Professional Documents
Culture Documents
Ian Hacking
Scientific Reason
Copyright © 2009 by the Institute for Advanced Studies in Humanities and
Social Sciences, National Taiwan University, Taipei, Taiwan (R.O.C.)
All rights reserved.
Ian HACKING
GPN 1009703212
ISBN 978-986-01-6115-1 (Hardcover)
This book may not be reproduced, in whole or in part, or transmitted in any form or by any means,
electronic or mechanical, including photocopying and recording, or in any information storage or
retrieval system without the prior written permission of the Institute for Advanced Studies in
Humanities and Social Sciences, National Taiwan University.
Contents
Editors' Foreword IX
Preface XV
Bibliography 158
Index of Subjects 168
Index of Names 173
Editors' Foreword
Jeu-Jenq Yuann,
Associate Professor, Department of Philosophy,
National Taiwan University and
Der-Lan Yeh,
Associate Professor, Department of Foreign Languages and Literatures,
National Taiwan University
Taipei, Taiwan,
July 2008
Preface
December 2007
Lecture 1
3 Sir Herbert Butterfield wrote one of the first textbooks of history of early modem
European science, based on lectures he got up for science undergraduates in
Cambridge University in 1948. The Origins of Modern Science, 1300-1800, London
and Cambridge: Deighton Bell, 1949. His chief authority was Alexandre Koyré. But
his fame is for political and social history of England; he despised "Whigs"—of
whom the philosopher cum historian David Hume is the best example —who wrote
about the past as a series of events that made sense chiefly in that they led on to and
validated the glorious present.
ON THE HISTORICAL ROOTS OF SCIENTIFIC REASON / 5
3. "TheEuropean Tradition"
The words here, with which I do the turning, are objects and
methods of reasoning. They are, in themselves, anodyne. The
objects with which the mathematical style concerns itself are
often called, by analytic philosophers, abstract objects, such as
numbers, shapes, and groups. The objects with which the
taxonomic style concerns itself are, for example, the species and
genera of systematic biology, not mere classifications of living
Thus Crombie in that talk I heard in 1978. He had not quite fixed
on his terminology of "styles of scientific thinking"—as you see,
he at first spoke of methods, rather than styles, but I shall keep
the two words distinct, in line with his later thought that styles of
scientific thinking are constituted by methods of reasoning and
objects of inquiry.
There is a more profound difference between this inaugural
paragraph of 1978, and Crombie's big book of 1994. Here he
speaks of late medieval and early modem Europe—roughly
speaking the time leading up to what other historians of
European science have long called the scientific revolution. In the
final three volume magnum opus, we have a tale going back into
ancient Greece where, for Crombie, a set of social institutions
created a cast of mind in which what we call the sciences began.
The massive work of 1994 is organized around the six styles, each
of which is presented as stemming from ancient antecedents. In
1978, as the above passage shows, he really thought of his styles
of scientific thinking as taking definitive shape much later than
that.
My chief innovation, in connection with Crombie's list, will
be to urge that his long-term styles of 1994 are punctuated with
what I call crystallizations, in such a way as to emphasize the
early modem period in European history. This may diminish a
certain tension between his earlier and his later discussion of
styles of scientific thinking.
Before becoming analytic, or worrying about the long or the
short term, we should honour our immediate take on the list of
six. Each item simply feels different, looks different, and may well
be different. Each may reflect the use of a different group of skills
for which the human mind is well suited.
Crombie's list of six is a template, that is, a convenient pattern
to follow. There are innumerable ways to carve up something so
extended as thousands of years of many civilizations of evolving
"sciences." There are many possible frameworks of analysis:
think only of Thomas Kuhn's paradigms, Imre Lakatos's research
programmes, Gerald Holton's themata-—he list goes on. The
value of any division must lie in how one uses it. None is an
ultimate and definitive framework in its own right.
I have sometimes tried to construct precise definitions of a
style of scientific thinking, which would validate Crombie's
template. That is a misguided enterprise. It is better to see what
one can do with his idea. A form of Occam's maxim provides a
good rule of thumb: the list of styles of scientific thinking should
not be enlarged beyond necessity.
Every style enumerated by Crombie is a living, evolving,
organism. He used his template as a way to organize the history
of science. He has been amply criticized, and some of the
criticisms strike at the heart of his historiographic project.10 I
refrain from criticism because I am not interested in his catalogue
for its own sake. My aim is rethinking the entire structure of
scientific reasoning from what I call a Leibnizian point of view. It
is a view that, as we shall see, can also be called anthropological
in something like the original sense of Kant's Anthropologie. I
mean a study of the human species, its innate powers, and its
6. Crystallization
Crombie's vision of the history of the European sciences
favours continuity. My instinct is exactly the opposite. I like to
14/ SCIENTIFIC REASON
Order of Things (Les mots et les choses: une archéologie des sciences humaines,
Paris: Gallimard, 1966.) That in tum goes back to the influential writings of Gaston
Bachelard (1884-1962), who took up a lot of space in Paris bookstores during the
1960s.
14 On the specific question of revolutions in the history of probability, see my paper,
"Was There a Probabilistic Revolution 1800-1930?", in Lorenz Krüger et aI., (eds.),
The Probabilistic Revolution, Cambridge: M.LT. Press, 1987, vol. I, pp. 45-58.
ON THE HISTORICAL ROOTS OF SCIENTIFIC REASON /17
history early in the 20th century.l5 There was a great vogue for
the word in Germany, including its compound Denkstil,
thought-style, or style of thinking. It figures in Oswald Spengler's
analysis of the decline of the West, published before and after the
First World War. It was very important, for example, to Karl
Mannheim's approach to sociology.16 Among the positivist
philosophers, we find Rudolf Carnap speaking of styles of
thinking, but it is not central to his analysis.17 The Nazis made a
great play with a specifically Jewish thought-style.l8
The crescendo of talk about thought-styles—in Mannheim,
Carnap, and the Nazis alike—occurred in the 1930s. That was the
decade of what we now regard as an enormously important book
by Ludwik Fleck, translated as Genesis and Development of a
Scientific Fact. That title omits the German subtitle, Introduction to
the theory of the thought-style and the thought-collective.19 Fleck's
idea of a Denkstil differs profoundly from Crombie's notion of a
style of thinking. Fleck is concerned with a way of thinking, a
way of finding out that is current in a specific community, a
thought-collective, at a definite time, and which, evolves, mutates,
or dies in a brief span of time. The thought-collective is a good
8. Self-authenticating styles
So far, so simple. Now I forewarn of the radical. I pass from
the historical to the metaphysical, from historical anthropology
to philosophical anthropology, and from bland description to an
attempt to undo ontological debates.
I contend that when it crystallizes, a style of scientific
thinking introduces new objects, and new criteria for the truth or
falsehood of statements about those objects. A style, with its
specific methods of reasoning, does not answer to any criteria
except its own. It is not good because it helps find out the truth in
some domain. It itself defines the criteria for truth—telling in its
domain. Thus in a certain sense each style is autonomous and
"self-authenticating". That certainly sounds radical, and that is
what I mean.
Few such surprising doctrines are all that new. In the heyday
of logical positivism, Moritz Schlick coined the slogan, "the
meaning of a statement is its method of verification." That was
quickly revised and then abandoned, but has more merits than
are usually assigned to it. Schlick meant a method of verification
appropriate to an individual statement, rather than very general
kinds of method. Without embracing his strict verificationism,
we assert that until there are methods of reasoning that bear on
the truth or falsehood of a scientific statement, the question of its
truth or falsehood does not arise. Meaning, as simplistic forms of
analytic philosophy have insisted, demands the possibility of
applying truth conditions. Taking advantage of Bernard
Williams' "truthfulness" that I shall soon explain, it can be said
that scientific statements come up for grabs, as true or false, only
when there comes into play a method for reasoning about their
truth.
Most ordinary statements are not like this. They have truth
22 / SCIENTIFIC REASON
9. Objects
Styles of scientific thinking introduce their own distinct class
of objects. Think of the abstract mathematical objects ('platonist'),
of the unobservable theoretical entities at the centre of the recent
debates about scientific realism, or of systematic biology with its
taxa. Each style is specific to it its own domain, but only because
it introduces the objects peculiar to that domain. It does not
create them—to say that would be foolish—but they have no
24 About 1985 Nelson Goodman said in public, "Ian, do you mean to suggest. .. "
[that the dinosaurs etc.] I said, no! I like it that the very first time this question was
posed to me, it was by the greatest pragmatist of his day, the greatest nominalist of
all time, the very man who is often taken to suggest, what he himself thought was
manifestly absurd, namely the ultimate reductio ad absurdum of any foolish version
of his own sound ideas.
ON THE HISTORICAL ROOTS OF SCIENTIFIC REASON / 23
25 Lorraine Daston and Peter Galison, Objectivity, New York: Zone Books, 2007.
26 Or a little before, for example Isaac Watts, Logic, London, 1724: "Objective
certainty, is when the proposition is certainly true in itself; and subjective, when we
are certain of the truth of it" (II. ii. § 8.)
27 Pierre Bourdieu, Méditations pascaliennes, Paris: Seuil, 1997; Pascalian
Meditations, Stanford University Press, 2000.
28 For my own reflections on science studies in connection with Bourdieu, see Ian
Hacking, "La science de la science chez Pierre Bourdieu," in J. Bouveresse et Daniel
Roche (eds.), La Liberté par la connaissance. Pierre Bourdieu (/930-2002), Paris:
Odile Jacob, 2004, 147-162.
ON THE HISTORICAL ROOTS OF SCIENTIFIC REASON / 25
11. Society
A style of scientific thinking answers to no criteria but its own.
This does not mean that it needs no support, as if it were some
sort of free-standing inhabitant of what Popper called the Third
World, a reincarnation of the headier postulates of Plato and
Frege. It is people who think. People reason. People find out.
People need to eat, and even if they are rich, they need people to
listen to them. Styles are enabled by institutions. Geoffrey Lloyd's
comparative study of ancient science in Greece and China,
mentioned earlier, attends closely to the institutional settings,
that allowed various of what he calls "styles of inquiry" to
flourish in one or the other society.
Styles of scientific thinking in the European tradition are just
as much social practices as Fleck's Denkstile. They are less local
and more enduring in part because they are built on
fundamental cognitive capacities. If they become extinct it will
not be by refutation, but by being abandoned. Or they can
become moribund. There is no better example than Greek
mathematics, which came to a full stop after Archimedes. People
wrote commentaries in Greek, Arabic and Latin, but no new
maths was created for a millennium. It was not so different in
China.
I have asserted that styles of scientific thinking are in a
certain sense self-authenticating. Perhaps this implies that they
cannot be undone "internally"; they cannot be called into
question from within, for they furnish their own canons of
correctness. But we can just lose interest, for all sorts of reasons.
Maybe people quit creating mathematics after Archimedes
because no one had any new ideas. Or the next generation just
got bored with those old fogeys—there was no pay-off anyway.
The abandonment of a style of scientific thinking could arise for
the most banal of reasons.
Herman Hesse's The Glass Bead Game is a parable for a society
built around an intellectual activity, wholly viable, maintained by
complex institutions.31 It is not unlike some aspects of the
mathematical sciences, but the core is music and it begins with
the art of fugue. In the end, the Magister Ludi, the veritable
Archimedes of the game, decides he must, in all humanity, leave
the institutions that make the Glass Bead Game possible. From
our point of view, such a system, described by a great and
inventive novelist, is just human enough to be fascinating, and
yet has the air of absurdity.
Hesse well knew that some of his contemporaries regarded
mathematics as a purely formal game. Nowadays we pay not
only pure mathematicians but also string theorists out of the
public purse. So far, the stunning conceptual innovations of
string theory have no empirical or testable consequences
whatsoever. There is always a little worry, in such fields, that the
purse will dry up, and in a generation or two the activity will
31 Das Glasperlenspiel, Zurich: Fretz and Wasmuth, 1943. The Glass Bead Game:
A Tentative Sketch of the Life of Magister Ludi Joseph Knecht, together with
Knecht's Posthumous Writings edited by Herman Hesse, London: Jonathan Cape,
1970.
28 / SCIENTIFIC REASON
come to an end.
Here are a few sentences from another source that you would
not expect. J. M. Coetzee, as it happens like Hermann Hesse, won
a Nobel Prize for literature. His Tanner Lectures on Human
Values present a lecture about animals by the Australian novelist,
Elizabeth Costello. She in turn supplements a lecture imagined
by Kafka in 1917. It is by Red Peter, an ape who has learned to
talk. We read that "seen from outside, from a being who is alien
to it, reason is simply a vast tautology. Of course reason will
validate reason as the first principle of the universe—what else
should it do. Dethrone itself? Reasoning systems, as systems of
totality, do not have that power."32 But maybe Costello could
persuade everyone to stop thinking that way.
From time to time there is, in Richard Bernstein's phrase, a
rage against reason.33 One such time was the late 1960s, when
science was seen as a tool of capitalist colonial war. Young rebels
tried to sack places of research. That tactic was foolish but right.
If you want to criticize a style of thinking, you cannot do it on its
own terms (for then you are merely thinking according to that
style, trying to do it better). You must destroy its institutional
base. For very different reasons, the cultural revolution of the
same epoch successfully destroyed a generation of Chinese
scientists. Reason was hated in China and America at about the
same time. Perhaps science could have come to a full stop,
forever.
I do not myself believe that is an option, because I suspect
that when people have uncovered ways of finding out, they will
continue to use them so long as they can do so. Only coming to
think, e.g., that there is nothing to do, after Archimedes, might
One thing I shall not consider, however, is the history of the concept
of truth, because I do not believe there is any such history. The
concept of truth itself—that is to say, the quite basic role that truth
ON THE HISTORICAL ROOTS OF SCIENTIFIC REASON / 31
Truth, then, has no history, beyond the fact that it is coeval with
the emergence of linguistic structures to convey information.
This conception is Aristotelian and Tarskian. The adjective 'true'
has many uses, but truth is a formal concept, essential to
semantics but with no semantics of its own. Thus it spans all
informative discourse, and has no genealogy.
Aristotle: "To say that that which is the case, is the case, and
that which is not the case, is not the case, is to say the truth."35
That is a blank, formal, assertion, conveying in passing the
fundamental fact that the adjective "true" primarily applies to
what is said, was said, or can be said. There is undoubtedly a
history of when human beings began to talk, to say things
informatively, to make what we can recognize as assertions. But
there is no further history of truth than that.
I read Aristotle's maxim as an early version of Tarski's equally
formal semantic theory of truth. Its scheme, " s is true if and only
if p", makes as plain as Aristotle did that the adjective applies to
sentences. The somewhat overblown theory of metalanguages
derives from that trifling grammatical fact. Tarski himself wrote
that his semantic theory appeared to be consistent with, and even
to express the core motivation for, every substantive "theory" of
truth, every theory which says what truth "is"—correspondence,
coherence or whatever. That is one way of saying that his own
theory is formal, and content-free. That means that, strictly
speaking, it is not a "theory" at all.
No philosopher dedicated more of his career to the concept of
36 I refer only to papers in the five volumes of collected essays published by Oxford
University Press.
37 Cambridge, Mass.: Belknap.
38 The longest chapter in my Why Does Language Matter in Philosophy?
(Cambridge University Press, 1975), is called "Donald Davidson's Truth".
39 Thus I re-express most of what I said about truth, in the two previous papers
mentioned in footnote 6, in terms of truthfulness.
ON THE HISTORICAL ROOTS OF SCIENTIFIC REASON / 33
emblemis Z.
15. Mathematics
Williams's schema can be applied to the sciences. He might
not have welcomed this. His Ethics and the Limits of Philosophy
kept the sciences distinct from humane questions about values.41
Let us see, however, how it can be done. Start with geometrical
relations. Who shall stand in for Thucydides? Here is the legend:
"A new light flashed upon the mind of the first man (be he Thales
or some other) who demonstrated the properties of the isosceles
triangle"—thus Kant.42 In that paragraph Kant waxes poetical
about entering "upon the sure path of science," "that royal road."
Kant calls this "the intellectual revolution" which was the
coming into being of mathematical demonstration. It was the
discovery of our capacity for mathematical proof, or rather proof
in geometry.
The iconic trailblazer of the crystallization of geometrical
proof is Thales, which is not to say that there was for sure any
such historical figure. The legend is that when X = geometrical
relations, Y = early in the sixth century BCE, and Z = Thales.
We need two things to understand styles of scientific thinking:
on the one hand, the study of mental capacities, and on the other,
the history of civilizations and of their institutions. What cultural
the first man who demonstrated ... " As if one man did the trick! It
takes a community for there to be a practice of proving. But there
is no other standard, of what is a correct proof, than proof itself.
Proofs are self-authenticating. Many a proposed proof has
proven to be fallacious, but the standard of validity or fallacy is
proof itself.
In the case of geometry, we have a new way of telling the
truth, and of ensuring that it is the truth-namely proof. In
quotation (4), Williams spoke of "operating in a new style." It is
probably fortuitous, but it is handy that Crombie and Williams
used that same mischievous word, "style".
the truth about X," at rather sharp points in time, you quickly hit
upon a figure of almost legendary proportions— Thucydides.
Rousseau/Diderot, Thales, Galileo, Boyle. For X = the species and
the higher taxa of systematics, legend congeals around Linnaeus.
For X = probabilities it is Pascal.
The doctrine of the hero in history has been totally debunked
by every historian of science in living memory. But there is still
truth in folklore. The fact that we have a legend, and a handy
emblematic trailblazer for a crystallization of each style of
scientific thinking, shows how tradition already recognizes that
each crystallized style is a radical innovation that can, for ease of
story telling, be associated with a giant. Each giant, to repeat, is
only a figurehead. I do not need reminding of this. An essential
part of the argument of The Emergence of Probability was precisely
that a new conception of probability seemed to spring up, almost
spontaneously, after 1650, in most regions of Europe. Pascal is
both the traditional emblem of this fact, and at the same time
merely a bit player.
19. Logic
Logic has come to mean deductive logic, perhaps as
formalized by first-order logic. There is an older tradition, in
which logic names the canons of reasoning. Peirce divided logic
into three parts, deduction, induction and abduction, or what in
the nineteenth century was called the method of hypothesis.
Peirce liked triads, and he was well aware how this triadic logic
itself fell under the mediaeval triad, the trivium, the bare
essentials of an education, namely grammar, logic, and rhetoric.
Deduction, induction, and abduction are not styles of
scientific thinking, in the sense of the present lectures. They are
surely founded on human cognitive capacities, although their
46/ SCIENTIFIC REASON
those who say that unobservable theoretical entities are real, and
those, from Auguste Comte to Bas van Fraassen, who deny it. In
systematics, some contend that the species are real, but not the
higher taxa. Others insist that genera, classes, orders are real, are
part of the natural order. And so on. Each ontological debate
takes place within its own scientific style. That is because every
style of finding out creates its own objects. We are on the verge of
a genealogical theory about the nature of the classic ontological
debates in the sciences.
.........
WHERE DO MATHEMATICAL
OBJECTS COME FROM?
1. Philosophical anthropology
53 Francis Galton, English Men of Science: Their Nature and Nurture, London:
Macmillan, 1874, p. 12.
WHERE DO MATHEMATICAL OBJECTS COME FROM / 51
2. Mathematics as a motley
I resolutely use the plural, "sciences" when I talk about their
philosophy, rather than talking about the philosophy of science.54
This practice is customary in French, and was common in 19th
century English, for example when William Whewell wrote three
volumes of History of the Inductive Sciences (1837) and two
volumes of Philosophy of the Inductive Sciences (1840). It is never to
be taken for granted that there is one definite thing, science, or
even one definite thing, technoscience. There is not even one
definite thing, mathematics; there is rather what Wittgenstein
called the motley of mathematics.55 Think how different is the
arithmetic that all of us learned when we were children, and the
proof of Pythagoras' theorem which many of us learned as
adolescents, or, what is tantamount to that theorem, the proof in
Plato's dialogue, the Meno, of how to construct a square double
the size of a given square. Think then of the mathematical idea
that Fermat had, when he wrote down what came to be called his
last theorem, and think of the proof ideas that lie behind Andrew
Wiles's discovery of a way to prove Fermat's last theorem.
Think next of the mathematics used in modelling the
behaviour of atoms. I recently became fascinated by very cold
3. Nature
We have good reason to think that the styles of scientific
thinking, which led me to speak of "scientific reason", are
grounded on "built-in" cognitive capacities. These are part of our
collective human inheritance. There is nothing relative or
historical or culture-dependent about them. (Or so I assume
without detailed argument.) Doubtless there have been
evolutionary pressures to favour these potentials, but do not take
for granted that they have any adaptive value for natural
selection. The ways in which they have come into being may well
be not for their immediate usefulness, but only as by-products of
genuinely functional adaptations in the development of humans
and animals. That is to suggest that the various cognitive
capacities essential to scientific reason may have emerged as
what Stephen Jay Gould and Richard Lewontin called
spandrels.56 Indeed there are scholars who urge some
explanation of the development of mathematical skills in terms
of evolutionary psychology, but I simply do not believe it. I do
not believe that the ability to do what we call mathematics had,
in the first instance, any adaptive value at all. Undoubtedly the
ability to navigate our three dimensional world is essential for
survival, and the ability to tell more from less, and even to count
a bit, would help birds, animals, and then people to get around,
but that takes us almost no way into the world of mathematics.57
56 Stephen Jay Gould and Richard C. Lewontin, "The Spandrels of San Marco and
the Panglossian Paradigm: A Critique of the Panglossian Paradigm," Proceedings of
the Royal Society of London, B, Biological Sciences 205 (1979): 581-98. Stephen
Jay Gould, "The Exaptive Excellence of Spandrels as a Term and Prototype,"
Proceedings of the National Academy of Sciences, 94 (1997): 10750-5. A spandrel is
an architectural feature, often decorative, that developed, as it were unintentionally,
out of a functional feature of the design of the building.
57 Thus despite the interest of the work by the cognitive scientist, Stephen Dahaene,
I cannot agree with the title of his book, The Number Sense: How the Mind Creates
54 / SCIENTIFIC REASON
who quite possibly was a historical figure, but perhaps the name
is simply a point about which legends have accrued. What is
striking is the extent to which folklore seems to have recognized
the need for such an accretion of myth. Kant's story is about:
(*) A shift in conceptions of what it is to tell the truth about
geometrical objects.
(**) This significant change took place in the 6th century BC, and its
emblem is Thales.
The new way to tell the truth about mathematical objects was to
prove relations between them, prove them a priori. Now why
should this ability be honoured, institutionalized?
6. Argumentative Greeks
Geoffrey Lloyd drew attention to an aspect of life in the
Greek city-state; Reviel Netz has elaborated on that theme.60
City-states were organized in many ways, but Athens is of
central importance. It was a democracy of citizens, who were all
male and none of whom were slaves. It was a democracy for the
few; not what today we call a democracy at all. But within those
few, there was no ruler. Argument ruled.
Athenians were the most consistently argumentative bunch
of self-governors of whom we have any knowledge. We read
Aristotle for his logic and not for his rhetoric. Greeks read him
for his rhetoric; his logic was strictly for the Academy. The
trouble with arguments about how to administer the city and
fight its battles is that no arguments are decisive. Or they are
decisive only thanks to the skill of the orator, or the cupidity of
the audience. But there was one kind of argument to which
66 My own ideas about the essence of the logicist programme, and its motivation,
are found in "What Is Logic?" Journal of Philosophy, 86 (1 979): 285-319
67 Der Krisis der europäischen Wissenschaften und die transzendentale
Phänomenologie, completed 1936, and translated as The Crisis of European Sciences
and Transcendental Phenomenology; An Introduction to Phenomenological
Philosophy, Evanston, Illinois: Northwestern University Press, 1970.
66 / SCIENTIFIC REASON
This is, without contest, the most important book of science studies
to appear since Shapin and Schaffer's Leviathan and the Air Pump.
69 I happen to be the person who told him about the book, so he very kindly
dedicated his review essay to me.
70 Steven Shapin and Simon Schaffer, Leviathan and the Air Pump: Hobbes, Boyle
and the Experimental Life, Princeton University Press, 1986.
68/ SCIENTIFIC REASON
field of science studies that came into being in the 1970s, but I
have never claimed to contribute to it, neither under the older
name of SSK, Social Studies of Knowledge, or a current name
such as SST, Society, Science and Technology. Latour was one of
the founding fathers of the field, inaugurating a basic on-going
programme of research, roughly speaking, the theory of scientific
networks. It is parallel to Barry Barnes's and David Bloor's work,
often referred to as the Edinburgh school or the strong
programme of scientific knowledge. Most workers in the field of
science studies regard me as at best an eccentric outlier, and I am
quite happy with that perception. Thus whereas it is entirely
fitting for Latour to pick the two most important books of science
studies of the past 21 years, I am not qualified to do so.
Nonetheless I would pick the same two books as Latour has.
What I should say is that, for my purposes, in developing a
philosophical anthropology of scientific reason, Netz's book is
(without contest) the most important book to come out of science
studies since Shapin and Schaffer's Leviathan and the Air Pump.
Latour is expressing the parallel thought, that for his purposes
these are the two most important books to come of science
studies in the past 21 years. Among his purposes is network
theory, and he takes full advantage of a pun; he calls his critical
study, "The Netz-Works of Greek Deductions."
71 Reviel Netz and William Noel, The Archimedes Codex: Revealing the Secrets of
the World's Greatest Palimpsest, London: Wiedenfield and Nicholson, 2007, p. 26.
WHERE DO MATHEMATICAL OBJECTS COME FROM /71
72 "Leibniz and Descartes: Proof and Eternal Truths,"' a Dawes Hicks Lecture to the
British Academy, 1973, published in its Proceedings, and reprinted in my Historical
Ontology, Harvard, 2002, 200-226.
WHERE DO MATHEMATICAL OBJECTS COME FROM /75
(**) The significant change took place in the 9th century, and its
emblem is Al-Khwarizmi.
73 The idea is ancient, but the label, in connection with mathematics appears to be
modem. It derives from the Paul Bemays, after whom von Neumann-Bernays set
theory is named. "Sur le platonisme dans les mathématiques," L 'Enseignement
mathématique 34 (1935), from a lecture given in 1934, and translated by Charles
Parsons for Paul Benacerraf and Hilary Putnam. Philosophy of Mathematics:
Selected Readings, 1964, 2nd ed., Cambridge: Cambridge University Press, 1983,
258-271.
74 Matiere a penser, Paris: Odile Jacob, 1989, Conversations on Mind, Matter, and
Mathematics, Princeton: Princeton University Press, 1995.
WHERE DO MATHEMATICAL OBJECTS COME FROM / 79
79 For a sketch of how to go on, see my course of lectures at the College de France on
line at http://www.college-de-france.fr/default/EN/all/ins__pro/ pI157460409944.htm;
especially "Demonstration" and especially "La stablilité des styles de penseé
scientifique. "
Lecture 3
............
1. Recapitulation
I shall begin by recalling the central novelties of the first
lecture. There were two starting points, both of them adaptations
of ideas developed by other scholars for other purposes. One was
associated with the historian of science, A. C. Crombie, from
whom I took the notion of a short list of enduring styles of
scientific thinking. The second was taken from the philosopher
Bernard Williams, the concept of truthfulness. I agreed with
Williams that truth does not have a history, but that truthfulness
does; indeed, that there are genealogies of truthfulness.
Crombie was an Australian who pursued his career in
Oxford, England. He was primarily a scholar of European
sciences in the high Middle Ages and the Renaissance; his hero,
like that of many other scholars, was Galileo. But contrary to
popular wisdom, he thought that the so-called "scientific
revolution" of the seventeenth century was not an autonomous
event of that period, but rather an evolution of developments
that began to flower in the 12th century. At a time when the
90 / SCIENTIFIC REASON
(**) This significant change took place in the Y century, and its
emblem is Z.
truth.
I noted in the second lecture that we should not be too
scrupulous in mimicking Crombie's sentences. With mathematics
all goes smoothly; there are pre-Ionian statements on geometrical
objects about which "Thales" can change the conception of what
it is to talk truly, namely by proving. But when we tum to what
happened in the time of Galileo, when there arose the idea that
God wrote the world in the language of mathematics, he urged a
new way of thinking about Nature. Yes, at the time of Galileo
there was a shift in conceptions of what it was to tell the truth
about X = motion, but the shift was far wider than that.
3. Recapitulation of 3 cautions
Before proceeding, I shall repeat three observations made in
previous lectures.
(a) A caution. Styles of scientific thinking are not sciences or
scientific disciplines, and they are not mutually exclusive.
Evolutionary biology uses lots of (1) mathematics, (2)
measurement and experimental exploration, (3) hypothetical
modelling and analogy, (4) taxonomy, (5) probability and
statistics, and yet it is our most viable example of a (6)
historico-genetic science. Most modem sciences use most of
Crombie's styles of scientific thinking.
(b) A cognitive conjecture. I postulate that each of Crombie's
80 Ian Hacking, "Language, Truth and Reason," Rationality and Relativism, ed. M.
Hollis and S. Lukes, Oxford: Blackwell, 1982, 48-66. " 'Style' for Historians and
Philosophers," Studies in History and Philosophy of Science 23 (1992), 1-20. Both
reprinted in Historical Ontology, Harvard, 2002.
94/ SCIENTIFIC REASON
philosophers of nature.
Crombie characterized his second style as "experimental
measurement, and exploration of more complex observable
relations." Whereas only a few civilizations have developed
much that I would dignify by the name "mathematics" I believe
that all peoples have engaged in various activities worth
dignifying by such words as experiment and exploration. That is
true not only of people but also of animals and birds. Until
recently, birds have been thought to have "bird-brains," but now
we realize that crows, for example, are pretty smart, and perhaps
even in the league of the higher apes. I doubt, however, that I
would dignify any avian activities as measurement; perhaps that is
a species-specific, peculiar, that is, to human beings. But a great
many human civilizations have engaged in measurement. In the
beginning, many peoples seem to have used parts of the human
body as the standards for transportable measuring devices, for
example the royal foot, the priest's arm from elbow to finger tips,
or the length of an exemplary human thumb. I am disinclined to
see any distinct crystallization, any discovery of a new human
potential, in the long development of human curiosity,
experiment, exploration, or measurement.
5. Measurement
I could well be wrong about this. Perhaps there was a distinct
crystallization around measurement, the second clause in
Crombie's second styles of scientific thinking. That crystallization
would be the realization of transportable units. But since the
original units seem to have been eminently transportable, human
feet and thumbs, there may have been only a gradual passage
rather than a sharp break to standard transportable units.
Nevertheless, a primary use of measurement is in planning
96 / SCIENTIFIC REASON
81 The "primitive language" of the builders with which Wittgenstein opens the
PhilosophicalInvestigations may be primitive for them, but someone has to have cut
the slabs, for which they are calling, to a measure. I suppose that could be done
without speaking, but it does remind us of the immense social and cultural
complexity that even the simplest intelligible tale of human talk presupposes.
82 I first realized that transportation is the essence of measurement when visiting
Churchill Eisenhart, at the old U.S. Bureau of Standards in Gaithersburg, Virginia.
(The Bureau is now the U.S. National Institute of Standards and Technology.)
THE LABORATORY STYLE OF THINKING AND DOING / 97
83 Bishop Butler's phrase, that probability "is the very guide in life" is well known;
less well remembered is the fact that the phrase comes early in the Introduction to
his Analogy of Religion, Natural and Revealed (1736).
98/ SCIENTIFIC REASON
7. Architectonic reasoning
I mentioned in the first lecture that there is a certain tension
between the early (1978) and final (1994) exposition of Crombie's
doctrine of styles. In the beginning the focus was on his own field
of specialization, the sciences of early modem Europe, that is, the
high middle ages (12th century) and on in to the Renaissance.
Later, he dwells on the continuity, and he tends to trace
everything back to the argumentative world of the Greeks. He
does, however, point to an event which he treats as relatively
gradual, but which, in rather different terms and organization, I
shall promote to a crystallization.
Crombie makes this move on page 1087 (!), in the second
volume of his book. There he states his view of what happened in
a single sentence:
This is, without contest, the most important book of science studies
to appear since Shapin and Schaffer's Leviathan and the Air Pump.
86 Steven Shapin and Simon Schaffer, Leviathan and the Air Pump: Hobbes, Boyle
and the Experimental Life, Princeton University Press, Princeton, N.J., 1985.
THE LABORATORY STYLE OF THINKING AND DOING / 105
88 Charles Bazerman, Shaping Written Knowledge: The Genre and Activity of the
Experimental Article in Science, Madison, WI: University of Wisconsin Press, 1988.
THE LABORATORY STYLE OF THINKING AND DOING /107
91 "Physicists create a new state of matter at record low temperature. Joint release
by the National Institute of Standards and Technology and The University of
Colorado, 7/13/95, still to be found on-line.
THE LABORATORY STYLE OF THINKING AND DOING /113
was doing, and protested vehemently. That man was that old
curmudgeon, the philosopher Thomas Hobbes. Given what I
said about Stillman Drake and Galileo as a builder of apparatus
in Chapter II, you can, if you prefer, use Galileo as the icon not
only for the Galilean but also for the laboratory style. Galileo had
to contest with Cardinal Bellarmine, who said nothing to make
us reflect on, now. Boyle had Hobbes.
S&S show that Boyle's work was contested at the very outset,
on both scientific and philosophical grounds. The figurehead for
the confrontation was the other man in the subtitle of their book,
Hobbes, Boyle and the Experimental Life. Hobbes was the aging
philosopher, the veritable creator of the European theory of the
state. Hobbes's book that inaugurated modem political science,
Leviathan, furnished part of the title of the book by S&S about the
air pump. Now Latour, like all sensible people, sees Hobbes as
the author of Leviathan, and as such, the spokesman for a new
age, a new kind of political society. In Canada, and most of the
English speaking world, every freshman studying political
science has to try to read Hobbes, which is by no means easy, for
the language now seems archaic or unintelligible to children
weaned on television. Latour brilliantly read S&S as displaying
the way that so-called modernity began, with a division between
the social and the natural, with one body of expertise and
intervention pertinent to one, and a quite different one germane
to the other. Hence his subtitle for the book, "Between science
(Boyle) and politics (Hobbes)". Latour rejects the distinction in
favour of what he calls Cosmopolitics, and he campaigns for a
Parliament of Things. Here I have a much more modest project in
hand, and make much more modest use of Hobbes.
In 1660, Boyle was 33 and Hobbes was 72, which happens to
be my age. If I disagreed, on fundamental principles, with a
brilliant, well-endowed, and well-supported young man of 33,
114/ SCIENTIFIC REASON
Boyle was the trailblazer for the laboratory, and we are still doing
what he did, building apparatus to purify or produce new
phenomena. Sometimes the phenomena are created in order to
test standing theories, and sometimes the phenomena precede
any theoretical understanding. Even testing theories is not the
usual point, but rather the aim is to achieve more understanding
of the predicted phenomenon. Sometimes there are what in
THE LABORATORY STYLE OF THINKING AND DOING /117
95 See my review in the London Review of Books, 10 May, 2007, which explains
some of the reasons why I think this book is so important. Online at
http://www.lrb.co.uk/v29/n09/hackOl_.html.
96 Peter W. Milonni, The Quantum Vacuum: An Introduction to Quantum
Electrodynamics, Boston: Academic Press, 1997, p. 104.
97 New York: Simon and Schuster, 1985, p. 104.
THE LABORATORY STYLE OF THINKING AND DOING /121
98 This lecture was prepared for a day-long workshop on Science, Technology and
Society. Professor Ruey-Lin Chen, who was my host at this workshop, wrote me
that "Some STS friends suggest that I should ask you to talk a little about your
'imagination' of East Asian techno-science at the end of your lecture to the STS
community." What I say here about state biotechnology in China is adapted from
what I said in response to this request.
122 / SCIENTIFIC REASON
.........
REALISMS
AND ANTIREALISMS
1. Plurals
You might expect this lecture to be titled "Realism and
antirealism," but I want to emphasize that many very different
philosophical doctrines bear these names. In the end, I shall focus
on realisms and antirealisms that are connected with the sciences,
but even there we have "scientific realism" about theoretical
non-observable entities, and types of realism called "Platonism"
in mathematics that would be better called mathematical realism.
Although my interests end up being narrow, it is important to
glance at a larger canvas, so I shall begin by mentioning debates
about which I have, here, nothing constructive to say.
I do not mean to imply that nothing can be said. This lecture
was presented at Soochow University, at a time when it had a
three year research group at work, under exactly this heading,
"Realism and Antirealism." As I understand it, quite a few types
of realism and antirealism were under discussion, so that my
emphasis on realisms in the plural was unusually germane.
130/ SCIENTIFIC REASON
101 Rudolf Camap, " Empiricism, Semantics, and Ontology,"~ Revue internationale
de philosophie, 4 (1950): 20-40.
REALISMS AND ANTIREALISMS /131
merely the class of all just entities, but something that itself exists,
independently of any acts, arrangements, decisions or whatever
individual items you may think of. It is not merely an ideal for
which we may strive, but something that actually exists, even if
we never in fact attain it. This is one kind of realism. It is
ontological in character; it is about what exists, or to pound the
table, what really exists. This is not a body of dispute that yields
readily to semantic ascent, although doubtless Carnap's tolerance
is called for.
Times have changed. I spoke to one well-known British
philosopher who writes about realism, and asked if anyone still
spoke of realism as the opposite of nominalism. He tartly replied
"no" in the rather abrupt manner of a scholar accustomed to
dismissing the tedious questions of inept undergraduates. What
then did he think that realism/antirealism debates are about? See
the section about Michael Dummett below.
I do not believe that this philosopher's brusque answer was
justified. It may be regrettable, but philosophers of various types
will go on posing ontological problems, and say that they are
debating realism versus nominalism. Nominalism is very
properly taken to be one kind of antirealism. But since I am
concerned with the sciences, I shall, so far as is possible, avoid
this kind of realism and antirealism, one that can be expressed in
terms of universals. That is why I headed this section, "Not
realism versus nominalism."
3. Universals
Sometimes the abstract, real, entities that a realist claims to
exist are called universals. Hence what is at issue in such
realist-nominalist debates is sometimes called "the problem of
universals." That is a scholastic way of putting things, which
REALISMS AND ANTIREALISMS / 133
103 R. E. Nisbett, The Geography of Thought: How Asians and Westerners Think
Differently ... and Why, New York: Free Press, 2003. For a summary of criticisms of
Nisbett's type of results, see Geoffrey Lloyd, Cognitive Variations: Reflections on
the Unity and Diversity of the Human Mind, Oxford: Clarendon Press, 2007.
(Reviewed by me in the London Review of Books, I November 2007.)
104 William von Humboldt's systematic ideas about the heterogeneity of languages
is to be found in his famous monograph, On Language: On the Diversity of Human
Language Construction and Its Influence on the Mental Development of the Human
Species, Cambridge, Cambridge University Press, 1999. (Translated from Über die
Verschiedenheit des menschlichen Sprachbaus und seinen Einfluss auf die geistige
Entwicklung des Menschengeschlechts, 1836). Wenzel, however, attends to a text
that is less well known, a letter written in 1827 to Jean-Pierre Abel Rémusat, who
had been professor of Chinese at the College de France in Paris since 1814, the first
chair of Chinese in Europe. Lettre á M. Abel-Remusat: Sur la nature des formes
grammaticales en general et sur Ie genie de la langue chinoise en particulier, Paris:
Dondey-Dupré, 1827. (Reprinted by "Elibron Classics" 2005.) Wenzel provides
references to German and other scholarship on this letter.
REALISMS AND ANTIREALISMS / 135
problems."105
105 For other senses of "linguistic relativity", see encyclopedia articles such as the
one by C. Swoyer in the Stanford online encyclopedia, http://plato.stanford.edu/
entries/relativism/supplement2.html, or L. Boroditsky, "linguistic relativity" in L.
Nadel (ed.), Encyclopedia af Cognitive Science, New York: Wiley.
REALISMS AND ANTIREALISMS / 137
the thing and turns into its very body: what at first was
appearance becomes in the end, almost invariably, the essence and
is effective as such! Only a fool would think it was enough to point
to this misty mantle of illusion in order to destroy the world that
counts as essential, so-called "reality"! We can destroy only as
creators! - But let us not forget this either: it is enough to create
new names and estimations and probabilities in order to create in
the long run new "things".106
One main point of this aphorism is, as its heading states: Only as
creators. A sub-theme then must be that we can undo a named
idea only by creating some positive concept in its place.
Deconstruction for its own sake is self-indulgent play. It is,
however, for the other thought that I single out this passage:
Unspeakably more depends on what things are called than on what
they are. And: It is enough to create new names and estimations
and probabilities in order to create new "things".
I happen cautiously to agree with Nietzsche when it comes to
names of kinds of people. I have used this quotation in a paper
stating my current position on an entirely different interest of
mine, namely the classification of people and the interaction of
classifications with the people themselves.107 But, subject to
qualifications again, I am not inclined to this way of thinking
when it comes to names of things. Needless to say, people are
"things", but it is the way in which people conceptualize
themselves, and other people conceptualize them, that makes the
difference. I am, if you want a label for me, very much a
materialist about non-sentient things, as will become more clear
towards the end of this talk. So Nietzsche's aphorism resonates
106 Friedrich Nietzsche, The Gay Science: with a Prelude in Rhymes and an
Appendix of Songs, translated by Walter Kaufmann from the 2nd ed. (1887), (New
York, Vintage Books, 1964), § 58.
107 "Kinds of People: Moving Targets," Proceedings of the British Academy 151
(2007): 285-318.
138/ SCIENTIFIC REASON
108 Chuang Tzu, The Inner Chapters, trans. David Hinton, Washington, D. c.:
Counterpoint, p. 7. For the record, Par Lagerkvist (1891-1974), Nobel Prize for
literature 1951, published his autobiographical novel Guest of Reality (Gäst hos
Verkligheten, Stocklholm: Aldus/Bonnier 1967) in 1925. For a discussion of Chuang
Tzu even more out of context than mine, see, Mark Berkson, "Language: The Guest
of Reality-Zhuangzi and Derrida on Language, Reality, and Skillfulness," in P.
Kjellberg and P. J. Ivanhoe, Essays on Skepticism, Relativism, and Ethics in the
Zhuangzi, Albany: State University of New York Press, 1996. 97-126.
109 Pierre Hadot, The Veil of Isis: An Essay on the History of the Idea of Nature,
Harvard University Press 2006, p. 14.
REALISMS AND ANTIREALISMS / 139
111 Bertrand Russell, The Problems of Philosophy, London: The Home University
Library, 1946, 84. Kant's question is stated in The Critique of Pure Reason,
translated by Norman Kemp Smith, London: Macmillan, 1929, 56 (B 20)
112 Ibid., p. 100.
REALISMS AND ANTIREALISMS /143
That is how I feel, and it is even how I feel about the proposal
that realism/antirealism debates are "mickey-mouse". Rorty said
that to me 20 or 30 years ago. It annoyed me then, but even so I
was inclined to agree with it, and hid from myself the fact that I
agreed with him. I am even more thus inclined today. I am also
inclined to make more use of Nelson Goodman's epithet, and to
speak of irrealism, using it to express a certain indifference to
traditional realist and anti realist debates.
114 From the back cover of the paperback, Richard Rorty, Philosophy as Cultural
Politics. Philosophical Papers IV, Cambridge University Press 2007.
115 Ibid., p. 133.
116 Arthur Fine, "The Natural Ontological Attitude," in his The Shaky Game:
Einstein, Realism and the Quantum Theory, Chicago: University of Chicago Press,
1986.
146 / SCIENTIFIC REASON
the ontological debates that beset the sciences, the various types
of objects rejected by antirealists, the pound-on-the-table objects
asserted to exist by realists, are all by-products of the styles of
scientific thinking by which they are introduced.
I do not think that realism is dead. I am not interested either
in killing it off or in giving it life support. I am curious why so
many kinds of realism keep on thriving. But here I restrict my
query: why do realism debates continue to thrive among
philosophers who reflect on the sciences, although they do not
matter at all to scientists at work on their fields of expertise?
117. R&I, p. 2.
REALISMS AND ANTIREALISMS /147
118 "On not being a pragmatist: eight reasons and a cause," in Cheryl Misak (ed.),
The New Pragmatists, Oxford, 2007, 32-49.
148/ SCIENTIFIC REASON
The experiment has been replicated by Jerôme Fatet; see his online seminar dated 26
January 2005, « Edmond Becquerel: La naissance de l'actinomètre electrochimique ».
122 Daniela Monaldi pointed this out to me, after reading what I wrote about
Becquerel in a preprint, "Another new world is being constructed right now: the
ultracold," Berlin: Max Planck Institute for the History of Science, 2006.
154/ SCIENTIFIC REASON
bent light they deliver us for free. Compare the cosmic rays,
amazing sources of cheap high-energy beams, used even in the
1930s for discovering sub-atomic particles. There we can
manipulate the beams, but not the sources.
I ended the paper perhaps a little too flippantly. I suggested
that one can still be a sort of phenomenalist about gravitational
lenses, regarding them as theoretical entities that admirably fit
and explain the phenomena. I even made such a suggestion
about black holes-that we are always going to be in a Duhemian
position with respect to those, that all we can do is "save (solve)
the phenomena." That is because we are only spectators: in fact
in the case of black holes, we are not even spectators, we are
spectators of the havoc we think they wreak.
I was getting carried away, as you will see from the fact that
this paper about astrophysics ends by quoting poetry. I should
have stopped a little earlier in the paper, saying simply the
gravitational lensing was, at the time, breaking news in
observational astronomy, of immense potential, but that the
experimental argument for realism would never apply. Thus, the
strongest argument for realism could not be invoked. I did not
mean positively to assert anti-realism about gravitational lenses. I
never intended to claim that they are definitely not real! To
anyone who thought I implied we ought not to believe they are
real, I apologize. I meant only that we have more compelling
reasons for asserting the existence of polarized electrons than of
gravitational lenses.
BIBLIOGRAPHY
Galton, Francis. English Men of Science: Their Nature and Nurture. London:
Macmillan, 1874
Gould, Stephen J. and Lewontin, Richard C. "The Spandrels of San
Marco and the Panglossian Paradigm: A Critique of the Panglossian
Paradigm." Proceedings of the Royal Society of London, B, Biological
Sciences 205 (1979): 581-598.
Gould, Stephen J. "The Exaptive Excellence of Spandrels as a Term and
Prototype." Proceedings of the National Academy of Sciences 94 (1997):
10750-10755.
Hacking, Ian. "Leibniz and Descartes: Proof and Eternal Truths." In
Dawes Hicks Lecture on Philosophy British Academy, pp. 3-16. London:
Oxford University Press, 1973.
____ . The Emergence of Probability. Cambridge: Cambridge
University Press, 1975; second edition, with a new "Introduction
2006." Cambridge: Cambridge University Press, 2006.
___ . "What Is Logic?" Journal of Philosophy 86 (1979): 285-319.
____ . "Language, Truth and Reason." In Hollis, M. and Lukes, S.
eds., Rationality and Relativism, pp. 48-66. Oxford: Blackwell, 1982.
____ . "Was There a Probabilistic Revolution 1800-19307" In Krüger,
L., Daston, L. J. and Heidelberger, M. eds., The Probabilistic Revolution,
vol. I: Ideas in History, pp. 45-58. Cambridge, Mass.: M.LT. Press,
1987.
____ . "The Participant Irrealist at Large in the Laboratory." British
Journal for the Philosophy of Science 39 (1988): 277-294.
____ . "Extragalactic Reality: The Case of Gravitational Lensing."
Philosophy of Science 56 (1989): 555-581.
____ . "Artificial Phenomena." British Journal for the History of
Science 24 (1991): 235-241.
____ . "The Self-Vindication of the Laboratory Sciences." In
Pickering, A. ed., Science as Practice and Culture, pp. 29-64. Chicago:
The University of Chicago Press, 1991,.
____ . "'Style' for Historians and Philosophers." Studies in History
and Philosophy of Science 23 (1992): 1-20.
BIBLIOGRAPHY/ 163
INDEX OF SUBJECTS
B c
Bachelard, Gaston 15, 16,90 Carnap, Rudolf 18, 97, 109, 130,
Bardeen, John 117 132,160
Barnes, Barry 68 Changeux, Jean-Pierre 77-78, 80,
Bazerman, Charles 106, 159-160 161
Beck, Ulrich 14, 159 Chemla, Karine 20,84-86, 160
Becquerel, Edmond 152-153, 159 Cho, Hung-Wen 117
Bellarmine, Robert 113 Chomsky, Noam 38, 41-42, 54, 80,
Bernstein, Richard 28, 160 102-103, 160
Bloor, David 68 Chuang, Tzu 136, 138, 140, 160
Bose, Satyendra 111-112, 117, 121, Coetzee, John Maxwell28, 161
157 Cohen, Bernard 1. 102-103, 161
Comte, Auguste 48, 141
Connes, Alain 78-80, 161
174/ SCIENTIFIC REASON
41-42,62,65,76, 101-103, M
164-165
Hyder, David 26,62, 164 Magruder, Kerry 12, 165
Mannheim, Karl 18
I McClelland, James L. 39, 166
Meadows, Arthur J. 12, 166
Iliffe, Robert 12, 165
Miller, James G. 134
J Milonni, Peter W. 120, 166
R U