This article was downloaded by: [George Washington University]

On: 02 January 2015, At: 02:59

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Art Journal
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/rcaj20

Conventionalism in Henri Poincaré and Marcel

Duchamp
Craig Adcock
Published online: 02 Aug 2014.

To cite this article: Craig Adcock (1984) Conventionalism in Henri Poincaré and Marcel Duchamp, Art Journal, 44:3, 249-258,
DOI: 10.1080/00043249.1984.10792553

To link to this article: http://dx.doi.org/10.1080/00043249.1984.10792553

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Conventionalism in Henri Poincare
and Marcel Duchsmp
By Craig Adcock

he mathematical and philosophical

T
1912)
writings of Henri Poincare (1854-
were among the major influences
on the development of Marcel Du-
champ's art and thought.' Through
Downloaded by [George Washington University] at 02:59 02 January 2015

Poincare, Duchamp learned the basic

principles of at least three major
branches of modern geometry: n-dimen-
sional geometry, non-Euclidean geome-
try, and topology.' So important were
these geometries to Duchamp that he , .
incorporated numerous references to
them in the iconography of his major
works: the Bride Stripped Bare by Her
Bachelors, Even (the Large Glass),
1915-23 (Fig. I), Tu m', 1918 (Fig. 2),
the ready-mades, and the satellite works
associated with them. For Duchamp,
Poincare's philosophical discussions of
the conventional nature of geometry
were a way of reinforcing his own spec-
ulations about the provisional nature
of aesthetics.
Many of Poincare's most basic
insights into complex areas of mathe-
matics involved applications of geome-
try. His brilliant work in his "qualitative
theory of differential equations," for
example, was predicated on topology.
His work on automorphic functions of
one complex variable (what he called
o
Fuchsian functions) and his work on
analytic functions of several complex
variables (Abelian functions) also in-
volved the application of geometrical
techniques to problems that might
otherwise have proven recalcitrant. -, j
Poincare argued that invention was
often a matter of choice, involving put-
ting things together that did not seem to
belong together: "Among chosen combi-
nations the most fertile will often be
those formed of elements drawn from
domains which are far apart." 3 He
recalled that some of his own most
important insights had involved just
such unlikely combinations. In Science
and Method. he writes that he had made
breakthroughs in the study of Fuchsian
functions because he had suddenly seen
relationships between the transforma-
Fig. 1 Duchamp, The Bride Stripped Bare by Her Bachelors. Even (the Large
tions he was using to specify differential
Glass), 1915-23, oil and lead wire on glass, 1091/ 4 x 69 1/ 8" . Philadelphia Museum
equations and those that occurred in of Art, Bequest of Katherine S. Dreier.
non-Euclidean geometry. He also recalls

FIl/l1984 249
 Fig. 2 Duchamp, Tu m', 1918, oil on canvas with long brush attached, 27 1/2 x 122W'. Yale University Art Gallery, Bequest of
Katherine S. Dreier.
that on another occasion, but with the geometry, a contradiction would occur invention of human reason: "these con-
same sense of immediacy, he had real- between the two Euclidean theorems ventions are the work of the free activity
ized that "the arithmetic transforma- from which they had been translated. of our mind, which, in this domain,
tions of indeterminate ternary quadratic But since "these translations are the- recognizes no obstacle. Here our mind
forms were identical with those of non- orems of ordinary geometry and no one can affirm, since it decrees; but let us
Downloaded by [George Washington University] at 02:59 02 January 2015

Euclidean geometry.:" doubts that the ordinary geometry is understand that while these decrees are
That Poincare could use geometries free from contradiction," questions imposed upon our science, which, with-
such as topology and non- Euclidean regarding the legitimacy of non-Euclid- out them, would be impossible, they are
geometry and obtain useful results was ean geometry should not arise.' The not imposed upon nature."?
important to his world view. Not too important part of this argument for Poincare believed that the source of
long before his time, only the postulates Poincare was that non-Euclidean geom- the general laws of mathematics and
of plane and solid Euclidean geometry etries, because they could be interpreted science lay in the rational human intel-
were believed to have any validity. The in terms of the unquestioned proposi- lect; that they were rational was what
development of such mathematical con- tions of ordinary Euclidean geometry, gave them their rigor and their essential
structs as complex numbers, quater- were no longer just empty displays of usefulness. But Poincare also asked
nions, n-dimensional geometry, and logic but could be concretely useful in whether, if made up in the mind, they
non-Euclidean geometry during the applied mathematics. He was quick to were then also capable of being applied
course of the nineteenth century had point out that he had obtained impor- in random fashion. No, he said: "Experi-
forced mathematicians to reexamine a tant results by applying Lobachevski's ment leaves us our freedom of choice,
number of their fundamental assump- geometry to the integration of linear but it guides us by aiding us to discern
tions. No longer could they believe that differential equations." the easiest way."!" In other words,
mathematics represented a true picture experiment in the real world orders and
of the world-that mathematical prog- oincare' s understanding of geome- directs one's choice of first principles.
ress was a matter of uncovering the
hidden laws of nature. They were forced
P try influenced his philosophical
insights. Because one geometry could be
Some are useful in discovering the
world, and some are not. Science pro-
to admit that certain aspects of mathe- translated into another and because ceeds by discovering and making use of
matics are the constructs of human rea- there was no way of determining if one those conventional hypotheses that are
son.' One could no longer deny the exis- geometry was more true than another, productive and by discarding those that
tence of such mathematical entities as he argued that geometry was conven- are unproductive.
n-dimensional spaces with n > 3 or of tional. Poincare is the father of philo- Poincare pointed out that some inves-
non-Euclidean spaces in which Euclid's sophical conventionalism. From such a tigators had gone too far in what they
parallelism postulate does not hold. position, there are numerous ways of took to be the implications of the con-
Such spaces may not have physical ana- describing the world; anyone way can- ventional nature of scientific principles:
logues, but they are mathematically no not be said to be more true than any "they have wished to generalize beyond
less real than three-dimensional Euclid- other-s-one way can only be said to be measure, and, at the same time, they
ean space. more useful than another under a cer- have forgotten that liberty is not license.
Some of Poincare's most important tain set of circumstances. Such argu- Thus they have reached what is called
mathematical discoveries involved ments can lead to arbitrariness, as Poin- nominalism." Poincare suggested that
methods of interrelating various geome- care himself was well aware. Since he such thinkers should "have asked them-
tries. In Science and Hypothesis. he believed that there was an external selves if the savant is not the dupe of his
describes, in accessible terms, a kind of world and that it could be discovered own definitions and if the world he
dictionary through which Euclidean and through science, he eschewed extreme thinks he discovers is not simply created
non-Euclidean geometries could be conventionalism, or "nominalism," as he by his own caprice."!' If science did
translated from one to the other. He called it. operate according to the random inven-
explains that if the theorems of a non- Poincare argued that there were a tion of descriptive models, it could never
Euclidean geometry were translated number of different kinds of hypotheses. tell us anything about the natural world
according to the terms of his dictionary, Some were verifiable facts and could be and the objects that occupy that world.
as one might translate a passage from considered truths; some were useful in "Still," Poincare added, "the things
one language into another, one would organizing research approaches but themselves are not what [science] can
arrive at the "theorems of the ordinary were unverifiable; and some were "dis- reach, as the naive dogmatists think, but
geometry.:" Moreover, such transla- guised definitions" or "conventions." only the relations between things. Out-
tions would necessarily be consistent: if These last, conventional hypotheses side of these relations there is no know-
a contradiction were to occur between were most likely to be encountered in able reality.':"
two theorems of the non- Euclidean mathematics and science. They were the Geometry and mathematics were

250 Art Journal

 ways of articulating the relationships research was free from intuitive ap- structure of the art world. How did
between objects of perception. They proaches they were deluding themselves. progress in art proceed and what was the
were not the same as experimental facts, Innovation requires intuition: "Pure role of theory or aesthetics within that
but they were useful in picturing the logic could never lead us to anything but process? On the one hand, Duchamp
world. Poincare argued that because the tautologies; it could create nothing new; had become disenchanted with practical
axioms of more than one kind of geome- not from it alone can any science matters of aesthetics: from personal
try could be shown to be consistent, one issue."!' Poincare thought that mathe- experience, he had found that such
could no longer believe that Euclidean matics and the other sciences progressed things could be arbitrary and guided by
geometry was the one true way of by something more than syllogistic suspect taste." On the other hand, he
describing space and its inhabitants. He arguments from first principles: they was intensely involved with making con-
pointed out that progressed by intuitive invention. One of crete contributions to art, especially
the sources for original creation was during his early years. He was like a
experience no doubt teaches us "generalization by induction, copied, so scientist making discoveries in a labora-
that the sum of the angles of a to speak, from the procedures of the tory but not knowing how to unify his
triangle is equal to two right experimental sciences.':" facts within an encompassing theory.
angles; but this is because the
Through his reading of Poincare,
triangles we deal with are too lit- Duchamp found a paradigm for articu-
uchamp's philosophy also shares
tle; the difference, according to
Lobachevski, is proportional to the
D characteristics with positivism. In
his interview with Pierre Cabanne,
lating his art problems; science provided
a metaphorical schema for defining the
surface of the triangle. Will this Duchamp pointed out that process of making art.
perhaps become sensible when we
Downloaded by [George Washington University] at 02:59 02 January 2015

operate on larger triangles or the Viennese logicians worked out

oincare's system of doubt does not
when our measurements become
more precise? The Euclidean ge-
a system wherein everything is,
as far as I understand it, a tau-
P imply a lack of faith in science or in
the ability of humans to do science. That
ometry would thus be only a provi- tology, that is, a repetition of
one could not decide on the ultimate
sional geometry." premises. In mathematics, it goes
from a very simple theorem to a truth or falsity of certain kinds of hypo-
What this provisional status implied for very complicated one, but it's all in theses did not mean that one had to
Poincare was that "one geometry cannot retreat into arbitrariness. Duchamp's
the first theorem. So, metaphysics:
be more true than another; it can only be system had a similar kind of a reason-
more convenient. ,,14 tautology; religion: tautology;
ableness about it. His system was careful,
everything is tautology, except
Poincare's philosophy developed out black coffee because the senses are plotted, consistent, and subtle. What
of his mathematics. In particular, his emerges from Duchamp's doubt is a
in control! The eyes see the black
attitudes were conditioned by his studies careful skepticism-a skepticism that
coffee, it's a truth; but the rest is
involving the relationships among dif- always tautology." also characterizes the best scientists.
ferent kinds of geometry. One could What one sees in Duchamp, as in
invent the principles of geometry and Duchamp's reference here to the Vienna Poincare, is a healthy willingness to
develop them through logic. These prin- Circle is suggestive. Philipp Frank question-a process that leads to new
ciples were conventions and disguised recalled that some of the most pressing solutions. It was in this regard that
definitions, but, nonetheless, they were questions that concerned him and his Poincare was most important to Du-
"drawn from experimental laws.':" In associates in Vienna, a group that champ. It was a matter of one genius
certain ways, Poincare's position was a included Otto Neurath and Hans Hahn, reinforcing another genius. It was a mat-
precursor to logical positivism. In involved the relationship between exper- ter of the best of human endeavor
Science and Method. he argues that imental facts and scientific hypotheses: (science) reinforcing the best of human
"In our opinion, the man who bridged endeavor (art).
a demonstration truly founded the gap successfully was the French Poincare articulates his discussions of
upon the principles of analytic logic
mathematician and philosopher Henri the provisional nature of different kinds
will be composed of a series of Poincare. For us, he was a kind of Kant of geometry by demonstrating their
propositions. Some, serving as
freed of the remnants of medieval scho- interrelatedness. By showing the con-
premises, will be identities or defi- lasticism and anointed with the oil of nections between, say, metric geometry
nitions; the others will be deduced modern science.'?" and projective geometry, between Eu-
from the premises step by step. But
One of the apparent discrepancies or clidean and non-Euclidean geometry,
though the bond between each problems in the development or progress and by then showing their various rela-
proposition and the following is
of science concerned the role of scien- tionships with topology, Poincare could
immediately evident, it will not at
tific generalizations. Material science show that how one chose one's geometry
first sight appear how we get from
seemed to advance steadily with respect was a matter of convenience. Duchamp
the first to the last, which we may
to the gradual accumulation of knowl- wanted to demonstrate the conventional
be tempted to regard as a new
edge and facts acquired through experi- nature of aesthetics-to show that how
truth. But if we replace succes- ment. But at the same time, grand the- one chose one's art was also a matter
sively the different expressions
ories or .large-scale theoretical schemata of context.
therein by their definition and if
this operation be carried as far as
were likely to be proven false and were Tu m', 1918 (see Fig. 2), clearly
then discarded by the scientific commu- shows that approach. The painting can
possible, there will finally remain
nity. If a theory was soon to be thrown be taken as a demonstration of the prin-
only identities, so that all will
out, why advance it in the first place? ciples discussed by Poincare, and the
reduce to an immense tautology. What was the role of hypothesis in the philosophical implications of the paint-
Logic therefore remains sterile
advancement of science? An important ing for art are similar to those that
unless made fruitful by intuition."
part of Poincare's philosophical inquiry Poincare articulated for science. Tu m'
In his discussion of intuition in The addressed the latter question. is susceptible to interpretation from the
Value of Science. Poincare writes that Duchamp was faced with a similar point of view of several different geome-
even when scientists thought that their kind of problem within the operational tries. The cast shadows or projected

Fall 1984 251

 shadows (ombres porteesi are refer- ever ... I thought that, by simple
ences to projective geometry. Perspec- intellectual analogy, the fourth
tive is a subcategory of projective geom- dimensional could project an object
etry, and part of the subject matter of of three dimensions.... "The
Tu m' is perspective. On one level, Bride" in the "Large Glass" was
Duchamp was dealing with a traditional based on this, as if it were the
two-dimensional picture surface that projection of a four-dimensional
represents a three-dimensional space. object."
This is already n-dimensional geometry In Duchamp's thinking, Tu m', the
with n = 2 and n = 3. But Duchamp Large Glass, the notes, the ready-
was also interested in situations where rnades, their cast shadows, and their
n > 3. Such situations are considerably geometrical implications were all tightly
more complex. interconnected.
Because it could be thought of as an The same may be said of their philo-
analogy for projection, the concept of a sophical implications. Tu m' is a meta-
cast shadow was important in Du- painting and perhaps metaphysical. Du-
champ's thinking about n-dimensional champ may have intended his shadows as
geometry. Part of the imagery of Tu m' Platonic references; he may have wanted
consists of shadows cast by three of his to suggest that the work of art was a
ready-mades: the Bicycle Wheel, the shadow of a shadow. If so, the reasons
Downloaded by [George Washington University] at 02:59 02 January 2015

Corkscrew, and the Hat Rack. The would not have been Plato's, but
Corkscrew has survived only as a Duchamp's, More precisely, they would
shadow on Tu m', The shadows of the have been Duchamp's as developed out of
three ready-mades are outstretched Poincare. By 1918, Duchamp had come
across the surface of the canvas with the to feel that painting was largely empty
Bicycle Wheel on the left and the spiral Fig. 3 Ducharnp, "Shadows of semblance. He was interested in the idea
of the Corkscrew leading out from the Ready-rnades," photograph taken in behind the work of art, not in order to
axle of the wheel to the center of the Duchamp's studio, 33 West 67 Street, affirm the first principles of a valuational
painting. A sign-painter's hand emerges New York, 1918. Collection, Mme aesthetics but in order to cast doubt on
from the handle of the Corkscrew and Marcel Duchamp, Villiers-sous-Grez. those first principles. Tu m' clearly
points towards the shadow of the Hat involves geometry. The purpose of this
Rack on the right. These shadows are involvement is less clear, but the connec-
Duchamp discusses the "cast shadows of
analogies: if a three-dimensional object tions between various geometries and
ready-rnades.t'" The inclusion of this
casts a two-dimensional shadow, then by their conventionalist implications are
note in the Green Box indicates that he
analogy a four-dimensional object what Duchamp used in the unification of
intended that the imagery both of Tu m'
would cast a three-dimensional shadow. his art. He had learned the geometry and
and of the Large Glass involve shadows
In one of his notes for the Large Glass the philosophy from Poincare. In Du-
cast from ready-mades. A photograph
in A l'Infinitif, Duchamp used that anal- champ's descriptions of art, as in Poin-
taken by Duchamp in his New York
ogy: "The shadow cast by a four-dimen- care's descriptions of the world, the rela-
apartment in 1918 (Fig. 3) reinforces
sional figure on our space is a three- tionships between objects were more
this idea. The forms of the shadows
dimensional shadow. "22 He then refers accessible than the objects themselves.
are very similar to the flat shapes of
to a passage is Esprit Pascal Joulfret's
the Bride in the upper panel of the
a
Traite elementaire de geometrie quatre
Large Glass. rojective analogies need not be
dimensions: "See Jouffret, Geom. of 4
dim., page 186, last 3 lines." In those
The implications of n-dimensional pro-
jection were explained by Duchamp on
P confined to discussing the interrela-
tionships of dilferent n-dimensional Eu-
lines, Joulfret suggests that "in this
several occasions. He told George Heard clidean spaces. Explanations of non-
regard [conceptualizing the fourth di-
Hamilton and Richard Hamilton: Euclidean geometry can also involve
mension]," one would do well to "con-
sider the horizontal shadow that attaches anything that has three-dimen- interdimensional analogies such as flat-
itself to you as you walk along in the sun sional form is a projection in our beings." The flat-being represented by
and that, long or short, wide or narrow, world from a four-dimensional the shadow of the Bicycle Wheel in Tu
repeats your movements as if it under- world, and my Bride, for example, m' may be a reference to non-Euclidean
stood you, although it is only an empty would be a three-dimensional pro- geometry. The circular rim of the wheel,
sernblance.?" Joulfret uses the analogy jection of a four-dimensional the circumference of the circle, if rotated
of the cast shadow to introduce a discus- Bride. All right. Then, since it's on around one of its diameters would gener-
sion of flat-beings. Since it is difficult to the glass it's flat, and so my Bride ate a sphere, which is an analytic Euclid-
envision higher-dimensional spaces, it is a two-dimensional representa- ean surface that can serve as a model for
might be useful to envision lower-dimen- tion of a three-dimensional Bride, the elliptic non-Euclidean geometry of
sional spaces. By imagining what it who also would be a four-dimen- Riemann. The wheel would generate a
would be like to live in a flat, two- sional projection on a three- figure much like the one on the right-
dimensional plane, one could better dimensional world of the Bride." hand side of a diagram from Jouffret's
appreciate the relationships between Traite (Fig. 4).28
In his interview with Pierre Cabanne, The diagram accompanies a descrip-
three-dimensional beings and four-
Duchamp connected this kind of discus- tion of Riemann's geometry. The draw-
dimensional spaces.
sion of projective geometry with the ing might also serve as an illustration for
In Tu m', three-dimensional ready-
notion of cast shadows:
mades become flat, two-dimensional Poincare's explanation in Science and
shadows. They become, in a sense, flat- Since I found that one could make Hypothesis. In order to discuss Rie-
beings. In one of his notes for the Large a cast shadow from a three-dimen- mann's geometry, Poincare introduces
Glass included in the Green Box, sional thing, any object whatso- the notion of "beings with no thickness."

252 Art Journal

 He then argues that if such beings were
cognizant and if they lived on the sur- Fig. 2.
face of a sphere, they would not invent
ordinary geometry: "First it is clear they ]' A
will attribute to space only two dimen-
sions; what will play for them the role of C
the straight line will be the shortest path
from one point to another on the sphere,
that is to say, an arc of a great circle; in a 0
word, their geometry will be the spheri- E
cal geometry." He goes on to say that E'
"Riemann's geometry is spherical ge- E
ometry extended to three dimensions.,,29
The shadow of the Bicycle Wheel in 0' f
E'
Tu m' is essentially made up of one-
dimensional line segments. They can be
thought of as projections from a two-
dimensional, non-Euclidean surface like
the one in Jouffret's illustration. The
A' I
left-hand side of Jouffret's diagram,
Downloaded by [George Washington University] at 02:59 02 January 2015

which resembles a bicycle wheel, can be

taken as a projection of the right-hand Fig. -I Illustration from E. Jouffret, Traite elementaire de geometrie aquatre
diagram. The sphere, in its turn, can be dimensions. p. 16.
taken as a projection from a three-
dimensional, non-Euclidean space.
Geometers often point out that there is
a conceptual correspondence between a
three-dimensional, non-Euclidean Rie-
mannian space and a four-dimensional
Euclidean hypersphere." Each can be
thought of as a sphere with an extra
dimension. Poincare had this correspon-
dence in mind when, in the introduction
to his discussion of his Euclidean-non-
Euclidean dictionary, he said that people
who were used to thinking about four-
dimensional geometry would have no dif-
ficulty in extending the two-dimensional
models of non-Euclidean geometry into
three dimensions."
Evidence that Duchamp was inter-
ested in the relationships between
Euclidean and non-Euclidean geometry,
specifically in regard to Poincare's use
of Lobachevski's geometry, is provided
by his Unhappy Ready-Made. 1919
(Fig. 5), and his original photograph of
the book hanging from a Paris balcony
(Fig. 6). In the latter, no "geometry" is
visible at all. Duchamp chose the dia-
grams as ready-made geometry and Fig. 5. Duchamp, Unhappy. Fig. 6. Duchamp, Un retouched
then retouched the photograph with Ready-made, 1919, from the photograph of Unhappy Ready-made. c.
compass and ruler. What he chose is Box-in-a-Valise, Paris, 1941-42. 1919. Philadelphia Museum of Art,
significant. The drawing depicts orthog- Philadelphia Museum of Art, The Marcel Duchamp Archives, Gift of Dr.
onal circles and demonstrates one of Louise and Walter Arensberg William A. Camfield.
Euclid's theorems." Various Eucl idean Collection.
theorems and their corresponding dia-
grams often begin discussions of raral- defines straight lines as being either Euclidean geometry." Thus, the non-
lelism in Lobachevski's geometry.' Spe- straight·lines passing through the center Euclidean geometry is consistent if the
cifically, Euclidean orthogonal circles of the fundamental plane or arcs of Euclidean geometry is consistent. Poin-
are often used by way of comparison in circles cutting the fundamental plane care's model is the one that he later
explanations of Poincare's method of orthogonally. Angles remain angles. described in more general terms as a
demonstrating the consistency of Loba- Distance becomes logarithmic. Two tan- "dictionary" in Science and Hypothe-
chevski's non-Euclidean geometry. gent arcs that intersect on the funda- sis," If one were to translate the dia-
Poincare first used this approach in the mental plane are parallel. What Poin- grams of orthogonal circles that Du-
Context of his work on automorphic (or care's system amounts to is a way of champ used for his ready-made geome-
Fuchsian) functions." relating the two geometries: he shows try book according to Poincare's rules,
Poincare begins by defining the fun- that the axioms and theorems of Loba- one would arrive at an arrangement that
damental plane as a circle. He then chevski's geometry are special cases of looks very different: the Euclidean dia-

Fall 1984 253

 length. In general, this flexible and
inextensible figure cannot be dis-
placed without leaving the surface;
but there are certain particular sur-
faces for which such a movement
would be possible; these are the
surfaces of constant curvature.
He goes on to point out that such sur-
faces can be derived for both branches of
non-Euclidean geometry: "The two-
dimensional geometries of Riemann and
Lobachevski are thus correlated to the
Euclidean geornetry.?"

oincare's remarks about a canvas

P bring Tu m' to mind. The buckling
around the trompe l'oeil tear suggests
that the two-dimensional plane surface
of the canvas has become a two-dimen-
sional curved surface. This kind of
Downloaded by [George Washington University] at 02:59 02 January 2015

deformation can be taken as a reference

to non-Euclidean geometry. The shad-
ows can be taken as references to the
figures projected onto that surface. The
points of an n-dimensional configuration
can be projected onto an n - I dimen-
sional configuration, as when three-
dimensional objects are projected onto
two-dimensional perspective or isomet-
ric drawings. In a less complete sense, a
two-dimensional shadow is a projection
of a three-dimensional object.
In one of his notes for the Large
Glass. first published by Matta in his
magazine Instead, Duchamp says that
"after the Bride," he wanted "to make
pictures using cast shadows." Tu m' was
apparently one of the results. In the
note, he goes on to say that by means of
n - I dimensional shadows projected by
n-dimensional objects, "first on a plane,
second on a surface of such or such
curvature, third on several transparent
surfaces, ... one can obtain a hypophy-
sica I analysis of the successive transfor-
mations of objects (in their form con-
tour). ,,39 The reference here to both
plane and curved surfaces suggests that
Duchamp was thinking in terms of both
metric and projective geometry and in
terms of both Euclidean and non-
Figs. 7 and 8 Duchamp, 3 Standard Stoppages. 1913-14, assemblage: three Euclidean geometry. Depending upon
threads glued to three painted canvas strips, 5V4 x 47W', each mounted on a glass the surface chosen-plane, positively
panel, and wooden templates shaped along one edge to match the curves of the curved, negatively curved-the result-
threads; the whole fitted into a wood box, 111/ 8 x SOW'. New York, Collection The ing geometries could be quite different.
Museum of Modern Art, Katherine S. Dreier Bequest. In another of his notes, this one from
the Green Box, Duchamp talks about
the shadows cast by "2, 3, 4 ready-
grams would be transformed into non- constant curvature. Poincare explained mades 'brought together.''' He dis-
Euclidean diagrams. Duchamp may this concept thus: cusses the shadows as being three-
have intended the deformations caused Consider any figure on a surface. dimensional ready-mades "having be-
in his book by hanging it from a balcony Imagine this figure traced on a come" two-dimensional projections of
as references to such transformations. flexible and inextensible canvas ready-mades. "Take these 'having be-
The curved pages may have been refer- applied over this surface in such a comes,' " he says, "and make from them
ences to the curvature involved in non- way that when the canvas is dis- a tracing without of course changing
Euclidean geometry." placed and deformed, the various their position in relation to each other in
The two-dimensional models for non- lines of this figure can change their the original projection.'?" Duchamp
Euclidean geometry require surfaces of form without changing their traced the shadows of the ready-mades

254 Art Journal

 onto the surface of Tu m' with pencil. he iconography of Tu m' involves in this discipline two figures are
This technique (calquer) involves the
"projection" of three-dimensional ob-
T both the philosophy of art and the
philosophy of science. Part of Poincare's
equivalent every time it is possible
to have one correspond to the other
jects onto two-dimensional surfaces. questioning of the foundations of geom- by means of a continuous defor-
The French term is interesting because a etry concerned the sources of axioms mation, whatever the law gov-
calque is not only a "tracing" but also and theorems. Were they experimental erning the deformation may be,
an "imitation" or "close copy." The truths? He argued that "we constantly provided that continuity is main-
ready-mades come out of the molds of reason as if geometric figures behaved tained. Thus, a circle is equivalent
mass-production techniques as close like solids," but such apparent connec- to an ellipse 01' even to any type of
copies. The calque made from the tions with what we see in the world do closed curve, but it is not equiva-
shadow of a ready-made is itself a com- not justify an assumption that geometry lent to a line segment because the
monplace reproduction: a shadow can be is experimental. He goes on to say that segment is not a closed figure. A
cast by "any object whatsoever." "the properties of light and its recti- sphere is equivalent to any convex
In a metaphorical sense, the fourth linear propagation have also given rise surface whatever, but it is not
dimension may have been for Duchamp to some of the propositions of geometry, equivalent to a torus because in
a device for flattening works of art into and in particular those of projective the torus there is a hole and there
shadows-"three times removed from geometry, so that from this point of view is none in a sphere. Let us imagine
the true." Images on picture surfaces one would be tempted to say that metric a pattern of any kind and the copy
and shadows on picture planes are n- geometry is the study of solids, and of this pattern drawn by a clumsy
dimensional projections of n + I dimen- projective, that of light. ,,41 But there is a draftsman. The proportions are
sional objects. They are geometrical fundamental problem with such ap- distorted, straight lines drawn by a
Downloaded by [George Washington University] at 02:59 02 January 2015

"sections" in the same sense that Alberti proaches: if geometry were based on trembling hand have undergone
used the term in his discussion of per- experiment or measurement, it could distressing deviations and result in
spective. In Tu m', Duchamp is dealing never be exact. Because there are no disproportional curves. From the
with a two-dimensional picture surface, rigorously rigid solids in nature, an point of view of metric geometry,
a Renaissance mirror surface as it were, experimentally based geometry could and even from that of projective
and three-dimensional objects projected only be approximate." geometry, the two figures are not
onto that surface, both as shadows and Several aspects of Tu m' involve the equivalent; but on the contrary
as perspective renderings. In the center approximate nature of measuring de- they are equivalent from the point
of the painting, the conventional symbol vices and the distinctions between met- of view of analysis situs."
of a hand emerges from the handle of ric geometry and projective geometry. Topological operations are carried
the Corkscrew and points towards both Poincare explained that out as if the bodies undergoing the
a shadow and an unconventional per- transformations were made of rubber.
spective arrangement. Overlapping the metric geometry is based on the
So long as continuity is maintained, a
traced shadow of the Hat Rack is an notion of distance; in it, two fig-
geometrical figure can be stretched into
open-ended, transparent box. One end of ures are considered as equivalent
any number of configurations. A circle
the box is made up of a white rectangle when they are "equal" in the sense is topologically equivalent to a square
drawn in perspective. From each corner which mathematicians assign to
and a torus is topologically equivalent to
of this rectangle, two curved lines are this word. Projective geometry is
a coffee cup, because if the circle were a
drawn parallel to the picture plane with based on the notion of the straight rubber band, it could be stretched into
the templates of the 3 Standard Stop- line. For two figures to be consid-
the shape of the square, just as a rubber-
pages (Figs. 7 and 8). The volume ered as equivalent in projective sheet torus could be stretched into the
suggests a curved region of space. geometry it is not necessary that
shape of a coffee cup. With these points
The double edges drawn with the tem- they be equal; it is sufficient that in mind, any number of Duchamp's "de-
plates are perhaps meant to cast doubt they correspond to each other by formations" can be interpreted as topo-
on measuring devices. They suggest means of a projective transforma- logical transformations. The flat and
the provisional nature of geometry. Fur- tion; that is, that one be the pro- curved pages of the geometry book can
ther, they suggest that there is an ad jection of the other."
be taken as topologically equivalent con-
libitum freedom involved in choosing Classical perspective is subsumed figurations, or homeomorphisms, as can
between alternatives. under projective geometry. In one of his the flat and curved surface implied by
Tu m' becomes an essay in making- notes, Duchamp says that "by perspec- the trompe l'oeil buckling on Tu m.
in art making, hypothesis making, and tive (or other conventional means) the From a topological point of view, the
aesthetic judgment making. Tradition- lines, the drawing, are 'strained' and curved line segments in 3 Standard
ally, the picture plane had been an area lose the nearly of the 'always possible' Stoppages (see Figs. 7 and 8) are equiv-
in which reality took place; it had been a with moreover the irony to have chosen alent to straight lines or to each other.
window opening onto a space con- the body or original object which inevi- As the individual pieces of string fell
structed according to the principles of tably becomes according to this perspec- through space, they underwent any
perspective-a subcategory of pro- tive (or other conventionj.?" In a per- number of transformations. Topologi-
jective geometry. Duchamp tore a hole spective. transformation, lengths and cally, the transformations were ad libi-
through that area. Similarly, the inven- angles are "strained" and "inevitably tum or a son gre. Such variations inter-
tion of non-Euclidean geometries had become" according to the principles of ested Duchamp because of their philo-
torn a hole through the intellectual the transformation. Duchamp's "other sophical implications. He believed that
structure of reality. The Euclidean convention" may have been topology the meaning of a work of art was elastic
space that had been depicted in pictures because, in it, transformations are more and ad libitum: "An oeuvre by itself
was called into question by the invention nearly "always possible." doesn't exist, it's an optical illusion. It's
of these geometries. Truth, geometric Poincare's remarks about metric and only made to be seen by the people who
reality, was ripping apart. It could be projective geometries introduce an ex- look at it. The poor medium is only
held together only by conventional con- amination of analysis situs, a geometry gratuitous. You could invent a false
structs, by safety pins. now called topology. He explains that artist. Whatever happens could have

Fall 1984 255

 been completely different."46 For Du-
champ, aesthetics, critical interpreta-
tion, art history, and artistic intention
were topological.
The line segments in 3 Standard
Stoppages are topological homeomor-
phisms in one dimension. At the lower
left of Tu m' are three curved lines
drawn with the templates of the 3 Stan-
dard Stoppages. They are at the edges
of three overlapping curved surfaces
that suggest the configurations that
would have been generated by the
curved line segments of the strings as
they fell through space. These three
surfaces are topological homeomor-
phisms in two dimensions. The three
"Draft Pistons" in the upper panel of the
Large Glass (see Fig. I) are also topo-
logical homeomorphisms in two dimen-
sions. Duchamp hung flat, two-dimen-
Downloaded by [George Washington University] at 02:59 02 January 2015

sional pieces of cloth in front of a win-

dow and allowed the wind to distort
their shapes. He then photographed the
pieces of cloth (Fig. 9) and used the
forms as templates for determining the
irregular holes in the epanouissement of
the Bride. As a next logical step in this
n-dimensional sequence, Duchamp's
ready-made Traveler's Folding Item
(Fig. 10) can be taken as a reference to
topological homeomorphisms in three
dimensions. The work looks like a rub-
ber-sheet cube. If taken as a metaphor
for a topological solid, it would become a
three-manifold with boundary. The two-
dimensional deformed surfaces of the
three "Draft Pistons" can be taken as
analogues for the deformed surfaces
that would be generated by the one-
dimensional deformed lines of the 3
Standard Stoppages displacing them- Fig, 9 Duchamp, "Draft Piston," 1914, photograph, 23 1/ 8 x 19 11/16" . Collection,
selves through space. The Traveler's Mme Marcel Duchamp, Villiers-sous-Grez.
Folding Item can, in its turn, be taken as
an analogue for the configuration that
would be generated by the two-dimen-
sional "Draft Pistons" displacing them-
selves through space. The next step
towards a displacement into the fourth
dimension would be, in Duchamp's
term, "hypophysical."
Traveler 's Folding Item is a type-
writer cover and thus associated with
writing. Both the three " Draft Pistons"
and the 3 Standard Stoppages are also
associated with writing. In one of his
notes, Duchamp refers to the " Draft
Pistons" as "alphabetic units," and it is
through them that the "commands" of
the Bride are telegraphed to the Bache-
lors in the lower panel of the Large Glass.
Duchamp explains that the Bride's
instructions have "their alphabet and
terms governed by the orientation of the
three Draft Pistons. ,,47 Thus , communi-
cation in the Large Glass proceeds by a
topologically deformed sign system . In Fig. 10 Duchamp, Traveler's Folding Item, 1916, H . 23". Sarasota, Fla., John and
another of his notes, Duchamp suggests Mable Ringling Museum of Art, Gift of Mary Sisler Foundation/Mrs. William T.
that geometrically deformed symbols Sisler.

256 Art Journal

 could be constructed using the 3 Stan- beauty was the last thing that he had matics and the philosophy of science-
dard Stoppages. He says that one could had on his mind when he chose them." and put them together with aesthetics.
"compose a schematic sign designating But these works were far from uninter- His connections were metaphors for the
each of [the so-called abstract words in a esting. He said that he had thrown the importance of relationship over any cate-
Larousse dictionary]." This sign could ready-mades into the face of the art gory of individual objecthood. Given cer-
"be composed with the standard stops. ,,48 public as a challenge and that now they tain relationships, the objects could be
Duchamp's desire to construct a "topo- admired them for their aesthetic beauty. anything whatsoever. Duchamp used
logical language" reflects his attitude This challenge was not just a joke, a conventionalism.in his arrangements to
concerning meaning: he believed that schoolboy blague. It was a way of ques- show that "everything could have been
meaning could be stretched into any tioning the underpinnings of art criti- completely different." His strategy rep-
number of configurations, as if it were cism and art history. Duchamp said that resents a substantial connection between
made of rubber. "taste is momentary," that if one waited art and science. Duchamp said that "we
Topological interpretations are in fifty years, taste changed. He argued don't speak about science because we
keeping with Ducharnp's interests in n- that "if one is logical, one doubts the don't know the lan~uage, but everyone
dimensional and non-Euclidean geome- history of art. ,,53 speaks about art." 6 Duchamp himself
tries. The various geometries can be Ducharnp's doubt began early in his was an exception. He knew a great deal
interrelated, as Duchamp had learned career. He remembered being turned about the language of science. Through
from Poincare: non-Euclidean geometry around by the small mindedness that Poincare, he had learned the philosophi-
is a metric geometry, and metric geome- had caused the rejection of Nude cal implications of advanced geometries,
try can be subsumed under projective Descending a Staircase from the Inde- and he used what he had learned to
Downloaded by [George Washington University] at 02:59 02 January 2015

geometry; projective geometry can be pendents exhibition in 1912: speak about art. And because what he
n-dimensional, and non-Euclidean geom- said has been so influential, those geom-
Cubism had lasted two or three
etry can be n-dimensional; Riemann's etries and the world views that they en-
years, and they already had an
geometry is commensurate with topol- gendered, have had fundamental effects
absolutely clear, dogmatic line on
ogy. Poincare points out that "[analysis on art, art theory, and the revisionism of
it, foreseeing everything that might
situs] gives rise to a series of theorems art history.
happen. I found that natvely fool-
just as closely interconnected as those of
ish. So, that cooled me off so much Notes
Euclid; and it is from this set of proposi-
that, as a reaction against such I Duchamp was probably familiar with at least
tions that Riemann constructed one of
behavior coming from artists four of Poincare's general works: La Science et
the most remarkable and abstract theo-
whom I had believed to be free, l'hypothese, Paris, Flammarion, 1902; Science
ries of pure analysis.?" Topology can
I got a job. I became a librarian et methode, Paris, Flammarion, 1904; La Val-
also be n-dimensional, as Poincare
at the Sainte-Genevieve Library eur de la science, Paris, Flarnmarion, 1908;
reminds us: "there is an analysis situs of in Paris. 54
more than three dimensions. ,,50 Dernieres pensees, Paris, Flarnmarion, 1913
Shocked into reevaluating the premises (English translations: The Foundations of
Science: Science and Hypothesis, The Val-
oinca re demonstrated that one's involved in interpretation and aesthetic
P choice of geometry was conven-
tional. Duchamp demonstrated that
judgment, Duchamp must have been
attracted to the alternative intellectual
ue of Science. Science and Method, trans.
George Bruce Halsted, New York, The Science
Press. 1921; Mat hemat ics and Science: Last
one's choice of art was conventional. tradition represented by mathematics Essays, trans. John W. Bolduc, New York,
Through conventional means, he trans- and science; it was in his new job in the Dover, 1963).
formed objects that were not art into art library that he probably began reading
objects. He did it by placing ready- Poincare and found in him a kindred 2 In addition to Poincare's books, the major
mades within the art context and by spirit. sources for Duchamp's mathematical knowl-
Duchamp called art "une dedale edge were two works of Esprit Pascal Jouffret:
then saying that they were not works of
illogique"-an illogical labyrinth." He Traite elementaire de geometrie a quatre
art. In spite of his disavowals, they
dimensions et introduction a la geometric a n
became works of art. Poincare argued believed that one lost oneself within the
dimensions. Paris, Gauthier- Villars, 1903; and
that one could not know objects; one laybrinth, retraced one's steps, and com-
Melanges de geometrie a quatre dimensions,
could know only the relationships ing around new corners that obscured Paris, Gauthier-Villars, 1906. Jouffret's books
between objects. Duchamp concurred. old vistas, perceived new ones. Poincare are straightforward mathematical textbooks
When asked what determined his choice argued that the approach to mathemat- and, although they contain discussions of
of ready-mades, Duchamp replied that ics and science also involved intricate numerous geometrical operations that Du-
it pathways: truth was hidden within a champ made use of, they were probably not as
labyrinth, but, nonetheless, one provided important for the development of his philoso-
depended on the object. In general,
oneself with maps, with hypotheses, and phy as were Poincare's books. For more
I had to beware of its "look." It's
proceeded inward. Both Poincare and detailed discussions of Duchamp's sources, see:
very difficult to choose an object,
Duchamp were aided by the philosophi- Craig Adcock, Marcel Duchamp's Notes from
because, at the end of fifteen days,
cal thread of conventionalism: it pro- the Large Glass: An N-Dimensional Analysis,
you begin to like it or to hate it.
vided the scientist and the artist with a Ann Arbor, UMI Research Press, 1983. pp.
You have to approach something 29-39; and Linda Dalrymple Henderson. The
skepticism that prevented mistaking a
with an indifference, as if you had Fourth Dimension and Non-Euclidean Geome-
corner of the labyrinth for its center; it
no aesthetic emotion. The choice of try in Modern Art, Princeton, Princeton Uni-
provided them with an abiding doubt
ready-mades is always based on versity Press, 1983, pp. 117-30.
about the veil of certainty that seems to
visual indifference and, at the same
fall across human consciousness too eas- 3 Poincare, Foundations ofScience, p. 386.
time, on the total absence of good
ily and too quickly.
or bad taste." 4 Ibid., pp. 387-88.
Duchamp took Poincare's advice and
Duchamp claimed that he had chosen put things together that did not seem to 5 For a discussion of these philosophical changes,
the ready-mades precisely because they belong together. He chose ready-mades see: Morris Kline, Mathematical Thought
were visually neutral-he neither liked and put them together with art objects; from Ancient to Modern Times, New York,
them nor disliked them-and that their he chose intellectual disciplines-mathe- Oxford University Press. 1972, pp. 1023-39.

Fall 1984 257

 6 Poincare, Foundations of Science. pp. 56--60; broadcast by the BBC, third program in the 43 Poincare, Mathematics and Science: Last
Nikolai Ivanovich Lobachevski's most accessi- series "Art, Anti-Art," 1959, quoted in Arturo Essays. pp. 57-58.
ble essay is his Geometrische Untersuchungen Schwarz, The Complete Works of Marcel
44 Duchamp (cited n. 22), p. 36.
zur Theorie der Parallellinien, Berlin, 1840; a Duchamp, New York, Abrams, 1970, p. 23.
translation by George Bruce Halsted (Geomet- 26 Interview with Cabanne (cited n. 19), p. 40. 45 Poincare, Mathematics and Science: Last
rical Researches on the Theory of Parallels) is Essays. pp.58-59.
27 In his interview with Cabanne, p. 39, Duchamp
included in Non-Euclidean Geometry by
credits Gaston de Pawlowski with being a pop- 46 Dore Ashton, "An Interview with Marcel
Roberto Bonola, New York, Dover, 1955, after
ularizer of the fourth dimension who had Ducharnp," Studio International, 171 (June
p. 268; Bernhard Riemann's most famous
explained "that there are flat beings who have 1966), p. 246.
paper is his "Uber die Hypothesen welche der
only two dimensions, etc." and who had "ex-
Geometrie zu Grunde liegen," Abhandlungen 47 Duchamp (cited n. 22), p. 36.
plained measurements, straight lines, curves,
der Koniglichen Gesellschaft der Wissen-
etc." Pawlowski's only mention of flat-beings 48 Ibid., pp. 31-32.
schaften zu Gottingen, 13 (1868), pp. 1-20.
occurs in his book, Voyage au pays de la
49 Poincare, Mathematics and Science: Last
7 Poincare, Foundations ofScience. p. 60. quatrieme dimension. Paris, Eugene Fasquelle,
Essays. pp.58-59.
1912, pp. 27-28, in the context of a discussion
8 Ibid., these were the Fuchsian functions men-
of non-Euclidean geometry: flat-beings living 50 Ibid., p. 43.
tioned earlier.
on the surface of a sphere would believe that
51 Interview with Cabanne (cited n. 19), p. 48.
9 [bid., p. 28. the angles of a triangle would add up to more
than 180 degrees, etc. Jean Clair, Marcel 52 See: Duchamp's interviews with Kuh (cited n.
10 [bid.
Duchamp ou Ie grand fictif: Essai my thana- 21), pp. 91-92; with Jeanne Siegel, "Some
Downloaded by [George Washington University] at 02:59 02 January 2015

II [bid. lyse du Grand Verre, Paris, Galilee, 1975, has Late Thoughts of Marcel Ducharnp," Arts
discussed Duchamp's dependence on Pawlows- Magazine, 43 (December 1968-January
12 [bid.
ki. Both Henderson (cited n. 2), pp. 119, 128, 1969), p. 21; and with Francis Roberts, "I
13 Ibid., p. 63. nn. 7, 31, 33, and Adcock (cited n. 2), pp. Propose to Strain the Laws of Physics," Art
33-34, argue that Clair's case for Pawlowski's News. 67 (December 1968), p. 62.
14 [bid., p. 65.
being a major influence on Duchamp is over-
53 Interview with Otto Hahn, "Marcel Du-
15 [bid., p. 125. stated. Pawlowski was without doubt part of
champ," L'Express (Paris), no. 684 (July 23,
Ducharnp's intellectual background, as he him-
16 Ibid., p. 483. 1964), p. 22. The original passage in French is
self says, but Duchamp's knowledge of geome-
worth quoting at greater length: "Le gout est
17 [bid., pp. 214-15. try is far more sophisticated than anything he
momentane, c'est une mode. Mais ce que ron
could have found in Pawlowski.
18 Ibid., pp. 215-16. considere comme une forme esthetique est
28 For an interpretation of Jouffret's diagram in debarrasse du gout. On attend done cinquante
19 Pierre Cabanne, Dialogues with Marcel
relation to non-Euclidean geometry, the Bicy- ans, et la mode disparait. Les choses prennent
Duchamp, trans. Ron Padgett, New York, Vik-
cle Wheel, and the "Oculist Witnesses" in the alors un sens. En fin de compte, c'est une
ing, 1971, p. 107.
Large Glass. see: Craig Adcock, "Geometrical entourloupette: une autre forme de gout. Ce qui
20 Philipp Frank, Modern Science and Its Philos- Complication in the Art of Marcel Duchamp," ne l'etait pas sur Ie moment Ie devient plus tard.
ophy. New York, Braziller, 1955, p. 8; Frank's Arts Magazine 58 (January 1984), pp. 105-9. Si on est logique, on doute de I'histoire de
discussion was brought to my attention by John I'art. ... Le public est victime d'un veritable
29 Poincare, Foundations ofScience. p. 57.
Philip Paul, "An Analysis and Evaluation of complot ebahi. Les critiques parlent de la
Henri Poincare's Cosmology, Epistemology 30 See, for example: H. S. M. Coxeter, Non- 'verite de l'art' comme on dit 'la verite de la
and Philosophy of Science," Ph.D. Disserta- Euclidean Geometry. Toronto, University of religion.' Les gens suivent comme des moutons
tion, Marquette University, 1969, p. 165. Toronto Press, 1965, p. 12. de Panurge. Moi, je n'accepte pas, c'est inexis-
tant. Ce sont des voiles inventes, Cela n'exista
21 ln his interview with Cabanne (cited n. 19), p. 31 Poincare, Foundations ofScience. p. 59.
pas plus qu'en religion. D'ailleurs.je ne crois en
31, Duchamp pointed out that the "'Nude
32 This is pointed out by Henderson (cited n. 2), p. rien, car croire donne lieu a un mirage."
Descending a Staircase' had been refused by
160. The diagram that Duchamp uses shows
the Independents in 1912." He added that the 54 Interview with Cabanne (cited n. 19), p. 17.
tangents drawn from an axis through the points
rejection was an occurrence that "helped liber-
of intersection. Such tangents are equal. 55 Interview with Robert Lebel, "Marcel Du-
ate me completely from the past, in the per-
champ, rnaintenant et ici," L'Oeil (Paris), no.
sonal sense of the word. [ said, 'All right, since 33 See, for example: Bonola (cited n. 6), pp.
149 (May 1967), p. 20.
it's like that, there's no question of joining a 250--64.
group-I'm going to count on no one but 56 Interview with Ashton (cited n. 46), p. 247.
34 Between 1882 and 1884, Poincare published
myself, alone.' " Throughout the rest of his life,
five landmark papers on automorphic functions
Duchamp remained suspicious of taste. In his
in Acta Mathematica. For this particular
Craig Adcock is Assistant Professor of
interview with Katherine Kuh, The Artist's
result, see: Acta Mathematica I (1882), pp.
Art History at The Florida State
Voice: Talks with Seventeen Artists. New University. Tallahassee.
1-62, esp. p. 8 and p. 52; reprinted in Henri
York, Harper & Row, 1960, pp. 91-92, he said,
Poincare, Oeuvres. Paris, Gauthier-Villars,
"I consider taste-bad or good-the greatest
1954, II, pp. 108--68.
enemy of art."
35 See: Kline (cited n. 5), pp. 916-17.
22 Marcel Duchamp, Salt Seller: The Writings
of Marcel Duchamp (Marchand du Sel), 36 Poincare, Foundations ofScience. pp. 58-60.
ed. Michel Sanouillet and Elmer Peterson,
37 Henderson (cited n. 2), p. 160.
New York, Oxford University Press, 1973,
pp.89-90. 38 Poincare, Foundations ofScience. pp. 58-59.
23 Jouffret, Traite (cited n. 2), pp. 186-87, my 39 Duchamp (cited n. 22), p. 72.
translation.
40 [bid., p. 33.
24 Duchamp (cited n. 22), p. 33.
41 Poincare, Foundations of Science. p. 64.
25 Interview with George Heard Hamilton and
42 [bid., pp. 64-65.
Richard Hamilton, "Marcel Duchamp Speaks,"

258 Art Journal