Norman Foster graduated from the School of Architecture and City Planning, part of the University of Manchester, in 1961. In his final year thesis, he critiqued public spaces and squares found in London, Bath, Greenwich, Oxford, San Gimignano, Siena and Verona

Norman Foster: The nature of enclosure in urban areas

Thesis from the past

For architects, their profession entails a lifetime of learning, pursuing evolving or changing skills and concerns. Each commission brings new matters to be researched and mastered. Students may look forward to their thesis and dissertation as marking the terminal climax to their studies. But once architects, they may look back on them as mere punctuation points or launchpads on a path of further learning. They are also reminders of the changing concerns of architecture and architectural education, giving useful perspective on their past and continuing development.

Considering the longevity of buildings, approaches to architectural design vary with surprising volatility. So, as with architecture generally, it is usually easy to guess the decade of student theses. Their changing approaches reflect and give insight into changes in society and technology, or in architectural fashions of form and theory. So, the theses and dissertations of architects of current renown give valuable perspective on their early interests and inspirations and reveal how, and how much, these may have changed and matured.

They also give insights into the precedents from which our present emerged, raising questions as to how pertinent our current concerns may prove? Will the quest for sustainability, for instance, lead to the pursuit of longer-term values and greater completeness and stability in design concerns so they date on a slower cycle?


The creation of enclosure is a basic human need, and architecture is the art of enclosing space, whether internally or externally. Space, although intangible, has an existence in its own right and is one of the raw materials of urban design. By urban we mean an environment in which natural surroundings have become dominated by man-made surroundings.


The conditions that bring buildings into a visual relationship may vary greatly – the sociological grouping of an Ashanti village, the humanist theories of Renaissance Italy, the political and aesthetic implications of Haussmann’s Paris and the commercial competition, site values and technological influences in New York, to mention but a few.


Since the 19th century, however, our traditional concepts of space have been shattered and we are in desperate need of formal criteria of urban space. There are a number of factors that have produced our present inheritance of overwhelming complexity and visual disorder. In an age of commuting and suburban strip development, our cities are growing larger every day as the process of urbanisation continues. In the past, a relatively small range of traditional building materials, less specialised human needs and a slower rate of growth achieved a unity and correctness, which is admired but seldom recreated in today’s terms. Large-scale developments when they did occur were based on theories related to a more leisurely way of life, with the accent on stationary or slow-moving observers. Today, large-scale projects are realised in a short space of time and take into account new forms, materials and techniques to satisfy an ever-increasing variety of human needs.


Combined with all this are our new rates of speed, both horizontally and vertically. A new scale is evolving and judging by our present man-made environment, some radical rethinking of design principles is needed, based not on the mystic intuition of the artist, but on a reasoned and analytical study of the problems involved.


One of the difficulties of analysing the concept of spatial sensation is that of imagining the sensation without actually being subjected to it at the time. There are so many factors involved that very little help is given by representation of spatial order in a different medium. It is possible to describe it in words with accompanying photographs and drawings, but this conveys an inadequate idea of space and, more important still, does not give one the experience of space. A scale model only adds to the confusion by introducing an element of a different scale and order. The musician has a score, the film editor has a ‘storyboard’ and the dancer has floor-pattern diagrams. The architect and designer, however, have only the means of producing two-dimensional patterns in space and a series of selective sketches or photographs.


These may convey enough information for an architectural appreciation, but are inadequate for the spatial representation of vision in motion. Although the perception of the visual world consists of a succession of separate images, the overall experience is a continuous and connected sequence. It should be noted that we do not need to be consciously scrutinising our spatial environment to be affected by it; similarly, we do not have to be consciously listening to music for its effect to be noticed. As Moholy-Nagy writes: ‘space experience is not a privilege of the gifted, but is a biological function. The biological bases of space experience are everyone’s endowment, just like the experience of colours or tones. By practice and suitable exercises, this capacity can be developed’.


Space may be said to be defined by the relationship of surfaces and objects. The surfaces are basically two-dimensional plane forms, limited in effect to the space they help establish, although in reality, they may be part of a larger object when seen in a different context. Objects may be thought of as three-dimensional convex forms existing in a larger space than that which they help to define, and relatively smaller in size than surfaces. Surfaces may be broadly classified in terms of their position with regard to the space they define. Thus, the side planes may be considered within the scope of this study to be the wall planes or facades of buildings defining the space (although they could be walls, massed vegetation, landforms or building masses). The three-dimensional objects, which have a definite modifying effect upon the spaces that contain them, are considered as sculpture and trees in the examples chosen for detailed study later.


The degree to which space gives a sense of enclosure depends on the relationship of the surfaces (and objects) and to what extent the views out of it are controlled. For example, enclosure is most complete in the cloister and weakest where buildings are separated by large gaps and bear little formal relationship to each other. A number of other important factors are involved such as the treatment of the surfaces themselves and the junctions of wall and floor planes. 


Norman Foster: The nature of enclosure in urban areas

This resultant space-body, which is the reverse of the plastic body of a building, is essentially one that is measurable, even if it is not subject to the same exact definitions as a room. Winfried Leonhardt interprets the famous urban space group at Nancy as a series of axially related space-bodies and the Piazza di Sant’Ignazio, Rome, as a single large space-body extended by two small cylindrical ones in adjacent corners.

Norman Foster: The nature of enclosure in urban areas

The scale of the space or its size measured in relation to the dimensions of the human figure is a basic factor and may qualify what sort of space experience, if any, results from a given relationship of surfaces and objects.


Frederick Gibberd, writing on civic spaces in Town Design, says, ‘No plastic space-body can be formed when the floor area is so great that the walls of the surrounding buildings bear no relationship to it. There is a practical limit on building height, but none on floor area, and we find that countless urban spaces fail because they are so large that the buildings appear to stand on the edge of the space. The walls and the floor of the space become disassociated and there is no sense of spatial enclosure.


‘Judging from the size of many urban spaces designed today, it is obvious that there is still confusion between the relative importance of floor area and spatial volume. A larger space can be less impressive than a small one; in fact, beyond a certain point, the larger the square the less impressive it is. It is only in plan that we can appreciate a large area; in actuality we look across the space at buildings, and when we do the greater the void around us, the less we feel the sense of spatial enclosure.


‘There is an excuse for making a civic space large when it is to be used for demonstrations of military might or political unity, although it is better to accept it as a two-dimensional floor rather than as a volume …’


Obviously, the ratio of the height of the enclosing elements to the dimensions of the ground plane is of vital importance. For example, take the simple case of a medieval cloister or collegiate square. If the enclosing elements are moved closer together or further apart, then the character of the square and consequently the spatial experience will vary considerably, depending on the relationship of the wall plane to floor plane. A point will be reached at the upper limit when the space as such breaks down and a sense of being enclosed is no longer experienced.


In this context, it is interesting to compare Piccadilly, Manchester (800 feet by 750 feet) with Albert Square, Manchester (220 feet by 440 feet). Compared with Albert Square, which is quite well-defined as a space, Piccadilly fails to evoke any response at all, although the height of the surrounding buildings is roughly the same.


According to Camillo Sitte, the minimum dimension of the square ought to be equal to the height of the principal building in it, and its maximum dimension ought not to exceed twice that height unless the form, the purpose and the design of the buildings will support greater dimensions. Werner Hegemann and Elbert Peets say that to see a building as a whole the observer ought to be separated from it by a distance equal to about twice its height, which means that it should be seen at an angle of 27 degrees. He also adds that if the observer wants to see more than just one building, a civic centre group for example, he should see it at an angle of about 18 degrees — a distance of about three times its height.


Sitte also claims that the average dimensions of the great squares of the old cities are 465 feet x 190 feet and reminds us that the Champs Élysées is 465 feet wide by way of comparison. At this distance, objects which are about half an inch in size may be seen under reasonable daylight conditions. 


This may mean that the maximum size of the squares was based on ‘face-to-face’ recognition for slow-moving or static traffic. Increased speed of the observer raises a new set of visual problems as the effect is to reduce the scale of the space. Conversely, increased haziness or diffusion of light will be another factor as it has been noted that ‘persons who are adapted to making ordinary judgements of distance in the city are usually misled by the extent of distances in the desert, mountains, on water, or from a plane. Generally, aerial distances are poorly estimated by such persons because they are unfamiliar with the visual cues present in the situation for space perception.


Ernö Goldfinger notes the effects of new rates of travel: ‘Two entirely different categories of experiencing the new spatial order will now appear. The one of persons moving up to, say, three miles an hour, ie, walking; the other of persons moving at a rate of 80 and more miles an hour. If we consider that a normal person can register about 12 separate images in a second, the fundamental differences between the two sensations become clear: solid screens relationship, etc.’ [sic]


The effect of varying the scale of a given relationship of surfaces and of objects should also be considered. For example, in the previous case of the collegiate square, what would be the effect of doubling the height of the enclosing elements and also doubling the distance between them. Gibberd in Town Design likens a civic space to the room inside a building, and remarks: ‘but far greater liberties can be taken with the former in departing from the rectangle than with the latter. In an indoor room, the eye almost instantaneously sweeps over the planes and the continuous lines of the cornice and skirting to make the shape at once apparent, but in the outdoor room the walls are so far apart, and the lines are so broken, that irregularities are hard to spot. Walking about in the civic space at Aylesbury, one finds it almost impossible to tell that the numerous juxtaposed planes only once form a right angle; but reduced to the scale of a room in a building, such a space would appear chaotic’.


This experience is borne out by traversing a number of irregular spaces and then later examining the plan shape. This suggests that an increase in scale has an effect of simplifying the form of the space. Philip Thiel finds the basis for this in gestalt psychology and quotes Rudolf Arnheim in Art and Visual Perception, who states: ‘any stimulus pattern tends to be seen in such a way that the resulting structure is as simple as the given conditions permit’. A large-scale relationship (of surfaces and objects establishing a given space) will require a greater amount of attentive energy for comprehension than the same relationship on a much smaller scale. Therefore, the scantier the perceptual data, the more strongly can the perceptual tendency towards simple form assert itself.


So far, the actual form of the space has not been taken into account. Goldfinger suggests three basic forms (1) Regular (Place des Vosges, Paris); (2) Irregular (San Marco, Venice); and (3) The combination (a Japanese homestead of the Nara period AD 710 to 784 combining a regularity of building elements with free landscaping).


Paul Zucker contends that certain basic types of square occur again and again with minor variations. He lists these (1) Closed square – space self-contained; (2) Dominated square – space directed; (3) Nuclear square – space formed around a centre; (4) Grouped squares – space units combined; (5) Amorphous square – space unlimited.


DG Thornley proposes two main systems of assembling buildings and spaces in urban areas. These are:(1) The ‘formal’ or ‘continuous perspective’ type (Circus – Brook St – Crescent complex at Bath or the place complex at Nancy) and (2) The ‘informal’ (Stonegate, York). The regularity, directness and clear-cut perspective effects of type (1) are contrasted against the variety and effects of overlapping planes in type (2).


Using a categorisation proposed by Prof Gyorgy Kepes of MIT, Theil draws the comparison between a structured coherent, geometrical shape and an open expanding, free shape. The suggested archetypes are the egg and the tree respectively.


Basically then, spatial form may be said to range between the irregular and the regular, or the formal and the informal, and each extreme would appear to have its own spatial characteristics. The overall quality of a given space, however, will be a result of the relative separate contributions of the defining surfaces and the modifying effects of any object within the space. When the number of variables such as scale, dominants, solid to void, rhythm, articulation, silhouette, colour, pattern, texture, etc etc are taken into account, it will be seen that there is an infinite number of variations between the two extremes. It will also be realised that the character or quality of a space, defined by a fixed relationship of surfaces, can be changed simply by altering the form quality of the surfaces themselves — in other words by manipulating these variables.


One of the objects of this study is to examine some of the more fundamental spatial forms of varying size, bearing in mind the factors already mentioned. The physical organisation of the space has been noted under a series of headings and the photographic commentary has been prepared separately to enable it to be used with the following descriptive passages. The virtues of observation and analysis cannot be overemphasised; it is only in the light of such experience that an adequate design vocabulary can be built up and a sensitivity to the subject developed. A brief visit to most New Towns will confirm the need for a deeper appreciation, or even an awareness, of the principles involved.

The Great Quadrangle — Christ Church, Oxford

The Great Quadrangle of Christ Church is square in shape and the view out is rigidly controlled in all directions. Of all the examples chosen, it is the most open and has the least degree of enclosure. Its proportion of wall plane height to floor plane dimension is 7 and the overall dimensions of the square are 280 feet x 291 feet. The spatial character has been described as a majestic openness and it certainly has an austere and restrained dignity; an almost monumental air.

As a space, it has several subtleties, despite what may appear at first to be an over-simple conception. Whether these are accidental, intentional or the result of alteration is unimportant for the purposes of this examination. The defining surfaces themselves are almost massively unbroken masonry, except for an inconsistent and at times random rhythm of two small-sized window openings. Running through all the minor irregularities is a superimposed arcade. This is the dominating rhythm and becomes a stepping-stone, as it were, taking the eye around the space in an even, measured progression. The articulation of this arcade is of a small enough scale not to break up the flatness of the defining planes from a long view, but is sufficiently well enough defined to give interest and modelling on closer approach.

The relative simplicity of the wall surfaces would appear to emphasise several effects. First, it lends importance to the main openings; because of the solidarity of the walls the entrance gateway exerts a greater power of attraction (an effect that has been noted at Verona). Second, it tends to focus attention on the space itself, or the vertical feature, depending on the position of the observer. And third, it combines with the insistent rhythm of the parapet to reduce the effect of what is quite a disruptive silhouette.

The silhouette is most important in a space of this size. The smaller the space, the less important the silhouette becomes as one’s awareness of it diminishes. In passing, the space-defining effect of the row of pinnacles to the abbey should be noted. Their rhythm forms a positive extension to the wall plane below.

Tom Tower is the major accent in this space and dominates it by sheer dimension. It has a finely modelled and interesting form that contrasts with the enclosing facades and invites the eye to pause and speculate on its shape. In a space of this scale, vertical features become increasingly important, because they can be grasped in their entirety from within the space and seen in their true relationship with the other elements of the composition. In relation to the dimensions of the quadrangle, Tom Tower would appear to be ideally proportioned. Apart from being the major element in a long view from the corners, one is also aware of it when moving around in the space. This of course is before it eventually loses its significance as a vertical feature and, on closer approach, its lower part behaves as a space-defining surface in the normal way.

In views towards the tower one is aware of the strong symmetrical organisation of the space. The same is less true on entering from the street, because the abbey in the corner and the plane of a spire beyond demand attention in the absence of a centralised feature. On approaching this corner a stronger feeling of enclosure becomes apparent, but once within its confines the effort of extra height is not noticeable. Its influence is confined to the approach from one side only (oblique entrance is rendered impossible by the lawn).

It is interesting to note the influence of the floor plane on the spatial character. The dominating impression is of a flat plane of grass. In actual fact, there is a stepped walk, with regular projections, about 18 feet wide and 3 feet above the general level. The grass itself is symmetrically subdivided by gravel paths with a central circular feature. Without doubt, an unbroken floor plane from wall to wall would increase the clarity of the spatial composition. Nevertheless, the present arrangement is satisfactory, but on a smaller scale, it would most probably be quite distracting.

This quadrangle is a basic, simple shape with no secondary spaces defined within it, the central feature is too insignificant to have any nuclear or dividing effects at all. The elements of contrast are the vertical features and the greater mass of the abbey. It is thought that if more interest had been created in the facades, or if the unity of the composition had been unbalanced by changing the treatment of one or more sides, then stronger and more powerful elements would have been necessary to restore the fundamental need of contrast and rescue the space from dullness and monotony.

Queen’s College quadrangle is formal, completely symmetrical and almost square on plan. As can be seen from the photographs, it is enclosed on three sides by buildings of equal height and on the fourth side by a lower arcade, surmounted by a cupola at its centre. There are two circulation routes: a main one along the path through the centre and a secondary one along the ‘outside’ of the space within the arcade. This arcade is a strong unifying element in the sense that it is common to three sides of the space. Its dominating rhythm is echoed above in the windows of the flanking buildings.

There is thus a unified back-cloth for what is really the ‘key’ building of the space – this is Hawksmoor’s Library and Chapel. Its rhythm, proportions and scale are quite different from the rest of the composition and give it a statelier, more dignified character. This effect is reflected in the character of the quadrangle, depending on the position of the observer, because in some respects this space has a dual personality.

The main axis from the street through to the second quadrangle relates to the two vertical features. These are the only verticals in what is a horizontally composed enclosure and their power to catch the eye is quite strong. From the street entrance, the view of the clock tower and pediment framed in the gateway is beautifully composed and has an almost picture-postcard quality.

Moving in this direction through the quadrangle, the impression is of a well-proportioned space with a dignified and stately character. It has, however, an intimacy quite different from the more open arrangement at Christ College. The wall planes are roughly the same height, but the overall dimensions of the space are 147 feet x 142 feet. The proportion of floor plane to wall plane at Queen’s is 3 compared with 7 at Christ College. It is not possible to make a direct comparison, because although they are both square, the surface treatments vary considerably. Nevertheless, dimensions and relationship of floor plane to wall plane are basic considerations and the surface treatment a secondary one. No amount of variation to the wall surfaces at Christ College  could simulate the same degree of enclosure as this example.

On entering the space, one’s first impression is of an immediate spatial expansion; this is typical of most squares in Oxford, more noticeably so in Christ College where the space is so much larger and the contrasts greater. This question of surprise is a most important factor because the main streets in Oxford are not small in the same way as the Venetian alleyways leading up to the Piazza San Marco (and the ‘explosive’ effect of space there is well known). One always has the feeling of climbing through a hole in the wall to gain access to the Quads, and in the process, the opposite wall or feature is framed with little or no modelling, appearing almost as a backdrop. This is the effect of distance before the dimensioning characteristics of perspective come into view.

After the first effects of spatial expansion, the vision is dominated by the chapel and hall, with the arcade on either side running up to it. Such an arcade would normally have a strong impelling quality, but this is lost to a certain degree by the dimensions of the space and its effect on the angle of vision. One would imagine that if the space were rectangular and the colonnades filled the long sides, then the visual pull towards the key building would be much stronger, because the angle of vision would have taken in the combined perspective effects and the key building as well. It could conceivably be argued that from this approach the space forms a setting for the chapel and hall.

Looking out from within, the arcade is most deceptive. There is no impression of spatial expansion, as the top of the buildings or sky cannot be seen; only the opposite facade is visible and it appears extraordinarily close. There are no clues to the distance unless one’s view takes in the low arcade to the street side. The size of the space is instantly grasped and the space sensed even if it is not experienced. From this oblique view, the cupola really is a three-dimensional piece of sculpture defining its own particular space.

Moving through the square in the opposite direction produces a completely different set of spatial experiences. The effect of the confined entrance passage, contrast of space and focus on vertical feature are quite typical. The sense of enclosure is not as marked, however. The formal quality of the space is marred by the overlapping plane and disruptive silhouette of the High Street beyond. On closer approach, this is cut off by the natural angle of vision. Fortunately, this conflict is not enough to reduce the visual attraction of the cupola, which, from being a flat plane at a distance, takes on a strongly modelled three-dimensional quality on closer approach and then fades out of the field of view to give way to the colonnade as a space-defining surface. At this point, the rhythmic qualities of the colonnade take effect and from being detained by the cupola, the eye has an unconscious tendency to roam around the space before one is plunged into the entrance ‘maw’ of the gateway. The strong attraction of such a pool of shadow is here weakened by its proximity to a rhythm of such shadows.

The division of the floor plane at Christ College and here dictates the spatial experience along a pre-arranged route and thus exercises a degree of control over it. The small-scale planting, incidentally, in Queen’s is to be deplored. The flat expanse of proverbially flat lawn is a tradition that makes for a loose sort of urbanity, and to mar the junctions and emphasise the path by flower beds is most unsatisfactory.

Norman Foster: The nature of enclosure in urban areas

Queen’s College Quadrangle, Oxford

The second quadrangle at Queen’s is a rectangular space completely enclosed on all four sides by facades of near equal height. Three sides are continuous and quite irregular, and form a U-shape which is completed by the slightly higher facade of the Library. These two enclosing elements have their own distinct character and the character of the space is a result of their relationship and relative contributions. 

In terms of enclosure, thin space appears to be more tightly contained than the other three examples. It measures 130 feet x 96 feet and the proportions of floor plane dimensions to wall plane height are 2.5 to 3.5. Apart from an almost hidden link with the colonnade of the first quadrangle, views out are completely restricted. The character of the space is dour and rather severe.

Although the library is articulated from the rest of the composition and it has a grander scale, it is not a key building in the same way as the chapel and hall is to the previous square. Architecturally it is not nearly so distinguished: it has a weak and hesitant quality that is the result of an equal relationship of solid to void and proportions that are a clash of the horizontal and vertical. But this apart, the building is seldom seen as a facade in formal relationship to any two sides. The entrance is through a relatively insignificant doorway in the middle of one of the long sides and there is no preparation or feature for the eye to focus on as in the previous example. Thus, the Library is seen obliquely in the role of a defining plane and not as a climax.

Norman Foster: The nature of enclosure in urban areas

There is little contrast between the library and the U-shaped block; the colour, texture and surface modelling are about the same, and the negative qualities of the former have already been mentioned. From some viewpoints at the extremities of the space and just outside it in the entrance to the colonnade of the first square, there is a satisfying effect in the regular window rhythm of the U-shaped block. It is, however, tentatively suggested that moving about in the space itself, this effect is not generally noticed because of the limiting dimensions; instead of the perspective effects resulting from such a rhythm, they appear more as a flat and regular pattern.

To summarise then, this quadrangle has a different spatial quality from the previous examples. This quality can best be described as a lack of directional emphasis; a static feeling of space. This is something apart from the degree of enclosure and can probably be attributed to an overall similarity of the wall planes and the relative lack of those contrasts evidenced at Verona and the other two quads in Oxford.

Norman Foster: The nature of enclosure in urban areas

Shepherd Market — Mayfair, London

Shepherd Market is basically rectangular, quite small in scale and of modest dimensions (a mere 110 feet x 45 feet). With its pedestrian approaches, strong sense of enclosure and almost domestic scale, it makes an interesting comparison with present-day conceptions of urban shopping precincts. 

Scale and height of the enclosing buildings are remarkably consistent bearing in mind the sort of alteration that has taken place over the years. Predominantly, the buildings are of a Georgian vernacular, except for the street facade defining one side of the space, which is Victorian. This is four storeys high instead of the normal three, and it also has a different scale. The proportions of floor plane dimension to wall plane height are 1.2 and 2.5. 

Including the approaches, the composition has a character of accumulation and almost haphazard variation. This is evident in the variety of shopfronts, signs, window blinds and stuccoing; all within a basically simple form. Entrances, whether pedestrian or vehicular, always have a visual stop and never become endless corridors without a goal in sight. If the approaches had been consciously designed to contain the view, exactly the same devices of curving streets and setbacks would have to be employed. 

If the composition of the space is broken down into its elements, it will be seen that basically there are two parallel facades of unequal length; the long one bonding obliquely and the short one at right angles to become Shepherd Street. The third defining plane is the strongly modelled Victorian corner, articulated from the square by Shepherd Street itself; the remaining defining plane is broken up and separated from the other facades by the pedestrian alleyways.

There is a strong sense of enclosure within the square that varies according to the movements of the observer and the views selected. Looking towards Shepherd Street, the feeling of enclosure is not as marked. There are probably several reasons for this. The most elementary one is that it is further away than the main parallel facades (assuming of course that one is within the confines of the facades). Secondly, the presence of Shepherd Street separates it visually and the suggestion of distance is heightened by part of the facade disappearing obliquely away from the main space. As the end of the square is approached in this direction there is a breakdown of enclosure and space looks out, as it were down Shepherd Street.

Despite all the variety of detail, the standard Georgian window proportion is repeated in nearly every facade and becomes a unifying link. From within the square, enclosure is most complete in the direction of the pedestrian alleyways. Although at the entrance to these circulation routes, there are interesting views out and the visual attraction of another series of spaces to be explored. In a space such as this where there are no predetermined circulation routes as at Oxford, the curiosity value of what lies around the corner should not be overlooked. The quickening of pace to solve the mystery is especially noticeable when walking in the cathedral precincts of York or Winchester.

It would be interesting to observe the effect of an unbroken Georgian facade over the shops at one or both ends. It is suspected that it would alter the apparent proportions of the space and create a greater degree of enclosure. As it is now, the entrance slot of the alleyway and the articulated Victorian corner tend to break up the enclosing planes vertically and reduce their apparent width. The recessed wall marking the pedestrian alleyway certainly leads the eye past the two planes defining the corners and opens out the space in that direction.

In the sense that the enclosing elements are not continuous, this square can be likened to the Piazza Dante, Verona. The introduction of a linking device similar to the arch at Verona would probably modify the space completely and reinforce the enclosure where it is most needed. Spatial expansion is conditioned by immediate past experience and it is interesting to note that the square appears larger when entered from the long pedestrian way than from the parallel and wider Shepherd Street. For their purpose, the size of the space and related approaches would appear to be ideal. With only a few people, they appear pleasantly crowded and even a café has been crammed, Italian-style, onto the pavement. More important still, the spaces never appear desolate, even when empty, which is a lesson to be digested by most of the ‘shopping parades’ of the New Towns.

Norman Foster: The nature of enclosure in urban areas

Piazza del Campo — Siena

The Piazza del Campo is roughly shell-shaped and is set between three low hills upon which the city is built. In 1218 a masterplan was drawn up which determined the shape of the present campo and the architectural character of the buildings. The paving was laid in 1413 and the piazza has remained largely intact from the Middle Ages. 

The undulating wall as a means of modulating space is a baroque device bringing unexpected movement and flexibility. In this theatre shape, the eye is powerfully drawn into the centre of the space, and in Siena full advantage has been taken of this tendency. The piazza is entirely dominated by the slender tower of the Palazzo Pubblico. Whether by accident or design, the buildings immediately adjoining the tower are lower than the others, so that it rises free and untrammelled for practically the whole of its height. Despite its height, the dimensions and contours are such that the space is large enough for it to be seen and yet small enough for it to dominate.

The bowl shape is emphasised by the pattern of brick and marble grids, which radiate star fashion from the lowest central point. The maximum dimension across the piazza is 310 feet and the straight side is 475 feet long. The enclosing facades vary in height from 40 feet to 97 feet, giving an average variation from 4.5 to 6.5 as the proportion of wall plane height to floor plane dimension. The continuous floor plane is broken only by a fountainwhich is partly recessed, and a line of bollards. These bollards, which are massively in scale, measure out the space in an even progression and give an immediate key to its size. Normal perspective effects of a fairly consistent line of windows are only really apparent from the curved perimeter. As a result of the slope of the floor, the curved facade appears to hang in space when viewed from within the hollow. This effect is heightened by the precise junction of vertical and horizontal planes.

Variations of window rhythm, silhouette, articulation and colour of the crescent-shaped facade become relatively unimportant on this scale and do not detract from its broad curve. These variations only become really apparent in photographs and it is quite surprising to see how irregular it really is in plan. On a much smaller scale, these irregularities would become readily apparent, an effect that has been observed in an earlier section of the study.

Contrasts are simple and dramatic. The darkness and almost claustrophobic restriction of the entrant street compares forcibly with the sunlit openness of the piazza and the tower with its background of the blue sky. This tower is a clear statement of verticality in a horizontal composition and its form is visible from most of the approaches. Apart from the natural enclosure created by the walls, it exercises its own attraction within the space. The degree of enclosure is quite remarkable when the dimensions of the piazza are considered. From within the hollow of the straight side, the slope up to the curving facades and restriction of the views out are instrumental in preserving this sense of enclosure. As a space, ‘il campo’ has a grandeur and breadth of scale that never oppresses, but absorbs the collection of cars, coaches and pedestrians that linger on its perimeter.

Norman Foster: The nature of enclosure in urban areas

Piazza della Cisterna and Piazza del Duomo – San Gimignano

The Piazza della Cisterna and the Piazza del Duomo are irregular medieval spaces in the hill town of San Gimignano. They are punctuated by a series of fortress towers, which in the 13th and 14th centuries numbered some 70 and must have given the town a Manhattan skyline. The towers, of which there are now 14 left, are two-scale elements in the sense that they make a positive statement of the town when seen from a distant approach and yet are easily comprehended from within.

The Piazza della Cisterna is triangular in shape and its maximum dimensions are 150 feet and 200 feet. The enclosing buildings vary in height from 40 feet to 60 feet, giving an average maximum floor plane dimension to wall plane height of about 3.5. There is a random variation of small window openings in what are dominatingly solid walls. This basic lack of rhythm is emphasised by the irregularity of the forms and silhouettes of the enclosing, and also the haphazard minor articulations. Towers, which vary between 100 feet and 130 feet in height, can in several instances be comprehended from both spaces.

By comparison, the Piazza del Duomo is completely irregular and dissolves into a number of articulated planes and levels of varying shape and size. The piazza itself is smaller, although the facades are about the same height. If this space has a key building then it is the Duomo, purely on account of the contrast of its symmetrically organised and elevated facade, which is visually separated from the rest of the composition. As in the larger piazza, the entrances are so narrow and slot-like that the views out are almost entirely restricted.

Within these two spaces, there is a third and most important space defined by the loggia. This vital junction between the two spaces is marked by a concentration of verticality: almost a gradual build-up of the forms along the length of the facade. The vertical entrance slot to the adjoining piazza is also within the only right angle to be found in the composition.

These two spaces are perhaps best enjoyed by the pedestrian in motion and offer an unfolding sequence rather than a centralisation of space as at Siena or the Circus at Bath in its original form. The approach is winding and confined and the spatial expansion of the Piazza della Cisterna unexpected. Formal perspective effects which have been noted to some extent in most of the other examples are missing and would be defied in any case by the two-way slope of the floor plane. The aforementioned right angle would appear to be important  in perceiving the shape of this space, because despite the irregularities the triangular configuration of the defining planes is quite easily grasped. There is a remarkable degree of enclosure that varies according to the path of the observer.

Enclosure is weakest in the direction of the angle formed by the two long sides of the piazza. It is difficult to say whether this is the result of a lack of containment at this point or the false perspective effect induced by the converging facades. However, this is offset to some extent by the visual interruption of the 650-year-old wall-head. From some angles, this becomes a focal point in a space defined by the surrounding buildings. There is a satisfying clarity in the way that the paving follows the contours of the slope to be sharply defined by a vertical wall plane. From within the piazza, this slope and the cluster of vertical features lead one almost unconsciously through the narrow corner and into the adjoining piazza. 

Before actually entering the piazza, however, one crosses into the loggia, a kind of space within a space. At one point there is a similar effect to that observed within the arcade of the first quadrangle at Queen’s College, Oxford. Namely, the sky is momentarily cut out of view and the facades of the buildings appear almost as a backdrop, with little or no meaning in terms of spatial distance. Because of the interruption of this loggia, there is the sense of a separate space rather than a continuation in different terms to the previous one. After the initial experience of spatial expansion, the Piazza del Duomo leaves the impression of a series of informal and picturesque groupings rather than a formally defined space such as the second quadrangle at Queen’s College, Oxford. This makes a most interesting comparison because the dimensions are nearly identical. When the drawn plan is studied it becomes apparent that the form is not nearly so irregular as it appears from within the space. On a larger scale, these effects would tend to be ironed out.

Despite all the irregularities that have been mentioned, San Gimignano has a basic unity that is not confined to its two piazzas. Running through all its picturesque articulations there is a unity of architectural expression and an absolute consistency in the use of materials. San Gimignano begins and ends on a hill and is a truly urban environment in refreshing and clear-cut contrast to the countryside around it.

Norman Foster: The nature of enclosure in urban areas

The Piazza Dante – Verona

The Piazza Dante, Verona, is a rectangular space defined by a number of buildings around it. On first examination, it is difficult to understand why this piazza creates so marked an impression as a space and not just as a collection of buildings, which is what it really is in one sense. Because, despite the variety of architectural expression, one’s immediate and post impressions are of a piazza. The average height of the buildings is 60 feet and the main dimensions are 249 feet x 111 feet; these give proportions of floor plane dimension to wall plane height of 2 and 3. As a space, it is marked by a strong degree of enclosure and has a lively, distinctly urban and human character. A random and almost picturesque quality distinguishes this piazza from the more formal examples chosen for study.

The key to the spatial qualities of this space would appear to be, paradoxically, unity and variety. There is variety everywhere: rhythms of windows vary from facade to facade; scale ranges from the domestic to the civic; silhouette changes from the traditionally embattled to the classically pedimented; pattern and colour vary from building to building and even the tower is placed asymmetrically. The key to the whole spatial composition, which ties this veritable riot of ideas together and creates a unified order, would appear to be the arches.

They become a dominating rhythm linking every adjacent building mass and reinforcing the sense of enclosure when space is most likely to ‘leak away’. In some places, there are double arches, one behind the other, and from several viewpoints the effect of a whole series of overlapping planes becomes visible. 

The buildings themselves sometimes line up exactly with the boundaries of the space and sometimes flow past and out of the piazza to become part of the street leading to it. In a sense there are almost two meanings to this spatial composition. One is a simple formal space defined by a series of planes, not joined but in a definite relationship to each other. The other is an informal, almost picturesque quality of the sculptural massing of the individual buildings. Views in and out of the square, for instance, show similar effects of overlapping planes as in some parts of Chester and York.

The large and deeply shadowed openings of the arch form give vigorous effects of light and shade, solid and void, penetrating the interior with outside space and bringing a degree of unity and integration. This is a device that Lutyens used to great advantage. There is the sense of the square and its surroundings being connected in a chain and not being a series of disconnected structures.

The principal buildings of the space would appear to be the Prefettura and Casa Nova del Comune (now Palazzo del Comune), occupying the sides and in this position visible in their entirety from most positions within the space. The immense power of a deeply shadowed opening to attract attention can be seen particularly in the Casa Nova del Comune. Compared with Christ College quadrangle the tower is relatively unimportant, as it can only be grasped in its true relationship as a vertical feature from the approaches or extremities of the space.

The statue in the centre of the square is related to the main axis and the tower, and provides a visual stop, framed by the entrance arches mentioned previously. Within the main space itself, it exerts a certain nuclear attraction, but it remains a point of interest within the piazza and does not in any way dominate it. It is interesting to note that what appears to be the more visually dominating building, the Casa Nova del Comune,  is also the only building with a symmetrically organised facade. It is difficult to say whether this dominance is the result of a contrast of formality in an informal setting or the power of its architecture, which has already been noted. 

The floor plane is flat, neutral and unbroken, except for a regular pattern of paving joints. In this case, the simplicity of treatment would appear to be of primary importance, as the floor plane runs right through all the irregularities and brings them together in a direct relationship.

Apart from the architectural contrasts already mentioned, the spatial sequence of the confined street against the space of the piazza, and the leisure of its café life against the clamour of the market in the adjoining Piazza delle Erbe, should also be noted. The space has a static quality in that there is no perspective progression for the eye to a definite visual dominant; even the street approach onto the statue eventually dissolves into a series of overlapping planes. It is suggested that this space is so satisfying when compared with some of the other examples, because it is completely integrated into its surroundings (note the relationship of the Scaliger Tombs nestling in the shadow of their parent church) and within the confines of a clearly defined simple space there is a large degree of informality and consequent picturesque quality. If we agree with Aristotle that ‘the function of cities is to make men happy’, then we have much to learn from the Piazza Dante.

Norman Foster: The nature of enclosure in urban areas

The Circus – Bath

The Circus is part of a sequence of architectural spaces constituting the 18th- and early 19th-century development of Bath. Two of the three radiating arms relate by long straight streets to Queen Square and the Crescent. The design was reputed to be based on the Colosseum, but resemblance would appear to end with the superimposed Roman order. The composition takes the form of three equal parts with the centre line of each road on the centre line of a terrace. Consequently, the views inside the space are symmetrical and the views from each road into the space are closed by curved walls. Distance from wall to wall is 300 feet and the enclosing terraces are 40 feet high, giving a floor plane dimension to wall plane height of 7½. 

The dominant object is a group of four mature trees, but drawings show that originally the space was cobbled from wall to wall. This landscaping makes up the floor plane to the extent of destroying its relationship with the wall plane. This is to be deplored and has been commented upon by at least one authority on Bath – ‘The trees should be literally extirpated if need be by dynamite, so that justice may be done to the work of Bath’s Architectural Masters’.

The regular repetition of the columns and attendant rhythms give to the size and also to the form or shape of the space. Even when the view is partly obscured by trees, the shape is inferred by the measured progression of columns disappearing on one side and reappearing on the other.

It is interesting to speculate on the spatial quality of the Circus as it would be without the trees. The regularly repeated rhythm of the facades, with no visual dominants, would produce a balanced and static effect of space, relieved only by the entrance voids. That is unless some object in scale with the space were placed in the centre as at Amalienborg Plads in Copenhagen. Such a device exerts a natural attraction in space and its sphere of influence is confined by the surrounding buildings. Goldfinger notes ‘a vast expanse without any spatial meaning is capable of becoming a tightly knit entity around the magnet of a central feature’ and likens it to a crowd around an orator.

In its present state, the space varies according to seasonal differences in appearance of the trees. In winter they are seen as a semi-solid filter and the eye perceives the full extent of the space through the textural and perspective effects of a veil. In summer, however, they are a dominating three-dimensional mass, completely obscuring any view of the opposite side. Despite this visual interruption, the curve is sufficiently well perceived to communicate the shape from any one observation point. If the trees completely filled the central space it would take on the character of a curved street, defined by trees on one side and by buildings on the other, with the view closed at either end by the curve of the facade. The approach from Gay Street is quite long and steep, and the only visual stop are the trees within the Circus. Otherwise, they produce an unresolved conflict in what should be a wholly urban environment.

Royal Naval College and Queen’s House – Greenwich

The couple of buildings at Greenwich (1610–1814) were originally part of Wren’s Palace scheme to include Queen’s House and King Charles Block. The Hospital, or Royal Naval College, as it is today, has since been separated from Queen’s House and its setting, by a main road. Its essential formality from the riverside has been retained, however, and the informal separation of one element of the composition from some viewpoints makes it a more interesting example for examination.

Within the space defined by the Queen Anne and King Charles blocks, there is little marked sense of enclosure. Views out of the space are relatively unrestricted; as well as an open cross axis, the whole of one side fronts onto the Thames with only a screen of railings in between. The flanking blocks are 72-foot-high in relation to an almost square space 255 feet across. This space is more a forecourt or setting for the composition of the cupola and colonnades of the Hospital. The main axis of the architectural composition is related to a central statue and continues out beyond the colonnades to Queen’s House.

This first space makes more sense from the river or the opposite bank; because then it can be seen as part of a sequence of spaces and the actual progression and diminution towards Queen’s House is at once apparent. Although it is not thought of in this context when moving about within the space, it is interesting to note that we apprehend intellectually that it is symmetrical from the evidence offered to our eyes. This effect becomes more marked when we move into the second space within the colonnades, which is further defined by a change in level of some seven feet. Moving about between the colonnades and across the space presents a series of oblique views which are ever-changing, but always related and unified by the imaginary axes of which the intellect is aware.

From some oblique viewpoints, the row of columns on one side appears as an undulating or corrugated wall and the opposite side as a screen of widely spaced columns. However, the role of the colonnade as a space-defining surface is never in doubt, even when the wall behind is opened up to reveal views into courtyards. The grand and dignified march of those columns diminishing towards Queen’s House is a positive and geometrically formal statement. Despite a distance of some 450 feet, the almost flat plane of Queen’s House is just as much a third side to this space as the Cathedral is to the Piazza San Marco. There are no formal indications of the distances involved, which is in sharp contrast to the concise definition of space by the colonnades.

These two colonnades have an almost magnetic quality and would appear to define a field of enclosure that varies according to the observer’s movements. On looking out of the space at the ends, the enclosure dissolves into the open spaces beyond.

Looking inwards from the ends, however, raises a fundamental issue of this spatial arrangement. In one direction the line of columns marks out a space and leads the eye out beyond this space to Queen’s House. Any gaps in the pictorial composition are filled by the colonnade to either side of the house. The form of these elements is all the more precise as they are sharply defined in white against the dark background of rising ground and trees.

There is thus a focus on the axial layout, and any argument that seeks to justify a stronger climax than Queen’s House does not alter what is basically a very satisfactory spatial arrangement.

The same cannot be said of the view in the opposite direction. Although the space has been clearly defined and the view out framed by the twin 150-feet-high cupolas, it lacks a focus and has an unresolved quality. The space itself measures 350 feet x 115 feet and the facades are 39 feet high; a proportion of floor plane dimension to wall plane height of about three.

One of the more unusual aspects of this group is the way that Queen’s House merges into an informally tree-studded landscape, rising to Greenwich Park. As a result of this, its formal relationship with the rest of the composition is not easily grasped from the road and more oblique viewpoints. The gradual build-up of levels from the space adjoining the river right up to Queen’s House is most important and is further echoed in the hill behind with its silhouette of the Royal Observatory. The same device of a rising floor is used in stage sets to give enhanced depth to the stage. The presence of human beings and their momentary occupation or use of the space, constitute the very essence of its meaning. Unfortunately, however, at its sightseeing busiest, circulation is debarred from the area between the colonnades, which only tends to emphasise the more theatrical aspects of this composition.

In the light of these examples, it would be worthwhile to summarise and expand on the introductory passages of this study. First, however, the way in which space is actually perceived should be briefly considered.

JJ Gibson notes the following means to which the mind is believed to infer a world of three dimensions from a series of visual ones. They are: texture perspective, size perspective, linear perspective, binocular perspective, motion perspective, aerial perspective, blur perspective, relative upwards location, shift of texture density, shift in amount of double imagery, shift in the rate of motion, completeness of outline, and transitions between light and shade. It will be seen that most of these are based on a static retinal pattern and operate in the case of a stationary observer.

Thornley notes the following means by which distance or depth is observed: 1) Overlapping planes and the related effects of aerial perspective and parallax; 2) Linear perspective; 3) Variations in the eye focusing mechanism in response to the different ranges at which objects are seen; 4) True stereoscopic vision; 5) Angular adjustment of the eyes operating in a similar manner to that of a rangefinder; 6) Differential displacement of the images on the retina. 

Factors 3) 4) 5) and 6) are classed as ‘kinaesthetic’ factors in that the operation of the mechanism of the eye, as it adjusts itself, gives the indication of the relative depth of different planes. This process operates on a subconscious level and is compared with factors 1) and 2) which depend on what is termed the ‘visual’ sense and partly involves a conscious evaluation dependent upon previous experience.

The observer’s awareness of space has been noted by MD Vernon as developing thus: 1) An immediate experience of spatial extension; 2) An immediate estimation of spatial distance (ie, (a) if further away than (b)); 3) Immediate perception of depth or relief in objects near at hand.

Reduced to the simplest of terms, spatial experience may be said to be the result of two components. The first would be MOTION and would take account of 1) Direction (up, down, left, right); and 2) Rate (velocity, acceleration, stops). The second component would be PHYSICAL GROWTH of the space and would include the following factors: 1) Size and scale; 2) Shape or form; 3) Relationship of surfaces and elements defining the space; 4) Treatment of the defining surfaces, ie, rhythms, solid to void, articulation, silhouette, colour, pattern, texture, etc.; 5) Secondary elements, being objects or ‘furnishings’ within the space, ie, trees, campanile, sculpture; 6) Position, size, shape, colour and texture of these secondary elements.

In passing it should be noted that a visually perceived space form exists only in terms of a given condition of illumination. Qualitative and quantitative changes will occur in the perception of a given physical relationship of space-defining surfaces and objects according to variations of natural and artificial light (for example, Piccadilly Circus under differing conditions of sun, fog and night-time illumination).

It should be noted that the experience of space involves sensory perceptions other than the visual. For example, auditory, olfactory, thermal and tactile sensations are all relevant to the final establishment of space. The abstract associations of philosophical and historical values should also be mentioned. It is suggested, however, that these are of a qualifying nature and secondary to the visual sensation.

The effects of size and scale were discussed in part at the beginning of this study. It will be remembered that Sitte suggested a proportion of floor plane dimension to wall plane height of 1 to 2 according to the architectural treatment of the space. On the other hand, Alberti suggested a proportion of 3. Hegemann and Peets also mentioned distances from which buildings and civic groups could be viewed in their entirety. The dimensions of the examples discussed here only range from 110 feet x 45 feet to 310 feet x 475 feet, and the proportion of floor plane dimension to wall plane height varies from 1.2 to 7.

However, within such a limited group it will be seen that there are so many spatial characters involved, ranging from the intimate to the majestic, and also so many qualifying factors, that it is not only impossible, but undesirable to lay down an ‘ideal’. But this only serves to emphasise the need for a method of comparative analysis of as many examples as possible. In this way, with carefully chosen spaces over a wide enough range of scale, a pattern may evolve, empirically, which would take into account such important variables as the rate of movement of the observer and the practical limits governing enclosure. 

It should be noted that the use of a factor denoting the proportion of floor plane dimension to wall plane height is useful only as a guide to the physical organisation of the space, and does not, for example, imply any comparative degree of enclosure. Apart from any other significant factors, it takes no account of the scale of the space or the position of the observer within the space.

The distinction has already been noted between the extremes of spatial form ranging from the regular and formal to the irregular and amorphous, with an infinite number of variations in between. Within even the limited range of examples examined, it will be seen that different fundamental forms possess their own distinct spatial characteristics Christ College quadrangle compared with the Circus at Bath, for example, both have about the same maximum floor plane dimension and wall plane height). It would be necessary, however, to examine a far wider range of examples of varying size on a truly comparative basis, before any assertions could be made concerning the effects of form beyond the basic characteristics already mentioned in the descriptive passages.

It is not easy to find two spaces of similar form and size with different facade treatments; so consequently, it is difficult to determine, on a comparative basis, what effect variations of the surface treatment of defining planes have on the space itself. It is, however, possible to isolate the various devices, which are to some extent those of the static arts of architecture, sculpture and painting, and to some extent those of the arts of music or the cinema which have a temporal basis.

The use of LINEAR PERSPECTIVE, depending for its effect on converging lines and the regular diminution of objects of a known size, clearly indicates spatial extent. When it occurs in two or more planes intersecting at right angles, it defines spatial configuration with great clarity. Thus, a rhythm of regularly spaced elements, such as columns, arches or windows, for example, can be a dominant in its own right (in this context note the hypnotic attraction of regularly spaced lamp standards disappearing in the distance); or the progression which is implicit in this can be used to form a vista. In the first case, the proportions and rhythms of colonnades or other bay systems directly affects the spatial character, and these were the subject of much Renaissance theorising (Alberti, Palladio, Serlio, Vitruvius). The vista and climax were used ad nauseam by the Beaux Arts school of urbanists in Paris, but the school is capable of producing groups of seat architectural grandeur. If the sides are made to converge in long unbroken lines, there is a tendency towards a weak effect at the end of the vista. This can be avoided by stepping the composition (Greenwich from the Thames) or creating a related sequence of spaces (Nancy). The effects of natural perspective can also be exaggerated or minimised by arranging the sides of the vista in converging or diverging lines as in Piazza San Marco, Venice. In all of these groups, the effect of perspective is to produce a picture whose lines converge and draw the eye to a central point. Absence of a climax or visual stop at this focus usually results in a weak and unresolved quality. This climax need not even be a dominating building as such; at the Piazzetta San Marco, the flanking columns frame a contrasting expanse of a lagoon, with San Giorgio Maggiore in the left distance. 

Less precise in defining spatial extent and configuration are the effects of OVERLAPPING PLANES and aerial perspective. The ‘greying-off’ and loss of detail due to aerial perspective have been noticed over quite small distances. Also related is the effect of parallax which occurs between overlapping planes and objects, which according to the movements of the observer appear to move in relationship to each other. This would suggest an informal grouping of sculptural elements and serpentine approaches as in the Greek Acropolis. The nearest in character to this is probably the complex of buildings, comprising the Duomo, Baptistry and Campanile at Pisa, which presents an ever-changing relationship of forms. Similar effects can be seen in the accompanying photographs taken in the vicinity of the Radcliffe Garden in Oxford. The cliff-like face of the building on the right dissolves into an articulated defining plane with its own pronounced silhouette as the spire swings into view. On closer approach, this mobile-like relationship alters to include the bulk of the Radcliffe Camera, which eventually ceases to behave as a flat plane and becomes a three-dimensional reality.

A formal grouping can be apprehended intellectually and becomes a known quantity even though the observer may perceive a series of oblique viewpoints. This makes for an interesting comparison with the view from beyond the Camera in the opposite direction, which has little meaning in relation to the previous vantage points. The crest of the floor plane in the first photograph, which cuts off the base of the buildings in the background, would appear to emphasise their facades as planes suspended in space. A corresponding effect resulting from variations in the contours of the floor plane has been noticed at San Gimignano, Siena, and elsewhere.

It is not possible to define a space without creating a contrast of some description. At its most elementary there is the contrast between the space outside and that which has been defined. The more pronounced this difference becomes then the greater the impression of space. What would be a spacious marketplace in relation to the narrow alleyways of a medieval town, would appear as a constriction if entered from, say, a Parisian boulevard. It is significant that most of the geometrical and formal spatial organisations were set in an informal and closely textured urban environment, which was the result of growth-by-accretion (ie, 16th-century Rome and early 19th-century Paris.

The purpose of a contrast is to emphasise the essential qualities of both parties. Within the space itself any number of contrasts are possible: form, scale, solid to void, silhouette, colour and so on. Depending on the degree to which they are used, contrasts can add life, variety and informality to an essentially static space, or at the other end of the scale direct and dominate the space. Between these two poles there is an infinite number of subtleties in variations. Contrasts of form would appear to be the most dominating and easily grasped; the vertical feature in a horizontally composed group and the strongly modelled sculptural form against the foil of a regular colonnade for example.

The power of the large and dark or deeply shadowed opening to attract the eye is interesting as a contrast of solid and void. Regularly repeated, in the form of an arcade, it becomes a rhythmic extension of the space, but in isolation it can become dominant in its own right. Its suggestion of an inner reality invites the eye to pause while the mind speculates. Ruskin draws the following moral: ‘After size and weight, the Power of architecture may be said to depend on the quantity (whether measured in space or intenseness) of its shadow […] so that Rembrandtism is a noble manner in architecture though a false one in painting’.

Norman Foster: The nature of enclosure in urban areas


Objects or ‘furnishings’ within the space can be classified as those which give meaning, scale and interest to a space and those which divide it, or alter it in such a way that its spatial character undergoes a marked change. An example of the former would be the bollards in ‘Il Campo’, Siena, and a corresponding example of the latter would be the trees in the Circus at Bath. The extent to which these secondary elements modify the space will depend on their position, size, shape, colour and texture.


A parallel study of elongated space or streets was undertaken, but extends beyond the scope of the essay. However, several relevant observations can be briefly noted, although in detail these would be considered against a background that takes account of their length, width and height. In this respect it will be seen that it is comparatively simple to find streets of basically similar size with varying treatments of facade, or streets with similar heights but of varying widths. This enables a direct comparison to be made and compares favourably with the difficulties already mentioned in respect of squares. 


The two extremes noted were the spaces clearly defined by precise perspective effects and the richly articulated streets, usually of medieval origin, with their accent on overlapping planes and facade interest. The most satisfactory spaces were the most true to type. Bedford Place, in Bloomsbury, is a first-class example of an urban space relying on the effects of clear-cut perspective.


Even more satisfactory would be an example where the floor plane is unbroken by kerbs, and the junction of horizontal and vertical planes unchanged by railings and the like. The utmost clarity in this respect is to be observed in some of the smaller canals in Venice. The opposite type of space, with its emphasis on overlapping planes, could be likened at its extreme to moving through a congested market, and being presented with an ever-changing series of close relationships, as an unfolding sequence of varied stalls come within the field of vision. There are many successful examples that combine the effects of overlapping planes and perspective, the main streets of many well-preserved villages, for example. A similar combination is noted in the Piazza Dante, Verona.


The street with its emphasis on unified perspective effects is most satisfactory when it terminates in a visual stop (ie, Bath Street or the Piazzale degli Uffizi); although within certain limits of size the need for this would appear to diminish. The same is apparently not true of the space that depends for its interest on richness of articulation; such spaces usually create their own pockets of enclosure as they twist informally (ie, Stonegate, York and parts of Chester).


It is significant that most of our present-day spaces, whether they be streets or squares, fall within these two poles, without any of the clearly defined characteristics of either. The unity of the street has usually been destroyed by such variations as scale, height, silhouette and colour. But within the confines of a formal building line it lacks the variety of articulation and modelling, as well as clear-cut perspective effects.


The favourable response of the individual, tourist or citizen alike, to the more historical and picturesque spaces in this country, and such northern Italian tours as Bergamo, Mantua and Verona, would suggest the advantages of a smaller scale relationship of spaces than characterised by a degree of informality and articulation which has already been discussed. Despite the arguments that such places have evolved organically, it would seem rather narrow-minded to reject the creation of similar equally sympathetic spaces on such arbitrary grounds as suggesting that their appeal lies in historical associations or a picturesque mellowing. The spirit of such spaces could be recreated in today’s terms without resorting to a pastiche of their motifs. This does not weaken the case for the more ceremonial and open areas of social significance, which would punctuate the closer urban texture. These spaces would gain immensely by contrast, whether or not they were organised on a geometrically formal basis.


If we accept as fundamental the eventual separation of pedestrian and vehicular traffic, then the creation of an acceptable urban environment hinges on our ability to manipulate space. Such an ability can only be the result of a working knowledge of its physical organisation on a humanistic scale, that also takes into account the rates and means of travel.


More from Issue 03