TRANSPARENCY: Literal and Phenomenal

Colin Rowe and Robert Slutzky

“Transparency", "space-time", "simultaneity", "interpenetration", "superimposition", "ambivalence": in the Literature of contemporary architecture these words, and others like them, are often used as synonyms. We are familiar with their use and rarely seek to analyze their application. To attempt to make efficient critical instruments of such approximate definitions is perhaps pedantic. Nevertheless, in this article pedantry will be risked in an attempt to expose the levels of meaning with which the concept of transparency has become endowed.
According to the dictionary definition, the quality, or state, of being transparent is both a material condition-that of being pervious to light and air-and the result of an intellectual imperative, of our inherent demand for that which should be easily detected, perfectly evident, and free of dissimulation. Thus the adjective transparent, by defining a purely physical significance, by functioning as a critical honorific, and in being dignified with far from disagreeable moral overtones, becomes a word which from the first is richly loaded with the possibilities of both meaning and misunderstanding.
A further level of interpretation-that of transparency as a condition to be discovered in a work of art-is admirably defined by Gyorgy Kepes in his Language of Vision: "If one sees two or more figures overlapping one ar~other, and each of them claims for itself the common overlapped part, then one is confronted with a contradiction of spatial dimensions. To reso1ve this contradiction one must assume the presence of a new optical quality. The figures are endowed with transparcncy: that is they are able to interpenetrate without an optical destruction of each other. Transparency however irnplies more than an optical characteristic, it implies a broader spatial order. Transparency means a simultaneous perception of different spatial locations. Space not only recedes but fluctuates in a continuous activity. The position of the transparent figures has equivocal meaning as one sees each figure now as the closer now as the further one"1
By this definition, the transparent ceases to be that which is perfectly clear and becomes instead that which is clearly ambiguous. Nor is this meaning an entirely esoteric one; when we read (as we so often do) of "transparent overlapping planes", we constantly sense that rather more than a simple physical transparency is involved.
For instance, while Moholy-Nagy in his Vision in Motion continually refers to "transparent cellophane plastic", “transparency and moving light", and "Ruben's radiant transparent shadows",2 a careful reading of the book might suggest that for him such literal transparency is often furnished with certain allegorical qualities. Some superimpositions of form, Moholy tells us, "overcome space and time fixations. They transpose insignificant singularities into meaningful complexities.., transparent quality of the superimpositions often suggest transparency of context as well, revealing unnoticed structural qualities in the object".3 And again, in commenting on what he calls "the manifold word agglutinations" of James Joyce, or the Joyeean pun, Moholy finds that these are "the approach to the practical task of building up a completeness from interlocked units by an ingenious transparency of relationships"4In other words, he seems to have felt that, by a process of destruction, recomposition, and double-entendre, a linguistic transparency-the literary equivalent of Kepes' “interpenetration without optical destruction"-might be effected, and that whoever experiences one of these Joycean “agglutinations" will enjoy the sensation of looking through a first plane of significance to others lying behind it.

Therefore, at the very beginning of any enquiry into transparency, a basic distinction must be established. Transparency may be an inherent quality of substance, as in a glass curtain wall; or it may be an inherent quality of organization. One can, for this reason, distinguish between a literal and a phenomenal transparency.
Our feeling for literal transparency seems to derive from two sources: from cubist painting and from what is usually designated as the machine aesthetic. Our feeling for phenomenal transparency probably derives from cubist painting alone; and a cubist canvas of around 1911 or 1912 would serve to illustrate the presence of both orders, or levels, of the transparent.
One may be skeptical of those too plausible explanations of cubism which involve the fusion of temporal and spatial factors. As Alfred Barr tells us, Apollinaire "invoked the fourth dimension ... in a metaphorical rather than a mathematical sense";5 and here, rather than attempt the relation of Minkowski to Picasso, it has been considered convenient to refer to somewhat less disputable sources of inspiration.
A late C~zanne such as the Mount Sainte-l/ictoire of 1904-06 (Fig. 1) in the Philadelphia Museum of Art is characterized by certain extreme simplifications. There is a highly developed insistence on a frontal viewpoint of the whole scene, a suppression of the more obvious elements suggestive of depth, and a resultant contracting of foreground, middleground, and background into a distinctly compressed pictorial matrix. Sources of light are definite but various; and a further contemplation of the picture reveals a tipping forward of the objects in space, which is assisted by the painter's use of opaque and contrasted color. The center of the composition is occupied by a rather dense gridding both oblique and rectilinear; and this area, apparently, is buttressed and stabilized by a more insistent horizontal and vertical grid which introduces a certain peripheric interest.
Frontality, suppression of depth, contracting of space, definition of light sources, tipping forward of objects, restricted palette,oblique and rectilinear grids, and propensities toward peripheric development are all characteristics of analytical cubism. In these pictures, apart from the pulling to pieces and reassembly of objects, perhaps above all we are conscious of a further shrinkage of depth and an increased emphasis which is now awarded to the grid. We discover about this time a meshing together of two systems of coordinates. On the one hand, an arrangement of oblique and crowed lines suggests a certain diagonal spatial recession. On the other, a series of horizontal and vertical lines implies a contradictory statement of frontality. Generally speaking, the oblique and curved lines possess a certain naturalistic significance, while the rectilinear ones show a geometrizing tendency which serves as a reassertion of the picture plane. Both systems of coordinates provide for the orientation of the figures simultaneously in an extended space and on a painted surface; while their intersection, their overlapping, their interlocking, and their building up into larger and fluctuating configurations permits the genesis of the typically ambiguous cubist motif.
As the observer distinguishes between all the resultant planes, he may become progressively conscious of an opposition between certain areas of luminous paint and others of a more dense coloration. He may distinguish between certain planes to which he is able to attribute a physical nature allied to that of celluloid, others whose essence is semiopaque, and further areas of a substance totally opposed to the transmission of light. And he may discover that all of these planes, translucent or otherwise, and regardless of their representational content, are implicated in the phenomenon which kepes has defined as transparency.

The double nature of transparency may be illustrated by the comparison and analysis of a somewhat atypical Picasso, The Clarinet Player (Fig. 2), and a representative Braque, The Portuguese (Fig. 3) in each of which a pyramidal form implies an image. Picasso defines his pyramid by means of a strong contour; Braque uses a more complicated inference. Thus Picasso's contour is so assertive and so independent of its background that the observer has some sense of a positively transparent figure standing in a relatively deep space, and only subsequently does he redefine this sensation to allow for the actual lack of depth. With Braque the reading of the picture follows a reverse order. A highly developed interlacing of horizontal and vertical gridding, created by gapped lines and intruding planes, establishes a primarily shallow space, and only gradually is the observer able to invest this space with a depth which permits the figure to assume substance. Braque offers the possibility of an independent reading of figure and grid: Picasso scarcely does so. Picasso's grid is rather subsumed within his figure or appears as a form of peripheral incident introduced to stabilize it.
In the first we may receive a pre-vision of literal transparency, and in the other, of henomenal transparency; and the evidence of these two distinct attitudes will become much clearer if a comparison is attempted between the works of two slightly later painters, Robert Delaunay and Juan Gris. Delaunay's Simultaneous windows of 1911 and Gris' Still Life of 1912 (Figs. 4, 5) both include objects that are presumably transparent, the one windows, the other bottles. While Gris suppresses the physical transparency of glass in favor of a transparency of gridding, Delaunay accepts with unrestricted enthusiasm the elusively reflective qualities of his superimposed "glazed openings". Gris weaves a system of oblique and perpendicular lines into some sort of corrugated shallow space; arid in the architectonic tradition of C6zanne, in order to amplify both his objects arid structure, he assumes varied but definite light sources. Delaunay's preoccupation with form presupposes an entirely different attitude. Forms to him-e.g. a low block of buildings and various naturalistic objects reminiscent of the Eiffel Tower--are nothing but reflections and refractions of light which he presents in terms analogous to cubist gridding. But despite this geometrizing of image, the generally ethereal nature of both Delaunay's forms and his space appears more characteristic of impressionism, and this resemblance is further reinforced by the manner in which he uses his medium. In contrast to the flat, planar areas of opaque and almost monochromatic color which Gris invests with such high tactile value, Delaunay emphasizes a quasi-impressionistic calligraphy; and while Gris provides explicit definition of a rear plane, Delaunay dissolves the possibilities of so distinct a closure of his space. Gris' rear plane functions as a catalyst which localizes the ambiguities of his pictorial objects and engenders their fluctuating values. Delaunay's distaste for so specific a procedure leaves the latent ambiguities of his form exposed, without reference, unresolved. Both operations might be recognized as attempts to elucidate the intricacy of analytical cubism; but where Gris seems to have intensified some of the characteristics of cubist space and to have imbued its plastic principles with a new bravura, Delaunay has been led to explore the poetical overtones of cubism by divorcing them from their metrical syntax.

When something of the attitude of a Delaunay becomes fused
with a machine-aesthetic emphasis upon physical substance and
stiffened by a certain enthusiasm for simple planar structures,
then literal transparency becomes complete; and it can perhaps
be most appropriately illustrated by the work of Moholy-Nagy.
In his Abstract of an Artist Moholy-Nagy tells us that around
1921 his "transparent paintings" become completely freed from
all elements reminiscent of nature, and to quote him directly: "I
see today that this was the logical result of the cubist paintings I
had admiringly studied".6
Now whether a freedom from all elements reminiscent of
nature may be considered a logical continuation of cubism is
not relevant to this present discussion; but whether Moholy
indeed succeed in emptying his work of all naturalistic content
is of some importance, and his seeming belief that cubism had
pointed the way toward a freeing of forms may justify the
analysis of one of his subsequent works and its comparison with
another postcubist painting. Moholy's La Sarraz of 1930 (Fig. 6)
might reasonably be compared with a Fernand L6ger of 1926:
The Three Faces (Fig. 7).
In La San'az five circles comlected by an S-shaped band, two
sets of trapezoidal planes of translucent color, a number of near
horizontal and vertical bars, a liberal splattering of light and
dark flecks, and a number of slightly convergent dashes are all
imposed upon a black background. In Three Faces three major
areas displaying organic forms, abstracted artifacts, and purely
geometric shapes are tied together by horizontal banding and
common contour, in contrast to Moholy, Leger aligns his
pictorial objects at right angles to each other and to the edges of
his picture plane; he provides these objects with a fiat, opaque
coloring; and he sets up a figure-ground reading through the
compressed disposition of these highly contrasted surfaces.
While Moholy seems to have flung open a window on to some
private version of outer space, Leger, working within an almost
two dimensional scheme, achieves a maximum clarity of both
"negative" and "positive" forms. By means of restriction,
Leger's picture becomes charged with an equivocal depth
reading, with a value singularly reminiscent of that to which
Moholy was so sensitive in the writings of Joyce, and which, in
spite of the positive physical transparency of his paint, Moholy
himself has been unable to achieve.
For in spite of its modernity of motif, Moholy's picture still
shows the conventional precubist foreground, middleground,
and background; and in spite of a rather casual interweaving of
surface and the elements introduced to destroy the logic of this
deep space, Moholy's picture can be submitted to only one
reading.
On the other hand, through the refined virtuosity with which he
assembles post-cubist constituents, Fernand Leger makes
completely plain the multifunctioned behavior of clearly
defined form. Through fiat planes, through an absence of
volume suggesting its presence, through the implication rather
than the fact of a grid, through an interrupted checkerboard
pattern stimulated by color, proximity, and discrete
superimposition, L6ger leads the eye to experience an
inexhaustible series of larger and smaller organizations within
the whole. Leger's concern is with the structure of form.
Moholy's with materials and light. Moholy has accepted the
cubist figure but has lifted it out of its spatial matrix; L6ger has
preserved and even intensified the typically cubist tension
between figure and space.

These three comparisons may clarify some of the basic
differences between literal and phenomenal transparency in the
painting of the last fifty years. Literal transparency, we notice,
tends to be associated with the trompe l'oeil effect of a
translucent object in a deep, naturalistic space; while
phenomenal transparency seems to be found when a painter
seeks the articulated presentation of frontally displayed objects
in a shallow, abstracted space.
In considering architectural rather than pictorial transparencies,
inevitable confusions arise; for while painting can only imply
the third dimension, architecture cannot suppress it. Provided
with the reality rather than the counterfeit of three dimensions,
in architecture literal transparency can become a physical fact.
However, phenomenal transparency will, for this reason, be
more difficult to achieve; and it is indeed so difficult to discuss
that generally critics have been willing to associate transparency
in architecture exclusively with a transparency of materials.
Thus Gyorgy Kepes, having provided an almost classical
explanation of the manifestations we have noticed in Braque,
Gris, and Leger, appears to consider that the architectural
analogue of these must be found in the material qualities of
glass and plastics, and that the equivalent of their carefully
calculated compositions will be discovered in the haphazard
superimpositions produced by the reflections and accidents of
light playing upon a translucent or polished surface/
And similarly, Sigfried Ciedion seems to assume that the
presence of an all glass wall at the Bauhaus (Fig. 8), with "its
extensive transparent areas", permits "the hovering relations of
planes and the kind of 'overlapping' which appears in
contemporary painting"; and he proceeds to reinforce this
suggestion with a quotation from Alfred Barr on the
characteristic "transparency of overlapping planes" in analytical
cubism8
In Picasso's L'Arlesienne (Fig. 9), the picture that provides the
visual support for these inferences, such a transparency of
overlapping planes is very obviously to be found. There Picasso
offers planes apparently of Celluloid, through which the
observer has the sensation of looking; and in doing so, no doubt
his sensations are somewhat similar to those of a hypothetical
observer of the workshop wing at the Bauhaus. In each case a
transparency of materials is discovered. But in the laterally
constructed space of his picture, Picasso, through the
compilation of larger and smaller forms, offers the limitless
.possibilities of alternative readings, while the glass wall at the
Bauhaus, an unambiguous space, seems to be singularly free of
this quality. Thus, for evidence of what we have designated
phenomenal transparency, we shall be obliged to look
elsewhere.
Le Corbusier's villa at Garches, almost contemporary with the
Bauhaus, might fairly be juxtaposed with it. Superficially, the
garden facade at this house (Fig. 10) and the elevations of the
workshop wing at the Bauhaus are not dissimilar. Both employ
cantilevered floor slabs, and both display a recessed ground
floor. Neither admits an interruption of the horizontal
movement of the glazing, and both make a point of carrying the
glazing around the corner. But now similarities cease. From
here on, one might say that Le Corbusier is primarily occupied
with the planar qualities of glass and Gropius with its
translucent attributes. Le Corbusier, by the introduction of a
wall surface almost equal in height to his glazing divisions,
stiffens his glass plane and provides it with an over-all surface
tension, while Gropius permits his translucent surface the
appearance of hanging rather loosely from a fascia which
protrudes somewhat in the fashion of a curtain box. At Garches
we can enjoy the sensation that possibly the framing of the
windows passes behind the wall surface: at the Bauhaus, since
we are never for a moment unaware that the slat is pressing up
behind the window, we are not enabled to indulge in such
speculations.

At Garches the ground is conceived of as a vertical surface
traversed by a horizontal range of windows (Fig. 11); at the
Bauhaus it is given the appearance of a solid wall extensively
punctured by glazing. At Garches it offers an explicit indication
of the frame which carries the cantilevers above; at the Bauhaus
it shows somewhat stubby piers which one does not
automatically connect with the idea of a skeleton structure. In
this workshop wing of tile Bauhaus one might say that Gropius
is absorbed with the idea of establishing a plinth upon which to
dispose an arrangement of horizontal planes (Fig. 12), and that
his principal concern appears to be the wish that two of these
planes should be seen through a veil of glass (Fig. 8). But glass
would hardly seem to have held such fascination for Le
Corbusier; and although one can obviously see through his
windows, it is not precisely here that the transparency of his
building is to be found.
At Garches the recessed surface of the ground floor is redefined
on the roof by the two freestanding walls which terminate the
terrace; and the same statement of depth is taken up in the side
elevations by the glazed doors which act as conclusions to the
fenestration. In these ways Le Corbusier proposes the idea that
immediately behind his glazing there lies a narrow slot of space
traveling parallel to it; and of course, in consequence of this, he
implies a further idea-that bounding this slot of space, and
behind it, there lies a plane of which the ground floor, the
freestanding walls, and the inner reveals of the doors all form a
part; and although this plane may be dismissed as very
obviously a conceptual convenience rather than a physical fact,
its obtrusive presence is undeniable. Recognizing the physical
plane of glass and concrete and this imaginary (though scarcely
less real) plane that lies behind it, we become aware that here a
transparency is effected not through the agency of a window but
rather through our being made conscious of primary concepts
which "inte~0enetrate without optical destruction of each
other".
These two planes are not all; a third and equally distinct parallel
surface is both introduced and implied. It defines the rear wall
of the terrace and the penthouse, and is further reiterated by
other parallel dimensions: the parapets of the garden stairs, the
terrace, and the second-floor balcony (Fig. 10). Each of these
planes is incomplete in itself or perhaps even fragmentary; yet it
is with these parallel planes as points of reference that the
facade is organized, and the implication of all is of a vertical,
layerlike stratification of the interior space of the building, a
succession of laterally extended spaces traveling one behind the
other.
This system of spatial stratification brings Le Corbusier's facade
into the closest relationship with the Leger we have already
examined. In Three Faces Leger conceives of his canvas as a
field modeled in low relief. Of his three major panels (which
overlap, dovetail, and alternatively comprise and exclude each
other), two are closely implicated in an almost equivalent depth
relationship, while the third constitutes a coulisse disclosing a
location which both advances and recedes. At Garches, Le
Corbusier replaces Leger's concern for the picture plane with a
most highly developed regard for the frontal viewpoint (the
preferred views include only the slightest deviations from
parallel perspective); Leger's canvas becomes Le Corbusier's
second plane; other planes are either imposed upon, or
subtracted from, this basic datum. Deep space is contrived in
similar coulisse fashion with the facade cut open and depth
inserted in the ensuing slot (Fig. 11).

One might infer that at Garches, Le Corbusier had indeed
succeeded in alienating architecture from its necessary three-
dimensional existence, and in order to qualify this analysis,
some discussion of the building's internal space is necessary.
On first examination this space appears to be an almost fiat
contradiction of the facade; particularly on the principal floor
(Fig. 13), the volume revealed is almost directly opposite to that
which we might have anticipated. Thus the glazing of the
garden facade might have suggested the presence of a single
large room was parallel with that of the facade. But the internal
divisions deny this statement and instead disclose a principal
volume whose primary direction is at right angles to that which
might have been presumed, while in both principal and
subsidiary volumes the predominance of this direction is
conspicuously emphasized by the flanking walls.
The spatial structure of this floor is obviously more complex
than it appears at first, and ultimately it compels a revision of
these initial assumptions. The nature of the cantilevered slots
becomes evident; the apse of the dining room introduces a
further lateral stress, while the positions of the principal
staircase, the void, and the library all reaffirm the same
dimension. In these ways the planes of the facade can be seen
to effect a profound modification of the deep extension of space
which is now seen to approach to the stratified succession of
flattened spaces suggested by the external appearance.
So much might be said for a reading of the internal volumes in
terms of the vertical planes; a further reading in terms of the
horizontal planes, the floors, will reveal similar characteristics.
Thus, after recognizing that a floor is not a wall and that planes
are not paintings, we might examine these horizontal planes in
very much the same manner as we have examined the facade,
again selecting Three Faces as a point of departure. A
complement of Leger's picture place is now offered by the roofs
of the penthouse and elliptical pavilion, by the summits of the
freestanding walls, and by the top of the rather curious gazebo-
all of which lie on the same surface (Fig. 11, 14). The second
plane now becomes the major roof terrace and the coulisse
space becomes the cut in this slab which leads the eye down to
the terrace below. Similar parallels are very obvious in
considering the organization of the principal floor. For here the
vertical equivalent of deep space is introduced by the double
height of the outer terrace and by the void connecting living
room with entrance hall; and here, just as Leger enlarges spatial
dimensions through tile displacement of the inner edges of his
outer panels, so Le Corbusier encroaches upon the space of his
central area.
Thus throughout this house there is that contradiction of spatial
dimensions which Kepes recognizes as a characteristic of
transparency. There is continuous dialectic between fact and
implication. The reality' of deep space is constantly opposed to
the inference of shallow space; and by means of the resultant
tension, reading after reading is enforced. The five layers of
space which throughout each vertical dimension divide the
building's volume and the four layers which cut it horizontally
will all from time to time claim attention; and this gridding of
space will then result in continuous fluctuations of
interpretation.
These possibly cerebral refinements are scarcely so conspicuous
at the Bauhaus; indeed, they are attributes of which an aesthetic
of materials is apt to be impatient. In the workshop wing of the
Bauhaus it is tile literal transparency that Giedion has chiefly
applauded, and at Garches it is the phenomenal transparency
that has engaged our attention. If with some reason we have
been able to relate the achievement of Le Corbusier to that of
Fernand Leger, with equal justification we might notice a
community of interest in the expression of Gropius and Moholy-
Nagy.

Moholy was always preoccupied with the expression of glass,
metal, reflecting substances, and light; and Gropius, at least in
the 1920s, would seem to have been equally concerned with the
idea of using materials for their intrinsic qualities. Bolh, it may
be said without injustice, received a certain stimulus from the
experiments of De Stijl and the Russian constructivists; but both
were apparently unwilling to accept certain more Parisian
conclusions.
For seemingly it was in Paris that the cubist "discovery" of
shallow space was most completely exploited, and it was there
that the idea of the picture plane as a uniformly activated field
was most entirely understood. With Picasso, Braque, Gris,
Leger, and Ozenfant we are never conscious of the picture plane
functioning in any passive role. Both it, as negative space, and
the objects placed upon it, as positive space, are endowed with
an equal capacity to stimulate. Outside the Ecole de Paris this
condition is not typical, although Mondrian, a Parisian by
adoption, constitutes one major exception and Klee another.
But a glance at any representative work of Kandinsky, Malevich,
El Lissitsky, or Van Doesburg will reveal that these painters, like
Moholy, scarcely felt the necessity of providing any distinct
spatial matrix for their principal objects. They are prone to
accept a simplification of the cubist image as a composition of
geometrical planes, but are apt to reject the comparable cubist
abstraction of space. For these reasons their pictures offer us
compositions which float in an infinite, atmospheric,
naturalistic void, without any of the rich Parisian stratification
of volune. And the Bauhaus may be accepted as their
architectural equivalent.
Thus in the Bauhaus complex, although we are presented with a
composition of slablike buildings whose forms suggest the
possibility of a reading of space by layers, we are scarcely
conscious of the presence of spatial stratification. Through the
movements of the dormitory building, the administrative
offices, and the workshop wing, the first floor may suggest a
channeling of space in one direction (Fig. 15). Through the
countermovement of roadway, classrooms, and auditorium
.wing, the ground floor suggests a movement of space in the
other (Fig. 16). A preference for neither direction is stated (Fig.
17), and the ensuing dilemma is resolved, as indeed it must be
in this case, by giving priority to diagonal points of view.
Much as Van Doesburg and Moholy eschewed frontality, so did
Gropius; and ii is significant that, while the published
photographs of Garches (Fig. 19) tend to minimize factors of
diagonal recession, almost invariably the published photographs
of the Bauhaus (Fig. 18) tend to p]ay up just such factors. The
importance of these diagonal views of the Bauhaus is constantly
reasserted by the translucent corner of the workshop wing and
by such features as balconies of the dormitory and the
protruding slab over the entrance to the workshops, features
which require for their understanding a renunciation of the
principle of frontality.

The Bauhaus reveals a succession of spaces but scarcely “a
contradiction of spatial dimensions". Relying on the diagonal
viewpoint, Gropius has exteriorized the opposed movements of
his space, has allowed them to flow away into infinity: and by
being unwilling to attribute Lo either of them any significant
difference of quality, he has prohibited the possibilities of a
potential ambiguity. Thus only the contours of his blocks
assume a layerlike character (Fig. 18); but these layers of
building scarcely act to suggest a layerlike structure of either
internal or external space. Denied the possibility of penetrating
a stratified space which is defined either by real planes or their
imaginary projections, tile observer is also denied the possibility
of experiencing the conflict between a space which is explicit
and another which is implied, tie may enjoy the sensation of
looking through a glass wall and thus perhaps be able to see the
exterior and the interior of the building simultaneously; but in
doing so he will be conscious of few of those equivocal
sensations which derive from phenomenal transparency.
Le Corbusier's League of Nations project of 1927, like tile
Bauhaus, possesses heterogeneous elements and functions that
lead to an extended organization, and to the appearance of a
further feature which both buildings have in common: the
narrow block. But here again similarities cease, for while the
Bauhaus blocks pinwheel in a manner highly suggestive of
eonstruetivist compositions (Fig. 17), in the League of Nations
these same long blocks define a system of striations almost
more rigid than that at Garches (Fig. 20).
In the League of Nations project lateral extension characterizes
the two principal wings of the Secretariat, the library and book-
stack area, is re-emphasized by entrance quay and the foyers of
the General Assembly Building, and dominates even the
auditorium itself. There, the introduction of glazing along the
side walls, disturbing the normal focus of the hall upon the
presidential box, introduces the same transverse direction. The
contrary statement of deep space also becomes a highly
assertive proposition. It is chiefly suggested by a lozenge shape
whose main axis passes through the General Assembly Building
and whose outline is comprised by a projection of a auditorium
volume into the approach roads of tile cour d'honneur (Fig. 21).
But again, as at Garehcs, the intimations of depth inherent in
this form are consistently retracted. A cut, a displacement, and
a sliding sideways occur along the line of its major axis; and as a
space, it is repeatedly scored through and broken down into a
series of lateral references-by trees, by circulations, by the
momentum of the buildings themselves-so that finally,
through a series of positive and negative implications, the whole
scheme becomes a sort of monumental debate, an argument
between a real and ideal space.

We will presume the Palace of League of Nations as having been
built and an observer following the axial approach to its
auditorium (Fig. 22). Necessarily, he is subjected to the polar
attraction of its principal entrance. But the block of trees which
intersects his vision introduces a lateral deflection of interest, so
that he becomes successively aware, first, of a relation between
the flanking of fie, e-building and the foreground parterre, and
second, of a relation between the crosswalk the courtyard of the
Secretariat. And once within the trees, beneath the low
umbrella they provide, a further tension is established: the
space, which is inflected toward the auditorium, is defined by,
and reads as, a projection of the book stack and library. While
finally, with the trees as a volume behind him, the observer at
last finds himself standing on a low terrace, confronting the
entrance quay but separated from it by a rift of space so
complete that it is only by the propulsive power of the walk
behind him that he can be enabled to cross it (Fig. 23) With his
arc of vision no longer restricted, he is now offered the General
Assembly Building in its full extent; but since a newly revealed
lack of focus compels his eye to slide along this facade, it is
again irretrievably drawn sideways, to the view of the gardens
and the lake beyond. And should the observer turn round from
this rift between him and his obvious goal, and should be look
at the trees which he has just left, the lateral sliding of the space
will only become more determined, emphasized by the trees
themselves and the cross alley leading into the slotted indenture
alongside the book stack. If the observer is a man of moderate
sophistication, and if the piercing of a screen or a volume of
trees by a road might have come to suggest to him that the
intrinsic function of this road is to penetrate similar volumes
and screens, then by inference the terrace on which he is
standing becomes not a prelude to the auditorium, as its axial
relationship suggests, but a projection of the volumes and
plans of the office building with which it is aligned.
These stratifications, devices by means of which space becomes
constructed, substantial, and articulate, are the essence of that
phenomenal transparency which has been noticed as
characteristic of the Bauhaus, which obviously manifests a
completely different conception of space. In the League of
Nations project Le Corbusier provides the observer with a series
of quite specific locations: in the Bauhaus he is without such
points of reference. Although the League of Nations project is
extensively glazed, such glazing, except in the auditorium, is
scarcely of capital importance. At the Palace of the League of
Nations, corners and angles are assertive and definite. At the
Bauhaus, Giedion tells us, they are "dematerialised". At the
Palace of the League of Nations space is crystalline; but at the
Bauhaus it is glazing which gives the building a "crystalline
translucence". At the Palace of the League of Nations glass
provides a surface as definite and taut as the top of a drum; but
at the Bauhaus, glass walls "flow into one another", "blend into
each other", "wrap around the building", and in other ways (by
acting as the absence of plane) "contribute to that process of
loosening up a building which now- dominates the architectural
scene".9
But we look in vain for "loosening up" in the Palace of the
League of Nations. It shows no evidence of any desire to
obliterate sharp distinction. Le Corbusier's planes are like
knives for the apportionate slicing of space (Fig. 25). If we could
attribute to space the qualities of water, then his building is like
a dam by means of which space is contained, embanked,
tunneled, sluiced, and finally spilled into the informal gardens
alongside the lake. By contrast, the Bauhaus, insulated in a sea
of amorphic outline, is like a reef gently washed by a placid
tide.
The foregoing discussion has sought to clarify the spatial milieu
in which phenomenal transparency becomes possible. It is not
intended to suggest that phenomenal transparency (for all its
cubist descent) is a necessary constituent of modem
architecture, nor that its presence might be used like a piece of
litmus paper for the test of architectural orthodoxy. It is
intended simply to give a characterization of species and also to
warn against the confusion of species.
a+u77:04

Note
1． Gyorge Kepes: The Language of Vision, Paul Theobald, Chicago 1944,p.77.
2． Moholy-Nagy: Vision in Motion, Paul Theobald, Chicago 1947; pp157,159,188,
194.
3． Moholy-Nagy: op.cit.p210.
4． Moholy-Nagy: op.cit.p350.
5． Alfred Barr: Picasso: Fifty Years of His art, The Museum of Modern Art,
New York 1946; p.68.
6． Moholy-Nagy: The New Vision and Abstract of an Artist, Wittenborn and Co.,
New York 1947; p.75.
7． Gyorgy Kepes: op.cit.
8． Sigfeied Giedion: Space, Time, and Architecture, Cambridge, Mass. 1954;
pp.490-491.
9． Sigfried Giedion: op.cit.p.489; and Sigfried Giedion: Walter Gropius, Reinhold,
New York 1954; pp.54-55.