the availability to
freely perform variations, the possibility to experiment
while freely establishing conditions'.39' We have already
391 H. Blumenberg, Paradigmas para una metaforologia ( 1960) (Madrid: Trotta, 2003), 242-3.
(Originally published in German as Paradigmen zu einer Metaphorologie, in Archivfar
Beg:ri.Jfsgeschichle (Bonn: 1960 ]. [Tr. R. Savage as Paradigmsjor a Metaphoroiogy (Ithaca:
Cornell University Press, 2010)].

342
P H E N O M E N O L O GY O F M AT H E M AT I C A L C R E A T I V I TY

emphasized the fundamental freedom of mathematics,

its variational richness in the realm of possibilia and its
capacity to enter the dialectic of visibilia and invisibilium.
To some extent, that plasticity is due to another profound
counterpoint in mathematical creativity: the pendular
weaving between the metaphorical and the technical. Math­
ematics has not only been a producer of external metaphors
for the unfolding of thought, but often proceeds internally,
in its processes of creation, by way of vague and potent
metaphors, still remote from precise, technical delimita­
tions. On this note, it is curious that analytic philosophy
of mathematics should have proscribed the study of
mathematical metaphorics as something unbefitting of
exact knowledge, when its entire program emerged from
explicit metaphors - metaphors later enshrined as myths,
has Rota has pointed out [86] : atomism, absolute,
dualism, foundation, truth, etc.
A phenomenology of mathematical metaphorics
should draw on Blumenberg's extensive work392 in the his­
torical tracking of metaphors in philosophical language.
Starting from the fundamental (Husserlian) antinomy
392 Major monographs: Paradigmen zu einer Metaphorologie (1960) (Paradigms for a
Metaphorology [ 2010 ]), Die Legitimitdt der Neuzeit (1966) (1he Legitimacy ofthe Modern
Age [ 19 85]) , Die Genesis derkopern ikanischen Welt ( 19 81) (1he Genesis ofthe Copernican World
[ 1987]), Die Lesbarkeit der Welt ( 1979), Lebenszeit und We/tzeit (1986), Hohlenausgange (1989 ).
Minor essays (and gems): Schiffbruch mil Zuschauer. Paradigma einer Daseinsmetapher
(1979) (Shipwreck with Spectator: Paradigm of a Metaphorfor Existence [ 1996]), Das
Lachen der 1hrakerin. Eine Urgeschichte der 1heorie (1987), Die Sorge geht uber den Flufl
(1987) (Care Crosses the River [ 2010]), Matthiiuspassion (1988). A good presentation of
Blumenberg's work can be found in F. J. \Vetz, Hans Blumenberg. La modernidady sus
metdforas (Valencia: Edicions Alfons el Magnanim, 1996).

343
 S Y NT H E T I C P H I L O S O P H Y

between the irifinitude of the philosophico-scientific task

and the.finitude of human individualities, Blumenberg
studies the complex lattice of graftings and rupture s
between the 'time of the world' and the 'time oflife', where
mathematics plays the role of an exception. Reviewing
Husserl's Logic of 1929, Blumenberg emphasizes how the
master introduces the metaphor of a 'historical sedimenta­
tion as the product ef a static ideality ef dynamic origin' .393
The metaphor's apparent contradictoriness stems from
disparate appropriations of Plato (a static reading versus
Natorp's neo-Kantian, dynamic reading, taken up again
by Husserl), but superbly recollects several of the primor­
dial modes of mathematical creativity that we have been
emphasizing. Creativity proceeds within the 'time of life',
within what Merleau-Ponty called the operative body, but
is always extending itself to the 'time of the world' that
envelops it; sedimentation is therefore historical, and takes
place through a web of ideal positions that appear static,
but which emerge on a ground of dynamic, polar media­
tions. Indeed, mathematics - understood as the study of
the exact transits of knowledge ('dynamics') - constructs
partial invariants ('static ideality') in order to gauge the
obstructions and osmoses in transit, and then gathers the
various registers obtained ('historical sedimentation').

393 H. Blumenberg, 1empo de/la vita e tempo de/ mondo (Bologna: ii Mulino, 1996) , 391.

344
 P H E N O M E N O L O GY OF M A T H E M A T I C A L C R E AT I V I TY

In the incessant mediation between dynamics and stat­

ics, reality and ideality, world and life, the Fundierung
according to Rota (who also, in his own way, takes up
and reinterprets Husserl's ideas) combines two central
processes in the construction of mathematical webs:
Jacticity and Junctionality.394 For Rota, a mathematical
(quasi· )object - through its many forms, from the vague
metaphorical image to the carefully defined and delimited
technical (sub-)object, passing through intermediate
modes, examples and lemmata - should be, on the one
hand, pragmatically inserted into a context (facticity) ,
and, on the other, correlatively contrasted in that context
(functionality). M athematics, which lives synthetically
on both the factical level (contextualization) and the
functional level (correlation), appears irreducible to any
supposed 'objectuality'. The Fundierung studies how math­
ematical processes lace into one another, independently of
their analytical classification (via E or � ) . The processes
(or quasi-objects) are understood as poles of a relation
of stratification, with complex gradations between them.
Mathematics studies the transformation and splicing of
those various gradations, independently of any illusory
ultimate 'ground' on which they would rest. Beyond
certain supposedly stable and well-founded mathemati­
cal 'objects', the jactical andfunctional webs ef coupling

394 Palombi, La stella e l'intero . . . , 62.

345
 OF C O N T E M P O RARY M A T H E M A T I C S

between mathematical quasi-objects, in the broadest

possible sense (metaphors, ideas, processes, conjectures,
examples, definitions, and theorems), thus constitute the
true spectrum of mathematical phenomena.
A phenomenology that seeks to capture the transits
of mathematics in an adequately faithful and nonreduc­
tionist manner must take many polarities into account:
analytic decompositions and synthetic recompositions,
modes of differentiation and integration, processes of
localization and globalization, particularities and uni­
versalities, and forms of creativity and discovery, among
others. As we have seen in the preceding chapters, the
differential-integral back-and-forth is not only situated
on the epistemological level ('how'), but continuously
extended into the 'what' and the 'where' of the quasi­
objects at stake. The many strata/environments/contexts
of mathematics, according to Rota's Fundierung, respond
to a complex regimentation of the prefix trans-, on ontic
and epistemic levels, breaking the customary barriers of
philosophical reflection.
For a minimal conceptualization of the trans-, a consid­
eration of the following operations is required:

346
P H E N O M E N O L O GY O F M A T H E M A T I C A L C R EA T I V I TY

(A) Polarities (or 'adjunctions'):

• decomposition/composition
• differentiation/integration
• deiteration/iteration

• particularization/universalization

• localization/globalization
• residuation/potentiation
(B) Mediations
• oscillation
• mixing

• triadization
• modalization

• sheafification

Figure i5. Polarities and mediations in a general operativity ofthe trans·.

Analytic philosophy of mathematics excludes these media­

tions from its realms of investigation, which perhaps
explains its incapacity to capture the universe of advanced
mathematics. The first dual pair ('decomposition/compo­
sition') recalls the necessary and irreducible dialectic between
analysis and synthesis, a dialectic that runs through the
entire history of philosophy and mathematics. In turn,
the first mediation ('oscillation') recalls a necessary and
irreducible pendular variation of thought, forever stretched
between opposing polarities. The second mediation ('mix­
ing' a la Lautman) accompanies that inevitable pendular
oscillation with the awareness that we must construct
mixtures to serve as support structures for an extended

347
 S Y N T H E T I C P H I L O S O P HY

reason ('reasonability') , mixture being understood here in

the sense of sunthesis (as composition, which is reversible),
as opposed to sunchusis (as fusion, which is usually irre­
versible) . The second dual pair ('differentiation/integra­
tion') recalls one of the major, originary problematics of
philosophical and mathematical thought: the dialectic of
the Many and the One. The third dual pair ('deiteration/
iteration'), together with the third and fourth mediations
('triadization' and 'modalization') , constitutes one of the
great richnesses of advanced mathematics, something that
is usually invisible in elementary strata or from analytic
perspectives, but which, by contrast, is reflected with great
acumen in the original operative nucleus of the Peircean
architectonic [1u-1�.u] . Indeed, the Peircean empha­
sis on the rules of deiteration/iteration represents one of
Peirce's most profound contributions, whether from a
logical point of view (the rules codify the definitions of
the connectives) , or from a cognitive one (the rules codify
creative transfers of information) . Similarly, the systematic
Peircean triadization and its modal filtration secures the
plural richness of the pragmaticist architectonic.
The fourth and fifth dual pairs ( 'particularization/uni­
versalization' and 'localization/globalization'), together
with the fifth mediation (sheafification), answer more spe­
cifically to contemporary mathematical forms of thought.
In certain well-delimited cases, sheafification allows local
information to be glued together in a coherent fashion,

348
P H E N O M E N O L O G Y O F M AT H E M A T I C A L C RE A T I V I TY

and takes us to global quasi-objects that capture the transit

of the information in the sheaf's fibers. Mathematics,
then, is to a large extent concerned with calibrating the
calculable osmoses and obstructions in those back-and­
forths between local and global properties, in the realms
of space, number, structure and form. The sixth dual pair
('residuation/potentiation'), for its part, captures the rich­
ness of certain strata of mathematics with strong coherence
properties (like the calculus of the complex variable, or
topos theory), where the residues become (quasi-)objects
that are completely reflective of their environment, and
exponents of mathematics' visionary possibilities.
In all of these processes there is an enrichment of
the Peircean summum bonum - understood as the 'con­
tinuous increase of potentiality'395 - and an explosion
of mathematical creativity into the most diverse forms.
The metaphorical and analogical webs studied by Chatelet
[93-95] combine with precise modes of sedimen­
tation - examples, definitions, theorems - giving rise
to a sophisticated 'integral and differential calculus' of
mediations and gradations (a Fundierung) that orients
the evolution of mathematical thought. Demonstrative
395 For an excellent presentation of the place that a broad 'reasonability' occupies in
Peirce's system, oriented by the summum bonum, and for a study of its correlations
with sensibility, creativity, and action, see S. Barrena, La creatividad en Charle:; S. Peirce.·
abducci6n } ra;;,onabilidad, doctoral thesis, Department of Philosophy, University of
Navarra, Pamplona, 2003 (part of which has been published as S. Barrena, La ra::.6n
creativa. Crecimientoyjinalidad de/ ser humano seg<ln C.S. Peirce [Madrid: Rialp, 2007]).
Beyond Peirce himself, Barrena's works represent a remarkable contribution to the
understanding of creativity in general.

349
 O F C O N T E M P O RARY M AT H E M A T I C S

'rationality' is wedded to imaginary 'reasonability': in the

processes of articulating/mediating/gluing the various
'dual pairs' we have just mentioned, lies the discipline's
inventive malleability. What we are dealing with here is a
peculiar plasticity, which combines a joyous capacity for
movement (a facility of transit through the world of the
possible) and a careful handling of dynamic differentials
(a control of change in the realm of exactitude) . It is per­
haps in this connection between plasticity and exactitude that
the true strength of mathematics is found, a connection
that is crucial for the processes of mathematical creation.
Any perspective that neglects that plastic 'art of doing'
[327] neglects the living nucleus of the discipline itself.
Serre, who should no doubt be considered as one of
the major stylists of contemporary mathematical literature,
used to point out the importance of mixtures in math­
ematical creativity, and of the presence of a vital inventive
penumbra ' I work at night (in half-sleep), [which] makes
-

changing topics easier' [ 177] behind the supposed

-

luminosity of proof. It could be said, observing the almost

crystallographic luster of Serre's own work, that the great
mathematical creator laces his perpetual decantation of
the penumbra (the realm of discovery) with a rare capac­
ity to reveal/construct luminous crystals (the realm of
invention), on his zigzagging path. Indeed, it is remark­
able that Serre's limpid style, astonishingly smooth and
'minimal', should, in the author's own words, reveal itself

350
P H E N O M E N O L O G Y O F M AT H E M AT I C A L C RE AT I V I TY

as a 'wonderful melange' situated on penumbral ground.

In the same sense, many of contemporary mathematics'
crystallog;raphic gems emerge from the obscure grounds where
they are born: Grothendieck's motifs with their musicality
in major and minor keys [145-146], Langlands's letter
to Weil with its putative unseriousness and casual tone
[180-182] , Atiyah's I ndex Theorem with its excava­
tion of murky depths [208], Cartier's dream with its
shifting terrains in mathematical physics [226] , Zilber's
extended alternative with its obscure intuition of the
complex variable's logical behavior [257], and Gromov's
h-principle with its ground of discordant situations in the
penumbra of differential equations [263], amid many
other examples that we took up in part 2 of this book.
As we have seen throughout this chapter, advanced
mathematical creativity can be understood only through
perspectives that reflect the phenomenology of nontrivial­
ized mathematical transits: intertwinings in webs, nondu­
alist gradations, contaminations on a continuum, recursive
modal interlacings, dialectics of sedimentations and resi­
dues, partial osmoses between metaphorical images and
technical objects, global gluings over coherent local adum­
brations, factical and functional processes of intermin­
gling (Fundierung) , and systematic mediations between
polarities. In the realm of elementary mathematics, these
phenomena tend to vanish, owing to the reduced complex­
ity of the entities in question. On the other hand, from

351
 SYNTH E T I C P H I LO S O P H Y

the perspectives o f analytic philosophy (and regardless

of the realm observed, whether elementary or advanced),
these phenomena are also neglected, since they are usually
considered 'ill-defined', or impossible to define at all (we
hope to have shown in chapters 8 and 9 that this is not
the case) . Perhaps these two tendencies, together with t he
tendency (predominant in the philosophy of mathematics)
to study the elementary from analytic perspectives may
explain the scant concern that has, until now, been given
to the problematic of mathematical creativity.

352
C HAPTE R 11

MAT H E MATI C S A N D C U LT U RAL C I R C U LAT I O N

In this final chapter we will take up two remaining consid­

erations, with which we will try to round off our work. If,
in chapters 8-10, we have emphasized a turn to questions
connected with the 'how?' in contemporary mathematics,
and if, in chapters 4 7 we have reviewed a few emergences
- ,

('where?') of precise and delimited problematics ('why?')

in detail, in this chapter we will study, on the one hand, the
general interlacing by which contemporary mathematics
(1950 to today) can be distinguished from prior math­
ematics (a conceptual 'whatjhow/why synchrony, beyond a
'

merely diachronic cut happening around i950). And, on

the other hand, we will also examine the general position­
ing of mathematical thought within culture, and the ways
in which it naturally shares a frontier with aesthetics (a
conceptual geography of the 'where?') .
In chapter 1 we distinguished from a 'bird's eye' view
(a perspective both distant and evanescent) some features
that allow us to provisionally separate modern math­
ematics (from Galois to about 1950) from contemporary
mathematics (from around i950 to today). We will sum­
marize these characteristics in the following table (with
i-5 being implicit in Lautman's work [30] , and 6 -10
being implicit in the developments of contemporary

353
 O F C O NT E M P O RARY M AT H E M A T I C S

mathematics [41]) , and then explicate the synchronic

conceptual g;rounds that may help to explain characteristics
1-5 and 6 -10, as well as the diachronic cuts situated around
1830 (the beginnings of modern mathematics) and 1950
(the beginnings of contemporary mathematics) .
In what follows, in the first part of this chapter, we
will study some forms of internal circulation within the
conceptual realm of mathematics, which will help us to
intrinsically distinguish certain periods of mathematical
production; in the second part, we will study some forms
of external circulation in the general realm of culture, which
will help us better explicate certain modes of mathematical
creativity through suitable correlations and contrasts. In
an exercise of counterpointing, we will go on to pinpoint
the 'why' of modern mathematics' emergence and the
'why' of its later evolution into the questions, methods
and ideas of contemporary mathematics. We will discuss
the characteristics observed by Lautman ( 1-5), with regard
to both their positive side and their obverse. This will lead
in a natural way to characteristics (6 -10) which reflect
,

certain crucial features of the new conceptual grounds at

stake in contemporary mathematics.

354
 M AT H E M AT I C S A N D C U LT U R A L C I RC U LAT I O N

MATHEMATICS

MODERN CONTEM PORARY

( 1 830-1950) ( 1950-present)

l. complex hierarchization
,/ ,/
systems of mediations

2. sema n tic richness

,/ ,/
irreducibility to grammars

3. structural unity
,/ ,/
multiple polarities

4. dynamics
,/ ,/
movements of liberation/saturation

5. theorematic mixing
,/ ,/
ascents and descents

transits/obstruction
hierarchies/structures
modeling/mixing

6. structural impurity
,/
the arithmetical via the continuous

7 . ubiquitous geometrization
,/
arc heal geometric nuclei

8. schematization
,/
categorical characterizations

9.fluxion and deformation

,/
obverse of usual properties

10. reflexivity
,/
complex forms of self-reference
fluxions/altemations
schemes/nuclei
rcllection/sheafilication

Figure 16. Some conceptualfaatures that help demarcate modern and contemporary
mathematics.

355
 S YN T H E T I C P H I L O S O P H Y

Modern mathematics more o r less surfaced through the

work of Galois and Riemann, with the introduction of
qualitative instruments for the control of quantitative
problematics, from a point of view that is both positive
(transits) and negative (obstructions) . On the one hand, in
fact, the structural interlacing of the hierarchy of Galois
subgroups and field extensions allow us to control the
behavior of roots of equations (and allow us to ascertain,
among other things, the general unsolvability of quintic
equations); the topological properties of Riemann sur­
faces, on the other hand, allow us to control the ramifica­
tion of multivalent functions of complex variables (and
allow us to ascertain the inequivalence of surfaces like
the sphere and the torus, for example) . From its very
beginnings, modern mathematics has faced a clearly defin­
able, generic problematic: (A) the study efthe transits and
obstructions ef mathematical objects, with qualitative instru­
ments, associated with structural mediations and hierarchies.
The conceptual ground at stake is partially reflected in
characteristics i-5 mentioned above, and corresponds to a
genuinely fresh and novel focus of mathematical percep­
tion, one which has since rystematically opened itself to a
qualitative comprehension of phenomena and a reflexive
understanding of the very limitations (negation, obverse,
obstruction) of that partial comprehension.

356
 M AT H E M AT I C S A N D C U LT U RA L C I RC U LA T I O N

Th e complex hierarchization o f mathematics (I) due to Galois

and Riemann yields many of the richest branches of mod­
ern mathematics (abstract algebra, functional analysis,
general topology, etc. ) , but it is a process (or, better, a
collection of processes) that naturally tends toward satu­
ration on each level of the structural hierarchies in ques­
tion. For example, behind the profusion of semigroups
and groups, contemporary mathematics would perform
a schematization (8) that would open the gates to grou­
poids in general topoi, or, even more schematically, to
operads [235]. On the one hand, the notorious structural
unity (3) of modern mathematics (a unitary solidity that is
perhaps its most distinctive feature) is, in contemporary
mathematics, counterpointed by a sophisticated extension
of the unitary toward that dialectical unity's polar frontiers,
with an altogether original capacity [ 201, 231, 264,
etc.] for tacklingfluxions and deformations ( 11) of structures
which, it once seemed, could be understood only rigidly.
On the other hand, the remarkable semantic richness ( 2) of
modern mathematics, with the enormous multiplicity of
models arising in the period between 1870 and 19 30, in all
the realms of mathematical action (geometries, sets, alge­
bras, functional spaces, topologies, etc.) , would later yield
a reflexive vision ( 1 0) of that diversity, in tandem with the
construction of the instruments needed for a reintegration
of the local/differential into the global/integral.

357
 OF C O N T E M P O RA RY M AT H E M AT I C S

Sheaf theory, whose construction can be traced to the

period between 1943 and 1951 (a conceptual synthesis
represented in figure 13 (163], and whose diachronic
emergence we have already observed [163, �85]) ,
constitutes , for us, the decisive index that allows us to
capture the changes/delimitations between modern and
contemporary mathematics. Indeed, in both their conceptual
ground and their technique, sheaves effectively symbolize
one of the great general problematics of mathematics in
recent decades: (B) the study effluxions and deformations
( arithmetico-continuous, nonclassical) efmathematical quasi­
objects, with instrumerits ef traniference/blockage betweeri the
local and the global, associated with processes ojschematization
and self-reference. Of course, this problematic, which we
have surveyed from above, neglects other important
applied, calculative and computational aspects of con­
temporary mathematics. It is nevertheless clear that, to a
large extent, it includes aspects of the major mathematical
works of the second half of the twentieth century, and
central aspects of all of the works that we have carefully
reviewed in chapters 4-7, in particular.
Contemporary mathematics thus inscribes itself in a
fully bimodal spectrum, in Petitot's sense: simultaneously
physical and morphologico-structural. Indeed, as we saw
with Rota, the incarnation of mathematics' quasi-objects
is at once factical and functional (346] . Lived experi­
ence and knowledge occur in relative environments of

358
 M AT H E M AT I C S A N D C U LT U RA L C I RC U LATI O N

information transformation, and not o n absolute founda­

tions. Mathematical intelligence thus consists in modes of
knowledge-processing that lead.from in/formation to trans/
formation, modes including both the analytic dismember­
ment of information and the synthetic recomposition of
the representations obtained in correlative horizons.
The bimodal and the bipolar, which yield progressive
gradations and precise frontier conditions in transit, lead
to the natural mediations/mixtures characteristic of
mathematical knowledge. I n the incessant search to pre­
cisely and correctly determine the multiple boundaries
of mathematical quasi-objects, the great problematics of
modern and contemporary mathematics, (A) and (B),
answer to conceptual grounds that, with respect to certain
frontier conditions, are welldetermined: (A) takes up the
de/limitation of classes of classical structures, in a first
approximation that partiallyfixes certain coordinates in
modern knowledge, while (B) takes up the extra/limita­
tion of those classes, deforming and differentiating them
(locality) in order to then reintegrate them (globality),
in a second approximation that.frees certain variations in
contemporary knowledge.
The net result of the conjunction of problematics
(A) and (B), the situation in which we find ourselves
today, is a thorough, mathematical understanding of
certain relative universals, that allow us to combine mod­
ern mathematics' fundamental quest for 'universality'

359
 SYNTH E T I C P H I L O S O P H Y

with contemporary mathematics' modulation toward

the 'relative' (and its search for the invariants behind
the transit). Indeed, many of the prominent works that
we have been reviewing respond in a precise and delim­
ited manner to the constitution of webs ifrelative univer­
sals: the Langlands correspondence [181] , Lawvere's
'unity-and-identity of opposites' [191], Shelah's general
theory of dimension [197] , Connes's noncommutative
geometry [ 222] , Kontsevich's quantum cohomology
[ 235] , Freyd's allegories and intermediate categories
[243] , Simpson's reverse mathematics [248] , Zilber's
extended trichotomy [256) , and Gromov's h-principle
[263], among many other accomplishments. In all of
these cases, which should be seen as typical expressions of
contemporary mathematics, we see, to begin with, a full
assimilation of mathematical dynamics; secondly, a search
for ways of controlling that movement (that is to say, the
ways in which adequate frontiers can be defined), and
thirdly, a construction of delimited and technically well­
adjusted quasi-objects that, with respect to the dynamics
and frontiers just mentioned, act as relative universals:
the Langlands group [ 184) , adjoint functors [191 ] ,
dividing lines o f the M ain Gap [199] , the Lax-Phillips
semigroup [219], the Grothendieck-Teichmuller group
[225] , Cor-Split-Map functors [245], second order
canonical subsystems [249], Zariski geometries [ 256],
Gromov's smooth inequalities and invariants [ 260] , etc.

360
 M AT H E MA T I C S A N D C U LT U RA L C I R C U LAT I O N

From a metaphorical perspective, the transition between

modern and contemporary mathematics corresponds to
a process of liberation and variational amplitude, reflected
in the internal circulation of concepts and techniques
that we have been describing. A profound 'shifting of
the soil' [150, 278] has freed mathematics. Through
a continuous process, we have travelled a path efreason/
imagi,nation 'sprogressive amplification: the working of math­
ematics' 'soil' (the analytic/set-theoretic reconstruction
of mathematics) , the experience of the soil's 'shifting'
(relative consistency proofs a la Godel, relative mathemat­
ics a la Grothendieck) , the 'bimodal' understanding of
mathematical transit (the emergence of category theory,
the return of close ties with physics) , and the synthetico­
mathematical construction of 'relative universals'. The
height efreason allows us to contemplate the shiftings and
sedimentations that together compose the terrains of con­
temporary mathematics. What we are dealing with here,
of course, is the gestation of a new topography, whose
looming presence the philosophy of mathematics must
recognize without delay, and which will surely shatter its
rigid academic matrices.
Although the aforementioned modulations must be
inscribed on a continuum, we may note that this progres­
sive amplification is stretched across discrete counter­
pointed looms. On the one hand, theorematic mixing (5)
has led to the singular discovery of archeal geometric

361
 O F C O NT E M P O RARY M AT H E M AT I C S

nuclei, i n the currents of a ubiquitous geometrization (7)

by which the modern mixings can, to a large extent, be
governed from new perspectives. There is an evident
tension here between continuous mediations and discrete
nuclei (motifs, allegories, combinatorial groups and semi­
groups, Zariski geometries, etc.), by which it seems such
mediation can be controlled. On the other hand, modern
mathematics' structural unity (3 ) has widened its margins
to a possible union ofjluxion and deformation (9 ) where
the structural seems to be no more than (an admittedly
central) part of a total, complex dynamic landscape; in the
extension of mathematical structures' generic regimentation,
discrete leaps (quantizations) counterpoint continuous
deformations, giving a new form to Thom's aporia in the
current landscape of mathematics.
Another profound transition helps to explicate the
'why' of the demarcation between modern and contem·
porary mathematics. The rising power of the asymptotic
- counterpointing the fixed or the determinate in modern
mathematical logoi permeates many of contemporary
-

mathematics' major forms. From the emergence of many

classifier topoi and inverse limits with which we may
'glue' the classifiers together - which yields an asymp·
totic understanding of logic, elsewhere corroborated by
Caicedo's results in sheaf logic [283] to Gromov's
-

great asymptotic sweeps through Hilbert's tree [ 262],

passing through the 'covering approximations' of what

362
 M AT H E M AT I C S A N D C U L T U RA L C I R C U LAT I O N

we have called the Grothendieck traniform [320-322] or

through the richness of partial coverings of the real in the
quiddital approach of Atiyah, Lax, Connes or Kontsevich,
contemporary mathematics has been able to express and
control, with the utmost conceptual and technical power,
the crucial notion of a hiatus between fragments of knowl­
edge. The hiatus, understood as a cleft or a fissure - that
is to say, as a 'between' on the obverse of our conceptions
- in turn becomes an aperture to the unexplored, just as
it appears in Merleau-Ponty's Eye and Mind [338] . The
to ti einai (the essential of essence) [124], cannot be
described as a universal concept or 'entity', but precisely
as a generic form of hiatus, inevitably present both in
the world ('transitory ontology' - chapter 8) and in our
approach to it ('comparative epistemology and sheafifica­
tion' - chapter 9 ) . The partial, relative, asymptotic covering
of that hiatus is one of philosophy's primordial tasks,
and one that can now be embarked upon with an entire
series of concepts, instruments, methods and examples
belonging to contemporary mathematics.
Flux, shifting, and hiatus have always and everywhere
engulfed us. Novalis's vivid resurgence in contemporary
culture is not a matter of chance, nor is the recognition
of the contemporary relevance of Peirce's asymptotic
architectonic. Nor, moreover, does it seem to be mere
chance that the beginnings of contemporary mathematics
can be situated around the emergence of sheaf theory, a

363
 SYNT H ET I C P H I L O S O P HY

theory as sensitive as any to the shifting covering of local

obstructions. The enormous philosophical importance of
this mathematics in which wefind ourselves immersed is to a
large extent rooted in its extraordinarily rich conceptual
and technical arsenal, an arsenal built for an increasingly
careful study of flux, shifting, and hiatus. As Corfield has
stressed, we must ' not waste it' [100] .

* * *

The changes and advances in mathematics during the

second half of the twentieth century have been remark­
able. We have seen that these transformations correspond
to a gradual amplification of problematics (the modula­
tion from (A) to (B) [356, 358]) and an extended
capacity for treating, with new technical instruments,
deformations of mathematical quasi-obj ects, the glu­
ing together of the local and the global, geometrical
nuclei of representations, processes of self-reference and
schematization, relative and asymptotic channels, and
nonclassical structural graftings with physics. Behind all
these accomplishments, and others still unmentioned, we
may glimpse the permanent presence of a certain 'general
operativity of the trans-' (figure 15 [347]), in the back­
ground of contemporary mathematics. I t is interesting
to observe that, as we move away from the hackneyed

364
 M A T H E M AT I C S A N D C U LT U R A L C I RC U LAT I O N

rupture associated with 'postmodernism' i n the years

r960-1970, and approach, instead, a web ofentrances into
and exitsfrom modernity, a transversal traversal ofthe modern -
what we could call a sort of 'transmodernity' - the internal
circulations achieved in mathematics seem to anticipate
a natural external reflection in culture.
An apparent rift between modernity and what is unfor­

tunately called 'postmodernity' is marked by the brilliant

work of Deleuze.396 Nevertheless, aside from the rich­
ness of 'postmodernism's' founding documents, many of
the movement's subsequent, 'degenerate'397 theses - like
'anything goes', absolute relativism, the impossibility of
truth, the conjunction of the arbitrary, the dissolution of
hierarchies, or the celebrated death of knowledge, among
other extreme propositions - impede the critical and com­
parative exercise of reason/reasonability/imagination.

396 Among the immense primary and secondary literature surrounding Ddeuze, let us
pick out a few text!; that may be useful for the philosophy of mathematics: P. Mengue,
Gilles Deleuze ou le systbne du multiple (Paris: Kime, i994) emphasizes Deleuze's systematic
thinking of mediations, imbrications, and fluxes, the occurrence of which within
mathematics we have repeatedly emphasized here. L. Bouquiaux et al, Perspective.
Leibniz, Whitehead, De/ew:.e ( Pa ris: Vrin, 2006), studies the problematic of the multipHclty
of points of view ('perspective'), of how they can be partially reintegrated, and how
they can be used to act on the world. Without having drawn upon Leibniz, Whitehead,
or Deleuze in our essay, we have repeatedly taken up this problematic through the
Peircean architectonic, the theory of categories, and the processes of sheafification.
Duffy, Virtual Mathematics, is a collection of articles on Deleuze's philosophy and its
potential effects on the philosophy of mathematics. Collapse: Philosophical Research and
Development, ed. R. Mackay) voL 3, 2007 includes an important series of 'nonstandard'
articles on Deleuze, which, among other things, tackle a potential 'integration' of
differential Deleuzian constellations.
397 'Degenerate' should be understood in Peirce's sense of the term: being of diminished
relational complexity. This is the case with the theses mentioned above, which.flatten
the landscape of thought.

365
 O F C O NT E M P O RARY M A T H E M A T I C S

Leaving aside the traumatic, self-proclaimed POST o f their

alumni, what Deleuze and Foucault teach us is how to
transit, how to enter and exit modernity in many ways and
from many perspectives. What was awoken by the second
half of the twentieth century - an age of the trans-, if ever
there was one - is better called transmodernity, as Rodri­
guez Magda had already proposed in i987.398
Contemporary mathematics, which is impossible to
associate with 'post' -modernity, finds a natural habitat
in the transmodernity in which we are situated. On the one
hand, contemporary mathematics is able to distinguish
valences, relativize discriminately, constitute asymptotic
truths, conjoin the coherent, and hierarchize knowledge,
against the 'degenerate' 'post'-modern theses. On the
other hand, contemporary mathematics is traversed by
incessant osmoses, contaminations, syncretisms, multi­
chronies, interlacings, pendular oscillations, coherent
gluings, and emergences of relative universals, bringing
it together with transmodern processes.399 The richness of

398 R. M. Rodriguez Magda, 'fransmodernidad (Barcelona: Anthropos, 2004). Without

using the term 'transmodern', the works of Garcia Canclini and Martin-Barbero are
consonant with the constant exits from and entrances into modernity that Rodriguez
Magda detects. See N. Garcia Canclini, Culturas hibn'das. Estrategias para entrar y
salir de la modernidad (1989) (Mexico: Grijalbo, 2005), and]. Martin-Barbaro, De los
medios a las mediaciones. Comunicaci6n, cultura y hegemonia ( i987) (Bogot<i: Convenio
AndrCs Bello, 1998). The plural richness of Latin America (and, more generally, of the
Hispanic world, if we indude Rodriguez Magda) has served as a natural brake on the
'postmodern' currents that, curiously, tend to unijonn everything in difference.
399 Compare this situation with the following quotation from Rodrigue1 Magda, for
example: 'Tram.modernity extends, continues and transcends modernity; in it, certain
ideas of modernity return. Some of these are among the most naive, yet the most
universal, of modern ideas - Hegelianism, utopian socialism, rnarxism, philosophies

366
 M AT H E M AT I C S A N D C U L T U R A L C I RC U L AT I O N

contemporary mathematics, in its overwhelming technical

imagination, disposes of a multitude of signs/operators/
mediators for the subtle observation of transit. Indeed,
reversing our approach, if mathematics, as has often been
the case throughout history, can serve as an index for
forecasting the tendencies of an age, then contemporary
mathematics may serve as an introduction to the transmo­
dernity that is now enveloping it. Just as the Renaissance
may have been encrypted in Leonardo's p erspectival
and geometrical machines, as the baroque corresponds
to Leibniz's differential and integral calculus, as classi­
cism was forecast in Euler's serial manipulations, or as
Modernity is just a temple for the visions of Galois and
Riemann, Transmodernity may be encrypted in the techni­
cal plasticity of contemporary mathematics, symbolized,
in turn, in the figure of Grothendieck.
If these associations or 'predictions' turn out to be
more or less correct, we should nevertheless observe that,
unlike the periodizations of the history of art that have
been used in recent centuries - Renaissance, Baroque,
of suspicion, the critical schools / ... J all exhibit chis naivety. Through the crises of
those tendencies, we look back on the enlightenment proj ect as the general and most
commodious frame through which to choose our present. But it is a distanced and
ironic rewrn, one that accepts that i t is a useful fiction. [ ... ] The contemporary zone is
transited by every tendency, every memory, every possibility; transcendent and apparent
at once, willingly syncretic in its 'multichrony'. ! . . . ] Transmodcrnity is the postmodern
without its innocent rupturism ( . . . ). Trans modernity is an image, a series , a baroque
of fugue and sel r·reference, a catastrophe, a cwist, a fractal and inane reiteration, an
entropy of the obese, a livid inflation of information, an aesthetics of the replete and its
fatal entropic disappearance. The key to it is not rhe 'post·', or rupture, but the glassy
transubscantiation of paradigms. These are worlds that penetrate one another and turn
into soap bubbles or images on a screen'. Rodriguez Magda, Transmodernidad, 8-9.

367
 S Y N T H E T I C P H I LO S O P H Y

Classicism, Modernism, and Contemporaneity, epochs

in which the forms of artistic expression are all equally
complex - the 'epochs' of mathematics are tied, by con­
trast, to a very palpable and continuous augmentation ef
complexity throughout their historical evolution. In fact,
'advanced' mathematics, which we have defined so as to
include mathematical technique from Classicism onward
[26-28] , clearly increases in complexity over the
centuries. No doubt, the steadily increasing difficulty efaccess­
ing those forms of mathematical expression, in the degree
it has reached today, blocks a thorough comprehension
of mathematics in its entirety, from the most elementary
to the most advanced. The task of a serious philosophy
of mathematics should nevertheless be to break down
these barriers, and develop a conscientious reflection on
modern and contemporary mathematics. The work of an
art critic who is unaware of everything that has happened
in art over the last hundred and fifty years would be
considered quite poor. The situation should not be any
different in the philosophy of mathematics, and the easy
complacency of reflecting on logic alone (in the style of
the Oxford Handbook efPhilosophy efMathematics and Logic
[101-107]) should be shaken off at once.
Aside from the coincidences that may be observed
regarding epochal delimitations in the arts and mathemat­
ics, the proximity of the creative grounds of the artistic and
the mathematical has been emphasized throughout the

368
 M AT H E M AT I C S A N D C U LT U RA L C I R C U LAT I O N

history o f culture. Pierre Francastel, the great critic and

historian of art, has forcefully pointed out how math­
ematics and art should be understood as the major poles
ef human thought.4°0 Behind those modes of knowledge,
Francastel, for his part, has observed the emergence of
systems and creative webs wherein the real and the ideal,
the concrete and the abstract, and the rational and the
sensible are combined. 4oi
In Peirce's triadic classification of the sciences, math­
ematics are situated on the first branch (1), in the realm of
constructions of possibility. Aesthetics appears in philoso­
phy (2) , where, in the normative sciences (2), it occupies
a primary place (that is to say, in branch 2 . 2 . 1 ) . 'Art' as
such does not enter the realm of the sciences and is not
to be found in the Peircean classification, but it may be

4 0 0 'Art and mathematics are the two poles o f all logical thinking, humanity's major modes
of thought', P. Francastel, La realidadjigurativa (1965) (Barcelona: Paidos, 1988), vol.
1, 24.

401 'From the moment we accept the idea that mathematical and artistic signs respond
to intelleccualized knowledge and not only to simple sense data immersed in matter,
WE a lso admit the jnrervention of a logic, of a system, and notions of order and

.
combination, equivalence, relation, operation and transposition appear before us. [ . . J
Just as mathematics combines schemes of representation and prediction, in which the
real is associated with the imaginary, so the artist brings elements of representation
into confrontation with others that proceed from a problematic of the imagination. In
both cases, the dynamism ofa thought that becomes conscious of itself by expressing
and materializing itself in signs interlaces, overtakes, and envelops the elements
of experience and those of the logic of the mind. l · · · J Just like an, mathematics
possesses the dualistic character by which they both reach the heights of abstraction,
while remaining anchored in the real. It is in virtue of this that both mathematical
and plastic symbolism preserve thejr operative character', ibid., 125-6. The 'dualistic'
character remarked upon by Francastel should be understood as the process of a dual
intermingling of the real and the imaginary, over a relay (60] of interlacing signs. The
mediation of the relay imposes itself on the dualism of the positive and the negative,
of the greater and the lesser.

369
 OF C O N T E M P O RA R Y M AT H E M AT I C S

seen as being very close to a creative materiality of type

3.2.2 or 3.23 - material mediations for engendering sense
(classical art) or action (contemporary art) . In the Peircean
classification of the forms of knowing (and understanding
art here as a vital part of knowing, something we do not
find in Peirce) , mathematics and art likewise emerge as
clear polarities (1 versus 3.2.3) . A vision of the tree from
below thus provides us with a possible transit between
mathematics and art. If, metaphorically speaking, we
situate the tree on the page of assertion of the Peircean
existential graphs and view it from its recto (alpha graphs)
or from its verso (gamma graphs), we may transit between
the various realms of creation, with interstitial passages
between mathematics and art (dotted gamma cuts, singu­
lar points of ramification), but with blockages between
them as well (strict alpha cuts, restrictive delimitations).
The metaphorics of the tree and the graphs has a much
greater depth than may appear at first sight. The tree,
on the one hand, as a triadic tissue, refers us to processes
of iterative construction in culture, which urifold in time
and space. On the other hand, the graphs, as specular
images, refer us to processes of singular vision,Jolding back
on themselves and encoding information. In the weaving
between iteration and deiteration (which are the main
logical rules of the Peircean existential graphs), culture is ·

unfolded and refolded, in a permanent dialogue with its

main modes of creation (arts and mathematics, according

370
 M AT H E M A T I C S A N D C U L T U RA L C I RC U LAT I O N

to Francastel). Th e creative proximities between art and

mathematics, evident in the emergence of inventiveness,
reinforce one another in the general modes of knowing,
from a point of view that is formal, dual and latticial.
A proximity is nevertheless far from an identification.
The creative forms of mathematics and art retain their
differential specificities, and, though the polarities form a
remarkable space of mediations (like the two poles of an
electromagnetic field - as Chatelet reminds us [308]),
they must begin, first and foremost, by folding back upon
one another. The demonstrative realm of mathematics,
cumulative and architectonic (3) , naturally repels the
intuitive, destructive and visionary (1 . 2 ) realm of art.
In this manner, though the modes of creation in both
realms are in close proximity, the quasi-objects at stake
are extremely distinct. We are therefore faced with a very
interesting asymptotic geometry between mathematics and
art: orthogonal 'what?', dual 'how?' , and inverse 'why?'.
The breath of aesthetics permeates mathematical cre­
ativity on at least two levels, as detonator and as reg;ulator.
Referring to the artistic imagination, Valery writes in his
Cahiers: 'Imagination (arbitrary construction) is possible
only if it's not forced. Its true name is deformation of the
memory ofsensations',4°2 and referring to the imagination in
general, he speaks of the 'imaginary magnitudes - SPECIAL

402 P. \'alery, Cahiers 1894-1914 (Paris: Gallimard, 1990), vol. 3, 219 (Valery's emphasis).

371
 S Y N T H E T I C P H I L O S O P HY

efforts when there are displacements or tensions'.4°3 We

have seen how contemporary mathematics systematically
studies deformations efthe representations efconcepts. In that
study, the aesthetic impulse initially occurs as a detonator,
as a buttress (in Peircean firstness) for the elaboration of
a vague image or conjecture, which the mathematician
then submits to meticulous contrasts, through delimita­
tions, definitions, examples and theorems. In turn, in that
contrasting (submitted to forms of Peircean secondness
and thirdness), the aesthetic impulse occurs as a regulator,
as a functor of equilibrium, symmetry, elegance, simplic­
ity. There is a perpetual double circulation of aesthetic and
technical factors in the creative act in mathematics. But
Valery goes on to underscore the central importance of
deformations, displacements and tensions in imagination. We
have, for our part, abundantly emphasized the presence
of these movements in contemporary mathematics, whose
imaginative expressions are thus forged almost symbioti­
cally with the most demanding forms ('SPECIAL efforts') of
the imagination, following Valery.
In mathematical practice itself we can observe the
force of certain aesthetic tensions that, even if they do not
govern the discipline, at least determine the climate of some
of its fragments - counterpointing certain foundational
branches outlined in Hilbert's tree, certain impressions

403 Ibid., 220 (Valery's italics and small caps).

372
 M AT H E M AT I C S A N D C U LT U RA L C I R C U L AT I O N

o f correlative density i n Gromov's 'clouds' [263] . There

exists a sort of aesthetic meteorology, hiding behind the
variability of many mathematical phenomena. Gromov's
cloudiness is typical of contemporary mathematics, and
seems to be incomprehensible - or worse, invisible - from
elementary, or even modern, perspectives. The shattered
tree of Hilbert (complexity, undecidability) is a tree by
turns displaced, deformed, and stretched, with extraor­
dinarily dense nodes (the explosion and penetration of
complex analysis in the most unexpected domains of math­
ematics, for example), whose expansive force detonating
-

and regulating - becomes a new object of study. The visions

of 'cohomologies everywhere' in Grothendieck [146],
of 'groups everywhere' in Zilber [ 256], or ' metrics
everywhere' in Gromov [259], ultimately answer to a
new aesthetic sensibility, open to contemplating the local
variations of ( quasi-)objects through global environments
of information transformation. The aesthetic regulation
that allows the invasion of cohomologies, groups or
metrics be calibrated is decisive.
Turning to particular cases, we can observe a few
complex overlappings between aesthetics and technique
in contemporary mathematics. Many of the examples com­
bine a sort of romantic detonator (recall Langlands's excla­
mation about the 'romantic side of mathematics' [185])
and a regulatory transmodern scaffolding. The interlacing
ifromanticism and transmodernity is perhaps surprising at

373
 O F C O NT E M P O RARY M AT H E MAT I C S

first, but less s o when we notice that many great roman­

tics - Novalis, Schelling and Goethe in particular - have
taken up, with technical instruments far less robust than
the contemporary ones, extensive studies of the trans-. To
start with, two of the major programs of contemporary
mathematics, Grothendieck's motivic cohomology and the
Langlands correspondence, explicitly trace themselves
back to romantic impulses (Grothendieck's 'grand vision'
[163], Langlands's 'mathematics that make us dream'
[186]), and are combined with extensive scaffoldings for
regulating the transmodern deformations and shiftings
in play (EGA [166) , the functoriality associated with
the Langlands group [184]) . In a more restricted form,
Gromov's h-principle [ 263] is the result of an initial,
global, romantic intuition (the synthetic penetration of
the ideas of holonomy and homotopy into differential
realms) and of an extensive subsequent elaboration of
hierarchies and local, analytic conditions under which the
h-principle can be incarnated. The vision of mathematics,
according to Lawvere, meshes perfectly with the bipolar
tension between romanticism and transmodernity; in his
case, the romantic ground corresponds to dialectical intu­
itions, a gazing into abysses [190], while his reflection's
transmodern richness is played out in a multiplicity of
new mathematical quasi-objects (in particular, subobject
classifiers and sheaves in an elementary topos [195]) by
which one can gauge fluxions and deformations between

3 74
 M AT H E M A T I C S A N D C U LT U RA L C I R C U LAT I O N

opposites. Shelah explicitly (and perhaps polemically)

proclaims the primordial role of beauty in his understand­
ing of mathematics [200] , Zilber plunges into the abyssal
hiatus of interlacings between model theory and non­
commutative geometry [258] , Connes underscores the
necessity of knowing the heart of mathematics [226]:
strongly romantic sentiments, potentiated and reoriented
by the various migratory modulations of transmodernity.
The 'continuous increase of potentiality', the summum
bonum of aesthetics according to Peirce,4°4 lies behind all
of these examples. As we hope to have shown in these
pages, contemporary mathematics presents a remarkable
continuous increase efpotentiality and reasonability. I n full
harmony with the summum bonum, the mathematics of
the last few decades amplifies our technical, imaginative
and rational capacities. A sprawling beauty lies in the
work of our great contemporary mathematical creators.
A synthetic vision allows us to link together apparently
distant strata of mathematics and culture, helping us to
break down many artificial barriers. Not only can today's
mathematics be appreciated through epistemic, ontic,
phenomenological and aesthetic modes, but, in turn, it
should help to transform philosophy. And, as Goethe
tells us, we must not forget that ' the most important
thing, nevertheless, continues to be the contemporary,

404 See Barrena, La raz6n creativa. . .

375
 SYNTH E T I C P H I LOSOPHY

because it i s what most clearly reflects itself i n u s , and us

in it' (our epigraph). This is as true of mathematics as
of any other realm of human culture. We hope that this
Synthetic Philosophy efContemporary Mathematics may help
to make up part of the reflection that is demanded - in
honor of the human spirit - by one of the most astonish­
ing adventures of thought in our times.

376
I n dex of N ames
A
Abel, Niels Hendrik 24, 75, 76 Berger, Marcel 26on, 2 6 m
Airy, George Biddcll 231 Bernays, Paul 83n
Alexander, James Waddell 67, 233 Bernoulli, Jacob 81
Alunni, Charles 93n Betti, Renato 317n
Apery, Roger 225n Birkhoff, Garrett 59, 2oon, 253
Araki, Huzihiro 22m Blumenberg, Hans 330, 342,
Ardao, Arturo 331n 342-344
Argand, Jean-Robert 94 Bohr, Niels 318-319, 319
Aristotle 89, 329n Bolyai, Janos 2 6 m
Arnold, Vladimir 216n, 335 Bolzano, Bernard 79n, 2 4 9 , 25rn
Artin, Emil 85, l81n, 185n, 2 2 7n, Bombieri, Enrico 227n
303 Borel, Armand 42n, So, 91, 249,
Atiyah, M ichael 43, t36, 137, 250, 303
207-213, 215, 2 21 n, 2 2 2 , Borel, Emile 42n, So, 91, 249,
2 23, 275, 3 2 1 , 335, 341, 250, 303
351, 363 Botero, Juan Jose 16
Bouquiaux, Laurence 365n
Bourbaki 53, 76, n, 96
B Boussolas, Nicolas-Isidore 279n
Brentano, Franz 329n
Badiou, Alain 12n, 71, 88-90, 91, Breuil, Christophe 187n
95, 307n, 312n, 3 2 9 Brouwer, Luitzen Egbertus Jan
Baez, J o h n 9 8 , 125 57, 80
Bailey, TN. 184n Brown, Lawrence Gerald 223
Banach, Stefan 31, 135n, 221n, B rown, Ronald 9 8
250, 252 Burgess, John 103, 105
Barrena, Sara 349n, 375n
Barwise, Jon 3 2 2n
Batt, Koelle 93n c
Baudelaire, Charles 188n
Baxter, Rodney James 2 24n Caicedo, Xavier Im, 16, 283n,
Bayer, Pilar 178n 307, 362
Beltrami, Eugenio 220 Canclini, Nestor Garcia 366n
Benabou, Jean 325n Cantor, Georg 57, So, 89, 9 2 ,
Benacerraf, Paul 1 1 , 91, 104-105, 102, 121, 194, 322n
279n, 2 81, 2 9 7 Cardona, Carlos 1 6
Benjamin, Walter 3 7 Carlson, James 2 27 n

377
 I N D E X O F N AM E S

Cartan, Henri 135, 158n, 161-162,

D
2 1 01 2 1 1 n
Cartier, Pierre 15on, 2 24-2 26, 351 Broglie, Louis de 94
Cassin, Barbara 157n, 277n Dedekind, Richard 9 2 , 98, 140,
Cassou-Nogues, Pierre 1 6 303
Cauchy, Augustin Louis 74, 75, Deleuze, Gilles 37, 89, 93, 365,
76, 81, 94, 208n, 2 1 1 366
Cavailles, Jean 5 8 , 67n, 8 6 , 273, Deligne, Pierre 41, 76, 137, 138,
27911 i42, 143, 149
Cayley, Arthur 265 Demopoulos. William 103
Cech, Eduard 165 Desanti, Jean Toussaint 156
Cezanne, Paul 338 Desargues, Girard 79
Chang, Chen Chung 253 Descartes, Rene 89, 102, 318
Chatelet, Gilles 71, 93-95, 27911, d'Espagnat, Bernard 2311
308, 30911, 3 1 2 , 349, 371 de Lore11zo, Javier 17, 2911, 71,
Chern, Shiing-Shen 18m, 20911 76-78, 8311, 339
Chevalley, Claude 42n, 53 de Tte11ne, Andre 17411
Chihara, Charles 10, 103 Detlefsen, Michael 103
Chong, CT. 177n, 179n de Vries, Gustav 215n
Clark, Peter 103 Diamond, Fred 187n
Cohen, Paul 88, 89, 9 2 , 106, 201, Dieudonne, Jean 42n, 80, 135,
207 136n, 15on, 1 6 2 , 167, 168,
Colmez, Pierre 158 169, 256n, 341
Connes, Alain 43, 45, 137, 166, Dirichlet, Johann Peter Gustave
171, 21611, 218, 2 21 - 2 2 9 , Lejeune 18m, 185n, 220
2 3 3 , 2 3 6 , 257n, 258, Douglas, Ronald George 2 23
26on, 264, 2 8 2 , 303, Dreifus, Claudia 214
3 1 2 , 321, 335, 342, 360, Drinfeld, Vladimir 42, 137,
363, 375 2 23 - 2 2 4n, 259n
Conrad, Brian 187n Duffy, Simon 52n, 93n
Consani, Caterina 2 2 8 Durand, Stephane 186n, 187n
Cook, Roy 103 Dwork, Bernard 138
Corfield, David 911, 24, 71, 87, 97,
97-100, 99, 107, 126n,
184, 364 E
Courti11e, Jean-Fra11�ois 27711
Cruz, Alexander 17 Easton, William 88
Ehresmann, Charles 53, 67n
Eilenberg, Samuel 42n, 67n, 1 8 8 n
Einstein, Albert 1 4 0 , 1 6 6 n , 2 7 0
E isenstein, Gotthold 1 8 5 n , 2 19
Elek, Gabor 26m, 264n

378
 I NDEX OF NAME S

Engels, Friedrich 19on

Euler, Leonhard 24n, 27, 72, 74,
G
81, 225n, 227n, 280, 367
Galileo 232
Galois, E variste 21, 27, 33, 38, 40,
53n, 62, 70, 71, 76, 79, 8 2 ,
F
8 4 , 96, 102, 136n, 140,
143n, 15on, 155, 168, 176,
Fadeev, Ludwig 219 181, 182, 184, 2oon, 224,
225, 233, 272, 303, 353,
356, 357, 367
Faltings, Gerd 41, 137, 257n

Gauss, Carl Friedrich 24n, 27, 79

Faraday, Michael 94, 309n
Gelfand, Israel 42n, 21!
Feferman, Solomon 103, 307n,
Girard, Jean-Yves 45, 314-315,
322n
335-336
Fermat, Pierre de 2 1 , 24, 56, 62,
Glivcnko, Valeri 59
76, 79, 176, 187, 339
Godel, Kurt 16, 68, 75, So, 88, 89,
Ferreir6s, Jose 16
91, 92, 194, 196, 2om, 251,
Feynman, Richard 225, 230-234
334, 361
Field, H artry 10
Goethe, Johann Wolfgang vii, 3,
Fields, John Charles (Fields
36n, 374, 375-376
M edal ) 147, 176, 183,
Goldstein, Catherine 226n, 227n
207, 210, 2 1 1 , 229, 257
Grabiner, Judith 83
Filmore, Peter Arthur 223
Grassmann, Hermann 94, 95,
Floyd, Juliet 103
308, 309
Focillon, Henri 6m
Gray, Jeremy 107n
Foucault, M ichel 366
Gray, John 285
Fraenkel, Abraham 28, 31
Gromov, Mikhael 43, 234n,
Fra·isse, Roland 322n
259-265, 2 ]3, 282, 287,
Francastcl, Pierre 37, 6m, 369, 371
297, 298, 299, 300, 303,
Fredholm, Erik Ivar 208
305, 312, 320, 321, 323,
Frege, Gottlob 324
334, 342, 351, 360, 362,
Freyd, Peter 43, 44, 69, 125, 143,
373, 374
195n, 242-248 , 282, 296,
Grothendieck, Alexander 27, 42n,
303, 321, 336, 342, 360
43, 46, 75, 96, 99, 102,
Frey, Gerhard 187n
176, 178, 182, 188, 195n,
Friedman, Harvey 31, 248, 252n
201n, 202n, 203, 205,
Fuchs, Immanuel Lazarus 181
207n, 210, 211, 2E?, 216n,
Fukaya, Kenji 234n
218, 222, 224, 225, 226n,
227n, 233n, 235, 240,
243, 244, 253n, 254, 255,
259n, 26on, 263, 270,

379
 I N D E X OF NAME S

271, 272, 275, 280, 281, Hirzebruch, Friedrich 210

2 83, 285n, 286, 296, 297, Hodges, Wilfrid 254
298, 303, 306, 310, 313, Hodge, William Valance Douglas
320, 323, 324, 327, 330, 144
331-333, 3% 339, 340 , Holton, Gerald 12on, 317n
341, 351, 360, 361, 363, Hopf, Heinz 99, 223
367, 373 . 374 Houze!, Christian 285n
See also: Chapter 4, 133-172 H rushovski, Ehud 253, 256-258,
Grothcndieck, Hanka 134 274 , 333
Hugo, Victor 143n
Husserl, Edmund 87, 344-345
H

Haack, Susan n9
Hadamard, Jacques 327 I
Hahn, Hans 3 1 , 250, 252
Iagolnitzcr, Daniel 211n, 229n
Hale, Bob 103
Hallett, Michael 107n
Hamilton, William Rowan 94
Ham, Lorena 155n J
H ardy, G. H. 9n, 24n, 126n
Jackson, Allyn 134n
Harrington, Leo 248n
Jacobi, Carl Gustav Jacob 76, 171,
Harris, Michael 1 8 2
185n
Hasse, Helmut 1 8 1 n , 185n
Jane, Ignacio 103
Hausdorff, Felix 322n
J anelidze, George 70
Hayford, Harrison 258n
Jones, Vaughan 233n
H ecke, Erich 181n, 185n, 227n
Jordan, Camille 303
Heidegger, Martin 54, 56
Joyal, Andre 23on
H eijenhoort, Jean van 247n
Joyce, James 50
Heine, Heinrich Eduard 249, 250
Heisenberg, Werner 2 2 1
Hellman, Geoffrey 1 0 , 103,
106-107, 276, 277
K
Herbrand, Jacques 52, 53n, 61,
6':!, 68, 181 Kant, Immanuel 87, 89, 102
Heydorn, Wilhelm 134 Keisler, H. Jerome 253
Heyting, Arend 78n, 244n, 316 Keller, Joseph 217, 2 2 1 n
Hilbert, David 21, 27, 40, 53n, 61, Kitcher, Philip 71, 81-82
64, 76, 79, So, 96, 181n, Kleene, Stephen Cole 249
i85n, 218, 2 2 m , 223, Klein, Felix 256
2 28n, 247, 251, 261n, 263, Kline, Morris 71, 80-81
273, 300, 304, 306, 362, Knapp, AW. 184n
372, 373

380
 I NDEX O F NAMES

Kodaira, Kunihiko 232 Lawvere, William 43, 70, 78,

Kolmogorov, Andrei 216n 89, 96, 125, 14m, 143,
Ki'mig, Julius 31, 250 188, 188-195, 243, 244n,
Kontsevich, Maxim 43, 137, 290, 316, 321, 334, 341,
17m, 215-216, 224n, 360, 374
225n, 226n, 229n, Lax, Peter 43, 213-221, 2 75, 2 8 2 ,
23on, 23m, 232-233, 2 9 7 ' 3 2 1 , 335, 3 4 1 , 360,
234-236, 259n, 26on, 363
2 8 2 , 303, 3 1 2 , 3 2 1 , 335, Lazard, Michel 285n
342, 360, 363 Lebesgue, Henri 135
Korteweg, Diederik 215n Lefschetz, Solomon 85, 285n
Kreimer, Dirk 2 25 Leibniz, Gottfried Wilhelm 4, 27,
Kripke, Saul 283 79, 8 1 , 89, 232n, 2 95 , 318,
Kruskal, Martin 215n 365n, 367
Kuhn, Thomas Samuel 300 Lenin, Vladimir Ilyich 19on
Kummer, Ernst 185n Leonardo da Vinci 367
Leong, Y. K. 177n, 179n
Leray, Jean 162n, 285n
Levine, Marc 152
L Lie, Sophus 77, 166, 177, 179, 1 8 1 ,
184, 222n, 223, 225n,
Lafforgue, Laurent 183
232n, 235, 265, 272, 303
Lafontaine, Jacques 26on
Lindstrom, Per 44, 66, 274, 3 2 2 n,
Lakatos, Imre 71, 74-76, 81, 98
323
Lambek, Joachim 125
Lipschitz, Rudolf 2 64
Langevin, Remy 26on, 262n,
Llull, Ramon 318n
273n
Lobachevski, Nikolai 79
Langlands , Robert 43, 180,
Lochak, Pierre 137n
180-188, 194, 2 2 4, 248,
Lolli, Gabriele 8-9n, 9
297' 303, 310, 321, 341,
Lovejoy, Arthur 19 4
351, 360, 3]3 , 374
Lowenheim, Leopold 1 9 8 n , 322n
Lang, Serge 257
Lozano, Epifania 1 7
Laplace, Pierre-Simon 208n,
Loz, Jerzy 284
220
Lusin, Nikolai 91
Laumon, Gerard 1 8 2
Lautman, Albert 9 n , 10, 29, 4m,
49, 52-71, 83n, 88n,
93n, 111, 15rn, 156n, 177, M
206, 220, 22m, 223n,
246n, 275, 2 7 9, 296n, MacBride, Fraser 103
307n, 329, 332, 347, Mackay, Robin 17, 365n
353, 354 Mac Lane, Saunders 67n, 71, 79,
84-85, 188n, 246n

381
 I NDEX OF NAM E S

Maddy, Penelope 10, 71, 90-92, Nocther, Emmy 303

95, 103, 105 Novalis 193, 363, 374
Makkai, Michael 1 00
Mallarme, Stephane 89, 338
Mao Tse Tung 19on 0
Marcolli, Mathilde 2220, 225,
Olshanskii, Alexander 265n
228n, 229n
Oostra, Arnold 17
Marey, Etienne-Jules 410, 239n,
Oresme, Nicolas 9 4
24on
Ortiz, Fernando 206n
M argulis, Gregori 42n
Otte, Michael 317n
Martin, Alejandro 17, 85n, 366n
Martin, Donald 91, 92, 105
Martin-Barbero, J esus 366n
Maxwell, James Clerk 94, 309n
p
McCarty, David 103
McLarty, Colin 134n Pacholski, Leszek 254n
Melville, Herman 258n Palombi, Fabrizio 85n, 345n
Mengue, Philippe 365n Pansu, Pierre 26on
Merleau-Ponty, Maurice 93, 1510, Panza, Marco 16, 317n
17 7 n, 277-278, 280, 31 7n, Parker, Hershel 258n
320, 330, 337-339, 344, Pascal, Blaise 4, 27, 73
363 Patras, Frederic 71, 95-96
Milnor, John 420 Pavlov, Boris 219
Minkowski, Hermann 232 Peano, Giuseppe i96, 2y:m, 257n
Monet, Claude 50 Peirce, Charles Sanders 5, 10, 37,
Monte!, Paul 61 410, 44, 46, 73, llO- ll8,
Mardell, Louis Joel 257n 144, 149n, 157n, 166n,
Morel, Fabien 152 i74n, i77n, 286, 289, 291,
Morley, Michael 198, 2oon, 253n, 297, 299, 304n, 325 n ,
256n 348 -349 , 3 63, 3 65n,
Moschovakis, Yiannis 91, 9 2 369-370, 375
Moser, Jurgen 216n Perelman, Grigori 44, 5 6 , 137
Mostowski, Andrzej 2 49 Perry, Roberto 155n
Musil, Robert 50 Petitot,Jean 274-275, 295, 296n,
307n, 31 2n, 317n, 336, 358
Phillips, Ralph 214, 218-220, 275,
282, 297, 360
N Picard, Emile 285n
Pier, Jean-Paul 253n, 26on, 336n
Natorp, Paul 329, 344
Planck, N!ax 205, 225n, 232
Nelson, Edward 325n
Plato 4, JO, 54, 58, 63-64, 75, 88n,
Newton, Isaac 8 1, 217
89, 22 6n, 279, 329-330
Nicholas of Cusa 342

382
 I N D E X O F NAM E S

Poincare, Henri 21, 44, 56, 67, 127n, 135, 138-140, 143n,
76, 80, 181, 219n, 220, 150, 155, 168, 179, 18m,
231, 233n, 261, 327 205, 206, 208n, 210, 2 1 1 ,
Poisson, Simeon-Denis 94, 230, 2 1 8 , 2 2 1 n, 223, 224, 2 25n.
232, 232-233, 233 227n, 227- 2 2 8 , 231, 234,
Poizat, Bruno 253n, 254n, 255, 236, 250, 356, 357, 36 7
256 Rivoal , Tanguy 225n
Polischuk, Alexander 235 Robinson, Abraham 65, 75
P6lya, George 71 , 72 - 74 , 79, 8 1 , Roch, Gustav i35, 15on, 155, 208,
83n, 2 2 8n 210- 2 1 1
Pontrj agin, Lev 165n Rochlin, Vladimir Abramovich
Popper, Karl 74 260
Posy, Carl 103 Rodriguez Magda, Rosa Maria
Prawitz, Dag 103 366-367n, 367
Prevost, Jean 327n, 3 2 8 n Rosen, Gideon 103
Proust, ,\1arcel 5 0 , 1 60, 338 Rota, Gian-Carlo 71, 85-88, 107,
Putnam, Hilary 104-105, 279n 15m, 177n, 287n, 288n,
Pythagoras 23, 2 2 6 3 2 7 , 330, 343, 345, 346,
358
Rothko, Mark 50
Russell, Bertrand 62, 75, 99, 240

Q
Quillen, Daniel 2 1 on
Quine, Willard Orman 103, s
105, 169
Salanskis, Jean-Michel 325n
Scedrov, Andre 243n
Schelling, Friedrich Wilhelm
R Joseph 374
Schlegel, Friedrich 194
Ramanujan, Srinivasa 219n
Schmidt, Erhard 210
Rapoport, Michael 1 8 2
Schneps, Leila 137n
Raussen, Martin 1 7 7 n , 2 1 2 n,
Schwartz, Jacob 85
216n
Schwartz, Laurent 78, 135
Rayo, Agustin 103
Seidel, Paul 235
Reidemesteir, Kurt 233n
Serre, Jean-Pierre 27, 43, 158-159,
Resnik, Michael 13n, 1 5, 103
161, 162, 164n, 176n,
Rham, Georges de 144
176-180, 18m, 202, 210n,
Ribet, Kenneth 187n
2 2 2 n , 303, 3 2 1 , 332, 350
Riemann, Bernhard 21, 24n,
27, 40, 5 3n, 58, 64,
Serres, M ichel 178n, 338
74, 79, 82, 84, 98, 102,
Servois, Julien 329n

383
 I N D E X O F NAM E S

Shabel, Lisa 102 Tanselle, G. Thomas 258n

Shapiro, Alexander 134n Tao, Terence 260
Shapiro, Stewart 6, 9n, 10, 101n, Tappenden,Jamie 107n
102-105, 277, 280, 288, Tarski, Alfred 75, 194, 253, 265n,
296, 302 274
Shelah, Saharan 27, 35, 43, 45, Tate, John 226
105, 196, 196-203, 197, Taubes, Clifford Henri 229n
199, 200, 201, 202, 21on, Taylor, Richard 182, 187
253, 261n, 274, 282, 311, Teichmiiller, Oswald 171, 205, 218,
313, 321, 323, 334, 341, 225, 233n, 272, 275, 280,
360, 375 297, 306, 323, 360
Shimura, Coro 1 8 7 Tennant, Neil 10, 103
Siegel, Carl Ludwig 181 Thom, Rene 42, 96, 138, 146, 183,
Simpson, Stephen 3 1 , 43, 248, 209n, 236, 283n, 291, 316,
248- 253, 249, 250, 251, 325n, 362
252, 273, 296 , 321, 342 , Thuiller, Jacques 61n
360 Thurston, William 42n
Sinaceur, Hourya 325n Tierney, Myles 195n
Singer, Isadore 207, 211, 212n, 223 Todd, John Arthur 209n
Skandalis, George 2 26n, 227n Toeplitz, Otto 208n
Skau, Christian 177n, 2 1 2 n, 216n Trocmc, Andre 134
Skolem, Thoralf 198n, 32w Truesdell, Clifford 189n
Skorupski, John 103 Turner, Joseph Mallard William
Smale, Stephen 42n, 207n 50
Soibelman, Yan 234n, 235n Tymoczko, Thomas 71, 8 2- 8 4
Solovay, Robert 91
Soule, Christophe 259n
Spinoza, Baruch 89
Steenrod, :'\/ orman 285n u
Steiner, Mark 102, 103, 107n
Ulam, Stanislaw 85
Stokes, George Gabriel 234
Stone, Marshall 80, 98
Street, Ross 23on
Stuhler, Ulrich 1 8 2
v
Suslin, Andrei 210n

Valery, Paul 327-329, 37 1-372

Vassiliev, Victor 233n
Vaz Ferreira, Carlos 331, 340
T Verdier,] ean-Louis 235n
Villaveces, Andres 16, 85n, i97n,
Takagi, Teiji 1 8 1 n
i99n, 202n, 253n, 258n
Taniyama, Yutaka 187

384
 I N D E X O F NAM E S

Voevodsky, Vladimir 137, 147, Zariski, Oscar 139, 140, 143, 256 -
259n, 298n, 333 259, 310, 333, 360, 362
Von Neumann, John 59, So, 217, Zaslow, Eric 235
221, 222, 223, 341 Zermelo, Ernst 2 8 , 3i, 92, 1 2 1 ,
324-325
w Zilber, Boris 43, 195, 199, 253-260,
263-265, 271, 272, 274,
Weierstrass, Karl 249, 251
275, 2 8 2 , 297, 303, 312,
Weil, Andre 4i, 42n, 76, 78, 94,
3 2 1 , 333-334, 342, 351,
98, 138-139, 1 42 , 143, 158,
360, 3 73, 375
159, 167, 178, 180-181, 187,
227n, 2 2 8 , 333, 351
Weir, Alan 103
Wetz, FranzJosef 343n
Wey!, Hermann 80
Whitehead, Alfred North 365
Whitehead, John Henry Constan-
tine 201, 21on
Wilder, Raymond 13n, 15, 71,
79-80
Wiles, Andrew 41, 56, 187n
Witten, Edward 215n, 230, 231,
233n, 234n, 261, 262, 335
Witt, Ernst 178
Wittgenstein, Ludwig 16, 31, 90,
103, 216
Woodin, Hugh 92, 105
Wright, Crispin 103

y
Yang, Chen Ning 2 24n
Yoneda, Nobuo 69, 141n,
246-247, 29i, 298n

z
Zabusky, Norman 215
Zalamea, Fernando 29, 4i, 283,
304, 317, 325

385