What The Wall Does

In architecture the main purpose of the wall is to delimit a space and to support the roof.

Whereas supporting a roof is a particular architectonic and structural problem, delimitation is found notonly in landscape forms butalso in social behaviour between people. In both cases the wall defines territories, it divides two spaces, each of which has a different content. The built wall shows the actual way in which two diverse areas meet, thus interpreting the strength relationship between them. Concerning our main question — the architectural inside-outside relationship — this means the extent to which the wall draws exterior space inside or interior space outside. The wall’s architecture, in other words, is a concrete realization of the existential struggle between an ‘attacking’ exterior and a ‘secure’ interior and thereby acquires expressive importance (Fig. 143).

What is it in the wall’s appearance which conveys a message of the comparative strength of inside and outside space? And, in what way does this strength relationship effect its expression?

The Wall Themes

Strength relationship is understood as the degree of penetration. When we consider a wall that delimits an inside space, it may vary between expressing complete openness, thus inviting us to enter, or complete closure, which rejects us. The background of our reactions to this relationship is again dependent upon the expressions of motion, weight, and substance. In other words, a heavy wall will seem more closed than a lightone, a large door stimulates a spontaneous impulse to go through it, whereas the small, low door arrests our movement. Similarly, a ‘soft’ wooden wall conveys warmth and is inviting as opposed to a rough stone wall which is ‘cold’ and rejecting.

The expression of penetration is dependent on the relationship between three themes, each of which is a result of the interface between delimited and enveloped space. This interface affects the wall in its breadth, height, and depth. In other words, the significance of the wall as the mediator between the surrounding spaces lies in the tension between the wall and these surroundings (Fig. 144).


External forces which affect the wall: (a) in depth, (b) in breadth and (c) in height.

The depth indicates the relation between the spaces in front of and behind the wall. The height indicates the relation to the spaces above and below, which are the ground and the roof, and the breadth indicates the relationship to the spaces which meet the corners to the right and left.

In these encounters with the surroundings, the surface of the wall will achieve completely dissimilar expressions. These expressions vary on a fixed set of archetypes, which are implied by each of the surrounding’s three dimensions. In the following, we will briefly examine the themes of breadth and height. Subsequently, we will look in detail at the themes of depth which are directly concerned with the communication between inside and outside.

The Breadth Theme

The Vertical Tripartition

The extension of the wall is linked to the meeting between its own surface and the spaces to the right and left —in other words, the terminations of the wall to each side: its corners. Consequently, the wall’s terminations represent separate fields of energy on the surface. These fields decide whether the wall will continue or whether it will stop, and if so, in what way.

But the surface of the wall which is stretched between the corners is also divided into various fields of energy. Our attention to the corners is drawn to each side from a common starting point located centrally on the wall. This means that the effect of the termination in relation to the extension of the wall is read against a central field on the surface where the transitional areas on either side are interpreted as mediators between the corners and the central field (Fig. 145). This phenomenon, which describes a wall’s extension as the dynamic relationship between a central field and two peripheral fields, can be called the wall’s vertical tripartition.


The theme of breadth: the vertical tripartition.

This theme exists as an archetypical reference for the wall’s treatment of breadth and can be broken down and strengthened, depending on the desired expression. Every wall surface is comprised by these fields of energy. An example can illustrate this relationship. Windows are located in a straight, white wall (Fig. 146, a-d). In the first example, the window is located just to the side of the centreline of the surface (Fig. 146a). In the other example, the window is located so that its one edge is tangent to one of the corners (Fig. 146b). Both placements involve a tension that can be explained by the fields of energy which attempt to ‘press’ the windows in other directions. We see that this is correct, if the windows are moved, respectively, to the middle of the surface and a small distance in from the corner (Fig. 146 c and d). Now it is as if the tension is relieved, calm is restored. In the first case, the window corresponds with the centre of the surface; they are no longer at odds. In the second example, the window is moved into the transitional field between the centre and the corner. The corner’s field of energy remains in this way untouched and is allowed to exist as an independent element.


The theme of breadth: the three ‘fields of energy’, the middle section and the corners: a, b) The window in tension with the fields of energy, c, d) the window in correspondence with the fields of energy.

The Expression of The Breadth Theme

What is the reason behind our experience of the three energy fields in the surface of the wall?


The theme is by its structure symmetrical. This is a factor that can in itself be the basis for many explanatory models. Symmetry, according to Sven Hesselgren, is an image of fundamental order which we carry with us as a reference for all of our actions.1 This order emanates from the interface between our body and the force of gravity. We have earlier shown that phenomena such as balance and imbalance play a decisive role in maintaining our existence. We project this over to all of our surroundings, surface not excluded, as a reference for expressions and tensions.

A different and more anthropomorphic explanation takes symmetry to be a sort of image which we recognize from our own bodies. The shoulders terminate and secure our bodily extension, just as the corners of a wall terminate a surface. And the head designates the veritable centrepoint of our bodily structure, at the same time as it is our element of communication, just as the central section of a wall is the central location for the wall’s openings between inside and out.

The Centrum and Corners of a Space

It is likely that these more bodily explanations must be supplemented by a model that is based on man’s more general spatial need in relation to what the forms invite us to ‘do’ (see p. 19 ft) in other words, the relation between our perception of the wall’s elementary role as a spatial boundary and our existential need for space and house as something which we can conquer and fathom. The theme of tripartition reflects these needs projected on the delimiting surface of the space. The fact is that the centrum of a space is the place where we experience the essence of the space — the goal of our conquering, while the corners are where the space is given its perceptible form. In other words, the order of the space has a general existential meaning, while the space’s form tells us where we are and has therefore a more local and concrete meaning. In order to more thoroughly clarify this point, we wish to take a closer look at the centrum and corners of a space one at a time.

The Centrum of a space

Man will always seek a place in his surroundings where things achieve their main objective or essence.

What then is the essence of a space? It is its centrum. In other words, we first conquer a space when we are inside it. Specifically, this means that we are first in the space when we find ourselves at the midpoint between its corners. This does not necessarily mean its geometric centre. Rather it means that when one finds oneself in the periphery near the space’s walls and corners, the space is experienced primarily as another part which lies

in front of us, glides past us or comes towards us. It is at the centrum that we command the order of the space, which implies that the egocentric and spatial characteristics of the space converge.

The centrum has an existential dimension, in that the walls — the place where the conflict with the exterior takes place — are approximately equidistant from our person. Confronted with a house or space, an individual will therefore always seek to reach this space by projecting himself into its centrum. At the space’s periphery, the space will always remain a ‘there’ until we at its centrum can perceive it as a ‘here’.

But a centrum can also have an opposite role. Specifically, a centrum attains another completely different meaning at the moment it is conquered. After first having been the goal of our spatial perception inward, the centre now becomes the starting point of our expansion outward. At the centrum, the space is our personal space, which we survey and which is the same as our potential space of action."

The centrum of a space is then a place for the space’s communication with the surroundings both as a goal for our movement inward and as the starting point for our movement outward (Fig. 147).


The centre of the space: the goal for our movements inward and the starting point for our movements outward.

The Corners of a Space

Our relation to a space’s centrum has general applicability for all spaces. The conquering of a space’s centrum involves the experiencing of a space whether it is based on the point or the line, on repose or motion.

The corners, on the other hand, decide the form of the individual spaces. It is the corners or, in other words, the angles between walls which intersect, in addition to how many such corners there are and how large the distance is between them, which dictate whether the space is to enscribe a triangle, a square or rectangle, be polygonal, organic or circular, etc. In this way, the wall which creates the space is a function of the corners. Seen in this light, ‘solitary’ corners create invisible walls. This is what takes place when we stand in between the masts on a ship or between tree trunks in the forest and sense that we are enclosed as if in a room. We draw invisible boundaries from corner to corner (Fig. 148). But the inverse is also true, in that two wall planes which are located apart from each other are optically extended so that thay meet at the corner and thereby complete the space (Fig. 149).2


The space’s corners: freestanding columns act as corners in a space which is defined by invisible walls which span from column to column.


The space’s corners: freestanding walls are ‘extended’ thus creating corners which enclose the space.

Thus, corners and walls are mutually dependent on each other for the definition of a space. It is the corners which make the space a figure; it is the corners which tell us where we are.

The vertical tripartition, with its various fields of energy, can be explained as a result of the wall’s function as a spatial boundary. As a projection of the space in front of or behind the wall, the centre section will be important 153. The breadth motif: the largest windows in the centre (facade from Nancy, France).

in terms of accentuating the space’s centrum, while the corners on either side have an importance for the delimitation of the space.

Thus, the corner and centre areas have dissimilar meanings. The corners hold the space together. They are the requisite for the space’s individuality and ‘force of resistance’. The central area, on the other hand, is where the space communicates with its surroundings, i.e. the relationship between inside and outside (Fig. 150).


The theme of breadth and the relationship between inside and outside: the centre section opens up while the corners close.

Thus, the contrast between the corners and the central area as closed and open areas respectively, decides the theme’s expression — that which communicates what the wall ‘does’. Buildings with faҫades in which the corners are powerful and the middle section is open, provide then an immediate sense of both strength and publicness. The corners accentuate and close the interior space, while the centre section mediates and opens a dialogue between interior and exterior spaces. It is therefore common that the theme of breadth is especially emphasized in monumental architecture, such as churches, palaces, city halls and libraries (Fig. 151). These building types have a definite communal role which is expressed by these facades: On the one hand, they are to have open contact with the world, but at the same time, they must stand as a guarantee for the protection and stability of these common values.


The theme of breadth and the monumental building (Treasurer’s House by Ledoux, after L’Oeuvre et les reves de Ledoux).

The Motifs of The Breadth Theme

If the theme of breadth is studied further, it can be seen that the tripartition can vary between four fundamentally different motifs. Each motif helps either to accentuate or to weaken the basic meaning of the major expression. We can call these motifs, (a) the breadth motif, (b) the split motif, (c) the right motif, and (d) the left motif (Fig. 152 a-d).


The theme of breadth’s motifs and their expressions: (a) the breadth motif, (b) the split motif, (c) the right motif, and (d) the left motif.

In the first two motifs, either the middle section dominates over the corners, or the corners dominate over the middle section. In the two last motifs, the middle section is the same size, but varies between lying to the right and to the left of the surface’s centreline.

This means that the expressive nature of the first two motifs is a variation over the theme of opening/closing within the fundamental theme of symmetry, while the expressive nature of the second two motifs is characterized by the dissolution of the theme’s symmetry.

The Breadth Motif

In the first motif, the middle section dominates (Fig. 152a).

This domination can be accentuated in many ways: The windows in a wall can be larger and greater in number at the centre than at the corners (Fig. 153). The column spacing in a colonnade can be greater at the centre, while they are brought closer together nearer the corners (Figs. 154, 155). A plaster wall can be accentuated with rustication along the edges. Often, the entire middle section is also thrust forward, while the corners are held back (Fig. 155).


The breadth motif: the largest windoes in the centre (facade from Nancy, France).


The breadth motif: colonnade and corner resolution (Parliament building in Helsinki, Finland by J. S. Siren).


The breadth motif: The projected centre portico (Villa Rotunda by A. Palladio).

This motif emphasizes the public character of a building: the communication between inside and outside is increased. The motif is generous and receptive, the entire building expands outwards either by ‘pushing’ the corners to the sides, or by springing out in order to meet us.

The Split Motif

The opposite is true of the split motif. Here, the corners dominate toward the middle, and the open field is pressed together (Fig. 152b).

While the former motif pointed, toward openness and expansion, this motif points toward an increasing closure, either by squeezing the middle section together or by pulling it back in relation to the corners.

An example of the first is the city gate into the Spanish city of Lugo (Fig. 156). The powerful, round corner towers, which seem over-dimensioned in relation to the narrow section in the middle, express an almost threatening re-enclosure, which seems to squeeze us together. At other times, the corners can spring forth as protrusions, bay windows or towers in relation to a middle section which lies to the rear. This is the formal language of the castle in which corner towers protrude as active protectors of the castle walls.


The split motif: Entry towers (city gate in Lugo).


The side motif: Notre Dame du Haut (photo from Bolle-Redat, Notre Dame du Haut, Ronchamp).


The side motif: villa project from Downing, The Architecture of Country Houses).

The Side Motifs

Common to the next two motifs is that the open middle section is located either to the right or to the left (Figs. 152c, d, 157, 158).

In this way, the corners are also different. The one is strengthend while the other is weakened. Thus, our experience of outside and inside is different from what it is in the symmetrical variant. In the symmetrical facade, we were led into the space’s centrum. In the asymmetrical facade, we are led in indirectly. Out of pure impulse, we gravitate toward the strongest corner. Why? Because the strongest most clearly characterizes the interior as a delimited and secure place and thereby makes manifest the intention of our motion inwards. In other words, a localized, subjective centrum is created in relation to this corner, which, in addition to the general midpoint, attracts our attention. An asymmetrical placement between open and closed areas elicits therefore a more private character than does a symmetric placement.

It is obvious that the strength of the accentuated corner, and with it our impression of the interior space itself, varies according to the articulation employed. But in addition, the placement either to the right or to the left of the centre line will, of itself, influence our impression.

We have earlier emphasized the differentiation inherent in the relationship between our right and left sides.3 Rudolf Arnheim has shown that we project these characteristics onto what we see. That section which corresponds to our right side is immediately understood as being the strongest. An example of this is Heinrich Wollflin’s analysis of the composition of Raphael’s Sistine Madonna.4 He compares the original version with a mirrored image of the painting (Fig. 159 a, b). In the original painting, the composition is balanced in spite of the formal asymmetry which is brought about by the overdimensioning of the papal visage at the lower left. In a mirror image of the painting, the entire composition ‘tilts’ over to the right. The reason for this is the psychological weighting of the right, to which is added the papal visage.

159 a-b

The difference between the right and left side: The Sistine Madonna, (a) normal, (b) mirror-image (from Arnheim, Art and Visual Perception).

If we return to the facades, we will see that the movement inward is terminated more ‘naturally’ if the strongest corner lies to the right rather than to the left (Fig. 160). If it lies to the left, it is as if the visual image is about to re-establish the symmetry, as with the balance in Raphael’s painting. It is this differentiation which Kandinsky characterizes by referring to the right side as the direction home (nach Hause), in other words, that which leads to a terminated foundation. On the other hand, the left side is the direction toward the earth (zur Feme) which is the more free, open dimension.5


The difference between the right and left side: above, ‘strongest’ to the right, below, balanced.

The Height Theme

The Horizontal Tripartition

We have seen that a wall’s theme of breadth deals with spaces to either side of a centre. The theme is concerned with the wall’s expanse. The theme of height reflects the wall’s relationship to up and down. This involves the meeting between the wall and the earth and between the wall and the sky. In architecture, this is the same as the meeting between the floor and the roof (Fig. 161). Again, we can divide the wall into three fields of energy: the upperfield belongs to the roof, the lowerfield belongs to the floor, and the middle field mediates the transition between these two. This theme is referred to by Goethe as architecture’s ‘three original concepts’, specifically, ‘the base, the column (including the wall), and the roof’.6 We shall refer to this as the wall’s horizontal tripartition.


The theme of height: the horizontal tripartition.

If we imagine these fields to be of exactly the same size, they would nonetheless express different motion tendencies within the whole. While the upper field tends to seem lighter and to rise upward, the lower field tends to seem heavier and to sink downward. In the middle field, the tension between these tendencies is interpreted. Thus, it is the middle field which ‘decides’ whether the wall’s character as a whole is to be dominated by the upper or lower field, and whether they are to be linked or separated.

Again, a simple example can illustrate just how an element in the wall can be influenced by these motion tendencies relative to which field they are located. Again we consider a window set in a vertical, white wall.

In the first instance, the window is located in the middle of the wall, in the second, just under the central section, and in the third, just over the central section (Fig. 162 a-c). In the first example, the window is in a sort of ambiguous in-between situation: the upper and lower fields are the same size. In the second instance, the window is about to sink; it is being pulled toward the ground. In the third example, the window is about to free itself from the ground and rise upward.


The theme of height: three fields of energy’: (a) in tension, (b) rising, (c) sinking.

In addition, we see that the window creates motions in the wall itself, in that the window heightens the meaning of the middle field as the interpretive field of energy. The ambiguity in the first instance is due to the fact that none of the tendencies, either in the upper or lower fields, achieve dominance. In the second instance, the window draws the entire upper field down with itself, it becomes ‘heavier’, while the lowerfield is pressed yet more compactly together. An impression is formed that not only the

window but the entire wall seems to weigh downward. The opposite takes place in the third example, in which the lower field rises up together with the window, which in turn makes the upper section yet more narrow and thereby lighter’.

The Expression of The Height Theme

What is the cause of these experiences, and why are the three fields perceived dissimilarly in terms of their motion tendencies? We have all experienced the difference between up and down — down is the direction of the ground and the earth, up is the direction of the sky and air. The force of gravity teaches us what this implies; up is light and free, down is heavy and bound. Therefore, the wall that links these realities takes on a differentiated life, influenced by which of the two realms it is to meet.

Now, it seems that form can give two opposite motion impressions. The one tends downward, a result of the actual pull towards the earth which gravity exercises. The other tends upward and results as a purely experientially-defined contrast to the pull of gravity. We will attempt to describe these two motion impressions individually.

Gravity and Motion from Above Downward

Physically speaking, the ground draws all things to itself. A wall’s own weight is greatest toward the bottom and decreases upward. Its role as the bearer of the roof follows the same laws. It leads the weight of the roof downward. Also in terms of foundation, a free-standing wall must be tied into the ground in order to stand.

It is these measurable and empirical laws which explain why the tension in a figure on a surface always refers to a downward reality (Fig. 163). This downward reality, the ground on which we walk, is a requisite for our daily reality. A vertical rhombus which is shown alone in a composition seems to balance on a point: the point is the only contact that the form has with the ground — the invisible reference for the entire expression of balance. If the rhombus is positioned at an angle, it immediately seems out of balance: it is about to fall to the ground.


The balancing square.

This pull towards the earth also explains why a square has to be made a bit taller than it is wide if it is to appear as an accurate square (Fig. 164).7 This addition of height is necessary in order to counteract a flattening out of the form which will arise due to the impression of downward force. A form which follows this tendency, for example a horizontally oriented rectangle, will seem bound to the earth and at rest.


The corrected square.

The Experience and Motion from Below Upward

Forms which objectively speaking sink towards the earth will in this manner seem to be fighting against this pull by giving us the impression that they do not sink, but rather rise up from the ground (see p. 75 f). The tendency is experientially ratherthan physically determined. Vernon Lee offers the following description in an analysis of ‘the rising of the mountain’.8

What we are transferring...from ourselves to the looked at shape of the mountain, is not merely the thought of the rising which is really being done by us at that moment, but the thought and emotion, the idea of rising as such, which had been accumulating in our mind long before we even came into the presence of that particular mountain.

If we return to the wall, we will see the same tendency there. There are especially three conditions which determine such an impression. The first is the wall’s function as the support for itself and the roof. The wall lifts both itself and the roof, it holds itself up by rising from the ground in order to carry the roof.

The second condition is related to the fact that the upper section of the wall approaches the sky. The sky, as opposed to the ground which binds, is the infinite space which opens. The sky is for our experience freedom’s space, as opposed to the earth to which we are bound by gravity.

The third condition, which defines the expression of ascent, is existentially determined and involves a shared experience which has to do with all of the forms in our surroundings with which we identify ourselves. This is based on experiences which suggest that in order to live, we must defy gravity daily. Gravity will always draw objects and ourselves down, representing in the extreme an annihilation of existence. That something stands up against this force by rising up is therefore synonymous with survival — and survival implies freedom from the earth. A vertical wall which is higher than it is wide has therefore a completely different character to that of a horizontal wall: it is the manifestation of the rising — it is victorious and free.

The Motifs of The Height Theme

The described relation between the upper and lower parts of a wall as the lighter and heavier, respectively, is the canon for the experience of the wall’s vertical expression. In this light, this impression is also involved in establishing the experience or the relation between inside and outside.

However, the theme of tripartition can be interpreted in many ways. Four motifs can be seen as dissimilar in principle. We can refer to these as (a) the rising motif, (b) the sinking motif, (c) the split motif and (d) the opening motif (Fig. 165 a-d).

165 a-d

Theme of the motifs of height and their expressions: [a) the rising motif, (b) the sinking motif, (c) the split motif, (d) the opening motif.


The rising motif in a gable: City Hall in Greifwald, (from Pothorn, Das Grosse Buch der Baustile).

In the first two motifs, the middle field is of a constant size and domination.9 The difference between the motifs comes about as a result of the middle section being moved either above or below the wall’s centreline. The next two motifs come about as result of the variation of the middle section’s width, while its placement in the middle of the wall remains the same.

The Rising Motif

In the rising motif, the middle field is pushed upward in relation to the wall’s centreline. In this way, the lower field becomes the largest, the upper the most narrow, while the middle section has an in-between size. Read from above, we see that the weight increases as we approach the ground, completely in line with our experience of gravity. But read from below, we see that the surface lightens towards the top, the form rises from the ground. The expression of the rising motif lies in these two contrasting motions. A wall that rises up seems well anchored and heavy, while at the same time upright and free. It gives the observer an impression of both secure solidity and proud stature (Figs. 165 a, 166).

The Palazzo Medici in Florence (1444) is a classic example of the use of the rising motif (Fig. 167). The facade is divided into three storeys. The ground level has been given a closed and heavy character with large coarse rustication. Above, the wall has been made progressively lighter from simple stone outlines on the middle storey to a finer plaster surfacing on the upper. In addition, the windows are fewer and more simple on the ground level and are given a richer character on the upper floors.


The rising motif in a surface (Palazzo Medici by Michelozzo, from Koepf, Baukunst in Funf Jahrtausenden).

We can see that the rising motif also presents possibilities for penetration in that the upper part is open while the lower is heavy and closed. This means that the part of the wall which meets us on ground level stops us, while the upper sections belong to the interior spaces that in this manner open outward. Solidity and pride can therefore be traced to the wall’s own variation between solid and void: solidity because the lower sections make it difficult for us to penetrate, and pride because the interior spaces are elevated and open themselves outward ‘from the heights’.

Typical examples of such an effect are the palaces of the Baroque after Bernini (Fig. 168). The floor is often made of massive stone, which makes the building unaccessible and heavy. The residential floors above, on the other hand, are dissolved in large windows and often joined by high columnar orders. Therefore, the character is proud and upright with the princely chambers as towering symbols of the importance of the residents.


The rising motif as expressed in a plastic form (Main Libary in Vienna by Fischer v. Erlach Jr., from Koepf, Baukunst in Funf Jahrtausenden).


The sinking matif (Palazza all Colonne in Rome by B. Peruzzi).)

The Sinking Motif

The sinking motif represents the opposite of the rising motif. The middle field is now drown below the wall’s centreline so that the section above becomes the largest, while the lower section becomes the most narrow. The result is a sinking wall: both because the upper field pushes downward with its weight, and because the narrow bottom field will be perceived either as a base which is being pressed together against the ground, or as a base which is about to sink into the ground (Figs. 165b, 169).

This threat in the sinking motif is consciously emphasized in certain buildings. A typical example is Baldassare Peruzzi’s faҫade for Palazzo Massimi in Rome (1535) (Fig. 170). The rules for Classical bearing have been broken. Instead of letting the columns rise freely over a heavy base, now it is the columns which have been bound and pressed downward by a high and heavy wall above. Another example whose character lies in the tension between pressure and resistance, is the Artist’s House in Oslo, by G. Blakstad & H. Munthe-Kaas (1930) (Fig. 171). The building is dominated by a block-like top section with a brick fascia. This rests on white pillars which hold up a large baldachin which runs the entire length of the building. The baldachin tips effortlessly upward in contrast to the weight of the top which sinks downward.


The sinking motif (Norwegian folk architecture).


The sinking motif (Kunstnernes hus in Oslo by G. Blakstad & H. Munthe-Kaas).

The Split Motif

As opposed to the rising and sinking in which the middle field has a constant size and a varying position, the split and opening motifs are varied by the changing size of the middle field, while its positioning remains the same.

In the case of the split motif, the middle section is made narrow in relation to the upper and lower fields which both become large and dominating. The result is a middle section which seems to be pressed from both above and below in that the lower field seems to rise, while the upper field seems to sink (Fig. 165c). In the extreme, then, the effect can be both crowded and threatening. Crowded because the middle field separates contrasting forces, and threatening because the split appears to be on the verge of ‘snapping’ together at any time.

The split motif results in a closed faҫade. Such a wall terminates the motion at the ground level, while the upper wall sinks and closes. Two examples of the split motif can be mentioned. Many of Frank Lloyd Wright’s buildings have narrow, horizontal strip windows between downward sloping roofs and high wall planes (Fig. 172). In this manner, the interior is expressed as being on the verge of enclosure: It is as if the energy from within is ‘threatened’ by two opposing forces which give the entire relationship between inside and outside an unresolved tension.


The split motif (Project for Yahara Boat Club by Frank Lloyd Wright, from Kaufmann & Raeburn (eds.), Frank Lloyd Wright, Writings and Buildings).

Another example in the spirit of Wright is a brick entry motif which leads into a sports hall in Oslo (Fig. 173). Here too the middle section is comprised of a narrow band of windows. The main form is long and low, with a heavy character. The heaviness is further accentuated by the rustication of the lower section. The pressure from above is also emphasized, in that the windows seem to be forced down into the rustication. The door, which is located in the middle of the surface, acts as a contrast to the closed nature in the rest of the wall. As a whole, it is an expressive image of the solid/void dialogue.


The split motif (sports complex in Oslo).

The Opening Motif

In the opening motif, the middle section is broadened and made more dominating in relation to the narrow fields above and below. The result is a wall that seems to expand at the middle by rising and pressing the upper field further upward, while at the same time it pushes the lower field further downward. The expression is both rising and proud, but also opening and accessible for our penetration at the ground level (Fig. 165 d).

Two examples can be mentioned. The opening motif is often consciously used in Classical architecture as an entrance or facade motif. The University of Oslo, designed by Chr. H. Grosch in cooperation with Karl Friedrich Schinkel (1854), is dominated by the large temple front, its pediment borne by four Ionic columns (Figs. 174, 175). The columns rest heavily on a low compressed stair and rise up powerfully toward the roof with its ascending tympanum. The facade appears appropriately triumphal and upright.10 But it also seems receptive and open, as a public entrance must.


The opening motif (entrance to the University of Oslo by Chr. H. Grosch).


The opening motif (detail from the entrance the University of Oslo).

The opening motif can also be interpreted as splitting. A typical example is Giulio Romano’s house in Mantova (1544) (Fig. 176). In line with Mannerism’s taste for contrast, the entire faҫade expresses a struggle between ascent and descent. The descent of the lower storey is revealed in two ways: firstly, the square windows are located below the centreline of the base’s surface, and secondly, the cellar windows seem like lowered full-size windows.


The opening motif (Romano House in Mantova by G. Romano).

The rising character of the upper field is a result of the roof appearing as if it is about to be lifted off the wall below. This comes about through the interpretation of the frieze in the entablature as something non-bearing and open: The static reality of the field is reinterpreted by means of its garlands and round attic windows. Instead of resting heavily on the wall below, the roof is on the verge of floating overhead because the entablature is split in two. This conflict between ascent and descent is repeated in the architecture of the entire middle field. The field is divided by a row of high arches. Inside each arch are smaller windows, which are notably lower than the height of the arch. In this fashion, the arches rise in relation to the windows which sink: put another way, two motions in the same surface which make preparations for that which is above and below respectively.

The Depth Theme

Wall-form expressions may seem infinite, stretching from main, overall forms and their divisions right down to details such as texture and structure. Equally varied are the ways in which the inside-outside relationship is experienced. From this we understand that these experiences regarding both the walls’ breadth and height are dependent on how they are built in relation to such categories. For instance, the expressiveness of the city gate of Lugo (Fig. 156) as an example of the split motif, was dependent on the relation between main form (convexity), building system (massiveness), and material (blocks of stones). The same was true of the rising motif, which was exemplified by baroque palaces (Fig. 168) in which the expressiveness was dependent on the contrast between two building systems, i.e. the massive system of the ground floor and skeleton system of the top floors.

In the following study of the depth theme we will examine these categories in order to explain the expressiveness of the relationship between the spaces in front of and behind the wall. The categories will be divided into four themes: (1) the main form, (2) the building system, (3) the openings and (4) the articulation (Fig. 177, a-d). Each theme represents a principal set of motifs and variations representing the archetypes of the wall.

177 a-d

The themes of the wall: (a) main form, (b) construction system (massive-skeleton), (c) openings (doors — windows), (d) articulation, (dimensions — connections — divisions — textures — colours).

The first group deals with the wall’s main form (Fig. 177, a). This means how conditions such as height and width, slanting and curving influence the impression of the relative strength between inside and outside.

The second group covers building systems and deals with how wall forms are constructed (Fig. 177, b). Is the wall built as a solid slab, as a composite skeleton, or as a combination of both? Moreover, how will these systems affect our impression of the transition between inside and outside?

The third group concerns the openings (Fig. 177, c). Here we must differentiate between two types of openings, door and window, each of which affects our impression of the inside-outside relation differently. Both door and window are perceived as holes in the wall. Their effect, however, depends upon the form and profile of the hole, the location of the door or window area, and the frame around the hole (Fig. 177, d).

The fourth group, articulation, affects all three previous groups and concerns the importance of dimensions, the way in which various parts are joined together, the division of the wall, as well as texture and colours. In the following we shall begin by describing the expressions of motion, weight and substance conveyed in the main wall forms and continue with the wall’s building systems. In connection with the latter, we shall cover the impression given by the use of various types of articulation. The wall’s main form and its building system are to be seen as a ‘background’ for the openings. These consist of windows and doors, which will be the last group we shall examine, and these too will be seen in relation to various types of articulation.

Main Forms

A wall area, in principle, may be formed within eight different motifs (Fig. 178, a-h).

178 a-h

The main forms of the wall: (a) horizontal, (b) vertical, (c) flat, (d) convex, (e) concave, (f) straight, (g) leaning toward, (h) leaning away.

The first two are concerned with the relation between width and height, in that the wall’s main form is either horizontal or vertical. The next three motifs deal with the relation to depth, which covers the flat, the convex, and the concave main forms. The final three motifs deal with the slant of the wall, which means that it may be upright, lean toward us or away from us.

All eight motifs are first and foremost actual representations of fundamental motion situations, which we may characterize by using prepositions and words specifying directions. The first two describe a ‘follow along’ and an ‘upward’ motion respectively. The next three convey a ‘halting’, ‘advancing’ and ‘retreating’ motion. The last three depict a ‘neutral’ motion, a ‘tilting away’ and ‘downward’ motion, and a ‘leaning towards’ and ‘over’ motion. Assuming that we stand in approximately the same position in front of the walls, they will arouse motion impulses corresponding to these terms, which in turn create highly different impressions of the inside- outside relationship in depth. These, added to definite weight impressions and associated with corresponding environmental forms, will convey a specific overall expression for each of the eight wall forms.

The Horizontal Wall

The horizontal wall expresses weight against the ground. Its horizontality gives a compressed and compact first impression. It will, therefore, have a basically closed and delimiting character (Fig. 178, a).

The motion impulses aroused by this wall type will also increase its closed character. Because it stretches out horizontally the impulse is to follow along beside it in either direction. This phenomenon might be compared to a street that defines a space and as it stretches away before us entices us to follow along with it (Fig. 179). Such a space conveys no urge to pause, to turn and enter. A horizontal wall is, therefore, actually an obstacle, because the interior it hides is not our concern. It does not penetrate the wall to make its presence known but is led past us.


The horizontal wall and the motion alongside. (Brick country house by Mies van der Rohe, from Johnson, Mies von der Rohe).

A directional space invites us to enter through the ends. For this reason too, our attention will be drawn to either side as if seeking an entrance around the corner’, which is where the interior will meet us.

The Vertical Wall

Whereas the horizontal wall rejects and shuts out, rather like the invisible wall between two passing strangers, the vertical wall is communicative. There are three main reasons for this.

The first is its weight expression. The horizontal wall weighs down upon the ground and shuts off, whereas the vertical will always seem lighter because of its rising effect. This wall seems to lift itself upwards and open up verticality.

The second reason is the motion expressed in the vertical wall. The wall itself illustrates the vertical, which marks point and line. Whereas the horizontal wall spreads movements, the vertical rising wall collects them (Fig. 178, b). The horizontal wall draws attention to the corners at each end, while the rising wall concentrates attention around the centre of the area. And, the centre in an area is the open’ part, the place which communicates with the surroundings. Throughout architectural history we find many examples of the exploitation of characteristic differences in verticality and horizontally oriented walls. The Swedish architect, E.V. Langlet’s theories from 1867 are typical in this respect and are revealed in several contemporary houses.11 In most of them the vertical wall is also the open wall. It is here the main entrance is to be found and here stairs, verandas and balconies are concentrated (Fig. 180). The horizontal side wings, on the other hand, are more compact and inaccessible with fewer windows and larger areas of solid wall. Here the more private and closed off spaces are situated.


Horizontal and vertical walls in combination: the horizontal is ‘closed’, the vertical is ‘open’ (project for a villa by E. V. Langlet, from Nordin, Trabyggande under 1800-tolet).

The third main reason for the vertical wall’s communicative content is that, like a tower, such a wall is the image of the erect, standing figure. Not only does it attract our attention but also corresponds to certain anthropomorphic conceptions of ‘great men’. Just like two people who stand talking, the vertical wall concerns us directly and personally either as something threatening or conversational. The first may often be the case with tall houses in an open landscape. Like some dominating landmark they are the focus of all attention, whereas horizontal rows of houses will direct our attention on past and away.

The Flat Wall

The flat wall tells us nothing about the inside-outside relationship. It merely closes off (Fig. 178, c). A stiff and impassive plane, it is like neutral theatre backdrop where activities take place both in front and behind. It is just like a projection screen, according to the Norwegian architect Kjell Lund, and may be compared to drawings on white paper where: ‘the drawings and the paper constitute a whole — the black gives value to the white and the white gives value to the black’.12

By itself, the flat wall is just a background. In order to express itself, it is dependent on how the surface is treated and the openings are placed (Fig. 181).


The vertical wall and the vertical space (Medieval house from Amsterdam).

The flat wall in which stress and counter-stress are in balance is, therefore, a typical man-made product, maintains Hugo Haring. The curved wall, on the other hand, is related to nature itself.13 In what he says, Haring emphasizes the essence of what we see in nature’s own formations — in the mountain, in plant life, and in the human body. The curve constitutes the sum of counteracting forces in the life process itself. The convex curve restrains the dominance of forces from within, while the concave curve receives the dominating forces from without.

The Convex Wall

When facing a convex wall one visualizes the interior space behind it as strong and dominating. This impression is conveyed either in the form of an expanding movement toward us or as an enfolding movement protecting the interior space from us (Fig. 178, d).

In the first case, we feel as if the interior space itself resists our approach. We are kept at a distance, stopped by the interior’s own force.

An example of this effect is found in the large curved faҫade of the Norwegian Automobile Association in Oslo by F.S. Platou (1974) (Fig. 182). It is a corner building at the end of a wide street. Thus, the convex faҫade has a double effect, because the curve is not only a projecting goal for the street but also ‘splits’ the course of the street. Let us imagine ourselves walking along beside it. There is a feeling of being led by the form itself. The wall guides us around the corner, we say, and thereby indicates that the movement is not our own volition. This experience might be compared to the difference there is in moving around amongst the undulating walls of Reima Pietila’s student residence (1967) and in following along beside them on the outside. In the first case the walls seem to fit us like a glove. It is we and our own movements which generate the flowing motions of the space. Outside, on the other hand, our movements seem to be guided by the inside space itself, which has dictated the form along which we are following.


The convex wall (NAF Building in Oslo by F.S. Platou).183. The concave wall (Baroque palace in Munich by J. K. Schlaun)

The convex form, in other words, seems always to be filled with something. ‘Convexity tends to win out over concavity’, according to E. Rubin.14 Arnheim maintains that, like children, we perceive the convex as a solid and concrete thing. In contrast the concave is perceived more as a background for things.15

Consequently, the strength indicated by convexity corresponds to the form’s own substance. In this manner, the convex surface of a stone represents the delimitation of that inner weight which is the nature of stone. Similarly, the curve of a tree trunk, a muscle or a column, will display the essence of its form — for the tree, sap, for the muscle, blood, and for the column, an inner cohesive core providing structural integrity. Therefore, a convex wall will be characterized not only by an outward expansion but also by an inward-looking concentration.

The Concave Wall

When we face a concave wall, the feeling is one of being received/the interior space seems to yield to our forward movement (see p. 105 f). As a consequence, the wall is a visualization of two expressions: the embracing and receiving, and the yielding and pliant (Figs. 178e, 183).


The convex wall (Baroque palace in Munich by J. K. Schlaun)

A person receiving us with open arms represents friendliness and security. Arms are clasped in order to hold one firmly and to give nearness and protection (Fig. 184). The cave is a space conveying a similar feeling. The harbour too is one of nature’s forms which depicts the protective aspect of the concave curve. The harbour is a haven of safety in stormy weather giving security by its very form. ‘The harbour in fact, represents one of the most commonplace types and in many languages the word ‘harbour’ is synonymous with safety and belonging’.16 In this sense, the concave wall expresses benevolent expansiveness — an outgoing gesture from the inside.


The wall as an embracing form (project for Piazza St. Pietro, Rome, in the shape of extended arms, from Hausmann, Baukunst des Barock).

But as stated above, concavity also depicts an inward penetration from the outside. Again the cave is a symbol of this. Its form may be interpreted as the result of an outside force — the solid mountain has yielded to a blow from without.

Seen in this way, the grotto with its spring is deeply meaningful as a symbol of concavity’s duality (Fig. 185). The cave-like form of the grotto reflects an intrusion whereas the water trickling forth is what this form ‘gives’ from within.


The grotto as a receptive and giving form (Fantana del Prigione in Rome, from Marton, Waters of Rome).

A combination of the convex and concave wall results in the broken or undulating wall. The undulating wall acts in its own particular way, in which the heavier and lighter parts balance each other in a continuous motion along the wall. This curving expanse reflects both interior and exterior space but is at the same time independent of both, because neither gets the upper hand and breaks through. As with the water’s surface, where each movement immediately exacts its counter-movement, the undulating wall keeps a dynamic balance between inside and outside.

Sigfried Giedion emphasized the undulating wall as one of architecture’s basic features or ‘constituent facts’.17 It dominated the Baroque, a typical sign of that period’s sense of a continuous spatial treatment. But, it is also used frequently in modern architecture, as seen in many of Alvar Aalto’s buildings (Fig. 186). His Finnish pavillion at the New York World’s Fair (1939) is a good example of the use of the undulating wall’s qualities.

An inclined wooden screen three stories high embraces the interior space in a freely drawn curve. The screen consists of three sections, each cantilevered over the other; at the same time the whole structure leans forward, intensifying thereby the impression of continuous movement.18


The undulating wall and the continuity along the surface (Finnish pavilion in New York by A. Aalto, from Joedicke, Geschichte der Modernen, Architektur).

Another variation of the undulating wall is to be found in the architecture of Paolo Portoghesi and Vittorio Gigliotti. Here, as a rule, the walls are interpreted as divided waves, with continuity assured by concave segments and arches with intervening open slits for doors and windows (Fig. 187). The result is an interior which presses forth between inwardly curving sections. In an expansive sense the interior is opened towards the outside. The inside-outside relationship, however, is not so uncomplicated as it was with the directional wall planes of the 1920s in which the overlapping of exterior and interior space was free and in equilibrium (Fig. 179). Portoghesi and Gigliotti restrain the exterior space. Tense curves simultaneously affirm and deny the impact of the surroundings. So, concavity is a necessary corrective to openness. Space is directed outward, but at the same time the walls protect against any inward movement. Hence, the walls’ function as a delimiting element is emphasized, but without expressing rejection, because within each concave form there is a receptive and pliant gesture of ‘welcome’ to the surrounding environment.


The undulating wall as a divided surface (Casa Andreis by Portoghesi/Gigliotti, from Norberg-Schulz, Existence, Space and Architecture).


The forward leaning wall (folk architecture, from Cornell, Byggnadstekniken).

The Slanting Wall

The vertical wall, just like the flat wall, balances between contrary motion tendencies. Thus, a wall section still upright, but which we know is about to fall, will convey an insecure, tense feeling. Will the wall fall our way and destroy us or the other way and we are ‘saved’? In these two intuitive reactions we find the expressions of the slanOed wall’s two basic variants (Fig. 178 g, h).

A wall that seems to tilt over us is threatening (Figs. 178g, 188). This sensation may be illustrated by the way people react to the leaning tower of Pisa (1174). Visitors very seldom pause immediately beneath the leaning side. They feel safe only when at a certain distance and preferably on the opposite side of the tilt. The wall that leans out over us is, therefore, seldom found in architecture. A well-known example, nevertheless, is Alvar/Aalto’s exhibition wall from the New York World’s Fair (see over) (Fig. 186). There the form’s boldness results from the way in which it leans out over the floor. With this solution Aalto exploited an ever-present technical problem in all exhibition architecture. Now all photographs could be clearly seen from below. He used this solution to create an expressive architecture in which the threatening aspect was converted to one of tense excitement. Imagine a corresponding wall but flat in form without.curves and horizontal divisions; we would feel ourselves to be in an uncomfortable and dangerous space. Aalto, however, by using a few simple effects transformed this basic impression. First the curves: they accentuate motions around us and alongside us and moderate the falling aspect by breaking it up. So, the sections: these divide the wall into three broad bands and thereby lighten the whole. This is further developed by the way the sections are composed of vertical struts against a pale background, which makes the whole seem to float above the floor. This last aspect is accentuated in that the wall stands all of three metres above the ground without visible supports or columns.

A wall leaning over us obviously threatens the space where we stand. On the other hand, the wall that tilts away from us no longer concerns us but threatens whatever is within the space on the other side (Fig. 178, h). Its expression, however, is completely dependent upon its articulation. It may be formed, in such a way that it is perceived as a heavy, massive form, such as we find in sloping stone bastions from the Renaissance (Fig. 189). Or, it may have a planar character like that of the shed roof (p. 365 ff). In both cases the form indicates a closed interior space. In the first case owing to the compact and inaccessible nature of the interior, the wall denies ‘penetration’ and encourages ‘scaling’. In the second case, the effect is that of a pitched roof which extends to the ground, thereby ‘turning its back’ on the exterior space (see p. 369 f).


The away leaning wall (Renaissance fortress, from Giedion, Space, Time and Architecture).

Building Systems: Massive System, Skeleton System, Infill System and Layer System

The wall’s main form is affected by various building systems. The expression of motion, weight, and substance in these systems can transform the impact of the main form.

By the term building system, we mean how the wall indicates the way in which it is built. In this context there are four motifs which predominate (Fig. 190, a-d):

  1. the massive system

  2. the skeleton system

  3. the infill system

  4. the layer system

190 a-d

The wall’s constructive system: (a) planar-wall system (cast wall — block wall), (b) skeletal system, (c) infill system (planar element in the skeleton — skeleton in the skeleton), (d) layer system (planar element in front of planar element — skeleton in front of skeleton — skeleton in front of planar element — planar element in front of skeleton).

In the massive system, the wall is built as a solid whole. This signifies that it is a compact mass from outer to inner surface and that all its parts are of equal value. In principle, there are two ways of building a massive wall. It may be formed as a solid mass [the moulded wall) or built of individual blocks laid one upon the other [the masonry wall). Equal distribution means that the structural capacity, both in terms of its own weight and that of the roof, is evenly spread throughout the whole of the wall.

In the skeleton system the wall is divided into separate units, each having a different role in transmitting vertical load. The basic unit of this system is a frame in which the lintel (the supported element) rest on two posts (the supporting element). In this system, therefore, support is guided in specific paths leaving the remainder of the area open.

In the infill system the supporting element is the frame. This system requires a main, primary skeleton to support the roof, while the opening is covered by a secondary wall, which may be either massive or skeletal. An infill wall presupposes that the secondary wall is subject to the main skeleton and that both are roughly on the same plane.

The layer system is composed of wall sections, juxtaposed in depth. Support may be carried out by either the front or rear sections or all sections together. There are four main variations in this system. If the wall has only two sections, one variant is plane behind plane, another is skeleton behind skeleton. The next two variations consist of a combination of skeleton and plane. In one, the skeleton is in front of the plane and in the other it is behind it.

The massive and skeletal systems are simple and unmixed, whereas infill and layer systems are combinations of the first two. In the inside-outside relationship this means first of all that all variations of the system are dependent upon the fundamental expression to be found in the massive system and in the skeletal system respectively. In the impression given by the infill and layer system there is the added factor of the way in which the elements are combined.

We shall see that the differences in expression between the massive and skeletal systems are determined by differences in support, which means that the relation between exterior and interior strength is affected by the systems’ vertical expression of weight. As for the infill and layer systems, there is, in addition, an expression of motion in depth as the eye is led inwards in different ways depending on the relationship between the layers.

In the following we shall first examine the infill and the layer systems in order to show how important their motion expressions are for the inside- outside relationship. Thereafter, emphasis will be on the massive and skeletal systems in order to find out how their expressions of weight affect our impression of the wall’s enclosure.

Following that we shall weigh the importance of articulation for these qualities: first in relation to the elements in the massive system and subsequently to those in the skeletal system.

The Infill System and Layer System

The infill wall indicates a balance between inside and outside (Fig. 191). This is because the framework signifies opening from the outside inward, while the infill element that is flush with the skeletal framework signifies a closed off interior. This is particularly apparent when the infill section is solid and contrast determines the opening-closing effect.


In the infill wall, exterior wall meets the interior wall in the same plane. The infill element belongs to the interior space while the surrounding framework belongs to the exterior space.

In the layer wall, this balance between inside and outside is substituted by a stage by stage motion inwards. This means that the outer wall, which belongs to the exterior, and the inner wall, which belongs to the interior, are combined in such a way that they delimit the space independently. Furthermore, the buffer zone between them may vary in depth and thus ‘participation’ in the spatial effect depends on whether the wall sections are adjoining or separated (Fig. 192, a-c).


In the layered wall, the outer and inner wall lie in parallel planes. Thus, the tension between them is a factor of whether they are, (a) adjoining, (b) pulled apart, or (c) formed independently of one another.

The following examples will illustrate how the inside-outside relation varies according to which of the layer motifs is used.

Plane on Plane

A modern example of a faҫade in which two solid planes are juxtaposed is Robert Venturi’s project for a town hall in North Canton, Ohio (1965) (Figs. 193, 194).


Plane in front of plane (pro-ject for a city hall in an Ohio city by R. Venturi, after Venturi, Complexity and Contradiction in Architecture).


Plane in front of plane (Kresge College by Ch. Moore, photo by M. Baer).

The outer plane is a large free-standing screen with a wide entryway and an enormous arch as the only hole. Behind this is the next layer, the office building wall with a plain, regular window system. The walls are separated to allow for a walkway between them. This solution responds to the inside-outside relationship on all of three levels. The frontal facade, with its large simple figures, is meant to relate to the more distant environment, the city’s squares and streets. The walkway forms an intermediate layer between these and the building itself as it stands drawn back with its uniform window pattern.

Skeleton on Skeleton

The next variation is one skeleton placed in front of another (Fig. 195).


Skeleton in front of skeleton (Piazza d’Italia by Ch. Moore, from A.D. 5/6, 1980).

This has been much used in Gothic (Fig. 196) but also in Classical architecture, both of which exploit the possibilities of this motif to give a layered opening into a wall. An example is Andrea Palladio’s Palazza Valmarana, Vicenca (1566) (Fig. 197). Giant pilasters in the forefront, extending from plinth to cornice, determine the scale of both the street and the house. Behind the pilasters are two storeys in which the upper storey appears to be supported by lesser pilasters partially recessed behind the greater. Behind these again but still on the first floor lies yet another skeletal wall of Tuscan pilasters supporting a recessed mezzanine. Finally, even further in are deep reliefs and windows forming the fourth and final layer.


Skeleton in front of skeleton (Romanesque wall system, from Koepf, Baukunstin Funf’ Jahrtausenden).


Skeleton in front of skeleton (Palazzo Valmarana by A. Palladio from Koepf, Baukunst in Funf Johrfausenden).

The purpose of these recessed layers is the same as in Venturi’s project, that is, to attain a gradual transition between inside and outside. But, the effect is different. One plane section in front of another results in an additive juxtaposition of the outer and inner plane, and this is very effective, especially when facing a large, open urban space as in this Ohio city. Superimposed skeletons, however, give a greater perspective depth. An example of this can be found in the narrow street facing Palazzo Valmarana. Here, the Palazzo’s main faҫade helps to widen and open up the street.

Skeleton on Plane

To continue with the next layer variation, the skeleton placed in front of the plane, we find that the effect created is a contrast between an open, public frontal wall and a closed, private rear wall. The variant is the most widely used system in architectural history. It was reintroduced for the first time since Antiquity by Leone Battista Alberti in his Palazzo Rucellai, Florence (1451) (Fig. 198). The system’s characteristics already appear in the classical temple (Fig. 199). With free-standing columns forming an open colonnade outside the massive cella wall, the duality stands out clearly. A closed inner world for the gods alone was surrounded by a wall of widely placed columns, which established contract with the outside world.


Skeleton in front of plane (Palazzo Rucellai in Florence by L.B. Albertl).


Skeleton in front of plane (Hadrian’s Temple in Rome, from L’Orange/Thiis-Evensen, Old- tidens bygningsverden).

The contrasting effect of the variant is often emphasized in the wall’s articulation itself. In Filippo Brunelleschi’s Foundling Hospital facade, Florence (1419), it is emphasized by the way the strict, precise and dark arcades facing the square clearly belong to the public sphere. In contrast, the massive wall behind, with its ‘unsystematic’ openings indicates another and more private world within the walls (Fig. 200).


Skeleton in front of plane (Foster Home in Florence by F. Brunelleschi).

Plane on Skeleton

We come finally to the fourth variation of the layered wall. In this combination a solid plane is placed in front of a skeletal section (Fig. 201). One of two things is necessary to enable us to comprehend the variant.


Plane in front of skeleton (from Bryggen in Bergen, Norway).

In the first place, the plane must be either perforated or transparent to allow us to see the skeleton behind. The latter is true of the curtain wall, which is in principle a glass front attached to an inner structural skeleton (Fig. 202).


Plane in front of skeleton (curtain wall enveloping stair tower, Werkbund, by W. Gropius, from Giedion, Space, Time and Architecture).

The other prerequisite is that the outer plane be membrane-like in character or in other words seem thin and non-supporting. Once again the curtain wall is an example. Asa protective skin it cannot stand on its own but is completely dependent upon the inner framework. Just as the skin on our own bodies, every membrane-like wall will indicate an inner supporting structure.

The Massive System and The Skeleton System

We have seen that the massive wall is both supporting and delimiting, whereas the skeletal wall differentiates between delimiting and supporting. In the inside-outside relation this means that the massive wall lends weight to the spatial boundary itself and thereby closes the space (Figs. 203a, 204).


The planar system is both delimiting and supporting: (a) it encloss while the skeletal system separates between delimitation and support, (b) it opens.


The planar system as an enclosing wall (a block wall from Palazzo Pitti in Florence by F. Brunelleschi).

The skeletal system, on the other hand, concentrates support in a primary load-bearing system which frees and lightens the spatial boundary and thus opens the space (Figs. 203b, 205).


The skeletal system as an open wall (Snyderman House in Indiana, by P. Eisenmann).

Throughout architectural history, not only de facto but also symbolically, the weight expressed in these systems has been responsible for the opening and closing of the wall.

The Massive Wall and Closure

The closure effect of the massive wall has been emphasized in buildings in which the main purpose is to isolate and protect the inside space. The massive wall, therefore, is particularly associated with concepts of fortresses, city walls and prisons. The city wall should not only be strong in reality but strong in appearance as well.

A material protection and at the same time a psychological defense, it was meant to arouse fear in the enemy by its terrible aspect and give the combatant a sense of security so that he could defend himself and strike as well. In the Middle Ages fortifications were essentially defensive; in the Renaissance they were both defensive and aggressive.19

Furthermore, in the theory of the 1800’s:

The exterior of a prison should be formed in the heavy and sombre style, which most forcibly impresses the spectator with gloom and terror’20

The material we immediately associate with a massive wall is stone. Inherently, stone has the same quality that the massive wall in its extreme form tries to attain: heavy and closed, earthbound and immovable.

The Skeleton Wall and Opening

The skeletal system is the very symbol of open space. The skeleton has the characteristics of a tree: dynamic, continuous, and ever-growing, in contrast to the heavy inertness of stone. Like a tree, the skeleton ‘when built of wood …. gives a character of lightness’ (Fig. 206).21 The skeletal system in itself, therefore, is an expression particularly well suited to buildings with an expansive content. The villa exemplifies this type of dwelling and already in Antiquity was associated with long colonnades extending into the landscape (Fig. 207).22 Such elements express the nature of the villa, says Alberti, ‘where you have as much freedom as you have obstruction in town’.23


The openness of the skeleton (Swiss-style villa in Oslo, Norway).


The openness of the skeleton (painting of portico villas from Pompeii, photo by Alinari).

A similar desire for openness lies behind the use of the skeletal system in Gothic cathedrals. With walls broken up and with soaring columns it was built in spite of the stone. Its expansiveness lay in the spiritual truth which the building was meant to symbolize, a truth to be spread throughout ‘the entire world’ (see p. 57).

The open character associated with the skeletal system has the column as its most characteristic element. The column is regal it was said in Antiquity, ‘oedificiorum purpura’.24 In this, not least of all, lay the acknowledgement of the column as representing the public aspect. It is the column that opens a building by making the walls redundant, and it is the column itself that is the intermediary of motion between outside and inside (See also columns p. 195 ff). Typically, therefore, it is the entrance faҫade, which is most frequently decorated with columns. This was the part facing the world, the part to be seen by the public.

It is in this context that the development of the fictive skeletal system must be understood. From the Renaissance, as first seen in Palazzo Rucellai, buildings of special importance have been decorated with entire storeys of non-structural rows of columns and beams (Fig. 198). The skeleton was an attempt to give a more public appearance to key buildings in medieval Florence, a city accustomed to heavy, inwardly orientated palaces such as Palazzo Vecchio (1340) and Palazzo Medici (1444). These buildings derived from ancient towers, fortresses and city walls.25 The new buildings had their roots in the urbanism of Antiquity. This urbanization, brought about by the introduction of the skeletal system, is of great importance for both church and palace in the entire subsequent development of architecture. The skeletal system’s expressive possibilities attain perhaps their greatest perfection in Michelangelo’s Campidoglio project, in which mighty pilasters become an integral part of both building and square (See also p. 85 f) (Figs. 101, 102).


Piazza Navona in Rome is one of the places that most clearly demonstrate the way in which skeletal and massive systems may express openness and enclosure of buildings (Fig. 208).


Interplay between the public openness of the skeletal wall and the private closure of the massive wall (Piazza Navana, painting by P. Pannini, from Hausmann, Baulunst des Barack).

The piazza takes its form from a stadium built at the time of Domitian (first century) and is surrounded by buildings from both early Renaissance and high Baroque. All the private dwellings are massive and closed, compact plastered structures broken only by rows of plain windows. Inserted into this world of privacy are two churches, St. Agnese (1657) and Nostra Signora del St. Cuore (1450). Their rich systems of pilasters, attached columns, lintels and portals indicate another and more public world than that of the neighbouring buildings.

Flanking St. Agnese is Palazzo Pamphili (1650), which differs from both the private houses and the churches in its combined public and private character. The central section is opened up by pilasters while the wings retain the form of solid volumes. This partly public expression illustrates the building’s function, since the house was both the residence and the symbol of the papal family.

In the following we shall concentrate in greater detail on the massive and skeletal systems by examining the effect of articulation. It is a question of how articulation, which particularly concerns texture and weight expression, can influence the system’s basic interpretation of the inside-outside relationship.

We begin with the massive system and question the importance of thickness, surface treatment, building method, and transparency. We continue by examining the skeletal system, with the main emphasis on the articulation of columns and beams. Finally we shall study the importance of colour in relation to both systems at the same time.

The Massive System

The inward or outward expression of the massive wall depends upon its appearance of lightness or heaviness. In other words, if a massive wall seems solid and heavy, it will impart a feeling of being ‘impenetrable’ and stop us, while at the same time conveying an impression of great structural capacity. On the other hand, if it seems thin and light, there seems to be no difficulty in ‘breaking through’, and its load-bearing capacity seems reduced.

In the massive wall four conditions determine this weight expression: thickness, surface treatment, bujjding method’, and degree of transparency.


It a massive wall seems thin, it conveys a light and therefore ‘open’ feeling. The reverse is true if it seems thick (Fig. 209).


The plane’s thickness and the expression of weight.

A thick wall corresponds to something inert and closed in that thickness indicates compactness and thereby inner resistance. A thin wall, on the other hand, has more the character of a light film and as a result seems far more vulnerable.

Just what determines this impression of thickness?

The main form is important because of its overall effect on weight character and thus on how we feel the wall can be handled — this in the widest sense. As a result, a curved wall seems thicker than one which is straight, a sloping wall thicker than a vertical one, a low one thicker than a high etc.

The openings too are very important, because by the depth of their jambs the thickness of the wall can be clearly read. This will be treated in greater detail in the chapter on windows (See p. 253 f and p. 259 ff).

Furthermore, the material of the wall plane is important in expressing weight. Materials span all the way from natural materials such as stone, clay, and wood to processed material such as concrete, iron, glass or plastic. Some of these are moulded, as with concrete and plastic, others must be constructed, as is the case with stone and wood. Each group of materials has its own special character and, providing they are left in their natural state, give varying impressions of thickness.26 The main reason for this is to be found in our own experiences with the materials regarding both their use and their quality. Hence, a wall of iron, glass, plastic or laminated wood will immediately be thought of as thin and membrane-like, whereas a wall of concrete or stone strikes us at once as thick.

Experience with these materials, however, has taught us that thickness does not always correspond with weight. An iron wall with the same thickness as a plastic wall will always seem heavier. Similarly, a thick wall of concrete blocks will seem considerably lighter than the thinnest steel wall.


What then about the effect of the surface on the wall’s expression of weight?

There are four factors which determine this: texture, pattern, relief and colour.

Colour, as we have already said, will be treated in a later chapter (See p. 240 ff). Texture may be smooth, fine or rough, while the effect of various patterns depends on three interrelated conditions, namely: the pattern’s order, its range of motifs, and its plasticity. The latter is a question of whether the pattern is flush with the wall or recessed.

Building Method

All four factors, texture, pattern, relief and colour affect a wall’s expression of weight regardless of whether it is moulded or block constructed. Nevertheless, the building method in itself may cause different expressions of weight. To the eye, the moulded wall will have a smooth surface, while a masonry wall will reflect the pattern determined by the size of the blocks, the bonding, and the texture. If we compare two such walls, the block wall with its chequered surface will immediately convey a thicker and heavier impression than the moulded wall with its smooth surface. What is the reason for this difference? (Fig. 210).


The plane’s method of construction and the expression of weight.

Again we find a principal explanation in the phenomenon of how we relate to things. A smooth wall has no immediate connection with our world of everyday experience. It tells nothing of its weight or size. As a reference for scale it eludes us. Thus, frequently a smooth wall area, particularly if it is pale in colour, will seem larger than it really is. Optically it extends both upwards and outwards, which means that the surface effect is greater than the depth and weight effect.

The reverse is true of the block wall. Such a wall immediately seems denser and more compact. In contrast to the moulded wall the block wall consists of comprehensible individual parts to which we can relate (Fig. 211). A single block can be ‘read’. Its weight and durability can be evaluated by how it can be grasped or held, whether it is large and unmanageable or light enough to lift and control. These familiar units are piled one upon the other to form an integrated whole. In this way, the whole becomes comprehensible, which in turn means that the wall area is ‘identifiable’ by means of its individual parts (Fig. 212).


The plane’s method of construction and the expression of weight (half-timber construction with brick infill from Akershus Fortress in Oslo, Norway).


The plane’s method of construction and the expression of weight (Palazza Pitti, by Brunelleschi).

To simplify this analysis, we shall relate surface to each of the two wall types separately. In order to do this we assume certain factors to be constant, thus the main form of the walls described are all perpendicular and without openings.

The Moulded Wall

In the history of architecture there are three variations in particular which correspond to what we mean by the moulded wall: the stucco wall, the concrete wall, and the glass wall. The glass wall will be considered separately in the concluding chapter on massive walls.

As for the concrete wall and stucco wall, only the first is cast in a strictly technical sense, while the stucco wall merely gives that impression. Stucco is usually an outer covering of an inner composite wall of brick or natural stone. Strictly speaking, concrete walls are not all isotropic planes. Frequently they are reinforced with steel which technically puts them in a mixed category between a skeleton and massive systems.

In any case, the stucco wall and the concrete wall merely represent degrees of difference within the same expressive limits and in the following will be treated as if they were alike. We shall first examine them with different textural treatment and then with various types of patterns, in the latter case those which are flush with the wall surface and those forming reliefs.


Elias Cornell calls moulded walls neutral: ‘They serve as backgrounds for colonnades or create smooth sections between windows and doors’ (Fig. 213).27


The cast wall and the neutral expression (house for an employee by C.N. Ledoux, from Christ & Schein, L’Oeuvre et les reves de Lodoux).

This means that the effect is primarily determined by the figures seen in connection with these surfaces. Their neutral effect, however, will be judged more particularly by the roughness of their texture. The finer the texture the more neutral the surface in the sense that in itself it gives no indication of being thin or thick but depends on openings and other forms of treatment to reveal this. A rough texture immediately conveys greater strength in the wall itself; it seems thicker and therefore provides more closure.

Textural treatment may, therefore, be divided into three categories: smooth, fine and coarse. All three correspond to specific tactile characteristics and thus to degrees of differently experienced weight and ‘resistance’ in the inside-outside relation (Fig. 214 a-c).


Smooth texture and hord expression.


Three categories of texture: (a) smooth, (b) fine, (c) coarse.

Smooth Texture

The smooth wall surface seems hard. We experience this most frequently in glazed tiles and glossy paint (Fig. 214). Familiarity with the qualities of the material plays a part in one’s experience of it. Nevertheless, smoothness as a phenomenon has a universal quality. A smooth area seems to slip away when touched; one cannot ‘get hold of it’, it seems unapproachable and unassailable. Smoothness becomes a protective layer giving the inner part more meaning and a strength which seems impenetrable.

Fine Texture

The finely grained wall surface seems soft. In a way it is exposed and ‘open’; it does not seem rejecting to the touch. A fine texture is associated with porousness, thereby a warm and protected space. A space with smooth walls seems colder than a comparable space with finely textured wall surfaces.28

Yet, certain combinations of a smooth finish and a finely grained mat finish will give quite the opposite effect to the one described above.

This is because a smooth finish usually both reflects and shines. If the mirror effect dominates, the wall is opened up in that outside space penetrates further into the wall fabric. The same occurs in a glossy wall, because the light striking it is reflected and ‘imprisoned’ within the wall itself. A glossy finish in this case will be an opening instead of a closing factor (See also glass walls p. 189 f and the open floor p. 63 ff).

Examples of this effect in connection with a mat finish are found in neo- Gothic walls (Fig. 215). Areas of mat brick were often alternated with bands of dark, glazed bricks. These bands stiffen, strengthen and ‘reinforce’ the wall, both because they are darker and because they are harder than the rest of the wall. But, because they reflect the light and mirror the surroundings they give just the opposite effect. Glistening and shining, they break up the wall area and make it lighter, whereas the porous and mat brick areas in-between make up the massive structural substance of the building itself.


The interplay between smooth and porous banding in a masonry wall from the 1800s (All Saints by W. Butterfield, from Hersey, High Victorian and Gothic).

Coarse Texture

We have already mentioned that a coarse texture seems to give weight to the wall. A coarsley textured wall rejects, but in another way than does the smooth wall. Whereas the smooth wall protects something within, a coarse wall draws the inner substance to the surface in an almost ‘aggressive’ manner. We can be scratched by a coarse wall; it may hurt if we touch it. It represents an active resistance and so possesses its own power and weight. Such is the character of the parapet surrounding Oslo’s underground stations. A coarse surface, combined with sharply edged undulations in a low wall, effectively keep people away from these large openings in the street surface.

The rough wall, in other words, indicates its own inner substance. Erling Viksjo’s mixtures of concrete consisting of fine gravel and sand mottled with larger stone fragments are typical attempts to reveal the masonry wall’s innermost substance (Fig. 216). At one moment the stones seem to break forth from the wall’s interior, but in the next instant they seem imbedded in it. The wall’s inner core, therefore, is the precondition for the effect conveyed and thus for our perception of its existence. This inner core is the concrete itself, which bonds everything together into a heavy mass.


Coarse texture in E. Viksjo’s concrete mixtures (from the Norsk Hydro Building in Oslo, Norway).


In principle, the surface pattern on a plane wall falls within three categories (Fig. 217). The first has an abstract character dictated by independent patterns of lines, grids or curves. The second is figurative, based on representations of people, animals or things. The third is rooted in materials and building methods and indicates how a wall is built up of separate parts. We shall deal with this last category separately in the chapter on masonry walls (see p. 183 ff).


The three categories of pattern motifs: (a) abstract, (b) figurative, (c) material-determined.

To a certain extent all three categories may: (1) be organized horizontally, verticality or diagonally, (2) appear as a pattern flush with the wall itself or, (3) be composed of various plastic reliefs. Each of these motifs whether abstract, figurative, or constructive, will give the wall a different weight expression.

Pattern Order

Patterns made up of horizontal compositions have an inert and heavy character. The reference is the ground, not only because the weight seems to lie heavily on it but also because motion parallels it (Fig. 218).


Horizontal patterning and the heavy/dynamic expression (Casa unifamiliare by M. Botta, from A.D. 5/6, 1980).

As an example, Viksjo’s Hydro building in Oslo (1960) consists of two main volumes, the tall building itself and the base on which it stands. This base, of coarse polygonal concrete, slants outward. Both the roughness and the slant of the main form indicate the power and weight necessary to withstand the stress of the tall building above. The pattern of the base creates an important accent in the overall expression. The entire surface consists of alternating light and dark horizontal strips extending uninterrupted around the entire form. The pattern increases contact with the ground but also holds the walls of the base together as it they were taut bands restraining pressures from within (Fig. 219).


Horizontal patterning and the heavy/dynamic expression (detail from the Norsk Hydro Building in Oslo).

A vertical pattern helps to give the wall a freer character. A house is lighter if the dominating composition is vertical. It is only necessary to compare two smilar wooden houses, one with vertical and the other with horizontal siding (Fig. 220).


Houses with vertical siding (above) and horizontal siding (below).

Another example is Paolo Portoghesi’s house, Casa Papanice (1970), which is made up of concave wall planes divided by perpendicular strips of windows (Figs. 221, 222, 231). Vertical bands of coloured tiles cover the faҫades. The ascending bands are green, decreasing in number and becoming paler as they rise. The descending bands are blue, flecked with stripes of gold. These too decrease towards the middle, where they meet the green bands rising from below. The colour tones play a role in tying the upper and lower parts together,- the green tiles continue the growth of the soil upwards, while the blue and golden tiles are a downward extension of sky and sun.


Vertical patterning and the rising expression (Casa Papanice by Portoghesi/Gigliotti, from Norberg-Schulz, Alia ricerca dell’architettura perduta).


Horizontal patterning and the connecting expression (inside Casa Papanica by Portoghesi/Gigliotti, from Norberg-Schulz, Alia ricerca dell’architettura perduta).

In addition, the vertical bands underscore the independence of the individual wall planes. This again accents the entire system of wall divisions around the house and hence the rhythmic opening of the exterior. That this is international may be seen when compared to the interior walls, which are covered by horizontal tiled bands. Here the wall planes are linked together around the interior in contrast to the way the exterior is split; the interior is closed, the exterior is opened.

Pattern Images

All the patterns discussed above were flush with the wall’s own surface, neither cutting into it nor projecting from it. Further, the examples of horizontal and vertical ordering were all illustrated by walls decorated with abstract patterns. This means that had the same structures been shown with figurative or constructive patterns the surface expression would have been different. In other words, each of the three patterns applied to the same surface will convey a greatly varied weight expression and thus a different opening and closing effect between inside and outside (Fig. 223, a-c).


The effect of the pattern motifs on the wall’s expression of weight: (a) abstract patterns along and independent of the surface, (b) figurative patterns within the surface, (c) material patterns which weigh the surface downward.

Abstract patterns will convey lightness. They reveal nothing of how a wall is built as do constructive pattern motifs. In contrast to the latter, in which the pattern symbolizes the wall’s own essential substance and load-bearing nature and hence indicates weight, abstract patterns are perceived as a membrane-like surface on a massive wall behind. The motions indicated by the pattern therefore seem more to exist on a covering skin, their directions following along the wall mass and not within it.

A figurative pattern will also give the wall a lighter character. By figurative patterns we mean murals and mosaics which cover the entire wall with figures. The effect meanwhile, depends completely upon the figural representation. In this sense, a figurative pattern may give a covering, carpet-like, non-structural effect as do Byzantine mosaics. Or, the pattern may create ‘holes’ in the wall, thereby opening it up as in the case of Pompeian illusionistic architectural murals. The common factor in all types of figurative patterns is that in one way or another they will always convey the effect of optical depth into the wall.

In just the same way as abstract patterns, figurative patterns will also help to disguise the wall’s essential substance. To these optical effects must be added the meanings of what the patterns represent. In content these have varied throughout history. The modern abstract mural plays on spatial effects between pure colours and plan. Early Christian mosaics exploited the symbolic effects in various constellations of stylized human figures and ‘sacred’ colours, while in Antiquity’s architectural illusions the eye was drawn deeply into the space of large painted landscapes. Pompeian murals dissolve the wall as a substantial reality, so that one is led into a magnified and open illusionistic space which makes even the smallest apartments seem spacious and airy (Fig. 224). In Byzantine pictorial tradition secular and spiritual qualities are united. A church with walls covered in life-size figures of saints and biblical scenes is transformed into a vision of the divine sphere itself. The wall ceases to exist and becomes lost in a shimmer of brightly coloured tesserae against a ground of gleaming gold. ‘The (light) comes from above and in front, filters down over everything depicted, settling upon it, strewn like gold and silver dust from the top of the dome’.29 The space ceases to be a conceivable reality, says Otto Demus. It is not so strange that corners are ‘ignored’ in the compositional whole. ‘The sanctuary seems to revolve round the beholder; the multiplicity of the view forces him to turn round and round and this turning of his is imputed by his imagination to the building itself’ (Fig. 225).30 The figures, however, do not hover freely in an unstructured void. They are ordered in a hierarchic plan as in the successive stages of religious perception. At the bottom, the smallest figures form a floor level frieze, the middle level consists of over-life-size martyrs, while the topmost level is reserved for the archangels and apostles in colossal images. This ordering has nothing to do with the wall’s own structural system but concerns the ideological content in the mosaics. From a heavenly scale the figures are gradually reduced to human dimensions as they descend towards the believer himself.


Figurative pattern which opens up in depth (Pompeian architectural illusion, from L’Orange & Thiis-Evensen, Oldtidens bygningsverden).


Figurative pattern which dissolves the surface and which ‘ignores’ the corner (detail from Convent church in Daphni, from MacDonald, Early Christian & Byzantine Architecture).

Pattern Relief

The patterns already discussed may also occur as reliefs in the wall. If we imagine that these motifs are similar in form and content but recessed instead of flush with the wall face, the wall will change from having a light membrane-like character to having a heavier and more substantial quality. As surface patterns, the three types will vary in degree of lightness. In a relief it will be just the opposite. Now the wall will show varying degrees of weight. Here, again, the difference will depend upon whether the motif is abstract, figurative, or derived from the type of material used.

Most important, however, is the relation of the relief to the wall. In this, three situations predominate (Fig. 226, a-c): the relief may seem to be (1) attached to the wall (addition), (2) projecting out of the wall (convex), or (3) sunk into the wall (concave).

These three principles will, as a rule, occur in combination. For example, a relief projecting from its own slab will obviously appear to be fastened to the wall. In the case of the convex relief, more depends upon the contours and details, in deciding whether the relief is added to the wall or emerges from it. In addition to plasticity, texture and colour treatment must be considered. An additive relief in a colour other than that of the wall will seem even more prominent, whereas a convex relief in another colour will be given more freedom.

226 a-c

The three categories of relief motifs: (a) additive, (b) convex, (c) concave.

Despite these modifying factors we will study each type of relief and its influence upon the wall’s weight expression. We shall limit our examples to those with figurative patterns.

The Additive Relief

The additive relief conveys a neutral character to the wall to which it is attached (Fig. 226, a). The relief corresponds to the principle of figure/ground and depends on articulation for its effect, assuming that the relief is visually active in contrast to a more homogeneous and uniform background.

We find an example of this in the restored reliefs depicting the Via Dolorosa on the plaster wall in St. Olav’s Cathedral in Oslo. Following the principles of the Gothic wall system, these carved wooden reliefs are perceived as a superimposed layer on the wall behind. The relief, therefore, creates a transition between the observer and the background wall. Consequently, the reliefs and the background wall ‘do’ different things. The wall forms a physical boundary between the church and the world around. The Via Dolorosa reliefs also create a ‘wall’, but a purely spiritual one which holds the spectator within a world of religious perception determined by the meaning of the reliefs. It is the relief which is ‘active’, not only optically but in content as well, whereas the outer wall in both ways becomes a subordinate background.

The Convex Relief

The convex relief protrudes from the wall face.

This may give two different impressions (Fig. 226, b). In one the relief seems to be restrained by the wall itself. The relief is a part of the wall’s own fabric, its forms ‘undulating’ within the limits of the wall. The other effect is more as though the forms in the relief are about to free themselves from the wall. The wall seems about to loosen its grip on its own substance.

The overall impression becomes one of tension between these two extremes, between a modelled wall and a wall about to disintegrate. The two expressions are, in other words, degrees of difference within the same thing: an active and massive wall having its own substance, which it appears either to ‘part with’ or to ‘imprison’.

We see an example of this dualism in an Art Nouveau facade from 1896 (Fig. 227). These organic relief patterns swirl over the entire faҫade, seeming atone moment to fuse with the wall and at the next to curl outwards and away from it. The same type of motifs are used in the main form of the house, the frames and window openings. In this way we see that the relief, in motif as well, is part of the wall treatment. In other words, the relief has both its origins and its basis in the wall mass itself.


The convex relief (atelier in Munich by A. Endell, from Pothorn, Das Grosse Buch der Baustile).

The Concave Relief

The concave relief lies within the body of the wall. This means that in the relief an inner substance is laid bare, an interior that is kept under control by the wall itself (Fig. 226, c). Both the exposure and containment reveal a plastic and powerful reality, giving the wall an air of massive weight.

Egyptian wall carvings are a good example of the use of concave reliefs (Fig. 228). On both the exteriors and interiors of temple walls the Egyptians carved long series of figures and symbols. The human figures, plastically modelled, were as if sunk into the wall — everything was contained within the stone itself. This impression of great weight corresponded completely with the Egyptian ideological search for permanency. Egyptian building art has always had a strong monolithic character, revealed in its colossal dimensions, tightly spaced rows of columns, and slanting block-built pylons. In this atmosphere, the concave relief adds further emphasis to the stone massif’s meaning as the ‘captor’ of all motion and life.


The concave relief (wall section from Horus’ temple in Edfu, from Lloyd, Architettura Mediterranea Preromana).

The Masonry Wall

From what has already been said, we see that the weight expressed by the masonry wall depends upon the different characteristics of its individual blocks. These may be divided into three main categories according to size, method of joining, and surface.

Size of The Blocks

We judge a block to be large or small according to whether we can handle it or not. Thus, a wall built of small blocks will seem lighter than one built of large ones (Fig. 229). A brick wall, for example, is made up uf small easily handled elements. Nor does such a wall seem thick because the size of the bricks on the surface is spontaneously associated with a corresponding size in depth. In just the same way, the blocks that we consider large will also be thought of as having a corresponding depth, thus giving the wall as a whole a heavy, impenetrable character. An example of the deliberate use of difference in weight between small and large blocks is classicism’s corner solutions. Wall surfaces are either in plaster or brick. The corners, however, are frequently reinforced with powerful quoins, whose job it is to hold the form together.


The size of the blocks and the expression of weight.

Surface of The Blocks

The second factor which can affect expression of weight is the block’s surface. The block’s main form, its texture and colour treatment affect the wall’s character (Fig. 230).


Brick details (from Ungewitter, Varlegeblaetter für Ziegelund Steinarbeiten).


The surface of the blocks and their expression of weight: (a) flush with the surface, (b) protruding from the surface due to over compression, (c) protruding from the surface due to inherent character.

As was pointed out when dealing with the moulded wall, a coarse texture will give a heavier impression than does a fine, smooth texture. Similarly, a solid, dark wall will seem heavier than a pale one. The same weight difference will occur between an ashlar wall in which the stones are flush with the wall face and a wall in which the stones jutout (Fig. 23, a-c). In the first case, the stones appear to be controlled by the surface itself. In the second case the stones seem to protrude as though from some inner thrust. In the first case, the stones are neutralized, in the second they jut forth towards the onlooker. This convexity has two effects.

One is determined by the feeling of tremendous pressure on the stones from above. The scale of Frauenzuchthaus in Wurzburg (1809) must be seen in this light (Fig. 232). The entire ground floor facade consists of horizontal ‘tubes’, which seem to be squashed together by the weight of the upper storeys.


Base under pressure (Frauenzuchthaus in Wurz-burg, from Klopfer, Von Palladio bis Schinkel).

The other effect is determined by the way each stone behaves independently — swelling and protruding as a result of the individual stone’s ‘own weight’ (Fig. 233).


Rustic block wall from Peru.

In any case, whether the blocks project or seem squashed together, they will convey strength to the plastic wall.

Joining of The Blocks

The third factor affecting the masonry wall’s expression of weight is the joints. The contours of the blocks decide this and thus in turn the pattern on the wall face. The variety of patterns may seem endless, yet they are all found within two basic structures. One is amorphous, the other geometric. By an amorphous pattern we mean one that is irregular such as when the stones have different dimensions, forms, and contours. Typical of this is Antiquity’s polygonal walls, which seem to consists of a network of interwoven and broken lines (Fig. 233).

In the geometric block pattern there are two main characteristics (Fig. 230). First, the pattern is the result of an overall determining structure which orders the individual blocks. This is the horizontal, the vertical or the diagonal structure, each of which as we have shown earlier, gives the wall differing ‘weight’. The other characteristic is the precise fitting of each individual block. Each has its allotted position, and all are exactly fitted together to make the form determined by the governing structure.

From this we conclude that a geometric ashlar wall, providing the dimensions of the blocks are reasonably alike, will seem thinner than an amorphous ashlar wall. This is because a geometric wall indicates an even, organized weight transfer towards the ground. In contrast, the transfer of weight in the amorphous wall is more casual, which in turn seems to require greater thickness to prevent it from falling apart.

In short, the effect of what has been described above corresponds with what may be called abstract and rustic characteristics.

The Abstract Masonry Wall

Great care is taken in the working and joining of the blocks in an abstract wall. The blocks are in every way subjugated and controlled. Small stones will always seem more subjugated than large ones, because the latter are not so easily controlled by the human hand. The individual stones are also formed by man; they are both prepared and fitted, carved and polished. In addition, abstract walls are very often characterized by symmetrical bonding patterns, which clearly organize every stone. The abstract wall is often built to fit the space it surrounds. In the ancient Greek Treasury of Atreus (1350 B.C.), each block is hewn to fit into the even curvature around this domed space (Fig. 234). It is symptomatic, therefore, that in Roman architecture with its interest in interior space, the ashlar wall was avoided. Both formally and technically the Romans built their walls as a plastic skin to comply with the motions of the interior space


Blocks of stone adapted to the shape of the space (‘Clytemnestra’s Grave’, photo by Photo Marburg).

The Rusticated Masonry Wall

The character of the rusticated wall is determined in part by the nature of its material.31 In contrast to the abstract wall, the main impression is one of sovereign independence — neither human hand, the wall itself, nor space gain ‘the upper hand’ over the stone blocks. In the so-called East Gallery in Tiryns (1350 B.C.) there is a constant struggle between the spatial form and the blocks (Fig. 235). The spatial form rises upwards to make way for forward movement, while the downward stress of the individual, rusticated blocks makes the space seem about to collapse.


Blocks of stone in contrast to the shape of the space (‘The East Gallery’ in Tiryns, photo by Photo Marburg).

The most important aspect of the rusticated wall, however, is its association with nature. Its elements are as if taken straight from nature itself. And, nature is a given thing, independent and beyond the range of human control, Topera di natura’.32

The basic qualities of the rusticated wall are best illustrated by the Cyclopean walls in Myceneaen architecture. Around the royal citadel in Mycenae (c. 1250 B.C.) the outer wall is built of huge uneven blocks forming an amorphous pattern in which each blockswells out in a tautcurve. The entire structure takes on the air of a gigantic pile of stones in which the stones seem ready to fall apart at any moment. Those walls were said to be uncertain, which were made of stones of unequal angles and sides’, declares Palladio.33 This uncertainty connected with the amorphous wall increases both its frightening and natural aspect. Over the whole is an air of the primeval, of magic, a Cyclopean masterpiece and ‘the dream of the people of the forest’ (Fig. 236).34


Naturalized block wall (Student dormitory by R. Pietila, photo by Chr. Norberg-Schulz).

The Glass Wall

An opaque wall will always seem heavier than a transparent wall. A glass wall can carry no weight and in character is ‘non-existent’. The impression it conveys, therefore, is in a category by itself, quite apart from the massive wall’s closed character and the skeleton wall’s openness.

A clear glass wall is, of course, transparent but it also transmits light and has a mirror-effect. It is these three qualities together which determine the way in which the glass wall carries out the inside-outside connection.

In the clear glass wall, inside and outside seem to merge. The mirrored version of what is in front overlaps that which lies behind (Fig. 237). Transparency and mirror-effect, therefore, unite inside and outside like projections on a screen. In the following, we shall treat them separately in considering the influence of glass on the inside and on the outside.


Glass wall by day: the exterior is directed inside (photo from GA, 2, 1979).


The character of whatever lies behind a glass wall is transformed when the wall is removed. The reason is that the clear glass wall takes on a quality, not based on itself alone but in relation to what it picks up and transforms. And, of what is seen behind the glass wall, it is light and colour which the glass ‘catches’ and transforms. Light and colour, however, are always a part of things, a part of what we see in there. Light and colour accent the wall face and enwrap the figure. Through glass, which can never be completely flat but always wavy, all this is transformed. Colours glow, as in stones seen through water, light is broken up into points, and rays and things vibrate and come alive, all depending on how we focus our gaze.

Things seen through glass convey a feeling of distance. The motif appears to have a life of its own, different to what it would be if measured without glass in front. At the same time things seem very close. In a way they are within the glass itself, the way a landscape appears in a highly varnished painting. This duality lends an ephemeral air to what we see behind the glass; the glass makes whatever is inside seem to withdraw and come forward at the same time.


Within the glass is a transformed interior but also an illusion of the exterior. Glass mirrors the outside and the actual space is enlarged by its own reflection. Thus, in the glass wall, time and space are united as Sigfried Giedion says when referring to the Bauhaus building in Dessau:

… there is the extensive transparency that permits interior and exterior to be seen simultaneously… variety of levels of reference, or of points of reference, and simultaneity — the conception of space-time, in short.35


Effects in glass, which include the visual relation of strength between outside and inside, are dependent upon day and night, lightness and darkness. In the daytime they become two spaces in ever-changing balance. Exterior and interior struggle for mastery in that both are seen as overlapping parts. But, if the interior light is intensified, exteriorspace is weakened. Just the opposite takes place with strong daylight — now the inside disappears and the outside space invades the interior. Or, if the glass wall is transluscent but not transparent and also depending upon whether the glass is dark or light, the reflection of the exterior will be either strong or weak. At night, on the other hand, it is the interior that is drawn outwards (Fig. 238). The shining interior becomes a ‘gift’ to the night.


Glass wall by night: the interior is directed outside (house by G. Grung, photo by G. Grung).

The Skeleton System

We have already shown that the skeletal system opens the inside-outside connection and that its composite character, in principle, appears lighter than the isotropic wall. The light and open character of this system, however, just as for the massive wall, depends upon its main form and articulation.

The main form varies, but within a limited set of elements. These are determined by the skeleton’s construction, which can be interpreted as a certain combination of elements at four different levels (Fig. 239, a-d).


The motifs of the skeletal system: (a) column and beam, (b) frame (straight — arched — quadrangle), (c) row (colonnade — arcade — grid), (d) pattern.

The basic level comprises the primary elements, column and beam. The column is the vertical supporting member, and the beam is the horizontal supported member.

These in combination form the next level of the construction, the frame, in that two columns are joined by a beam. Frames may be formed in very different ways, but their variations all fall within three motifs primarily determined by the form of the beam. The first is the straight frame with a horizontal beam, the second is the arched frame, in which the beam may be semicircular, pointed or flat. The third motif, a variation of the straight frame is the quadrangle, in which beam, columns, and sill form an unbroken rectangle.

In the next stage several of these units are repeated to form a row. Based on the above forms we find three row motifs: the colonnade with straight beams, the arcade with arched beams, and the grid of equal quadrangles.

The fourth and final level is determined by the way the repetitive row motifs are integrated in a pattern within the same wall. These patterns may be formed in different ways. One of the repetitive motifs alone may cover an entire wall, as in the Greek temple (colonnade) and the modern curtain-wall (grid) (Fig. 240). In another variation the rows may be combined in the same wall as in Classicism’s layering of colonnade above arcade (Fig. 241). The repetitive motifs may be combined with massive wall area as, for example, in the Doge’s Palace, Venice, in which a solid wall is supported by lower rows of columns. The latter system may be reversed, as in the Palazzo Vidoni-Caffarelli, where columns rise above a massive socle. The systems may even be juxtaposed in depth etc. (Figs. 205, 242).


The skeletal system: grid (detail from the Bauthaus by W. Gropius, photo by Chr. Norberg-Schulz).


The skeletal system: pattern achieved by the combination of arcades and colonnades (detail from a building in Czechoslovakia).


The skeletal system: pattern achieved by the combination of a massive element and a colonnade (Palazzo Vidoni-Caffarelli by Rafael).

From the above description it becomes clear that two sets of elements are present in all combinations within the skeletal system. First of all come column and beam, followed by the frame motifs which they create and which in turn form the basis for rows and patterns.

In examining the skeletal system, in terms of opening and closing, it is necessary to analyse the archetypes that constitute columns, beams, and frames.

The Skeleton System’s Expression of Support and Motion

If we imagine ourselves facing a skeletal wall, it will be the weight and substance effect — just as for the massive wall — which influences our desire ‘to enter’. In the skeletal wall this means a unified directional expression, which is particularly determined by how we experience the system’s ability to carry a load (support expression) and the way the system directs our passage in (motion expression).

Support expression may mean that certain skeletons seem heavy and overloaded so we hesitate to enter. Others may seem strong and light so that we feel no ‘danger’ in entering (Fig. 243). Support expression is particularly determined by the system’s plastic character, which in turn is decided by how individual columns and beams are articulated and joined together. The question is, when will a beam best withstand stress: when it is pointed, arched, curved or straight? And, when does a column seem strongest: when it is high or low, straight or ‘swollen’, dark or light?


The supportive expression of the skeletal system: ‘insecure’ and ‘secure’.

The other factor influencing the inside-outside relationship is the way in which the skeletal system impels our movements. This motion impulse is largely caused by the intervening spaces in the system. It concerns the treatment of the columns, as to whether they are quadratic or round, rough or smooth, and also whether the form of the frame’s opening is wide, narrow or high, the beam straight or arched, etc. The repetitive motifs, too, are important in creating movement impulses in us, because any directional pull depends on the form, extent and rhythm of the row of columns.

Thus, sensations aroused by the skeleton system are determined by a combination of impressions. The support expression conveys security, while the system’s directional aspect invites motion impulses.

In the following we shall separate these two factors by first examining support expression, thereafter the motion inviting expression, and then apply these to column, frame, and beam respectively.

The Column

The Column’s Expression of Support

The column’s expression is to be found in its alternating pattern of rising and sinking. We recognize a corresponding interaction in our supporting function. In its broadest sense it reveals itself in our ‘Willensenergie’, strength of will, says Theodor Lipps: ‘Because within ourselves arises a corresponding action pattern and thereby a feeling of self which is a natural part of that same action’.36

Identification with the column is, therefore, an absolute necessity, because within the ability of the column to give support lies the presumption of safety. If the column gives way, both it and whatever it supports collapses and crushes us’. The tension in Milton’s Samson is ours; Samson and the column are as one:

…these two massy pillars with horrible convulsion to and fro He tugged, he shook, till down they came and drew The whole roof after them, with burst of thunder Upon the heads of all who sate beneath (Fig. 244).37


The inherent security of the column (The Death of Samson’ by G. Dore, from Dore, Bibelen i Billeder).

This organic interplay of strength in the column is clearly illustrated in Pier Luigi Nervi’s concrete forms. His column shafts clearly describe the dynamics of strength in the vertical, from downward stress to upward thrust. The principle underlying these colums is always the same. It is an integral part of the nature of concrete, says Nervi, in its plastic and static unity, ‘which because of the monolithic nature of construction in reinforced concrete, holds together the various parts of the structure’.38 The alternate rising and sinking is a continuous process in the column shaft itself between two different cross-sections, one in the upper and one in the lower part. ‘The transition between each section is obtained by joining with straight lines the corresponding points of the two extreme sections’.39 The most usual method is to expand the form at the bottom while compressing it at the top. The principle is most clearly revealed in the enormous pillars of the Palazzo del Lavoro in Turin from 1961 (Fig. 245). At the bottom the columns are in the form of a cross, at the top they are round with a diameter half that of the bottom. The entire form is a visible expression of balance and unity. The lower part expands to withstand lateral stress, and the top is concentrated to tie the roof ribs together (Fig. 246).


The form of a column (column from Pa-247. Column form (detail from Palazzo Canlazzo del Lavoro by P. L. Nervi, from Nervi, celleria in Rome). Aesthetics and Technology in Building).


The column’s dynamism (column and ribs from Palazza del Lavara in Torina by P.L. Nervi).

Expression of Support: The Column’s Three Parts

From this description we see that the column may be divided into three ‘energy sections’.

The uppermost is the column head or capital, which receives the load of roof or beam. The bottom part is the foot or base, which makes the transition to ground or floor. Between these two extremities stretches the third section, the column shaft, the intermediary of the rising and sinking action (Fig. 247).


Column from (detail from Palazzo Cancelleria in Rome).

In the following we shall examine the Classical column. We assume that all the examples support the same load. First let us look at the column shaft separately in order to study the importance such factors as cross-section, height, thickness and entasis (swelling) have for its expression of support. Then we shall examine how certain forms of capital and base extend this expression further to roof and ground.

Expression of Support: Round and Square Shafts

In comparing two column shafts having the same height and volume, but with round and square cross-sections, the circular shaft appears slimmer and stronger than the quadratic (Fig. 248).


Column form: round and square columns.

The circular shaft appears stronger because it is a closed form, concentrated inwardly upon itself. The curved surface conveys an impression of concentration, a circling around a central point, a lasting unity in itself’.40

In contrast, the square pillar is a composite form. It is easy to imagine that its surfaces and corners can ‘come apart’ because of the unequal distance to a central unifying core.

The round column’s concentration around a point conveys a slimmer impression than does the pillar. Its form circles unambiguously around the vertical line of the core, and verticality indicates upward thrust and height, which is the underlying principle in any form of growth energy.

This difference in dynamics between a round and a square column shaft has been exploited throughout the history of architecture. An example is the tall free-standing Gothic columns, which are almost always circular or spiral, never square. Had they been square, they would have lost their characteristic toughness. They would seem too weak and flimsy, rather like easily broken ‘sticks’ (Fig. 249). As another example we find that square columns with swelling are seldom found. The reason is evident. A pillar with a convex curved profile will seem about to sink, while a round column with the same treatment simply gains extra flexibility. An illustration of the first is Borromini’s use of the bulging pedestal, here to be understood as a small pillar, beneath a powerful round column (Fig. 250). The expression cannot be mistaken. The entire form appears to be pressed down as if by the column above.


Column form: the slender column of the Gothic style (Rouen Cathedral, from Frankl, Gothic Architecture.).


Column and ‘bulging’ pedestal.

Expression of Support: Short and Tall Shafts

The relatively short column appears to be squashed together by a combination of stress from the roof and resistance from the ground (Fig. 251, b). It reflects the two opposing forces which affect it verticality. The shorter the column, the thicker and broader it seems. Width increases the visual mass and with it the weight of the column.


Long column and short column.

If we imagine the same column but considerably taller, it will assume an air of independence (Fig. 251, a). Atall column, in otherwords has enough visual mass to give it its own centre of gravity. From this centre vectors will

spring in two directions, towards the ground and towards the roof. Whereas the short column appears tied down, the tall column gives an impression of freedom, the active conqueror of its surroundings above and below.41

Examples of the visual impact of this are the triumphal columns of Roman Antiquity. In their very form these soaring ‘heavenly arrows’ visibly express the divine elevation of emperors and heroes super ceteros morfales. Another illustration confirming this triumphal content is romanticism’s burial motif, the broken column. The column stump stands above the grave as an earthly symbol of life itself, which in its course stood forth in the full column’s vertical dynamism (Fig. 252).


The ‘broken’ column.

Expression of Support: Shaft Entasis

Entasis creates an even curvature in the shaft, which may be either above or below the column’s mid-point. ”

If the swelling is above, it means that the upper part of the column is thicker and therefore heavier than the lower part. We find this in both Cretan palaces from the seventeenth century B.C. and in modern piloti such as those of Le Corbusier’s I’Unite d’Habitation from the twentieth century (Fig. 253).


V-columns that support L’Unite d’Habitation (above), and V-columns that support the roof at Knossos (below).

The main effect of such columns is that of downward led stress, particularly if they carry a great weight. Here too lies the V-column’s expression. Because the entire mass balances on a small point at the bottom, an immediate impression of insecurity and danger is conveyed. It is to this the critics were referring in calling the proportions of ‘Gallerie des Machines’ (Paris, 1889), an error in judgement, ‘This lack of proportion produces a bad effect; the girder is not balanced; it has no base… The eye is not reassured’ (Fig. 254).42


V-column in the Calerie des Machines in Paris (from Giedion, Space, Time and Architecture)

Thus immediately we ‘reverse’ the actual support aspect and read the form from the ground up. In other words, we identify ourselves with the column’s own thrust power and ‘help’ it to withstand the load above. ‘Moving downward, the trusses become increasingly attenuated until they appear scarcely to touch the ground; moving upward, they spread and gain weight and power’.43

In the Classical column the swelling is below the middle point making it broaderat the bottom and narrower at the top. It follows, therefore, that the centre of gravity is closer to the ground. The column seems to push itself off the ground while simultaneously rising upwards. Fundamentally, the Classical column is characterized by its firm anchoring, corresponding to the actual dynamic surge in the column’s mass itself. The primary impression, however, depends upon two other conditions in the column’s main form. One involves the way in which the swelling continues on down towards the ground, the other concerns the height of the column.

If the column shaft diminishes as it nears the ground, the impression of pressure and sinking is intensified. Typical in this respect is the archaic Doric as exemplified at Paestum, where the column shaft seems ready to give way at any moment under the enormous pressure from above (Fig. 255). But this downward stress impression is also dependent upon the distance of the swelling from the mid-point. If it is right down at ground level as in many Egyptian column types, the sinking characteristic has a more self-contained content. Now it is the column’s own weight and not the weight it supports which appears to cause the swelling. The shaft takes on an almost ‘soft’ and self-burdened character, quite different from the tense will-power of the Doric (Figs. 256, 257).


Columns entasis (column from the ‘Basilica’ at Paestum).


Columns from the ‘Basilica’ at Paestum.


Column entasis: a typical Classical column.

The column’s height, furthermore, is important in the effect of entasis. If the column is short and wide, this swelling will accentuate the sinking character. In Fagerborg Church, Oslo, by H. Schytte-Berg (1903), the interior columns are formed as just such short, compressed volumes. The intention has been to create a Romanesque atmosphere, an interior hall which, in combination with the heavy roof trusses pressing down upon the surrounding walls, is a manifestation in stone of Lutheran gravity itself.

On the other hand, if the column is tall and slim as in the Corinthian order, entasis will accentuate its rising quality. For the Corinthian order Palladio prescribed a proportion of 1:3.44 The lower third of the column profile is perpendicular, giving it the appearance of resting drum-like on the ground. Above this point the column narrows gradually to a form characterized by an accelerating rise to meet the load above. In comparison, a similar column without entasis has a stiff and spindly look. A column with entasis appears, therefore, to be stronger than one without, precisely because it replies to the overhead load by swelling in compliance, while at the same time the accelerating upward thrust masters that same pressure (Fig. 258).


Column entasis: a typical Egyptian column.

This double content of downward stress and upward thrust contained in the swelling is fully balanced in the developed Doric column. This column, shorter and wider than the Corinthian, is heavier also and therefore visually expresses its ability to support a heavier load. The downward pressure is made visible in the shaft’s bulging entasis. At the same time, greater weight means greater strength in the column’s own support capacity. Seen thus entasis expresses not only downward pressure but equally the flexibility in the shaft’s counter thrust as it rises to meet the entablature and roof.

Expression of Support: Surface of Shaft

The column’s basic form is in itself not alone in conveying the visible aspect of its support capacity. Surface colour and articulation are additional factors. By means of different colour treatment a thick Doric column in its white marble splendour appears larger but at the same time lighter and thereby ‘weaker’ than the same column in a dark colour. The latter in turn seems reduced in size but at the same time gives an impression of greater compactness and support capacity. The reverse is true of a slim column. The graceful iron skeletons of the 1800’s for example, would appear thinner if painted black rather than white. The latter treatment would enlarge the skeleton and make it visually thicker.

Texture too affects the column’s expression of support. Rustication and rough cutting will as a rule give the shaft greater ‘weight’. In the squat stone pillars of Stonehenge it is precisely the coarseness of their surfaces that is of essential importance in the whole megalithic atmosphere. At the other extreme a smooth surface will also seem to strengthen the form. Now, however, the impression of strength is quite unlike the above, in that it is decided by a flint-like hardness. H.P. Berlage has interpreted just this effect in designing the columns in the main hall of the Amsterdam Stock Exchange (1902). The primary columns supporting the greatest weight are highly polished and ‘hard’. The secondary columns, on the other hand, have a rough surface compared to the others, giving them more texture and ‘softness’.

Treatment in relief may also change the column’s appearance of strength. The most common surface treatment of this kind is fluting.

Doric, Ionic, and Corinthian columns all have these verticality grooved shafts. Their effect is twofold. On the one hand, they may outline verticality in itself and thereby accentuate the upward thrust of the shaft form. On the other hand, they can be a factor in emphasizing exactly the opposite, which is the shaft’s own essential substance and weight. The dominating characteristic depends upon the thickness or slimness of the column and whether the fluting is deep or shallow. On the elegant Corinthian column the flutes are narrow and deep, separated by flat edges (Fig. 259). These grooves appear as deep cuts in the body of the shaft itself while the edges outline the actual surface (Fig. 260, a). This effect of concentrated vertical lines accentuates the entire uprightness of the form.


Column surface treatment example of Corinthian fluting.


Column surface treatment example of Doric fluting.

In the thicker Doric column the fluting is quite different. Here the grooves are broad and shallow, with sharp edges in between. It is as if this fluting does noe ‘penetrate’ the body of the shaft itself but remains an exterior decoration on the form’s mass (Fig. 260, b). This serves merely to accentuate its plastic character, which becomes quite clear when compared to similar columns without fluting such as those of the temple at Segesta (400 B.C.). The smooth columns here seem stiff and lifeless in comparison with fluted ones, whose softly shadowed grooves emphasize all the inner flexibility of the shaft.

Confirmation of the effect that arises with and without fluting is to be found in columns where the shaft is both fluted and smooth. Examples of this combination are already to be found in Antiquity, particularly in the slim Corinthian column. It is done by filling in the grooves in the lower third of the column shaft so that the surface becomes approximately even. A possible reason could be that the flute edgings were particularly exposed to wear and tear on the lower part of the column. This pattern, however, has been maintained even where columns are raised on high pedestals. Is the reason then of a visual nature? As we have shown above, the lower part has approximately the same diameter top and bottom. It is only when the fluting begins that the entasis is intensified.-The following impression is clearly conveyed by the different surfaces: the lower part becomes even heavier, the upper section even steeper. The resulting contrast emphasizes the dynamics in the column’s own expression of support, which is the struggle between sinking and rising (Fig. 261).


Column surface treatment with examples of combinations of smooth and fluted surfaces.

Expression of Support and The Column’s Head and Foot: Capital and Base

In the transition of column to ground and roof, the form given to these connecting elements is decisive for the effect of the column’s dynamics on its surroundings. In classical architecture this is a question of how capital and base are designed. Vitruvius stated three ways of forming these members: the Doric; the Ionic, and the Corinthian, each of which should correspond to a definite ratio of proportions in the column shaft (Fig. 262). These three orders formed the basis of an extensive architectural school of thought. The main content in his teachings was based on the wish to differentiate between various weight and strength relationships in the column’s visual support capacity. This differentiation has played an important role in traditional classical architecture, both in the interpretation of specific building tasks and in the formulation of their various parts and details (Figs. 263, 264, 265).45


Doric, Ionic and Corinthian orders (diagram after Uhde, Die Architekturformen des Klassischen Altertums).


Doric, Ionic and Corinthian orders as articulating elements: the major column is Corinthian, the inner columns are Ionic, while the outer columns are Doric (temple in Bassai, from Berve etc., Greek Temples and Shrines).


The column meets the ground (corner columns from Rome, photo by I. K. Barstad).


The column meets the ground (base from temple in Bassai, from Charbonneaux, Das Klassische Griechenland).

Expression of Support: The Doric Column

Of the three orders, as already stated, the Doric column is the heaviest in character.

The Doric order portrays masculine strength says Vitruvius:‘… The Doric column, as used in buildings … exhibits the proportions, strength and beauty of the body of a man’.46

It follows, therefore, that the Doric was a suitable style for temples dedicated to heroes and male gods. ‘The temples of Minerva, Mars and Hercules will be Doric, since the virile strength of these gods makes daintiness entirely inappropriate to their houses’.47

As a result, the Doric has been used through the ages in strong and rusticated buildings ‘ … such as city gates, fortresses, citadels, treasuries, arsenals, prisons, ports and similar military installations’.48

The form given to base and capital was intended as a part of this visual expression of weight.

The Doric capital comprises two parts. Immediately beneath the architrave is a square slab, the abacus, and beneath it the head of the column itself, the echinus, which is in the form of a convex circular cushion. The Doric column has no base in the usual meaning of the term. The fluted shaft is led straight down to the ground with no emphasis on the transition.

The form at both places reflects the balanced tension between the column’s upward thrust and downward stress. The lack of a base accentuates the power of both tendencies simultaneously. Viewed from the bottom up it is as if the column springs directly from the horizontal level, strong and uncompromising. Fluting emphasizes the dynamics of this growth. Read in reverse, the impression is exactly the opposite: the column is forced directly into the ground by the heavy pressure from above. This balance between rising and sinking also occurs in the fully developed Doric capital. In the column capitals of ‘Poseidon’s’ temple at Paestum (c. 450 B.C.), for example, the two tendencies overlap in the echinus itself (Fig. 266). The rising tendency is emphasized by continuing the shaft fluting part of the way into the neck of the capital and terminating it by a row of fillets (anuli) immediately beneath the cushion. The sinking impression is accentuated by terminating the capital a short distance down the shaft itself with a similar row of fillets. In this way the two tendencies overlap and are united.


Doric capital (classical type).

The curvature between neck and cushion is also decisive for this double tension in the capital. In the classical capital the curve is even and upright in contrast to the archaic capital, where it is broken by a deep groove immediately beneath the cushion (Figs. 255, 256). Here the column profile narrows sharply towards the top while the cushion bulges outwards over it like a slightly flattened inner tube. The cushion’s form emphasizes the effect of downward pressure. The shaft, on the other hand, accents the upward thrust in that the diminution of the upper half ‘sharpens’ and accelerates the verticality. Altogether, the archaic column is a greater visualization of a dramatic collision of opposing forces rather than a fusion of them, as we have seen in the fully developed classical column.

Expression of Support: The Ionic Column

The Ionic column takes its place midway between the heavy Doric and the slim Corinthian. It personifies the feminine qualities says Vitruvius. It is, however, the mature woman who is represented in the calm elegance of the Ionic. Whereas the Corinthian column depicted youthful feminine spontaneity, the Ionic was the style of the more serene goddesses.

The construction of temples of the Ionic order to Juno, Diana, Father Bacchus and other gods of that kind, will be in keeping with the middle position which they hold, for the building of such will be an appropriate combination of the severity of the Doric and the delicacy of the Corinthian.49

In the Doric order we saw how entasis was clearly visible as an interpretation of a powerful balance between thrust and stress. This struggle between survival and resistance is less apparent in the Ionic column. The impression here is of something accomplished and sure. First of all, the Ionic column is slimmer than the Doric and the diminution not as great.50 Combined with slimness we find a shaft rising unconcernedly upwards. In the second place, both Ionic and Corinthian columns have a base that divides shaft and ground and thereby blocks any further continuation. This ‘buffer’ absorbs the jolt of the shaft meeting the ground but at the same time, like a springy counter-power, initiates the upward thrust. The effect, however, depends upon whether the base appears to belong to the column or the ground. If the base is formed as part of the ground, it is as if the ground itself rises to push the column upwards, giving an overall rising effect. The reverse is true if the column pushes downwards — a sinking effect occurs (Fig. 267, a, b).


The base as a part of (a) the ground, and (b) the column.

This upward or downward spring-like action is made clear by the form of the details. In considering the classical Ionic, we see that the base is divided into two main parts. At the bottom is a square slab (plinth) which neutralizes the transition between level ground and the round base and column above. The round part consists of two convex mouldings (tori) above and below a concave groove (Fig. 268).


Ionic (Attic) base.

The bulging cushion-like mouldings illustrate the effect of downward stress. The lower ring, therefore, which receives the greater pressure, protrudes most. On the other hand, the concave groove depicts the way in which the form thrusts itself upwards. This means that the lower and deeper the horizontal groove is, the heavier seems the pressure from above. Inversely, the higher and more upright the groove, the stronger seems the ‘thrust-off’ towards the top (Fig. 269).


Ionic base with broad and narrow cavettos.

The Ionic capital conveys a ‘soft compliancy’. In contrast to the Doric order’s conflict-filled struggle against the weight above, the Ionic capital appears to yield to this pressure and settles obediently but resiliently against the roof-bearing architrave. Whereas the Doric capital draws attention to an overlapping of contrasting powers, the Ionic capital appears at first glance to be split by the pressure from above in the way its two spiral volutes curl outwards in opposite directions beneath the weight of the entablature. In older proto-lonic capital forms it is even more apparent that the volutes are a result of such a splitting in the upper part of the rising column shaft itself (Fig. 270). In the classical capital the form is seen as a resilient cushion in which the volutes form its parallel sides. The other two sides, which face the spectator, show a convex downward curl illustrating clearly and directly the strength of the pressure from above, a pressure identical to that which causes the capital to coil (Fig. 271). The entire capital thus reflects both the reception and resiliency in the column’s contact with the architrave and roof.


Pre-lonic capital (diagram from Robertson, Creek and Roman Architecture).


Ionic capital (classic type).

An interesting variation of the Ionic volute pattern, confirming the above description, is Borromini’s formulation of the columns in the interior of San Carlo alle Quattro Fontane (1641). In contrast to the other interior columns, whose volutes are of the usual reclining type, the eight columns carrying the dome have capitals with active upright volutes (Fig. 272). The reclining volutes terminate the verticality in the column shaft, while the upright volutes continue it further up and into the roof system’s ribs and arches.


Capitals from F. Borromini’s St. Carlo alle Quatro Fontane (Above with vertical scrolls and below with horizontal scrolls).

Expression of Support: The Corinthian Column

The Corinthian column is usually slimmer than the Ionic but otherwise has similar fluting and base. The capital, however, is different and in its form a direct response to the verticality in the shaft itself. Whereas both Doric and Ionic capitals illustrate the alternation of thrust and stress in the column’s support capacity, the Corinthian capital appears to state unequivocally upward thrust alone.

This is already apparent in the capital’s ornamentation. The entire capital is covered with rows of naturalistic acanthus leaves giving the form an organic and non-supporting appearance. The verticality in the shaft is interpreted as a free growth, unhampered by architrave and roof, ‘bursting forth’ playfully against the roof moulding. While both Doric and Ionic capitals are compressed or divided by the load above, the Corinthian thrusts on upwards as though having nothing to support (Fig. 273).


Corinthian capital in motion (drawing by G. Neumann, from Domus, 610, 1980).

This unfettered and dynamic impression is caused not only by the overlapping acanthus foliage butalso by the concave inward swing of the abacus and the diagonal outward swing of the corner volutes (Fig. 274). The volutes dynamically lead the entire capital out into space in contrast to both the Ionic and Doric capitals where the form is strictly imprisoned within the architrave above. In the concave grooves between the volutes, however, there are indications of another diametrically opposed motion, which seems to be caused by some external pressure. It is as if the form itself pulls back to make way for a further growth of the underlying foliage, which may climb over the abacus and even right up into the entablature itself (Figs. 275, 276).


Corinthian capital (classical type).


Corinthian capital (Rococo variation from the Bishop’s residence in Wurzburg by L. von Hildebrandt, from Forssman, Dorisch, jonisch, korintisch).


Corinthian variation (detail from Nicholai Church in Leipzig, from Klopfer, Von Pallodio bis Schinkel).

This non-supporting and vertical effect in the Corinthian is well exploited throughout the history of architecture. The Corinthian was considered to be the most ‘delightful’ of the three orders and was used in particularly important and prominent buildings. ‘It is reserved for particularly elegant buildings meant to be impressive by their nobility of character and splendid ornamentation’.51

This usage was due not only to the capital’s rich ornamentation but doubtless also to the entire vertical emphasis in the column’s form. Such a column portrays victory over all stress and resistance. It springs freely aloft as if in triumph. The triumphal column of Antiquity was, therefore, preferably Corinthian and supported no more than a statue of the exalted one. Thus the column’s inherent verticality became also a meaningful part of the entire monument’s content, which was the victor’s ‘triumph and power’ (Fig. 277).


Corinthian triumphal columns (reconstruction of Forum Romanum, from Dal Maso, Rome of the Coesors).

Expression of Support: The Relation Between Inside and Outside

In general, the heavier, shorter and more tightly formed the skeleton system is, the more closed off it appears.

The distance between the squat megaliths of Stonehenge may be just as great as between the clustered pillars in a Gothic nave (Figs. 52, 278). The diameter may also be the same, but their height and texture are vitally different. For this reason, Stonehenge ‘shuts out’ while Gothic ‘opens’.


Stone pillars at Stonehenge (from Hitchcock (ed.), World Architecture).

A column may also seem taller than it really is if the row in which it stands is taller than it is wide and is compared to one which is wider. We see, therefore, how the eight frontal columns of the Parthenon stretch skywards to a greater extent than the seventeen columns on each flank (Fig. 279).


Apparent column height and the effect of the length of the colonnade (Parthenon).

Various surface treatments of columns are also used to differentiate between inside and outside. In historicism it was quite usual to use coarse and closely placed granite columns in an outer wall while the same column in the interior was of highly polished and colourful, veined marble. The latter gave the form a composite and flamboyant character contrasting sharply to the more natural, exterior columns serving as buffers against exterior space.

The difference in the expression of support between a round column and square column is also intentionally used to convey the impression of varying motions in the wall. Let us imagine a comparatively high wall supported by three round arches, in one case by square columns and in the other by round columns (Fig. 280, a-c).


Round and square columns and the supportive expression of the wall: (a) rising from below, (b) sinking from above, (c) alternation between rising at the middle and sinking at the corners.

The square columns, which in their form are a part of the wall, passively receive the weight that seems evenly distributed by the rounded arches. If these columns are replaced by round columns, the sinking effect is countered by a rising effect. The result is a contrast between the upward thrust of columns and arches and the downward stress of the wall. If round and square columns are combined, the central arch being supported by round columns and the corners by square columns, the result will be a rising midsection and two sinking corner parts, which means that the wall opens up in the middle and shuts at the corners (Fig. 280, c).

The varying character of the classical orders is also used to differentiate between inside and outside. The Corinthian and Ionic orders are more open than the Doric. Again we take the University buildings in Oslo as a typical example. In the mid-section, where one enters, the columns are not only round but Ionic as well. The applied pillars on either side, on the other hand, are Doric. The columns are used to enhance the width in a case like this, in which an opening section is framed by closing sections.

This differentiation takes place in depth also, and in the same way. In Palladia’s Palazzo Valmarana, Vicenza (1566), the outermost giant pilasters are Corinthian (Fig. 197). These are festive columns opening the wall’s outer layer towards the world. In contrast, the innermost layer of pillars, really part of the wall, are Doric. In classical Athens this order is reversed. In both the Propylaea leading in to the Acropolis (431 B.C.) and the Stoa of Attalos facing the Agora (150 B.C.) the outer columns are heavy Doric — a protective wall against exterior space. Within, however, the columns are in the lighter Ionic (Fig. 281).

The orders are superimposed as well, thereby indicating differences in the wall’s stress pattern. One of the earliest examples of superimposition in such a context is Rome’s Colosseum (A.D. 82) (Fig. 282). The. building is a massive arched wall construction in four storeys, each storey having fictive skeletal frames around its arches. At the bottom is the Doric order, above it the Ionic while the two upper stories are Continthian. The ground storey is the lowest of the four, its columns are without pedestals and stand directly on the ground. The two storeys above are equal in height while the top storey is the highest. The entire fabric of the building may be read in even stages from the ground up. The heavy Doric base is surmounted by the more graceful and gradually lighter Ionic and Corinthian storeys with the slimmest version at the top. In symbolic form, the columns summarize weight differences in the supporting wall. The skeletal system delineates these differences and in so doing makes visible and expresses the rising in an otherwise homogeneous and uniform building. The amphitheatre’s heavy muscularity is transformed into something proud and erect, which at a glance prepares the spectator for what he is about to encounter within: hero and soldier in proud combat for emperor and people.


Variation of column orders, Daric outermost, lonic inner-most (Attala’s Stoo in Athens.)


The columnar orders in super-position: Doric, Ionic, Corinthian (Colosseum in Rome, from Koepf, Baukunst in Funf Johrtousenden).

The Column’s Expression of Motion

In our introduction to the free-standing column we pointed out that its expression of support concerned the plastic treatment of the column form.

This will also affect its expression of motion. Its size, for example, whether tall, short or thick, will arouse different motion impulses. The tell column with accent on verticality will convey an arresting rather than a directional impulse. With verticality and centre so strongly emphasized, a spontaneous feeling of reserve will arise in the passer-by (see the vertical wall p. 145 f). A thick column stops us in another way. Its sheer plastic mass brings us physically to a halt, in contrast to the short, normal column, which, because it is our equal’, allows for easier passage.

The column’s aspect of motion, however, is largely determined by the form’s outer surfaces. In consequence, the smooth and the rough column have different directional capabilities. We slip easily pasta smooth column but a rough one disturbs and ‘brakes’ our passage (See rough and smooth surfaces p. 171 ff). In this connection the cross-section is especially important. Consequently, in the following we shall limit our study to the most important motion effects conveyed by square and round columns.

Expression of Motion: Round and Square Columns

The round column lends freedom to the surroundings — a square column directs them. In approaching a round column we have a choice of either proceeding along and past it or along and around it (Fig. 283). In approaching a square column we encounter a mass which, if it does not stop us completely, leads us along beside it in a definite direction. The round column conveys freedom, the square column restricts it. This corresponds to their own forms. The round column is an independent individual, in all parts equal and self-contained. The square column, on the other hand, has sides leading along its centre and corners that point away from this centre. When it is rectangular, an inequality arises; the longer sides dominate the shorter (Fig. 284).


Round columns and freedom of motion (from The Ministry of Health in Rio de Janeriro by Le Corbusier, from Le Corbusier, Oeuvre complete 1910-65).


Square columns and directional motion (The Supreme Court in Chandigarh by Le Corbusier) from Le Corbusier, Oeuvre complete 1910—65).

Round Column and Square Column: Surrounding Spaces

Freedom and restriction are also the impressions made by the round column and the square column (or pier if rectangular), in the context of their immediate surroundings.

The round column is released from its surroundings because nothing can be joined to it. The square column on the other hand, has four sides to which walls and slabs may be added (Fig. 285). A round column in a wall stands out because the wall is straight and the column curved. The pillar is perceived as a part of the wall itself, or rather’ what remains of the wall’.52


Round column and square column and their relationship to other elements: (a) the round column is free, (b) the square column is attached.

This freedom and restriction also concerns the surrounding spaces. In principle, round columns lend freedom to space, both verticality and horizontally. We find an example of the first in the relation between column and floor slab in Le Corbusier’s pioneer projects. The columns have the same diameter from top to bottom and are without base or capital. There is nothing to prevent the columns from continuing on through the slab in both directions (Fig. 286, a). Nor do the slabs seem attached to the column shafts, they seem almost movable, as if they could be slid up and down the columns. If these same columns had been square instead, the overall impression would be one of additive isolated units (Fig. 286, b) The slabs between the storeys would appear to be whole units, unpenetrated but carried by superimposed supports.


The round column and the square column and their relationship to vertical space: (a) the round column is free, (b) the space column is additive.

Horizontally as well, in relation to its environmental space, the round column allows for freedom. A modern example is the office building of the Norwegian University Press in Oslo (1980) designed by the architectural firm of Jan and Jon (Fig. 287, a). The great columns which extend through all the storeys are surrounded by curving and sharply cornered walls, decorations and mouldings. Areas are created which are not just random parts in an undulating whole. They become specific places, because the columns themselves create focal points. If the columns had been quadratic, a situation of conflict would have arisen — a conflict between the curving freedom of the walls, on the one hand, and the columns’ directionality on the other (Fig. 287, b).


The round column and the square column and their relationship to horizontal space: (a) the round column sets free and focuses surrounding elements, (b) the square column leads and binds them.

Round and Square Columns In Combination.

Round and square columns may be combined in many ways.

They may be juxtaposed, as we have seen in the case of the Colosseum. This variant conveys a stage by stage inward motion from an open skeletal layer on the outside to a more closed pillar and wall system further in (Fig. 288, a). Round and square columns may also be superimposed as in the column mounted on a pedestal. This combination conveys verticality in the upper part and a leading horizontality below (Fig. 288, b). Round and square columns may also be combined to create a new form. This we find when two semicircular columns are attached to opposite sides of a square column(Fig. 288, c). This symbiosis encourages the impulse both to proceed along the unit and to go around it. A square column may also have its corners sheered at an angle or rounded, which in turn gives the double effect of directional and centralizing motion.


The round column and the square column combined: (a) round and square joined, (b) round and square on top of one another, (c) a cross between round and square, (d) round and square abstracted (cruciform column).

A special variation of the combined column is the cruciform column (Fig. 288, d). In section it is a cross, something quite apart from the above examples. It is not meant as a combination of round and square columns but rather as a fusion of their individual qualities; the column focuses, the pillar directs. Mies van der Rohe’s use of this type in his Barcelona pavilion (1929) shows that its combined qualities are to be seen as an accentuation of the building’s own spatial system (Fig. 289). This spatial system combines strictly ordered rows of supporting columns with expansively free wall planes connecting outside and inside. Just like the walls, the columns are to be interpreted as an arrangement of freely arranged horizontal planes. At the same time, however, they point inwards towards a unifying core at the crossing point. Thus, the qualities of both the column system and the plane system are revealed in concentrated form; the columns order the space and hold it together; the wall planes direct and spread it.


Cruciform column in directional space (Barcelona Pavilion by Mies van der Rohe, from Johnson, Mies von der Rohe).

Round Column and Square Column and The Relation Between Inside and Outside

Round and square columns differ in terms of their relationship to the concept of inside and outside. Let us take two examples, beginning with the main entrance to the University in Oslo (Fig. 174). This entrance is formed like a temple front with four free-standing round columns framed by a square column at each corner (see also p. 137f). The walls on either side are divided into a system of giant pilasters. The corner columns are meant as conclusions to the projecting side walls, which frame the round columns while at the same time creating the transition between these and the shallow wall pilasters on each side. Whereas the square columns and pilasters are seen as part of the delimiting wall, the round columns are intermediaries of the connection between inside and outside (Fig. 290). One passes easily around and past them. They form the actual transfer point for the continuous flow of people moving freely between the fore-court and the great staircase.


Round column as entrance motif (diagram of entrance to The University of Oslo).

A square column, however, does not necessarily hinder inward movement, it may also emphasize it. If a square column is slab-like it may either halt or guide movements, depending upon whether it stands across or parallel to the entrance direction. In Le Corbusier’s Supreme Court in Chandigarh (1953), piers paralleling the direction of entry are used to make the entrance inviting (Fig. 284, see p. 295). The building is long and narrow with the approach to its broad side. As we have shown earlier, such a horizontal form has a rejecting effect on further forward movement. Consequently, the architect creates a contrasting motion to this transverse breadth by freeing three of the colossal piers which support the roof. In this way one is led alongside the piers and straight into the mighty stairs and ramps within.

The Frame

The Frame’s Expression of Support

The frame is a complete figure (Fig. 291, a).


The frame and its tendency to deflect and buckle: (a) diagram of the phenomenon, (b) C. Siegel’s stiffened solution (Siegel’s diagram, from Siegel), Strukturformen der modernen Architektur).

In the combination of column and beam lies the expression of support. For all frames, the main problem in both vertical and lateral stress is to prevent combinations of the beam deflection and inward or outward buckling of the two columns.

A simple diagram by Curt Siegel illustrates this factor and its importance for the form of the frame (Fig. 291, b).53 The danger of beam deflection is avoided by giving the beam sufficient size and cross section in relation to the span and by using appropriate materials. As for the columns, the danger of cracking or splaying is prevented by firm foundations. To prevent failure when the foundation is immovable, the upper part of the column is strengthened by being made thicker in the transition to the beam.

Providing a frame with acceptable stiffness and load-bearing capacity has represented a major problem throughout architectural history (Fig. 292). A good example is the Gothic buttressing system (Fig. 293). Buttress form is impossible to understand unless directly associated with the roof and its lateral stress. In cross-section the Gothic church nave is seen as a frame construction in which the pointed arch and buttress together form the beam. The arch as a form and not least of all the pointed arch, is a direct answer to the problem of effective deflection deterrence in a stone construction with maximum span (see p. 228 f). The buttresses illustrate strategic points of transition and anchoring which transmit lateral stress. Tall, towering superstructures topped by pinnacles and finials supply extra weight and reinforcement at the points where arch and buttress meet. The buttress itself is enlarged in stages in the descent in order to offset diagonal stress.


A diagonally stiffened frame (proposed portal into The University Press in Oslo by Jan & Jon).


A diagonally stiffened frame (Gothic buttressing, from Macaulay, The Cathedral).

In a way, this support expression at the junctures above visualizes the entire Gothic ideology. Gothic construction was meant to depict triumph. This was a triumph overfalling and earthbound forces, which in concentrated form were imprisoned and conquered for the sole purpose of illustrating the very opposite — the ascending and heaven-bent.

In other periods, the wish was not to express triumph and conquest but quite the opposite — collapse and pessimism. This was done by weakening the skeleton at precisely the same points that would otherwise have been strengthened. Such tendencies were particularly seen in Mannerist architecture. In Giulio Romano’s Palazzo del Te, Mantua (1526) it is the architrave which is weakened. Here the keystones appear to slip down in the very centre where the sagging tendency is greatest (Fig. 294). In Andrea Palladio’s Loggia del Capitanio, Vicenca (1571), the architrave is broken at the same point, this time by the windows, which are pushed right up into the architrave (Fig. 295). The outer pilasters of Nicolo Salvi’s Fontana di Trevi, Rome (1762), are a Baroque example of a similar weakening of firm foundation. They are ‘broken’ just where they meet the solid stone on which they stand (Fig. 296). 299. The relationship between the beam and the columns (from Phleps, Vom Wesen der Architektur).


The deformed frame (Loggia del Capitaniato by Palladio, 295. The deformed frame (Loggia del Capitaniato by Palladio, from Vendetti, The Loggia del Capitaniato). from Vendetti, The Loggia del Capitaniato).


The deformed from (loggia del Capitaniata by Palladio, from Vendetti, The loggia del Capitaniato).


The deformed frame (sketch detail from Fontana di Trevi in Rome, by Salvi).

Expression of Support: Coordination of Beam and Columns

From what has been said already it is clear that columns and beam must be in accord if the motif is to appear secure. Thus, the beam and columns must be suitably proportioned to both the width and height of the frame’s opening seen in relation to our shared experiences with the materials used.

An example of this is to be found in a colonnade by Alberti in which the centre inter-columniation is greater than those on either side (Fig. 297). We see that the dimensions of the architrave are not scaled to the narrower side openings but to the central bay, where the tendency to deflection is greatest both visually and in reality. On such grounds, according to Rudolph Wittkower, Palladio could have criticized Baldassare Peruzzi’s fictive facade system for the cathedral in Capri (1515) (Fig. 298). There, the only support for the long transversal main beam are two small corner pilasters and no apparent support in the middle, a ‘necessity’ to prevent optical deflection.


The relationship between the columns and the beam (Doric frame, with larger intercolumnation at the centre, by L.B. Alberti, from Alberti, Ten Books on Architecture).


The relationship between the beam and the columns (Cathedral in Carpi by B. Peruzzi, from Wittkower, Architectural Principles in the Age of Humanism).

Herman Phleps maintains the same by pointing out the spontaneous relief one feels when a long and heavy beam is supported in the middle — even if this is statically unnecessary (Fig. 299). Frame constructions in many modern halls also lack such a reassuring corrective. Enormous concrete girders, technically strong enough because of inner reinforcement, are only supported at each end by thin pillars. To accept such a construction automatically we must be aware of the hidden reinforcement. This means, in fact, that general reactions to what is considered ‘good’ form have become dependent on knowledge (Figs. 300, 301).


The relationship between the beam and the columns (from Phleps, Vam VVesen der Architektur).


The relationship between the beam and the columns (a modern concrete parking garage).


The relationship between the beam and the wall (beam resting on glass, from Ekeberghallen in Oslo).

The Beam

The above examples show that in the coordination of frame elements the beam is particularly important for the expression of support. It is the beam’s form, dimensions and span which tell us whether or not the supporting columns are suited to their task —and not the other way round. Two mighty columns, therefore, may carry an undersized beam without any indication of danger and collapse. Danger is first apparent when a heavy, oversized beam is supported by columns that are too slender (Fig. 302, a, b).


The relationship between the beam and the columns: (a) over-dimensioned columns, (b) over-dimensioned beam.

In the following, therefore, we shall limit our study to the support expression of the beam. We repeat that we shall base this discussion on the straight beam in both its classical and modern variations and the arched beam with the pointed arch, segmental arch, and flat arch as its most typical variations.

Expression of Support: The Straight Beam

Of the various beam forms, the straight beam is both in appearance and reality the one most prone to deflection.

To offset this tendency and to convey a ‘secure’ impression, optical correctives have been used. The entire form may be given a slight curve, or particularly unsure points along the beam accentuated.

The first solution may be compared to the form of the Greek temple floor. Given a very slight convex curve the sinking tendency of the stylobate was counteracted and the entire structure seemed to rise upwards towards heaven. This same principle is also used in modern design, especially in the case of bridges with lengthy spans (Fig. 303). For reasons more visual than technical, the course of the bridge itself is raised to offset an otherwise slack and insecure line movement.54


The curve of a straight beam (bridge in Schwarzenburg, Switzerland, by R. Maillart, from Bihalji-Merin (ed.), Brucken der Welt).

The other solution concerns optical correctives at the beam ends, midpoint and upper and lower edges (Fig. 304).


The curve and correction points of a straight beam.- end points, midpoints, upper- or lower edges.

As we have already pointed out, the support expression of all straight beams depends upon the distance between the supporting points. If the span seems too great the horizontal line may be made more taut by extending the beam beyond the support points on each side. These extensions, both visually and in reality, will function as counter weights which relieve the stress in the middle (Fig. 305, a).


The curve and end point correction of a straight beam: (a) concrete beam by Siegel, (b) the Doric ‘ears’, (c) the Classical pediment.

This effect is a possible explanation of the so-called ‘ears’ of the Doric doorframe (Fig. 305, b). Combined with slightly inward slanting posts a plain rectangle is transformed into a vibrant curve with all the sturdiness of the Doric order. The classical triangular pediment may also convey the same effect (Fig. 305, c). The weight of the triangle’s raking cornices in falling on the outer ends of the straight cornice give this horizontal element an upward bent effect.

A triangle, in itself a stiffening form, gives an additional vertical effect to the whole composition. This theme has been used in endless variations, especially in the Renaissance and Baroque periods (Fig. 306). (See also the triangle as a roof form p. 335 ff).


The curve and midpoint/endpoint correction of a straight beam (door from Hatton Garden in London, from Amery (ed.), Period Houses and their Details).

Another method of counteracting this visual deflection is to emphasize the beam’s keystone (Fig. 307). This is often done in brick beams and very clearly so in Dutch buildings of the 1800’s in which the voussoirs (the wedge-shaped stones) are held in place by a larger keystone. Frequently this keystone protrudes both outwards and upwards in the form of large wedges and consoles (Fig. 308). This triangular form helps to lift the middle section, while at the same time the sturdy concrete gives visual strength to an otherwise weak point in the form. In this motif, it is quite evident that the slightest change in position and dimensions of the keystone can convey the completely opposite effect (Fig. 309). In buildings by both Nicholas Hawksmoor and Giulio Romano, enormous slipped keystones give a threatening rather than a rising effect.


The curve and midpoint correction of a straight beam (keystone).


The curve and midpoint/endpoint correction of a beam (brick warehouses from Amsterdam).


The curve of a beam and an excessively heavy keystone (from the Palace of Justice in Rome by G. Calderini, photo by Chr. Norberg-Schulz).

The third method of counteracting visual deflection is by the use of various mouldings along the upper and lower edges of the beam. The simplest method in this respect is the use of horizontal fluting on the beam’s underside (Fig. 310). This motif, used particularly in the Gothic period and in the 1800’s, gives the effect of removing part of the mass where the stress is greatest, while by arching the corners at the same time the entire form is given a necessary vertical lift.


The curve and lower edge correction of a straight beam (routed moulding).

Expression of Support: The Arched Beam

The arch is the solution to the problem of deflection. Its upward curve resists deflection and is the ultimate consequence of the correctives in the straight beam and the goal for forms and lines of articulation.

In a straight frame, verticals and horizontals are in contrast. Columns and beam are parts added to one another. In an arched frame the accent is vertical because the horizontal element is ‘cancelled’. Columns and beam are joined (Fig. 311, a, b). Thus, in an arched frame the supporting element will always dominate the vertical stress element. Basically, therefore, the arched frame is ‘secure’ and the straight frame ‘insecure’.


The supportive expression of a beam: (a) straight beam (vertical and horizontal addition), (b) arched beam (vertical and horizontal unification).

We find that the support expression of the arched frame is more dependent on the relation between columns and beam than is the straight frame. This relationship is in turn mainly determined by two aspects of the frame’s form.

The first aspect concerns the form of the opening as outlined by its surrounding borders. This form consists of two secondary figures, one within the arch opening at the top and the other within the limits of the columns below (Fig. 312, a-c). As already stated, arch forms are variations on the basic types of pointed, rounded and flat. The rectangle bounded by columns is dictated by the height and distance between these columns and will vary accordingly, but here too within certain fundamental variations: the horizontal, the vertical, and the quadratic rectangle. Thus, the same arch supported by very tall or by very short columns will convey quite different expressions of support. The first will seem ‘free’ and rising, the other heavy and sinking.


The supportive expression of an arched beam and the form of its opening: (a) arch section and column section, (b) the shapes of the arch section (pointed, semicircular, shallow), (c) the shapes of the column section (horizontal, square, vertical).

The other aspect of importance for the effect of the frame is the plastic form given to the arch and columns. Concerning the arch, as already noted, there ore three places in particular which are important: the top of the arch, the upper and lower edges, and the transition to the columns. In the following, the rectangle between the columns will remain unchanged so that we may study the primary effects of the arch form independently. We shall take as our point of departure the principal forms of arch openings, while the role played by various kinds of plastic treatment of both arches and columns will be considered in connection with individual examples.

Expression of Support: The Pointed Arch

Of all the archetypes of the arch, the pointed arch is the most ‘non-bearing’.

The principle behind the connection of column and arch in the Gothic pointed arch is vertical integration, in which arch opening and column opening merge to form an extended, unified whole (Figs. 293, 313).


The supportive expression of a pointed arch: the formal expression of the pointed arch is integrative.

In the Gothic nave verticality is already indicated by the slenderness and height of the columns. The columns become clusters of pipe-like perpendiculars interspersed with deep grooves, which convey a dynamic rising stretch of light and dark lines. There may be no sign at all of the transition to the arch as frequently seen in late Gothic. Most usual, nevertheless, is the spreading foliate capital, which with its organic character further emphasizes the growth of the form. Within the arch are two characteristics which extend this verticality, the steep sides and the point at which they meet.

The sides of the arches are seen as a continuation of the columns’ own structure, as if the columns themselves were bending in towards one another (Fig. 314).55 This unity has both technical and spatial reasons. Technically, the desire was for maximum vertical stress in order to alleviate lateral stress and the need of overly thick walls. Spatially, the capital marking the transition from column to arch could be placed far down the shaft without destroying the main form. In this way, narrow and wide pointed arches could be combined and still maintain common springing points and height.


The formal expression of a pointed arch: arch sides bound together (after Rykwert, On Adam’s House in Paradise).

The point of the arch itself emphasizes the verticality of the form. It seems to tear the very wall it supports (see window p. 263). The point also indicates that the arch does not stop within itself but continues on upward. The arch form, which is constructed by the overlapping segments of two circles, shows that its conclusion lies outside the frame (Fig. 315). The principle behind the continuity of a round arch is that it leads evenly back to the same level at which it began — the form ‘closes’. Just the opposite happens in the pointed arch were continuity is broken at the top point — the form ‘opens’.


The formal expression of a pointed arch: arch sides are parts of two circles which overlap each other.

We realize that the uncompromising upward rise of the pointed arch was not created as a load-bearer. The intention was rather to give a symbolic send-off, a continuation beyond itself of verticality, the very principle of the heavenly ascension. This concept was of such importance in sacred architecture’s use of the pointed arch that buttresses, the visible solution to the stress problem, were placed outside the spiritual space. The Gothic pointed arch, in other words, was not developed to visualize the realities of the forces of gravity but, on the contrary, its purpose was to defy this world, not to follow it.

Expression of Support: The Round Arch

Whereas integration is the keynote in the relation between arch opening and column opening in the pointed frame, addition is the keynote in construction of the round arch frame. This form is seen as a combination of a complete semicircle and a complete rectangle (Fig. 316, a). This means that the frame is composed of independent but balanced upper and lower elements (Fig. 317).


The supportive expression of a semicircular arch: (a) the formal expression of the semicircular arch is additive, (b), (c) the semicircular arch and column support, (d), (e) the semicircular arch and articulation (moulding, end points, keystone).


Semicircular arch and the angle of intersection with column field.


Semicircular arch (from courtyard in The Palazzo della Cancelleria, Rome).

The support expression of the arch is also in balance between to tendencies. In contrast to the accentuated rising aspect of the pointed arch, the true semicircle may be interpreted as either sinking or rising. Read from the bottom, the form may give the effect of a tense curve supporting the load from above (Fig. 316, a). Read from the top this same curve may appear to be an equal distributor of vertical stress. Consequently, determination of the dominating tendency depends on three conditions: the size of the load the arch appears to carry, the form of the columns, and the form and articulation of the arch itself.

An example of major stress on the arch is found in the openings carrying the clerestory walls in Ottonian churches (Fig. 318). The overly heavy wall presses downwards, and the arches distribute the motion towards the ground. In other arches which support no great weight, as for example in the interior of Brunelleschi’s St. Lorenzo, Florence (1421), the stress is vertical and motion is led upwards (Fig. 316, b).


Semicircular arch under load (detail from apartment building, Oslo).

The columns beneath the arch are equally important in the expression of support (Fig. 316, c). If a round arch is supported by square columns that must be interpreted as parts of the wall, the motion has a downward direction. If it is supported by round columns, which in themselves are independent vertical forms, motion is led upwards (see p. 212 f).

The articulation of the arch itself may also have a corresponding effect (Fig. 316, d, e). If an arch is smooth and unarticulated, it will seem compact and heavy, with downward pressure predominating. If carved profiles and mouldings parallel the curve of the arch, the accent is on its light and composite character, which adds vigour to a rising effect. Of equal importance are the springing points of the arch. Only an exact semicircle assures a balanced effect between rising and sinking. If the segment is made smaller, a break will occur in the transition to the columns, and the frame will seem pressed both downwards and outwards by the weight from the top. The opposite effect is achieved if the arch is extended verticality down towards the columns. The arch seems to rise in an involuntary response to the verticals beneath (Fig. 316, f).

Another treatment of the arch itself which is vital to its expression of support is the keystone. It is the last unit to be inserted when constructing an arch and assures the form’s static and functional integrity. It is pressed downwards as well as sideways against the arch flanks. One effect of this is that the arch sides seem to move in the opposite direction. On both sides they spring towards the burden at the top as if to hold it in a ‘vice’ (Fig. 319). In short, an arch with a keystone rises.


Semicircular arch and keystone (detail from apartment building, Oslo).

Until now we have considered the arch as a vertical intermediary between up and down. Its task, however, also concerns the form horizontally. Just like a bridge, it connects two sides by vaulting the gap between. The arch, again just like the bridge, contains a general spirit of triumph and victory. These aspects of the arch are exploited in a meaningful way in Antiquity’s triumphal arches (Fig. 320). These triumphal arches were imperial symbols of victory: the eternal power of imperator and empire made manifest.56 In these constructions bearing the emperor’s statue, the round arch was the central motif and indicated merely by its form the content of its message. The arch is usually decorated with accentuated keystones. These were embellished with sacred symbols and figures, a direct reference to the heavenly allusion of arch and vault (see p. 305 f). Above the arch proper, and thus ‘within’ heaven itself hover two celestial Victories pointing their long victory staffs directly up towards the keystone. Aided by these godesses of victory the entire arch seems to rise towards the keystone with its message of divine triumph.


Semicircular arch as triumphal symbol (reconstruction of Titus’ triumphal arch in Rome, from Fletcher, A History of Architecture).

Expression of Support: The Flat Arch

A typical characteristic of the pointed arch was the extension of the columns into the upper part of the opening and their integration into a dynamic form. In the round arch frame this relationship was in balance; motion between the two elements was variable — at times upwards and at other times downwards. In contrast to these, motion in the flat arch is an unequivocal sinking. It conveys a feeling of pressure from above (Figs. 321 a, 322). Of these flat arches, the Tudor and the Art Nouveau types are the most typical. Both are to be seen as deformations of the two arch types examined above. The Tudor arch has its source in the pointed arch and the Art Nouveau arch in the round arch (Fig. 321, c, d). The Tudor arch can be interpreted as a profane version of the sacred pointed arch. It was particularly widespread in England during the sixteenth century and frequently used in private homes and other secular buildings in an intended contrast to the more sacred pointed arch (Fig. 323). Earthly forces press down on the taut sides of the pointed arch causing the whole form to collapse. This collapse in the Tudor arch, however, is never total. The counter movement of its sharp point rising in the centre allows the arch to ‘survive’ despite the pressure. Considered thus, the Tudor arch points to a synthesis of vertical and horizontal, and as such is part of the transition from Gothic to the Renaissance.


The supportive expression of a shallow arch: the formal expression of a shallow arch is determined by a deformation of the semicircular arch (Jugend arch) or of the pointed arch (Tudor arch).


Shallow arch (detail from prison in Pittsburg by H.H. Richardson inspired by the ‘Bridge of Sighs’ in Venice. Photo from van Rensselaer, Henry Hobson Richardson and His Works).


Shallow arch (loggia with Tudor arches, from Koepf, Baukunst in Funf Jahrtausenden).

The Art Nouveau arch is typical of the new ideas emerging at the close of the nineteenth century. This ‘new art’ was thought if as being freed from earlier style-bound forms but nevertheless its sources were deeply rooted in history. The Middle Ages, Rococo and The Baroque were all forefathers to its special characteristics. Thus the Art Nouveau arch is a paraphrase of Romanesque heaviness and Gothic flamboyance (Fig. 324).


Shallow arch (entry arch in Jugend style, from machine hall at Sollern II/IV coal mine in Dortmund-Bovinghausen, by B. Mohring).

The Art Nouveau arch is a continuous unbroken form, as may be seen in Josef Hoffmann’s arched doorway (1898, Fig. 325). Hector Guimard’s work must be interpreted in the same way. His arches grow from their columns as plants from their stems to join in a lightly swaying downward curving crown of foliage (Fig. 326).


Shallow arch and plasticism (sketch of an entrance, by J. Hoffmann, From Tschudi-Madsen, Sources of Art Nouveau).


Shallow arch and continuation (room by H. Guimard, from Naylor, Hector Guimard).

This combination of suppleness and downward curves in these organic wrought iron forms is also expressed in a completely different context. In The Romanesque rusticated wall, the Art Nouveau arch is a visualization of the entire form’s suppleness. Much of Henrik Bull’s work confirms his understanding of this content. In the Museum of History, Oslo.(1902), he allows his arch around the entrance to be the gauge for the entire body of the building (Fig. 327). The entrance is set in the central part of the building, which has a rusticated basestorev. Above this the wall is opened by a large staircase window between two towers. The projecting section is finished off at the top by a curved attic decorated with the royal coat of arms. The arch above the main entrance is pressed flat by a severe and heavy moulding marking the transition to the window area. The articulation of the arch maintains this character. Beneath each end of the arch stands a pair of squat columns on a thick, broad console. Above them and between the edge of the arch and the bordering moulding are two owls, not erect and scowling but wide-eyed and with flapping wings (Fig. 328).


Shallow arch and tension (entrance to The Museum of History in Oslo, by H. Bull, from Thiis-Evensen, Henrik Bull).


Shallow arch and plasticism (detail from entrance to The Museum of History in Oslo, by H. Bull).

The large arched opening leading to the staircase is also pressed flat by the heavy moulding. The arch is squashed down in the middle and bulges out at the corners. In this opening stretched between two massive corner towers we find a wholeness full of contrasts, an expanding opening held in check by the two verticals. Even the royal coat-of-arms above is set in a distorted form. In this case it is a circle squashed by two tautly drawn mouldings.

It is clear that the Art Nouveau arch is used to illustrate the inside-outside conflict itself (Fig. 329). The lower part clearly shows the conflict that arises by inserting a large opening in the most closed part of a wall. It symbolizes exterior pressure, while the window area above reveals an inner expansion. At the very top, it is the weight of the roof which converts the form into a set of compressed curves.


Shallow arch and tension (diagram of entrance to The Museum of History in Oslo, by H. Bull).

The Beam’s Expression of Motion

We have seen how the visual problem of sinking has been solved in vastly different ways by various beam forms. With the same articulation we saw that the straight beam is the ‘weakest’, the pointed arch the ‘freest’, and the round arch in ‘balance’, whereas the flat arch is ‘compressed’.

The same beam forms also elicit highly varying motion impulses in us (Fig. 330, a-d). A frame with a straight beam will be more of a barrier between inside and outside than one with a curved beam. Within the arched variations, the round arch is the most open, whereas the pointed and the flat arch, each in its own way, lessen the penetrative effect. Why?


The expression of a beam: (a) straight and impaired motion, (b) semicircular arch and penetrating motion, (c) pointed arch and rising motion, (d) shallow arch and closing motion.

A straight beam will accent a horizontal motion crosswise to the penetrative motion. Just as in the case of the horizontal wall, and as we shall see later, the horizontal window (see p. 261), a motion paralleling the construction is indicated, a motion that interrupts our intended entry.

The arched beam seems to be lifted by the opening itself. The arch makes the entire frame rise to outline a vertical form, and this verticality indicates a direction which, as already stated, initiates direct communication between inside and outside.

Of the three arch forms, the round arch most clearly emphasizes the act of entering in itself. Our own head in the centre of the semicircular curve, which means that the opening outlined by the arch is caused by us alone and not by external forces.

In the pointed arch the self-centred quality found in the round arch disappears. Here the form points upwards and away from us. The opening is less directly concerned with us and our intended entry. The form is the result of something ‘up there’, something beyond our own action which is to be led inside.

In the flat arch the form is distorted in the opposite way. The weight supported by the arch presses downwards, conveying a potential closing of the entrance passage itself.

Expression of Motion: Colonnade and Arcade

The principal difference in the expression of motion between the straight and the arched beam is most clearly seen when the frame is repeated and becomes a row. As we have seen, it is the straight frame that forms a colonnade and the arched frame that forms an arcade (Fig. 331, a, b).


The expression in (a) a colonnade and (b) an arcade.

In a colonnade the beam becomes one continuous horizontal moulding which accents the principal motion that parallels the colonnade. This motion seems to take place behind and within the row of columns themselves, a characteristic which further accents the limitation in the form. Despite the apertures between the columns, a straight beam gives less direct communication between inside and outside.

The arcade consists of a row of individual openings in which the arches both separate and isolate each opening, which becomes a self-contained unit. Despite the horizontal extension, the arched beam gives greater communication between inside and outside.

Thus, an arcade indicates a penetrative motion, the colonnade a lateral motion.

A classical example of these differences combined in a meaningful way is the so-called Palladian motif (Fig. 332). The motif comprises three openings, the central one spanned by an arch and those on either side by straight beams. The centre area is used as the entrance, while the side usually consists of full-length windows. The side apertures, in other words, are intended to keep us within the columns, while the central area is for passage in and out.


Straight beam and arched beam combined (‘Palladian motif’, detail from the Basilica in Vicenza by A. Palladio).

The duality of the Palladian motif may be confirmed by ancient precedents. In the silver missorium in Madrid (388) the motif is given ideological meaning, in that the straight frame and the arched frame convey different depths of contact between the onlooker and the persons portrayed in the opening (Fig. 333).


Straight beam and arched beam combined (silver missorium from Madrid with emperor in the middle and his sons at either side).

Emperor Theodosius I commands the centre area beneath the arch. His sons sit on either side. The figure of the emperor is the largest, the sons are somewhat smaller. The arched opening emphasizes these dimensions. The person of the emperor appears to be led forth’ beneath the arched frame — he becomes more direct and inviting. The flat beam, on the other hand, holds the sons back in relation to the middle. These beams, with their lateral motion, point directly at the emperor, thereby accenting the ideological basis of the composition in the figure of the emperor itself. It is he who is the unifying force and the centre of the world (Fig. 334) (see also gables p. 352).57


Diagram of silver missorium from Madrid.


Arcade and colonnade in combination (diagram of motion patterns in St. Peter’s, Rome).

Differences in the expression of motion between arcade and colonnade are also used to distinguist between dissimilar spatial forms.

An example of this is the ancient complex of St. Peter’s, Rome (333) (Figs. 35, 335 a, b). The mighty nave leads directly to the high altar in the apse, which is the meaning and goal of the entire interior. The architraves above the columns are made straight in order to emphasize unequivocally the linear and extended length of the form all the way to its finish. The quadratic peristyle surrounding the courtyard in front of the church is in the form of an arcade. In the middle of the square is a well, the centre of the courtyard’s liturgical content. From this point of view the arcade is an answer to the central form itself, just as the colonnade was to the rectangular form. With its tall arches the space opens equally in towards the centre as if reflecting the rays radiating from the central point. At only two places is the continuity broken: in the east by the raised baldachin leading into the church and in the west by the three-arched entrance to the entire complex.

In this way the differences between straight and arched openings help to tie the two spatial forms together. The entrance opens directly out towards the incoming throngs. Once within the peristyle all attention is focused on the central well then led further in a straight line through the narrow baldachin and into the church nave. Here the straight architrave takes over and leads the eye directly into the curve of the apse. In this way all of four stages in a pilgrim’s progress are united. From the outside world one passes to the next stage of baptism and preparation and from there to follow the path of humility to the goal itself, the altar as the symbol of Christ.

It is obvious, of course, that these directional tendencies found in arcades and colonnades are merely a base from which each individual case will be weakened or strengthened by different kinds of articulation. The rows of columns may be long or tall, the intercolumniation equal orin a rhythmicsequence, or they may be combined and overlapped, etc. An important factor in this context is the column’s form. If the columns of a colonnade are round, the transverse direction is increased at the cost of the-lateral effect of the main form and architrave. If the columns are square, just the opposite takes place — the accent is on the barrier effect and the row will be perceived more as a wall with ‘holes’.

The architectural development of column rows, as we are able to follow it from the earliest basilicas, shows in fact that a rhythmic alternation of round and square columns was an important means of attaining varied spatial effects.58 In the inner arcade of St. Maria in Cosmedin, Rome (sixth century) (Fig. 336) a wide buttressing column with a square cross-section is introduced for every three round columns.59 These pillars both break the arcade’s natural transverse motion and emphasize the principal direction of the nave. At the same time they aid in dividing the space into three zones (Fig. 337). These divisions are functional, because each zone has a different ‘value’. The lowest zone is closest to the entrance and signifies entry or preparation. The middle zone, which is broken up by choir stalls and lecterns, is the area in which the congregation worships and gives praise. At the top is the choir zone itself, to be entered only by the clergy and used for the preparation of the sacrament.


Column alternation (St. Maria in Cosmedin).


Column alternation (plan of St. Maria in Cosmedin, from Roma e Dintorni).

The Colours

‘Plain drawing is an abstraction … because everything in nature contains colours. Only when the colours are rich is the form worthy of merit’.60

Through colour and light the world is made known to us. Colour is the device that indicates the structure of our surroundings, because everything is composed of colours.

Colours, therefore, embody specific meanings. In many cultures black and yellow symbolize sorrow. Red means danger, while green is identified with growth. On the water faucet hot is marked with red and cold with blue. In language too, colours convey particular meanings: ‘when I look at him I see red’, ‘at night all cats are grey’, ‘the black heart’, etc. These miscellaneous examples suffice to justify the following question. Is the perception of a colour’s meaning relative and conventional or does the individual colour have a given and unchangeable meaning?

If the question is to be answered, it is necessary to consider some of the widespread research that has been done on the significance of colour. Of the visual phenomena it is colour which has attracted the greatest attention, particularly in recent times. Colour studies may be divided into two areas. The first, already tackled by Newton in his ‘Opticks’ from 1704, concerns mainly the physical and chemical content of colours — what Goethe called the ‘atomistic’ view of colour. It is Goethe in particular who has studied colours from the other viewpoint. Using the term ‘Die sinnlich-sittliche Wirkung’, he is among those seeking to analyse the psychic effects of colour.61 In this connection it is maintained that certain colours and colour combinations will cause specific emotional states which, in certain situations, are decisive for our psychic and motor reactions. This has been experienced, not least of all, by artists throughout the ages, in that architects, designers, and textile artists have been absorbed by the principles of colours’ compositional and harmonious nature.

The Expression of Colours and The Importance of Composition, Association and Person

By comparing the viewpoints which have emerged in this field of research it is clear that especially three factors are considered to be particularly influential in determining the effect of colours.62 One is composition, by which is meant that the effect of a particular colour depends on whether it is modified or enhanced by neighbouring colours. Artists know that colours do not act alone but have their effect in concert with other colours. ‘Of different colours equally perfect, that will appear most excellent which is seen near its direct contrary: a pale colour against red, a black upon white, blue near yellow, green near red’.63 Modern researchers also, such as Erich Raab, stress the importance of interdependence in the effects of individual colours:

The pleasure aroused by a certain combination of colours depends upon the individual colour tones in combination, (2) the pleasure conveyed by the whole composition cannot be predicted on a basis of individual colours. The impression is also partly qualified by the gestaIt-effect and the quantity of the colour distribution.64

The second factor of importance in the effect of colour is association. This means that the effect of a colour will depend upon whatever it is compared with. This association with the environment is partly dependent upon conditioning and conceptual habits. The comparison may be limited to the colours alone as, for example, when one says that red stands for blood or fire while green signifies the fresh growth of nature. On the other hand, the effect may depend upon the form with which a colour is associated, whether it is abstract (circle, square, organic, etc.) or descriptive (nature, animal, human, etc.).65 M.J. Friedlander says accordingly: ‘Only coloured form exists and correspondingly, only formed colour’.66

The third precondition in the effect of colour is the person himself. This implies that personal characteristics, sex, age and status all take part in ‘filtering’ the meaning in colour. Aristotle maintained in fact that certain colours are preferred by specific personality types: red by the ‘bon vivant’, purple by the snob, yellow by the schemer, and blue by the intelligent.67 A modern researcher such as G.J. von Allesch doubts, furthermore, that colour effect can be understood at all without an exact knowledge of the recipient’s personality.68 Based on statistical investigations, Raab sees a direct correlation between certain colour combinations and the observer’s sex, as for example: ‘Men prefer the colour combination red/blue while women (but in a less pronounced way) prefer the combinations red/yellow and green/yellow’.69

The Expression of Colours As Motion, Weight, and Substance

In terms of the generation of their effects, colours do not differ from other phenomena. Just as in other architectural elements the effect of colour is a factor of varying preconceptions.

But do colours also have an existential expression?

Many researchers maintain that each colour is presumed to have its own particular expression, which is basic and invariable.70

If we look at our introductory examples, we find that there is an inner structural similarity in the feeling expressed and the phenomenon that is symbolized. Black is chosen for sorrow for just like black, sorrow is heavy, deep and rejecting.71 Yellow is chosen because it can be associated with our reactions to light and sun — and light is sacred in the sense that it describes that other world of which the deceased is now a part. Furthermore, physical heat is aggressive; it burns you and is intense in the way red is said to be. Blue, however, is cool and chilling just like cold water. The existential expression of colours is, therefore, once more to be found in connection with our basic physical reactions to the environment.

Rudolph Arnheim has summarized these expressions in the term, the dynamics of colours.72 By this is meant the inherent expression of motion in colour characterized by such terms and ‘distance’, ‘nearness’, extension’, ‘contraction’, ‘expansion’, etc.

There are particularly two conditions which give these expressions.

First, we find that colours differ in weight and size. Hue, saturation and brightness all posses varying degrees of ‘material form’. They are preceivedas having ‘weight’ and ‘size’, which is to say that thay are directly compared with other experiences of weight and physical reality.

Secondly, we may sense that colours have different temperatures. Some colours are described as ‘warm’, others as ‘cold’, qualities also instrumental in relating them to our experiences and to physical reality.

Colours are not only important as an aid in emphasizing details in a building’s structure (Figs. 338, 339, 340). As a dynamic phenomenon colours are also instrumental in giving a space motion and thus have a direct bearing on the relation between inside and outside. In connection with walls, it is important to explore the qualities of colours which convey varying impressions of weight, size and temperature.


Colour and articulation of plastic form (Egyptian temple in Edfu, from Uhde, Die Architekturformen des Klassischen Altertums).


Colour and articulation of the skeleton (half-timber building from Celle in northern Germany).


Colour and articulation of the surface (model of Schroeder House by G.T. Rietveld).

Hue, Brightness and Saturation

The expression of motion in colours is decided, in principle, by the interaction of three qualities.

First is the hue, which distinguishes the colour type. In general we recognize three primary colours: yellow, red, and blue. The complementary colours are orange, green, and violet.

The second is brightness, which describes hues according to their tendency towards white or black.

The third quality is saturation or density, which is a measurement of the strength or weakness of a colour.

In considering the relationship of these three qualities it will be found that saturation is independent of hue. Thus, all colours may be either strong or weak. Brightness, on the other hand, is more closely associated with the hue itself. Accordingly, yellow will tend towards white, while blue tends towards black. It follows that there is an inner relation in the qualities of pale and yellow and of dark and blue. In all cases the three qualities will affect each other mutually. Furthermore, if one of them is to be examined, the other two must remain constant. The following account is based on this necessary limitation.

Brightness and Saturation as Weight

It is the degree of brightness and saturation in a colour which makes the wall it covers seem heavy or light, larger or smaller.

There are, of course, other qualities in colours which may give a corresponding effect. As has already been said, hue is important in that yellow and blue also convey lightness and darkness. The same applies to the association evoked. A pale brown colour will, accordingly, seem heavier than a pale blue, since brown is ossociated with the earth and blue with the sky. Correspondingly, saturation and brightness convey a feeling of warmth or coldness. A pale and unsaturated colour, therefore, tends to appear cool and chilly while a dark colour tends towards closeness and warmth.

Degrees of weight and size, nevertheless, are the most typical impressions conveyed by differences in the brightness and saturation of colours. What effect then does this have on the inside-outside relationship? Pale and unsaturated colours share the quality of giving a surface the appearance of lightness and openness. Dark and saturated colours share in giving a feeling of weight and closure.73

The explanation of these fundamental effects is once again to be found in our common experiences. Both that which is pale and that which is unsaturated are by nature light and airy, having within them expansive and outgoing qualities just as light and day. That which is dark, however, draws together and becomes denser — it becomes smaller but also harder and more impenetrable, somehow like night. One’s nature is liberated by white and constrained by black. Pale colours, therefore, are ‘friendlier’ than dark ones. Dark colours inhibit our natural life develoment.

Darkness is without light and thus prevents the possibility of any meaningful activity in the world. It is first when man renounces the daylight clarity of the world and entrusts himself with uncertainty to darkness that it displays itself in its own character.74

Saturation and brightness, however, convey openness and closure in quite different ways.

An unsaturated colour, be it pale or dark, has about it an air of lightness which conveys an impression of opening in depth. A wall is given texture when covered by on unsaturated colour. The pigments are loosely connected to each other, one sees between and ‘behind’ them. The wall ‘opens up’ and is made lighter as in a three-dimensional curtain.

In a saturated colour, on the other hand, the pigments are pressed more densely together — the wall it covers becomes solid and strong, giving the appearance of a plainer two-dimensional expanse.

The lightness occurring with the use of a pale colour does not lie in a depth effect but in a weight effect. It means that this characteristic of weight or lightness lies always in and along the surface itself. A dark colour, therefore, will increase, but a pale colour lessens the area’s own expression of weight.

It is the degree of transparency and self-weight which tells. The same factors play their part in giving an area different dimensional effects. In this connection Goethe maintained that a dark object appears smallerthan the same object when light. He claimed that a black shape on a white ground appears to be one-fifth less in size than a white shape on a black ground. He points out, at the same time, the accepted fact that black clothes have a slimming effect while white have just the opposite.75 (See also relation of colourand column p. 204). Thus, basically, a light coloured house will seem larger and more open than the same house in dark colours, which in turn will appear more closed and ‘withdrawn’.

The same applies in the dividing of walls. A wall several storeys high, all in the same colour, degree of saturation and brightness, gives the effect of a stiff slab without transition to ground or roof. If, however, the storeys differ in the degree of brightness and saturation, a darker lower section and paler upper part will make the wall sit well against the ground while at the same time conveying a lighter effect as it rises upwards (see also rising effect p. 135). If the process is reversed, the wall will appear to be pressed downwards from above (see sinking effect motif p. 135).

In the next stage let us imagine a skeletal system fastened to the wall and presuppose the whole to be treated in the same colour but with skeleton and wall given different degrees of brightness and saturation.

In the first alternative the skeleton is pale and the wall behind dark. This combination gives an impression of depth. The light coloured skeleton expands forward and out — the dark wall behind contracts and draws inward. This shows the difference in the function of the two parts. The skeleton is active and expansive in contrast to the wall’s more closing-off role. The skeleton system, as the more prominent and public element, is emphasized as a part of the exterior space, while it is the private character of the wall behind which is accented.

In the second alternative, the skeleton is made dark and the wall light.

Now the two systems ‘meet’ in that the skeleton contracts and is drawn backwards, while the wall opens up and expands forward. The wall becomes independent and the skeleton appears to be ‘stuck on’.

We have seen that the effect of brightness and saturation on a wall’s expression of strength and weight differs in the case of height and that of depth.

The effect on height follows the principles described above in that the dark appears heavy and forbidding while the paler area seems lighter and more open.

Seen in depth, the reverse may take place. As shown above, pale hues will lend strength to wall elements without necessarily giving them a ‘heavy’ character. Dark colours, furthermore, will in certain circumstances appear as ‘holes’ in the wall surface and thereby open up the wall rather than close it, which the relation beween weight and dark would otherwise lead one to expect.

Classical palace and tenement architecture is a good example of how such variations in lighter and darker areas add to a rich interpretation of just this inside-outside relationship (Figs. 341, 342, 343).


Differentiated surface (Nobel Institute in Oslo). 343. Differentiated skeleton/background (apartment building in Oslo).


Differentiated skeleton/background (apartment building in Oslo).


Differentiated skeleton/background (apartment building in Oslo).

As presupposed above, the examples have had the same hue throughout. If this too is varied, the picture immediately becomes more complicated. If we visualize for a momentan imaginary wall in which wall and skeleton have the same degree of brightness and saturation but different colours, one colour may appear heavier and stronger than the other. This may be seen in the illustration of a church faҫade in Sutri, Italy (Fig. 344). The pilasters are in grey tuff, the wall itself in reddish tuff. The pilasters stand out immediately as the strong and supporting elements because the colour grey is associated with a harder stone, whereas the intervening walls are identified with a more porous and brick-like stone type. The impression in this case is different, governed as it is by association, but, it will be seen as well that colours contain essentially different weight values. With that we turn to the importance of hues in the relation between inside and outside.


Differentiated skeleton/background (church facade from Sutri, drawing from photograph).

Hues As Substance

While a colour’s degree of saturation and brightness first and foremost give a surface its character of varying weight and size, hue will be a primary factor in deciding the varying degress of heat and cold in the same surface.76

By this it is meant that certain colours cause definite psychic reactions which correspond in structure to our physical reactions to heat and cold. Heat incites and intensifies, whereas cold freezes and stiffens like something dead.

This psychic reaction has been measured as a purely physical condition. F. Birren, accordingly, has established that saturated and warm colours af feet the automatic nervous system in such a way that blood pressure rises and the pulse quickens.77 Subjectively one experiences a rise in temperature. With cold and unsaturated colours blood pressure sinks and the pulse slows. The temperature seems to fall.

J. Itten also reports similar reactions.78 Two identical work-rooms, one in blue-green and the other in red-orange, showed a difference in experienced temperature of as much as 3 to 4 degrees, although the temperature in both rooms was exactly the same (15 °C).

What effect then does this have on the wall’s open or closed character?

To find an answer we might compare our experience of heat and cold with our spontaneous reactions to ‘extroverted’ and ‘introverted’ people. A warm person invites us to approach while a cold person makes us stop and withdraw. Our reactions simply mirror qualities in the person himself. The cold person behaves as if he himself felt cold — unwilling, reserved and closed. The warm person, on the other hand, has a vitality that advances towards us. The same holds true for colours. Warm colours seem to invite us, cold colours keep us at a distance. The relationship may be determined dynamically, in terms of motion in depth: with us, before us, or from us. In the following we will examine the principal motion content in each individual hue and thereby indicate the influence of hues on the wall’s welcoming or rejecting qualities. We will follow Wassily Kandinsky’s descriptions. Among researchers and artists it is perhaps Kandinsky who most clearly has stated the visual possibilities of colour hues.79

Hues As Motion

For Kandinsky the dynamics in the contrast between warm and cold colours is ‘the first great antithesis’.80 The other contrast is to be found between white and black, between light and dark colours. The degree of lightness and darkness can lend nuances but not change the intensity of the dynamics in warm and cold colours. If the colours are pale, they become insubstantial or dissolved, they have ‘no resistance’. If they are dark they become denser, more concrete and assume weight.

Kandinsky maintains furthermore that the colours yellow and blue constitute the extreme on the scale of dynamics. The warm colours, tending towards yellow, will advance towards the spectator, whereas cold colours, tending towards blue, will retreat. Areas of colour will also vary in relation to themselves. Yellowish tones will show eccentric motion, bluish tones concentric. ‘Yellow easily becomes sharp and never shows depth. Conversely, blue can never shine forth and up’ (Fig. 345 a, b).81


W. Kandinsky’s colour-dynamic model (from Kandinsky, Concerning the Spiritual in Art).

In a motion context red and green are found between the two colour extremes yellow and blue. The colour red moves ‘within itself’ says Kandinsky — its power is to be found in a sort of contained potential: ‘…red does not have the irresponsible appeal of yellow … it glows maturely within itself and does not shine forth aimlessly’.82 The colour green also has its own inherent quality, but this is one of repose. The contrast of its complementary colours yellow and blue is balanced in a motionless calm. Pure green is the most restful colour lacking any undertone of joy, grief or passion.83 In contrast to grey, which Kandinsky considers a ‘hopelessly motionless’ colour because it has its source in the lack of tonality in white and black, green in different mixtures is able to break into either forward or backward motion because of its position between two active colours.

This is true too of the other colours. By mixing, primary qualities may be subdued or strengthened. With the addition of yellow, red will radiate as orange, if black is added to red it will sink towards brown. The latter colour is ‘immovable’.84 When blue is added to yellow, the yellow colour will be checked and ‘die away’.85

We see from this that colour hues are of great importance in the impression of a building’s welcoming aspect to the world outside. If all the colours of a house are in their purest form, with the same degree of saturation and brightness and without undue influence from their environment, they will ‘stimulate us’ very differently. The grey wall may give a deadly sensation, convey an insignificant, empty impression as if the house were unoccupied. A green wall may emanate a restful expression and give the house a sort of lived-in character. The red wall may impart a stirring and intense clang, a wall that attracts on the strength of its own supremacy. The yellow wall will shine brightly towards us like an invitation to a sunny place, whereas the blue wall will retreat in cool reservation. The white wall is immaculate. It may convey an expression of elegance, as in le Corbusier’s early villas, or of innocence and friendliness, as in the white idyllic seaside houses of southern Norway (Figs. 346, 347).


The ‘heavy’ inland architecture (Langbrata drawn by Th. Kittelsen, from TheodorKittelsen i tekst, tegninger og malerier, Oslo 1957).


The ‘lightweight’ coastal architecture (a house in Arendal, Norway).


The Window

The Window and Inside-Outside Relationship

The wall’s fourth and final theme is that of openings.

An opening in a wall occurs when the wall’s structural system is interrupted either in the form of a hole in a plane wall or a change of rhythm in a skeletal wall (Fig. 348).


Opening as a break in the grid system (sketch of a skyscraper by Le Corbusier, from Le Corbusier, Var bostad).

The aperture need not be a true one in order to be perceived as an opening. Blind windows or a false door are also openings (Fig. 349). The precondition is that the wall acts as a ground, while the openings stand out as figures.


Opening as blind window (from an apartment building in Oslo. Photo by T. Lange).

The window and the door are two types of openings which function very differently in the relation between inside and outside. The basic difference is that the window is meant to be looked through and to admit light, whereas the door is primilarily to be gone through.86 While the door is determined by its relation to what is outside, the window is the symbol of what is inside. Just as the eye, it expresses the interior’s outlook over exterior space, while as a light source it bears witness to the fact that light is necessary for the use of the interior.

This means that the window, regardless of form, size and location, will always be an expression of the interior to the world at large. Consequently, it is the windows which announce our mode of life, according to Loudon:

A cottage with one bedroom, a living-room and a bathroom ought to express these three rooms by using windows of three different sizes. Windows can also express differences in floor plans, the presence of stairs and other aspects of the plan.87

In other words, regardless of the reason for using a small or large window, it will by its size alone describe the relation of inside to outside. It is invariably the ‘struggle’ between interior space and exterior space which the window expresses, a question of whether the interior seems to be drawn outwards or whether it remains protected within the dividing wall.

In the following we shall examine the elements in a window composition which determine the window’s importance in opening or closing the space. Each element will be treated individually and in introducing them they will be described and given a technical/functional foundation. We shall then explain their potential expression in relation to the expression of motion, weight, and substance.

This study will be based on the window as a hole in a planar wall. The skeletal system will be discussed only in connection with the articulation of the window elements.

Window Motifs

Four elements in the window’s composition exert their effect either in combination or individually. These are (a) the opening in the wall, (b) the face in the opening, (c) the frame around the opening, (d) the space in front of the opening (the bay) (Fig. 350, a-d).


Windox motifs: (a) hole, (b) face, (c) frame, (d) bay.

The Hole

The primary element is the hole. Two main factors, form and profile, determine its importance in the inside-outside relation.

The form is based on three variations (Fig. 351, a-c). The first is the vertical window with upright orientation. Traditionally, the verticality effects a variation of three motifs all determined by the form of the lintel, whether it is a round arch, pointed arch or straight arch. The second form is the horizontal window, which in principle conveys a sideways motion. The third form is the centralized window, which is essentially neutral, marks a point, and may vary from a square to a circle.


Window openings: (a) vertical opening (circular arch — pointed arch — straight arch), (b) horizontal opening, (c) centralized opening (square — circular — polygonal).

Throughout architectural history the vertical window has been the type most used. It was preferred because it met two different factors: the limited width of the span in a post and beam system and the desire for the largest possible opening to admit maximum light (Fig. 352).


Vertical opening Grathic window (from Teyn Church in Prague).

The horizontal window is particularly associated with the Functionalism of the 1920’s. One of its origins was the desire to open space horizontally (Fig. 353). This was made technically possible by the introduction of reinforced concrete, allowing a greater span above openings.


Horizontal opening the functionalistic window. (From Villa de Monzie, 1927, by Le Carbusier, from Kenneth Frampton, Madern Architecture 1851–1945).

The central window is less determined by technical factors than both the vertical and the horizontal window. A well-known exception is Louis Kahn’s circular openings in his building complex in the capital of Bangladesh (1962—1974). Because of frequent earthquakes the round form was necessary to offset shear forces in the building mass (Fig. 354).


Centralized opening (from the Governmental Centre in Dacca by LI. Kahn, from Narberg-Schulz/Digerud, Louis I. Khan, idea a immagine).

The Profile

The profile outlining the opening can be straight or diagonal (Fig. 355, a, b).


The opening’s profile: (a) right-angled profile, (b) diagonal profile.

By a straight profile we mean one that is cut into the wall at right angles (Fig. 356). By a diagonal profile we mean that the sides of the form slant inwards to a smaller light aperture. Such a reveal can be formed in many ways. Typical examples stretch all the way from simple rounded-off forms as often seen in plastic wall constructions (Fig. 357), through deep and richly decorated stepped profiles found in Romanesque and Gothic niche openings right up to smooth, deep diagonal reveals such as those in Le Corbusier’s Notre Dame du Haut, Ronchamp (Figs. 358, 359 a-c).


Right-angled profile (from Notre Dame du Haut by Le Corbusier.


Diagonal profile: formed plastically (Casa Mild in Barcelona by A. Gaudi).


Diagonal profile: formed as simple angled surfaces (from Notre Dame du Haut by Le Corbusier).


Diagonal profile: (a) plastically formed, (b) stepped, (c) angled.

The Face

The next element in a window composition is the window face. This includes the frame and system of transoms and mullions breaking up the face. It also includes the covering, which may be of glass or other transluscent materials such as parchment, paper and certain types of stone. The same applies to coverings made of more solid materials such as shutters, Venetian blinds, and curtains, but also open grille and lattice systems.

The window face is a collective expression for all transparent or other types of covering which close the window opening.

As well as the structure of the fgce itself, its location within the opening is of importance. Whether alone or in combination the layers of the face will always be found within four extreme positions (Fig. 360, a-d): (a) inside the opening, (b) in the middle of the opening, (c) outermost in the opening, (d) outside the opening.


Location of window face: (a) inside the opening, (b) in the middle of the opening, (c) outermost in the opening, (d) outside the opening.

The Frame

The next element in a window composition is the frame.

The relation of the frame to the wall opening itself is decisive in the extreme variations. Here there are particularly three motifs which stand out (Fig. 361, a-c).


Location of the frame: (a) within the opening, (b) around the opening, (c) in front of the wall and extended past the opening (depth), (d) in front of the wall and over the opening (height), (e) in front of the wall and to the side of the opening (breadth)

The first is the frame that lies within the opening. In the second motif the frame is free of the opening in so far as it is attached to the wall around and outside the opening. In the third motif the frame becomes part of a larger skeletal system in front of the wall. In both the second and third motifs the frame’s location means that it may vary in size and form independently of the opening’s form. The frame is freed. This independence may be found in depth, height and width.

Within all three motifs discussed above, four possible interpretations stand out. Each of these is determined by the part of the frame which decides the visual effect (Fig. 362, a-d).


The frame’s components: (a) complete, (b) sill, (c) lintel (d) jambs.

These four variations may appear each in its own ‘pure’ style or in combination. By a pure interpretation we mean that the frame motif is seen against an otherwise plain diagonal or straight cut reveal around the opening.

In the first interpretation the frame is complete, i.e. the lintel, sill, and jambs, with no one part dominating, form an unbroken contour around the window.

In the next two variations, either the lintel or sill alone is the dominating element.

In the final variant it is the jambs which convey the main impression.

In the above examples, the basic variations were in their ‘pure’ form. The most usual, however, is combined interpretations in which the basis is a plain surrounding frame but in which lintel or still, together or individually, dominates the overall impression (Fig. 363).


The complete classical frame with an accentuated sill in the form of a straight, arched or triangular pediment (frame around a bust of Palladio, from Cevese, Polladio).

The Bay Window

When a window occurs as a frame, it will, at the same time, be perceived as a space. A frame describes a separate spatial volume with its own ceiling, floor and wall. Its boundaries are within the hole. The word ‘aedicula’, which indeed describes a frame, means a ‘little house’. If the frame is fastened to the wall around an opening, either as an independent element or as part of a skeletal system, it also suggests a potential space, but in this case one in front of the opening.

In spite of this, the window as space is considered to be in a separate category within window architecture. It appears in the form of a bay window. In the bay window both window face and frame are secondary elements of the main form. The bay window’s four variants, however, are completely determined by their relationship to both the window opening itself and to the wall in which it is located (Fig. 364, a-d).


Bay window motifs: (a) as a bulge in the wall, (b) out of a split in the wall, (c) as overlapping element in the wall, (d) as attached to the wall.

In the first variant the bay window appears an an outward bulge in the wall itself. In other words, the wall of the bay window is a part of the overall wall system.

The second variant is seen when the bay window form bursts through the wall, irrespective of whether the wall is built as an open skeleton or as a closed massif. In such cases the bay window seems to push its way through great slashes in the wall or out between pilasters and columns.

In the third variation, the bay window must be seen as an overlapping space that both protrudes from the wall and is drawn into it.

The fourth variant appears when the bay window is attached to the wall in front of the opening as an independent volume. One method of doing this is to place the bay window on its own support in front of the wall. Another way is to fasten it to the wall by the use of exposed consoles or with overlapping ‘wings’.

The Expression of The Window

The next question is: what do the hole, the face, the frame and the bay express in relation to motion, weight, and substance, and what is their importance for the way in which the connection between outside and inside is experienced?

The Hole

‘Houses with windows which are merely holes in the wall are like empty skulls’ (Fig. 365).88


Window without frame (Column house in Retz by E.-L. Boullée, from Arthaud, Dream Palaces…).

This was said by John Ruskin and reveals the essence of the effect of a window as a naked hole. A window that is only a gaping hole in the wall transforms the wall to a lifeless skin around a dead and empty interior. The hole is not the result of an interior force, rather it seems to have been punched in from the outside.

So, a house with gaping openings will always remind us of a ruin. The interior no longer seems like private property but rather a deserted, empty space — a common property. A space without window frames and glass is physically unprotected as well. Wind and weather play havoc with what the wall is supposed to protect. Thus, an empty window is merely a hole and influences the entire meaning of the wall. It loses its function of protection and becomes lifeless.

The Profile

The impression that the motion is from the outside inwards can be heightened or lessened by the opening’s profiles or reveals.

A cut at right angles to the wall emphasizes motion from the outside (Fig. 366, a). The strength of the wall is weakened, it ‘offers no resistance’. With a straight profile it is as if the wall’s own substance deflects the incision. The entire wall takes on a thin character, a stiff plane with no muscular strength.


The profile and the motion expressions of the opening: (a) the profile emphasizes motion from the outside, (b) the profile weakens the motion from the outside.

On the other hand, a diagonally cutopening will ‘resist’ motion from outside (Fig. 366, b). The narrowing of the hole itself shows that the wall is about to close. It is given added weight and substance as well, because the diagonal reveal conveys an impression of greater thickness than the wall actually possesses.89

In the diagonally cut opening the hole itself appears to lie deeper in the wall than does the right angle cut opening. It is less accessible and is protected within the wall itself. In some cases these qualities serve to differentiate between door and window. In Sudanese folk architecture windows are not only small but plastically rounded and deeply inset. The door, however, is often cut straight into the wall. Thus the door demonstrates its association with motion from the outside inwards while the window belongs to the protected interior.

The Form

We repeat that the window is immediately perceived as an expression of interior expansion towards the outside. We saw from the above that when the window is in the form of a naked hole, that impression was weakened, because the interior was dead and the exterior alive. Further distinctions resulted from the treatment given to the reveals. When we consider the importance of the form as interpreter of motion from the inside outwards, we find mainly two determining conditions.

One condition is determined by the form’s own expression of motion. By this is meant the various impacts revealed by the vertical, horizontal and centralized window (Fig. 367).


Window forms and the relationship between inside and outside.

The other concerns the effect of these forms on the wall’s own expression of motion, depending on their location and size. In combination they can strengthen or weaken one another.

The Form’s Own Expression of Motion

In the wall face we saw that the vertical form initiated arresting and inviting movements (see p. 145). The upright window will accentuate this motion coming from the inside and thus strengthen contact with exterior space. There are two main reasons for this. First of all, the form is that of a person standing and looking out. The other reason may be explained by the way the form conveys the possibility that one may enter. Considered in this way the vertical window is related to the door. This type of window, therefore, is frequently interpreted as a door partly filled in with panels, balustrades or wrought-iron railings (Fig. 368). This motif also appears in interiors. In the period of Historicism, apartment house windows were ‘opened’ right down to the floor with deep panel areas. The entire form is associated with French windows which, as a combination of something to both look through and walk through convey a direct and strong feeling of the interdependence of inside and outside (Fig. 369).


The vertical window and the open context between inside and outside (from an apartment building in Oslo).


The varying relationship between inside and outside when the windows have a dado and when the windows do not have a dado.


The horizontal window and the dividing expression between inside and outside (from southern Harildstad in Gudbandsdalen, Norway).

We pointed out above that the horizontal wall encourages lateral motion (see p. 143). As a window, the horizontal or oblong form will suggest a motion that cuts across the inside to outside contact. It is as if the people inside do not concern us directly; they seem to pass in front of us and past us (Fig. 379). Stopping is an exception defied by the form.

A wall with equal height and width, we maintained, conveys a neutral and reserved motion impulse (seep. 145). Asa window opening the impression becomes the exact opposite, because the centalized form conveys the feeling of a hole caused by a direct and penetrating motion from the interior. For example, a round window not only resembles an eye but by its form is a direct rendering of concentrated ‘peering out’ (Fig. 371).90


The centralized window and the directed expression (detail from house in Oslo by E. Collett).

Le Corbusier’s la Tourette (1959) is an example of a building in which all the three window forms mentioned are used (Fig. 372). From the interior they lead the eye towards the landscape in three different ways according to the type of space from which one looks out. The horizontal window indicates the corridor behind. As narrow bands along the surface they outline the boundaries around the building’s interior (Fig. 373). By way of contrast, the windows of the study rooms are small centralized peepholes, while the openings into the common rooms are larger sheets of glass in which verticality is accentuated by the vertical mullions. For the monks inside, the study windows frame a concentrated section of the landscape, the common room windows allow a total impression of the unity of heaven and earth, while the corridors carry the eye horizontally along the silhouette of the landscape.


The vertical and the centralized window (La Tourette, in Eveux, France by Le Corbusier).


The horizontal window (from La Tourette, in Eveux, France, by Le Corbusier).

The Form and The Wall’s Expression of Motion

From what has already been said, we know that the reason for the impression of motion conveyed by the form is to a great extent determined by the directionalities of the form itself. In this way, the verticality oriented window accentuates verticality, the horizontally oriented window stresses horizontally, while the centralized window emphasizes the point and thereby a right-angled motion in depth. Wherever the window is placed, these directional factors will influence the relation between window and wall. The design of windows, however, is also important for the way in which motion in the wall itself is to be perceived. ‘These holes are often the destruction of form, they must be made an accentuation of form’, says Le Corbusier (Fig. 374, a-d).91 If we consider the influence of the vertical window on the wall in which it is situated, we realize the importance of whether the window finishes in a pointed arch, round arch or a straight lintel.


Window form and the relationship to the wall: (a) pointed form tears the wall, (b) circular form causes both the rising and sinking of the wall, (c) the straight form can have a neutral effect, or (d) cause the sinking of the wall.

A pointed arch ‘slashes’ the wall above it, because the vertical line does not finish within the form but continues past it. The round arch supports weight from above and leads it equally to each side. In that sense the round arch indicates a more composite motion than the pointed arch. It expresses pressure from above in addition to a certain upward push from below. The semicircle, as part of a centralized form, emphasizes a horizontal motion through the wall from within. The pointed arch, on the other hand, exaggerates upward motion, thereby subduing motion through the wall and contact between inside and outside. In the Gothic cathedral there is a purpose in this. The entire interior is isolated and made more sacred; it is cut off from horizontal communication despite the skeletal wall and translucent glass. The symbolism of the vertical has determined the window form and is meant to create an eloquent image of the church for the outside world. If the window if finished off with a straight lintel, it means that its expression of motion is determined by width in relation to height. If the width is relatively great, motion through the wall is reduced, because the accent is on crosswise motion. This also applies when the width is so great that the

Interplay of wall and window will sometimes cause wall motion and window motion to accent one another, while at other times it leads to a contrast between them (Figs. 375, 376).


Window form and the relationship to the wall: the window and the wall system follow each other (from Palazzo Valmarana by Palladio).


Window form and the relationship to the wall: the window and the wall system are separate (from an apartment building in Oslo, photo by T. lange).

An example of the first is to be found in the vertical window, which accents a tall wall as seen in skeletal walls of the Gothic cathedral. Horizontal windows, in just the same way, can extend and add weight to the horizontal wall. This added weight is exploited by Michelangelo in the attic storey of St. Peter’s. The horizontal windows cause this storey to sit heavily on the tall pilasters below (Fig. 377, a). An extended, stretched-out look is also typical of Functionalism’s bands of windows, which give a horizontal accent to the motion of the entire wall.


Window form and the relationship to the wall: (a) contrast between the wall’s vertical and horizontal motions (diagram of St. Peter’s wall system), and (b) accentuation of the entrance and the motions of the interior space (diagram of the rose window in the medieval church).

An example of a contrast between window form and wall is the Gothic rose window, which suggests motion not within the wall but through it. All verticality in the faҫade design stands suddenly in sharp contrast to the rose window’s horizontal impulse. This latter is in keeping with the motion of entry below and creates a closer and more direct rapport with the length of the nave (Fig. 377, b).

The window’s location also affects the wall’s expression of weight. A horizontal window placed low in the wall increases the sinking effect and a vertical window high up increases the rising effect, while a centralized window conveys ambiguity (Fig. 378). A window form’s own particular motion expression may be lost, depending on where the window is placed in the wall. The horizontal window, as we have said, indicates a lateral motion crossing the outward impulse. People behind the wall are restrained by it — they follow alongside the wall. This effect, however, is reversed if the window extends down to the floor. Then it is the floor which becomes the directing element and in a continuous process leads the inside out and outside in.


Window form and the importance of the location in the wall.

The Face

The window’s form and profile showed the relation of the opening to the surrounding wall, whereas the role played by the window face mainly depends upon its relation to the opening in which it is located.

The wall face and window face are seen as the boundaries of two different spaces (Fig. 379, a). The wall face is the boundary towards the exterior and is perceived as the outer shell of the house. The window face, on the other hand, is a boundary relating to the interior, because through the window we glimpse the interior’s own life, which is held in check by the membrane of the window.


The window face is located innermost in the opening: (a) wall and surface exist as two independent layers, in which the exerior space dominates the interior space. The face can be further separated with contrasting patterns or coloun (b, c).

This means that the four choices of location for the window face — inside or in the middle of the opening, outermost in or outside the opening — convey highly different impressions of relative interior and exterior strength. In the following descriptions, wall thickness will be held constant.

The Face Inside The Opening

With a window face inside the opening the wall face and window face will be perceived as parts of two independent layers extending freely alongside one another (Fig. 379, a).

This phenomenon corresponds to what we have said about the hole (see p. 259). Through the opening we look into an inner detached world, which might be compared to watching swimmers glide past us when looking through a side window in a pool. The motif is interpreted in many ways. Interior curtains and shutters show the independence of the window face to the hole in the same way as when mullions and transoms create a crisscross pattern in the hole’s own form. (Fig. 379, a-c). The window face gives the effect of being part of an independent membrane around a separate interior behind the stiff wall. In addition, when the window face is sunk deeply in the wall, the depth of the wall is revealed in the opening itself. Thus, it seems that the hole is caused not by the interior but by a force from the outside which cuts through the wall to reveal an inside world bounded by the window face (Fig. 380).


The window face is located innermost in the opening: wall and face are separated — the wall becomes an outer skin (from Tribune Review Building in Pennsylvania by L.I. Kahn, from Giurgola/Mehta, Louis I. Kahn).

The Face In The Middle of The Opening

When the window face is placed in the middle of the opening, the effect is one of inside and outside meeting at the halfway mark. The window face seems to be pushed out from the interior but is halted half way by a counter movement from the exterior, all of which can be read in the opening’s depth (Fig. 381).


The window face is located in the middle of the opening: the exterior and interior are in balance.

To experience this, the face must be transparent and so reveal both the front and rear part of the embrasure. Or, that an accentuated frame around the face shows its position within the hole. Here again, the relation between the form of the hole and the way in which the face is located will determine whether an outward or invard motion ‘prevails’.

The Face Outermost In The Opening

In placing the window face outermost in the opening the interior space appears to extend right out to the wall face (Fig. 382).


The window face is located outermost in the opening: The interior is led out.

This can result in two effects. In the first, the entire volume of the building seems ‘overladen’. This is because the outer wall surface and the inner substance meet in a common face conveying a feeling of interior space trying to break out. The effect is aimed at in neo-Classicism (Fig. 383). According to its tenets a building should appear as a combination of stereometric masses. Herein lies the reason for large hipped roofs with no projection, plain wall faces, and Doric articulation. Windows are kept small and quadratic, usually flush with the wall. The location of the window face supports the architectural intention and gives the effect of a solid and ‘inflated’ building mass.


The window face is located outermost and the door is located innermost in the masonry wall (neo-classical house in Oslo).

The other effect of placing the window face outermost in the opening gives the wall, especially when the windows are relatively large, the impression of being a thin ‘skin’. The main reason for this is that wall and window face are seen as one and the same. Thus, it is not without reason that Functionalists placed glass windows outermost in the opening. The walls should appear as thin planes, a sort of non-supporting cloak wrapped around an open and expansive interior (Fig. 384).


The window face is located outer-most in the opening: the wall and surface are flush — the wall becomes ‘thin’ (from an apartment building in Oslo).

The Face Outside The Opening

The final variation, where the window face is fastened to the wall outside the aperture, gives the impression of a face belonging to the exterior space. In Sigurd Lewerentz’s brick architecture from the 1950s, glass faces cover the openings like ‘bandages’ and seem to hold back the interior space (Fig. 385). The wall flows freely behind them and thus seems in a way to have regained both individuality and strength (Fig. 386).


The window face is located in front of the opening: wall and surface are separated — the wall becomes ‘powerful’ (from S. Lewerentz’s ‘Klippan’ in Helsinki).


The window face is located in front of the opening: the exterior restricts the interior.

From what has been said hitherto we realize that variation in window face location can be a factor in the wall’s character as a whole and in the relation between inside and outside.

A neo-Classical faҫade from the 1920’s in Oslo is organized according to the vertical tripartition (see p. 119) (Fig. 387). The central area is divided into two storeys with small quadratic windows, their faces outermost in the opening wall. Doors have been placed in both corner sections. The French windows in the upper storey give onto narrow balconies, which rest on the frames surrounding the doors on the ground floor. All the doors are set deeply in the doorways.


The window face is located outermost while the door is located innermost in the masonry wall (neo-classical house in Oslo, by N. Beer)

The effect is immediate. The treatment of the centre area conveys an urge to expand. The interior seems to strain forward from within because of the windows’ centralized forms and the flush position of the glass faces. The corner sections, on the other hand, seem massive and heavy. An important reason for this is the deep doorways, which reveal the thickness of the wall. Another reason is the location of the doors themselves, which emphasizes the difference between door and window. In line with the outer wall face the window leads motion outward from within, while the deep-set doors reveal their purpose of leading us in from outside (Fig. 388). In framing the ground floor doors, two contrary motions are emphasized at the same time. The deep-set door conveys a penetrating motion, the frame conveys an emanating motion. This last element brings us to our next theme, which is the frame’s role in the way we experience the role of the window for the relation between inside and outside.


Pattern of motion in the facade of a neo-classical building in Oslo, see Fig. 387.

The Frame

A frame surrounds and creates the setting for a window face, just as four walls surround and create a room. This implies that the frame accents and emphasizes whatever it surrounds in relation to its environment.

The frame also draws a boundary between the window face and the walls around, which means that it isolates and protects what it surrounds. Thus, every frame has a double effect in both accenting and separating a motif (Fig. 389).


The window frame accentuates the opening and separates it from the wall (from Enerhaugen in Oslo, Norwegian Folk Museum).

This again leads us to perceive the framed motif as being more important than its surroundings. The frame, therefore, is used to accentuate important objects, people and situations. When one receives a gift of a drawing or a painting, it is a sign of appreciation to promise to ‘frame it’. A painting, according to Alberti, is an excerpt of a symbolic reality and like the window it frames another and more elevated world. In his well-known balcony painting, Edouard Manet concentrates and emphasizes a group of people by using the edges of the door to frame the figures (Fig. 390).


The window frame accentuates the framed motif (The Balcony’, painting by E. Manet, from Janson & Janson, Maleriets historie).

For us this means that a frame will accentuate what we see when looking in through a window. The frame, on other words, increases the importance of the inside space and in a way brings it towards the spectator on the outside.

This is clearly demonstrated when similar windows in the same wall are framed or unframed. In our example, the upper storey windows and the door beneath are merely unframed holes (Fig. 391). They show that the outside is forcing its way in, while the framed windows on either side accent an inside being led out.


A farm house with and without frames around the openings (after Venturi, Complexity and Contradiction in Architecture).

The Frame Parts and Their Expression of Motion

The expression of motion depends upon the actual composition of the frame.

From what has already been said we know that in relation to the opening, the frame’s location is essential for the interpretation. We have pointed out the three basic motifs which can be distinguished by this relationship (see p. 255 f). Each gives the frame a different degree of importance in guiding motion, depending upon whether it is placed within or outside the opening. If, however, its location is outside the opening, the frame’s variations are a function of a separate space in front of the wall. It is not until it is within the wall that the frame is determined by the interior space and so becomes its mediator to the outside (Fig. 392, a, b).


The motion expression of the frame components: (a) the frame in front of the wall, (b) the frame within the wall.

In the following, therefore, we shall limit ourselves to discussing the variations within this last motif and study examples in which the frame appears as a complete form or only as a head or sill. Each of these variations will result in essentially different ways in which the interior is directed outward. These differences are based on our perception of the frame as marking the boundary around a space. This shows that the interior by means of its frame communicates with the outside in different directions, because the frame elements describe ceiling, floor and walls (Fig. 393, a-c).


The motion expression of the frame components: (a) complete frame, (b) lintel, (c) sill.

The complete frame expresses a straight forward motion. The entire space, with ceiling, walls and floor, seems to reach towards the outside.

Windows in which the accent is on the head lead the ceiling out and up — the motion becomes vertical and rising.

If the window has only a sill, it is as though the interior space is being ‘emptied out’ and down whereby a sinking motion prevails.

There are nuances in these effects depending upon the location of the window in the wall. A completely framed ground floor window indicates direct communication between the spectator and the interior. If, however, the same window is in a higher storey, it is as if the frame accents a communication above the head of the spectator and this no longer concerns him. This feeling of distance and lack of association is increased if the window has an accentuated head but will decrease when the sill is emphasized. In the latter case it is rather that the interior is on its way down to meet the spectator (Figs. 394, 395).


The frame motifs and motion expression in relationship to the location in the wall.


A city house with the sill as the dominating motif (by F. Fumess, Philadelphia, Pennsylvania, from Venturi, Complexity and Contradiction in Architecture).

Frame Parts and The Wall

We realize that these variations may affect the expression of motion in the wall itself.

Let us imagine three identical windows above one another in a wall. These can alter the entire rising and sinking effect of the wall according to how the openings are framed. If all the windows have heads only, in principle the wall conveys a rising effect (Fig. 395, b), with sills only the effect is one of sinking (Fig. 396, c), while the wall with complete frames on all windows will be practically motionless (Fig. 396, d). If these variations are combined in different ways, the wall flight can be made to ‘pulsate’. Let us picture a complete frame on the lowest window and this window surmounted by two other windows with heads only. The overall effect will be one of a rising from the second window (Fig. 396, e). If the two upper windows are given sills instead, the motion is reversed. They now sink towards the complete frame at the bottom (Fig. 396, f). Similarly, a wall with a complete frame on the lowest window, one with a sill only at the top, and one with only a head in the middle will convey a rhythm alternating between immobility, rising and sinking. This is a more ‘traditional’ interpretation of the wall in the way it responds to the roof’s pressure and to its firm attachment to the ground (Fig. 396, g).


The frame and its motion effect on the wall’s expression: (a) windows without frames (neutral), (b) windows with lintels (rising), (c) windows with sills (sinking), (d) windows with complete frames (expanding),


windows with dissimilar frame motif (pulsating).

Again it is important to emphasize that the effects described may be given different interpretations and nuances depending on the articulation of the frame forms. Some examples follow. In a window with slanted sills the sinking effect is accented (Fig. 395). Correspondingly, we see that the rich crowning pediments of Classical architecture were the result of wanting to give the windows a rising effect. The most usual pediment types are the triangular and segmental, which each in its own way has a different rising effect. The triangular pediment continues the upward rise, while the segmental closes and rounds off the form. These qualities were used with intent and purpose in facade composition in order to provide changing rhythm both horizontally and verticality. An early example is Antonio da Sangallo’s and Michelangelo’s Palazzo Farnese (1534—1548) (Fig. 397). Within this strict block form, the different window pediments make the otherwise similar storeys rise stage by stage up towards the heavy cornice. On the ground floor the pediments are straight and give a flat, pressed down effect in keeping with the massive character of this part of the building. The next storey conveys a rising effect, but with alternating segmental and triangular pediments demonstrating in a way its indeterminate position between top and bottom. The entire top storey, on the other hand, rises upwards in a row of sharp triangular pediments in which even the bases of the triangles are broken by the arch of the window (Fig. 398).


The form of the lintels and the wall’s expression: alternation between straight pediments on the lower level, arched and triangular pediments on the middle level and triangular pediments on the upper level (Palazzo Farnese in Rome by G. da Sangallo and Michelangelo, from Linn, Storgardskvarteret).


The form of the pediments and the wall’s motion expression: diagram of the frame system at Palazzo Farnese.

The Complexity of The Frame

Finally we shall study an example illustrating the complex changes which occur in the inside/outside relation when several frames are combined and when they are placed both within and on the outside of the actual window opening.

Our example shows part of a window-wall section in a three-storey apartment house in Oslo from the 1880’s (Fig. 399). The windows are equal in size but framed differently. The lowest window (not shown) is set heavy wall, the two above are placed between high colossal pilasters. The colossal pilasters belong to the outside space, emphasizing the main motion in the exterior building mass and linking the roof and the ground floor. At the same time they are a part of the expression of the window opening, in that the pilasters seem to ‘draw aside’ to make way for the opening between them (Fig. 400). There is motion from the outside in the top window, because it has no frame and is stepped inwards into the body of the wall. The window beneath, on the other hand, opens up from the inside, because the head guides the interior out and up by means of its powerful lintel. In contrast to the outward and inward motions above, the deep-set ground floor window announces another and more withdrawn space, well protected by a sturdy wall.


Complexity of the frame: various frames combined in the same wall (from an apartment building in Oslo).


The form of the frame motifs and the wall’s motion expression: diagram from an apartment building in Oslo, see Fig. 399.

The composition above is taken from the Baroque and was used during that period to symbolize the building’s function. The public nature of the great pilasters indicated the owner’s noble status, whereas the different window frames were meant to show the type of room lying within. At the lowest level were shops and storerooms which, being at street level, required protection. The first floor, ‘piano nobile’, contained the reception and entertainment rooms with an extroverted and public reference. On the third floor were the more private bedrooms and servants’ quarters.

The composition of the facade was also a logical solution to a location facing a narrow, dark street. The pilasters ‘open’ the street space to either side, giving it more optical breadth while at the same time guiding the eye up towards the light (see p. 159). The latter also explains the form of the windows. The darkest areas at street level are repeated in the closed ground floor and the gradual lightening upwards corresponds to the rising effect of the middle storey, while the strongest light is ‘received’ by the top storey windows.

The Bay Window

We conclude by examining the importance of the bay window in our experience of the inside-outside relationship.

A bay window is to be understood as an expanding window. Its purpose is to appropriate exterior space both optically and in terms of light. Thus, the expression of the bay window is a question of whether the wall resists or gives way to this expansion from within.

The first bay window motif describes how the wall resists spatial expansion. The bay window pushes against the wall from within. The wall is bent by an inner power, which it controls and holds back, an expression giving the wall a character of both strength and flexibility (Fig. 364, a). The bay window and wall are in fact the same thing, and their encounter takes place within the body of the wall itself. The motif, therefore, is often found in buildings where the main emphasis is on the plasticity of the walls themselves. This is found in the Amsterdam school, in which architectural expression and the expression of mass are one and the same, so bay windows are formed as undulating membranes around expanding interior space (Fig. 401).92


Bay window as a bulge in the wall (from a factory at Toyen in Oslo).

In the next bay window motif connection between interior space and the wall is broken. The conflict between expansion and control is revealed by the way the bay window seems to burst through the wall opening. Inside space and outside space now have separate surfaces, each with its own purpose. The exterior space attempts to fill in the opening, while the bay window tries to open it (Fig. 364, b). A bay window of this type reveals the conflict in the architecture of the wall: a struggle between immobility and motion, pro et contra. In architecture this conflict is usually understood by exploiting contrasts in the composition. This may be found between curved and straight forms, as in cases where a straight wall directly encounters a curved bay window (Fig. 402). It also occurs in building systems where the wall is massive and the bay window skeletal (Fig. 403) and also between volumes where the bay window is slanted in relation to a straight wall.


Bay window extending from a split in the wall (from an apartment building in Oslo).


Bay window extending from a split in the wall (from an apartment building in Oslo).

In the third motif we see how the bay window describes a separate space which both breaks into the wall and springs out of it (Fig. 364, c). Whereas the first two motifs demonstrated that the bay window presses from the inside, this bay window overlaps both from within and without. We shall look at two examples. In ‘Sea Ranch’ designed by Charles Moore et al. (1965), a corner bay window projects from one of the wooden houses (Fig. 404). With its corners and slanted roof this bay window must be seen as an independent enclosure which both juts out of the interior space and breaks into it by taking a large piece of the corner. The other example is from the Maritime Museum in Oslo, designed by Trond Eliassen and Birger Lambertz-Nilssen (1974). A deep gash has been made in the corner facing the fjord and deep within it is a slanting bay window which projects from one of the sides. The bay window penetrates into the wall and projects from it. Both these examples derive their meaning from their seaside locations. The bay windows are a part of these windswept outposts in which the protruding part of the form is an actualization of defiance of the elements, while at the same time the retracted part shows the need for protection behind the wall itself.


Bay window overlapping the wall (from Sea Ranch in California by Ch. Moore et al. from Futagawa (ed.), MLTW).

In the fourth bay window type the wall passes freely behind the bay (Fig. 364, d). The bay window appears to be a box hung on the wall from the outside. Expansion from the interior is caught and imprisoned by an exterior enclosure fastened to the wall, which may be compared to Charles Moore’s ‘saddlebags’.

The wall in the first motif resembles a flexible membrane; in the second it gives the impression of being a contradictory narrowing element, and in the third of being an open element with neither wall nor bay dominating. In this last case, however, the wall is totally independent of the bay window. Therein lies the possibility of interpreting the wall as the superior element, as something stiff and strong in contrast to its lighter ‘appendage’ (Fig. 405). This point is more clearly expressed in the great mountain monasteries in Athos, Greece. These massive megalithic structures crown the island’s craggy landscape like unassailable bastions. The characteristic covered galleries stretch along the top of the wall just like open skeletal windows. These open passages and massive walls complement one another; the very massiveness of the walls is accentuated by the lightness of the galleries and their openness is in turn emphasized by the closed walls.


Bay window as a room attached to the wall (19th century buildings in Amsterdam).

The Entrance

The Entrance and Inside-Outside Relationship

The entrance is a thing through which one passes and belongs to the space outside (see p. 251 ff).

To go in is to experience entering, and in this lies an existential description of the transition itself — the distance between qualitatively different places — between inside and outside. It is by entering that one succombs both physically and mentally and ‘occupies’ the architecture with all its fundamental meanings. The symbolic value of entering and of the entrance is revealed in both the rituals and behaviour of most cultures. To carry the bride over the threshold, to bid welcome and farewell at the entrance is still customary in western European countries (Fig. 406). A glance at recently built residential areas shows the importance still attached to the entrance, even in a ritual-free society. Lamps, flower boxes, the old wagon wheel and ornate stone work accentuate broad, hardwood doors, all meant to give the residents individual identity. Formerly the entrance could illustrate the concept of eternity itself. ‘The Lord shall preserve thy going out and thy coming in from this time forth, and even for evermore’.93 A blood sacrifice was placed beneath the threshold to protect the interior, and the door casing was smeared with blood so that ‘the angel of wrath’ would pass by that house (Fig. 407).94


The welcome at the entrance (‘Velkommen’, painting by E. Soot).


Entrance surrounded by painted windows and wreaths (Sudanese folk architecture).

By comparing the most diverse buildings throughout architectural history we find certain specific motifs which through constant recurrence accentuate the very action of entering. They appear with varying degrees of strength, alone or together, and are found in both monumental and folk architecture.

These motifs concern two elements in the entrance design. The first is the opening itself and includes its form, profile, and the door. The other element is the surrounding frame and concerns the space around the doorway.

Doorway Motifs

The form of the doorway follows the same principles as those of the window. It differs, however, because the window may be allowed a freer form. The door depends more upon the human figure and usually appears as an upright rectangle. Just as for the window the head will vary, according to style and purpose, between straight, curved or pointed, and the reveals too will alternate between straight and diagonal.

The door face or door itself may vary in construction, material, and location to the same degree as the window face (Figs. 408, 409). Of particular importance, however, is the way the door opens (Fig. 410, a-g). A door can be opened in four ways: away from us, towards us, around us (revolving door) and sideways (sliding door). It may be either a single or a double door (two leaves). Likewise, the single door may be hinged on the right or the left but also at the top or bottom. Two familiar variations of the latter are the portcullis, flush with the wall and hinged at the top, and the drawbridge, hinged at the bottom and lowered from above.


Wrought iron gate (entry by M.L. Leray, from Battersby, Art Nouveau).


Solid wood door (apartment building entry door from Venice).


The doorway motifs: (a) away, (b) towards, (c) sliding, (d) double sliding, (e) double swinging, (f) lowered, (g) drawbridge.

Door-Casing Motifs

The motifs used as casings around the door-opening are all variations on the wall’s own archetypes. It is customary to take into account eight such entrance motifs, for which we shall use the following terms (Fig. 411, a-h). First is the frame or portal motif. This motif constitutes the primary form in the wall’s skeletal system (see p. 193). It runs the gamut from naive ornamentation around doors found in rural and folk architecture (Fig. 412) to mighty column and cornice combinations in Baroque church facades (Fig. 413). The next motif, the split or twin-tower motif, may stretch from simple gateposts leading into the gardens of private homes (Fig. 414) to monumental Egyptian temple pylons and tall west-front towers in Gothic cathedrals (Fig. 415). The third motif, the deep-set or niche motif, (Fig. 416) occurs on a smaller scale, as when the door sits deeply in the opening itself (Fig. 417) and on a larger scale when projecting side wings create deep courtyards in front of the opening (Fig. 418). The convex motif makes the opposite impression to the niche motif and occurs when the facade springs forth either as an undulation or a risalit. The dominating feature in the shelter motif is its varying roof forms but it may also take the form of tall towers. The latter is typical for entrances at the foot of church and palace towers. We meet the first type in both column-supported anterooms and suspended baldachins (Figs. 419, 420). The directional wall motif has the same effect as the horizontal wall form (see p. 143) and appears as low half-walls leading to the entrance in Frank Lloyd Wright’s houses or as high stone slabs leading straight into Mies van der Rohe’s pioneer houses (Fig. 421). The side tower motif is found especially within Romanesque architectural tradition and takes the form of a tower on one side of the entrance (Fig. 422). The last motifs, the path and stair motif have also been examined earlier in connection with floor architecture (see p. 89 ff) (Fig. 423). As shown, it is the stair motif which permits particularly rich variations ranging from the plain stoep in front of Dutch bourgeoisie houses to broad series of stairs in the Baroque style (Figs. 424, 425).


Door casing motifs: (a) frame, (b) split, (c) niche, (d) shelter, (e) directional wall, (f) side tower, (g) path, (h) stair.


Frame motif (from a farm in Gudbrandsdalen in Norway).


Frame motif (St. Maria in Campitelli in Rome by C. Rainaldi, from Sedlmayr, Epochen und Werke).


Split motif (double-tower motif) (garden entrance to Villa Borghese in Rome).


Split motif (double-tower motif) (Egyptian pylons in front of the temple at Edfu).


Niche motif (entrance facing street intersection in Nancy, France).


Niche motif (entrance facing Damstredet in Oslo).


Niche motif (terrace between two wings into the Cultural Centre in Risør, Norway by E. Anker & A. Hølaas).


Shelter motif (covered entrance in front of Bergamo Cathedral).


Shelter motif (canopy in front of Stazione Termini in Rome by L. Calini and E. Montuori).


Directional wall motif (from the Barcelona Pavillion by Mies van der Rohe, from Johnson, Mies von der Rohe).


Side-tower motif (project from the 1850s by A.J. Downing, from Downing, The Architecture of Country Houses).


Street motif (pavement in Bergamo).


Stair motif (stair up to St. Marino in Montemagno).


Stair motif (from an apartment building in Oslo).

The Expression of The Entrance

All the motifs that have been mentioned can be traced to both practical and technical considerations. Thus we see that the size of the door is often determined by the number of people to pass through (see difference between main entrance and auxiliary entrance) while the form may be the result of what is to be transported through the door (Fig. 426). Similarly, the frame or casing is the result of the need for added support and stiffening around the opening itself, while the baldachin solves the problem of protection both for those entering and for the interior space from sun and rain, while stairs are necessary to allow movement from one level to another.


Form of the door opening: form adopted to mule with packs (photo by Norberg-Schulz).

Historical usage shows that they have an inherent value in themselves which far exceeds the purely practical. The reason is that the motifs reveal basically different ways of entering.

Concerning the form of the door opening, we have in connection with the window, drawn attention to the fact that breadth, height, and arch form take part in interpreting ‘the speed’ of the entering motion (see p. 259 ff). The door itself and the direction in which it swings contribute in turn to the degree of strength experienced in this motion relation between inside and outside. The phenomenon may be explained thus: the door swinging outwards brings the interior space out towards us, the inward swinging door leads the outside space inside from us, and the sliding door opens between us and the interior space. In the same manner the portcullis and the drawbridge bring the interior space down on top of us.

The casings around doors may be described in a similar manner. The frame motif conveys the feeling of entering through something, the split motif of passing between something, the deep-set motif of entering into something, the baldachin motif of passing under something, the directional motif of entering by following alongside something, the side-tower motif of entering by passing by something, the path and stair motifs, respectively, of entering on something and entering up or down.

Thus, different types of motion can be compared to prepositions in language, words that describe our physical orientation in relation to concrete objects in our surroundings. Regardless of cultural background, all of us have experienced different types of movement. They are universal and as a result immediately recognizable. This means that consciously or unconsciously we transfer our experiences to the motif we see. In other words, we enter’ mentally before doing so physically (see also p. 29). The motifs, therefore, have been important in the preparation for entry, a process that may begin long before one reaches the entrance itself. For just this reason several motifs in combination are frequently used in the same faҫade in order to interpret the various stages on the way towards the goal. The Gothic cathedral may serve as an example. From a distance its soaring twin towers announce the entrance to the church. Upon arrival at the square in front of the building the pointed arch around the door ‘takes command’, and finally it is the arched niche which concludes the outside space.

We have shown earlier that certain types of motion, embodied by certain motifs arouse in us corresponding sensations (see p. 29 ff). Buildings throughout acrhitectural history appear to show that this expressive content has been understood and exploited.


‘Build high the door and broad the gate. The king of honour cometh here’.95 The quotation shows that not only the form of the door opening but also its relative size conveys a definite impression of the relation between inside and outside (Fig. 427). A large door opening is both generous and public; generous because it lays bare the interior and extends towards the person entering and public because the opening seems calculated for the passage of large crowds or socially important people. The height of the door is particularly important for the impression of size. If a door is low and one must bend to enter, a basically submissive movement is created. This is the expressive content of the low door. In Japan, entrances to tea-houses frequently measure no more than a square metre and are placed about 60 cm above the ground. One enters on one’s knees, bowing at the same time, an action intended to put the visitor in a state of humility. The same impression is conveyed by the lowslit through which one enters Emanuel Vigeland’s mausoleum in Oslo (1926). One bows one’s head, not just to the space within but equally to the artist himself, whose urn is immediately above the door.


The size of the door opening: the largest frame is adapted to the faҫade’s public image, while the smaller is adapted to private expression (cliff grave from Petra, from Springer, Handbuch der Kunstgeschichte).

The door itself may be seen as a movable part of the surrounding wall. In this sense the way it swings is an actualization of the relative dynamic strength of inside and outside. Thus, the outward-swinging door gives an impression of rejection; the interior space is shoved at us and makes us back away. Conversely, an inward-swinging door gives substance to our movement inwards; it lends added action and follows up this movement. This characteristic differs according to whether the door is hinged on the right or left. If it is hinged on the right, entry is a strong action in which the stronger hand opens the door and the interior space is actively shoved ‘to the side’ (Fig. 428 a,b). If it is hinged on the left, the active element is reduced because now one encounters the strong part of the space. If we imagine two identical spaces with the entrance on one wall, on the right and left respectively, it will seem more difficult to enter from the right than from the left. The reason is to be found in the spaces’ own elements of strength, which in the right-hand case is what one meets and which offer more opposition to entering than if one enters from the left (Fig. 429 a, b).


The door and its opening: (a) hinged on the left, (b), hinged on the right.


The door and its location: (a) on the left, (b) on the right.

The portcullis lowered from directly above conveys a threatening impression. Even more rejecting is the drawbridge, because it threatens the entire space in front of the opening in just the same way as the slanting wall (see p. 151 f). Medieval fortresses usually had both partcullis and drawbridge which made an effective rejection, both visually and in reality, doubly certain and quite in accord with the intention of these complexes.

Door Casings

We have pointed out that casing motifs emphasize the wall’s own themes. The same applies to the sensations which they arouse. In the following we shall look at some examples which illustrate this relationship.96

The Frame Motif

We have maintained earlier that the casing or frame was both the setting and an accentuation of whatever was within it. As an entrance motif this means that it accentuates the person about to enter. The frame/casing thus has an idealizing content. That this is fully understood is shown in the triumphal arches of Antiquity, which, detached from architectural connections, bear witness to emperor’s res gestae (see p. 231) (Fig. 430). The same may be said of the enormous portals in Baroque facades, which opening out towards the city to announce ‘the church triumphant’ (Fig. 413). All in all the casing and the portal are classic motifs in buildings that have their origin in socities holding clearly idealistic views of mankind.


Frame motif (Constantine’s triumphal arch in Rome).

The Twin-Tower Motif

The twin-tower motif, as shown earlier, conveys an impression of dignity and strength. This is because the motif is a visualization of an almost claustrophobic feeling which arises whenever we find ourselves between elements expressing power (Fig. 431). Vertically is such a powerful symbol, its structure indicates a strength which we immediately recognize with our own bodies. For this very reason the Phoenicians called the mountainous cliffs on either side of the Straits of Gibraltar, the Pillars of Hercules.97 Similarly, the columns on either side of the entrance to the Temple in Jerusalem were called ‘Jachin and Boas’, meaning stability and strength.98 In the split between the towering flanks man becomes ‘small’. Thus, the motif is much used in sacred architecture. Between the pylons of Egypt’s temples, stylized mountains joined together by the arc of the sun, man as a person is powerless, at the mercy of the gods (Fig. 432). The same is expressed in the towered west walls of medieval churches. Here the motif evokes humility beneath a heaven stretching above and between the two tower peaks.


Split motif (double-tower motif) (rear entry to Notre Dame du Haut by Le Corbusier).


Lowered doors in a medieval fortress tower (from Caboga, Die mittelalterliche Burg).

The Niche Motif

The niche motif conveys a feeling of being both received and embraced (seep. 149f). The basis of this quality is that the motif conveys the feeling that architecture yields before us. In other words it illustrates the intrusion of exterior space, which in turn means that interior space yields to the space before us. Basically, the motif conveys an intimate touch, which may explain its revival as an entrance motif in many mass-produced houses. The Baroque cour d’honneur has, therefore, been compared to a welcoming gesture. Lorenzo Bernini said that he recognized the arms of man in the wings of palaces, and Emil Langlet wrote that the wings of the Norwegian Parliament, like ‘outstretched arms’ were meant to welcome the ‘representatives of the people’ (Fig. 183).99

The Shelter Motif

The shelter motif has a protective air and conveys an impression of being the building’s offering gesture to or>e who enters. The shelter is a visualization of an anteroom giving a three-dimensional preparation to the act of entry (Fig. 433). This preparation may consists of creating hesitation before entering, a psychological ‘halt’ as in the anteroom of the Pantheon. But, the preparation may also convey the feeling of being led in, as beneath Le Corbusier’s baldachin over the entrance to Villa Stein (1927).


Shelter motif (the shelter is created by a bracing arch, from the Bergamo Cathedral).

The Directional Wall Motif

The directional wall motif leads us forward with a feeling of security but also of dependence. This is because the motif is based on asymmetry and the contrast between the vertical and the horizontal. One ‘keeps close to’ the vertical. This is emphasized in Mies van der Rohe’s brick house project (Fig. 434). The relationship between the large, open space and the wall leading straight to the entrance makes entering an almost forced action. Without the wall one is lost in an open space and feels drawn to follow its uncompromising and unbroken course right up to the entrance door.


Directional motif (Barcelona Pavilion by Mies van der Rohe, from Johnson, Mes van der Rohe).

The Side-Tower Motif

The side-tower motif is essentially the same as the directional wall motif, but despite this it differs in the way it dramatizes verticality and thus the feeling of security. In placing the entrance beside this accentuated point where the vertical line and the ground meet, the entrance motion is given a surprising and contrasting effect. This effect is one of the reasons the motif was so popular in the romantic period. The side-tower’s expressive content is exploited in modern architecture as well, as in Le Corbusier’s Notre Dame du Haut (Figs. 358, 435). A conflict in plastic form is created by the wall’s upward slant from the tower and the decline of the heavy roof towards it, but all of this is kept in balance by the tower. Thus, it is safe both to be close to the tower and to enter along side it.


Side-tower motif (Schroeder House in Utrecht by B.T. Rietweld).


The roof as experience (photo Aftenposten).

The Path and Stair Motif

Path and stair-motif. Earlier we have pointed out that the path-motif is a visualization of one’s own action radius. The path leads but at the same time expresses an independent action. By means of the path one can conquer the world, says Bollnow. In other words, the path is active and purposeful as well as goal orientated. We find this conquering quality accentuated in Egyptian temple axes which lead into the interior without interruption. Corresponding features are found in Baroque architecture and similarly in the ramp leading into Le Corbusier’s Carpenter Centre (1963) (Fig. 39).

Stairs are a path dramatized. Their diagonal direction suggests the same tension as does their function, which is to connect two different levels, up and down. Herein lies their content as well, because of the different value attached to up and down in our action radius. If stairs lead us upwards from below, we expect to reach something important. If, on the contrary, we are led downwards, our movements take on an air of ‘being tolerated’. For this reason stairs convey a feeling of both contact and distance. The stair-motif, therefore, is particularly emphasized in monumental architecture, an architecture aimed at responding to common interests and the elevation of common ideals (Fig. 424).