[Contents]--[Abstract]--[1]--[2]--[3]--[4]--[5]--[6]--[7]--[8]--[References]

3. The city as an area of spatial experience

This chapter studies one particular type of environment - namely the city. In the first section of this chapter the navigational task in the city is described as a special case of the navigational task outlined in section 2.4. This section also describes tools commonly used to help navigating the city environment.

The following section reviews the work of Kevin Lynch who studied navigation in cities extensively. He describes several elements of the city image which can be seen as another mental representation of spatial knowledge. Lynch's work focuses on defining guidelines that help in designing cities that are easy to navigate.

Further communicative and social aspects of the environment are described - in particular the use of space as ordering principle and how the environment serves as information carrier about itself. In a separate section social aspects of space are outlined. Social issues play an important role in the use of an environment.

The final section in this chapter describes a detailed ontology of city elements which support navigation. This ontology will be used in a later chapter to define a similar ontology for the Information City.

3.1. The navigational task in a city

Instead this section concentrates on navigation in general by walking, by using transportation infrastructure like taxis, cars or subways and the like. In a real city all those means of navigation are in a way restricted as is car navigation. Even pedestrians have their special routes and cannot simply walk through houses. Navigation in the city still is a planning task which tries to find a path to a location. There are two cases to distinguish in this task: either the location is known or it is not.

To know a location in a city means knowing its "address". As was mentioned already in section 2.2.3. it is possible to point out locations exactly by formal systems of referencing. An address is an example of such a formal way to point out a location in a city.

Finding a location requires additional knowledge or infrastructure

An address commonly consists of a street name or number and a house number in that street. Addresses define a global reference system for an environment. Interestingly an address alone is not sufficient to reach that destination effectively. The reason is that an address, for example "Möllwaldplatz 5", tells nothing about where this Möllwaldplatz is. Additional information, like: "near the Favoritenstrasse" or "in the 4th district", restricts the search space.

The actual navigational planning in city navigation requires survey knowledge about the city structure, about how to reach areas quickly and efficiently and so forth. This knowledge includes bus routes, typical areas where traffic jams occur, where to find a taxi, estimations about distances and travel times to decide which means of transportation to choose and so forth.

Making mistakes in navigation leads to the problem of how navigation can be supported in real life. In confusing areas it would take a considerable time to acquire a correct image of the environment only by navigating it. Often it is not necessary to develop an environmental image of an area since people navigate the area only a few times or need to reach only one particular location in this space.

Maps and route-sketches

People very early developed tools to support navigation, for instance maps. Map-making allows to represent spatial relationships on paper that can be duplicated and handed to other people. This art or skill allowed easy sharing of navigational information. Whoever is able to read the language of the map is able to get information about the environment.

There are two basic types of maps. One of them illustrates how to reach a single location. This map is like a route descriptions. The other type of map describes a whole area. Route description maps are strongly influenced by the mental image of the person drawing the map, as this person tries to transfer route knowledge from her mental representation of the area to paper.

A typical example of such a map are route sketches people commonly give to each other when describing the route to - for instance - a party. An example of such a map is shown in figure 4 (next page). It looks almost ridiculously uninformative on the first glance but it contains all information needed to find the place when supplemented with spoken information.

The additional description in this case is: "Drive into the village till you reach the only traffic light. Then turn half right and then go on for about the same distance." The map contains a circle representing the traffic light and shows how the streets separate at the traffic light.

This map gives no indication on distances however. The person drawing the map normally goes by car so the distance to the traffic light was perceived as being not far. If I had to walk this distance I certainly would constantly think I had made a mistake because after 10 minutes I still would have seen no traffic light. In this case a reference landmark should be given like: "About half way there you pass a gas station".

Fig. 4: A very simply route-map.

As this example shows such route description maps are very close to verbal or textual route descriptions. Most of these maps are only a supplement to a verbal route description. The map in figure 4 is an extreme example of drawing a cognitive map on paper. It leaves out all except the most important information. Such maps often act only as a reminder. Sometimes they are useful only to provide scaling information.

Reorienting oneself during navigation

In many cases even using a map and a route description is not sufficient to reach a destination on the first try. Instead people tend to ask other people about the right way. This is done for one part to check if they still are on the right path and on the other part to solve unclear points in the map or description. A third motivation for such a checking can be to realign the map to the environment.

Kevin Lynch describes that people in earlier times often used such a "progressive approaching" to a location. This type of city navigation was dominant, because people had a tendency to stay in the close vicinity of their living and working places because of the unavailability of public transport and of good maps:

Imaging and referencing by district was very common in older cities, where quarters and their populations were relatively stable, isolated and distinctive. In Imperial Rome, addresses were given solely by small defined districts. Presumably, arrival at such a district allowed one to proceed to one's final destination by personal inquiry. [LYNC60, p.130].

A typical journey through an medieval city might have worked like in the following part of a story. Here the protagonist looks for an old friend. To make this task not too easy this old friend has changed his name and his profession:

"I seek a man of the sea," said the prince. "A captain. His name is Jan Olvegg, or perhaps he is now known as Olvagga. Do you know him?
(...)
"Ask after Jannaveg the sail maker on the Street of the Weavers."
"This is how Jan is now known?"
The other nodded.
(...)
Dressed as a beggar himself, it did not seem out of place for him to stop and speak to the hunchback with the begging bowl.
"Greetings, brother," he stated. "I am far from my quarter on an errand. Can you direct me to the Street of the Weavers?"
(...)
"That way." The man gestured with his head. "The third street you come upon, turn there to the left. Then follow it past two streets more, and you will be at the Circle of the fountain before the temple of Varuna. Coming into that circle, the Street of the Weavers is marked by the Sign of the Awl."
(...)
He left the line and circled the fountain to where the Sign of the Awl hung upon a pillar. He moved up the Street of the Weavers.
Three [more] times did he ask after Jannagga the sail maker. (...) "
-- from Roger Zelazny, "Lord of Light".

City navigation in unknown areas in most cases is a process involving several sources of information. Those sources of information take the form of simply route sketches, route descriptions, maps, information in the environment, like the Sign of the Awl, and simply asking other people. This type of navigation is typical for tourists - they make use of tourist map, which point out sights and landmarks but often are distorted, hand-drawn maps from the hotel receptionist and crude route descriptions they get when asking people for the way.

People traveling an area only seldom will perhaps never acquire a coherent mental representation of the area and will almost always rely on diverse sources of information. Navigation as described in this section therefore relies mainly on knowledge about how to get additional information and how to use transportation to reach a certain destination.

Conclusion

These thoughts about navigation can be summarized into five navigational strategies in cities [PARU89], [BaBD93]):

• Identifier Strategy. This strategy associates a unique identifier or description with each entity of interest and permits to recognize it. By itself this strategy degenerates to exhaustive search.
• Path Strategy. This strategy gives a procedural description of how to reach the target. Depending on topology in a space there may be only a unique path to a destination or multiple paths.
• Direction Strategy. This strategy tries to avoid exhaustive search and cuts the search space into smaller parts. Whereas the path strategy uses references that have a local origin, like the possible exits at a certain location, the direction strategy uses global references.
• Distance Strategy. This strategy reduces the search space to a circle around a location. This circle is defined using the distance relation defined on the topology of the space in consideration. Every space allowing a path strategy also allows a distance strategy as distances can be defined over paths.
• Address Strategy. This strategy defines an orthogonal set of coordinates. These coordinates can be searched independently.

3.2. Elements of the city environment according to Lynch

3.2.1. Paths

Some paths are linear features connected to a special activity like a shopping mall and the like. This type of path not necessarily is a path traveled frequently. Occasionally path are important only because of their structural qualities, like their narrowness or width.

The identity of a path is distinguished by various sensory inputs - most of them visual. So a special facade characteristics can be a very important aspect for the identity of the path. Also a certain style of houses might give a path a unique feeling. Interestingly though planting might reinforce this image of the path, textures on the ground seem to be less important according to Lynch. Ground textures can enforce sections in paths and places however. Several examples of this are given in Cullen's "Townscape" [CULL91].

Proximity to a special place or feature of the environment often increases importance of a path. A good example is the Kärntnerstraße in the center of Vienna. It is close to the major landmark (see section 3.2.5.), St. Stephen's cathedral. It is a major shopping area and it is a place of social importance. Since car traffic has been banned from this area it has become an even more important meeting-place for people.

Directionality of paths

Paths commonly show directionality. A path by definition is a trajectory traveled to "reach a certain point". Often paths show directionality as a "gradient" in distinguishing features along the path. A gradient is a regular shape in a quality, which is cumulative into one direction. Those gradients are often topographical in nature but other example can be found in roads in old cities which get narrower towards the center.

A gradient sometimes allows estimating one's position in respect to the length of the path. In cities built based on a regular grid, streets often are numbered. In this situation the gradient is expressed in the numbering of crossing streets. A similar type of gradient is the numbering of houses. Often houses are numbered ascending as one leaves the city. This again is a feature often found in "naturally grown" cities with a strong center.

Such a numbering scheme additionally can show an absolute quality as in the case of many American cities. There each block starts with a number that is multiple of hundred. This form of gradient even allows to tell the exact number of blocks to walk to reach a destination.

Patterns in paths

Paths seldom are straight lines. They often follow linear elements in the environment, for instance streets. These tend to split, to cross other streets or to change direction. Paths using such streets lead to distortions in the mental representation of the path and to mistakes when estimating distances or alignment.

A crossing of paths form a node in the environmental image (see section 3.2.3.) These crossings are not only distinguished elements of the city image but serve as major referents along the path and provide distinctive structures. Path intersections are always points of decision. At those points it is important to visually support this decision and not to break the path or the gradient in the path.

An examples of how to enforce path continuity is the use of street names. Unfortunately street names often are changed after major crossings. Still the continuation of the path can be supported visually by directional cues, like the building structure, visual identity of the path, or by planting. When a crossing has more than four exits those cues must be strong enough to give the observer a distinctive feeling of path continuation to be useful.

Even paths in a regular grid are sometimes hard to distinguish. Examples again are the very regular centers of many American cities. Here often the street names (or rather numbers) are the only means of deciding which path one is on. A clear visual signal (vista) set at the end of the paths can help considerably to keep oneself oriented. An example is a street leading to an obelisk which visually is very distinct from other streets. A similar example is a set of streets leading to a central point of the city - like the streets leading to the "Arc the Triomphe" in Paris. Often such streets form a star-like, very centralized pattern (see figure 5).

Fig. 5 - A star-pattern of paths around the "Arc de Tromphe" in Paris.

Another frequent pattern is the circular street-pattern of naturally grown cities. Such a pattern can be seen in the older parts of Amsterdam (see figure 6 on the next page). About every third of these circular "streets" has a canal (gracht) in it. This distinguishes this "gracht" enormously from the other streets in the circles. There are three prominent grachts called the Herengracht, the Kaizergracht, and the Prinsengracht which in this sequence are outside of each other. They are major grachts and as the name indicates they have been places where people of different financial resources have lived. This shows in the look of the houses in these grachts and defines a gradient leading into the center. The gradient in this example forms not along a real path but rather in the sequence of circles crossed by the paths leading out of or into the center.

Vienna shows a clear circular pattern as well (see figure 8). As this pattern is very useful for navigating the city the old house-numbering scheme in Vienna showed the type of path one was in: Round house number signs are used in streets paralleling the "Ringstraße" which is a circular path around the very center. Streets with rectangular house number signs are perpendicular to the Ringstraße. Unfortunately this scheme is no longer used, but on many old houses the old signs still can be seen.

Paths can form a "melodic pattern", which has been described by Lynch:

There is a final way of organizing a path or a set of paths, which will become of increasing importance in a world of great distances and high speeds. It might be called "melodic" in analogy to music. The vents and characteristics along the path - landmarks, space changes, dynamic sensations - might be organized as a melodic line, perceived and imaged as a form which is experienced over a substantial time interval. Since the image would be of a total melody rather than a series of separate points, that image could presumably be more inclusive and yet less demanding. [LYNC60, p.99]

Fig. 6: Map of central Amsterdam.

This section describes the concept of paths, which are the most dominant elements in the environmental image. Paths are routes connecting locations and the first features learnt in an unknown environment. Paths make use of linear elements in the environment, for example streets. The importance of a path is increased when important locations are in the vicinity of the street it runs through. Paths commonly show a common character of buildings and directionality (gradients). These gradients allow people to estimate their location on the path. Many cities show typical street structures like stars or circular patterns.

3.2.2. Edges

Another typical example of an edges is a line of houses that look different from the other houses in the vicinity. Such a configuration acts as a strong visual border but normally does not form a real border as the street system still continues between those houses. Such edges represent seams along which information or material is exchanged between areas.

Edges, like paths, sometimes have clear directional qualities. People living in their vicinity will commonly refer to this directionality when giving directions. A typical example is the directionality of a river. Directions are likely to be given in terms of "upstream" or "downstream" near rivers.

Edges are the second linear element of the city image. Edges often are real borders in the environment. Paths can become edges when seen from a different perspective.

3.2.3. Nodes

The extension of a node can be small as in the example where two paths form a crossing. More often big place or squares act as nodes. When looking at environments of very large extensions, for instance nations or a continent even a whole city can act as a node. The observer heightens his attention at nodes and those places and nearby elements are perceived more intensely.

Nodes can be introvert or extrovert. Introverted places give the observer the feeling of "insideness" of the place or a feeling of "closure". Extrovert places also convey that feeling in a way but they provide a feeling for directionality.

Nodes in transport systems

Nodes play an important role in every transport system. Most transportation systems have predefined stops and thus can be seen as paths leading from node to node. An example for this are subway systems where the environment is invisible during the ride or where the observer can perceive only parts of it during the stops.

As subway systems are not bound to ground-level structures they are a detached world on their own that is connected to the "real city" by virtue of the subway stops. Therefore they are sometimes perceived as a second level of reality in the city. An example of how the user of a subway system still can be given a certain feeling for the environment she cannot see during the ride is realized in parts of the Boston subway system: Photos of sights near the subway stops are shown in the subway stops (see figure 7 on the next page).

These photos evoke a closer connection between the (underground) subway stop and the upper-level environment. Subway stops normally are placed at important nodes in the city which, because of the presence of the subway stop, gain additional importance.

Nodes are point-like elements in the city image although entire areas can be perceived as nodes. They often occur at the intersections of paths or edges and are positions where people heighten their attention as these nodes are points of decision. Nodes therefore are perceived more intensely. Nodes play an important role in urban transport systems as for instance subways are accessible only in certain spots in the city. These positions commonly form nodes in the city image.

Fig. 7: Pictures create a connection to the upper-level world in the Boston subway.

3.2.4. Districts

Interestingly districts are the only type of element of the city image that may be mainly non-visual. A harbor area is likely to have very distinct sounds and smells than typical railroad, airport or recreational areas. Districts can be distinguished by a functional concentration. An example is an area where there are many banks as in the example of a financial district or many shops of one kind as in the Street of the Weavers example.

Boundaries of districts

Many districts show a clear boundary in the form of a path or an edge, others exhibit a boundary in the form of a slow change in character. In that case the outline of the district is somewhat fuzzy.

Hard boundaries are very common in older cities, where they often have historical origins like city walls or the limits of the city at a certain time. Growth of the city beyond those boundaries frequently brought a significant change in building structure. This change heightens the contrast between the two adjacent districts. The former city walls and green-belts in front of those walls often become rings of parks or important circular paths around the "old city". In this case the separation between the two districts becomes even clearer as it consists of both a change in building structure and of a circular path (see also section 3.2.6.).

In case the boundary of districts shows a strong directionality there is often a second path or edge close-by that features the opposite directionality. An example is the Viennese "Ringstraße". It is in the position of the former city walls and is paralleled by the "Lastenstraße", which is located at the former end of the green belt (Glacis) between the city walls and the next buildings. Today both of these streets are partly one-way streets leading around the city in opposite direction. The former green-belt between the two streets has been filled with representative buildings (parliament, town hall, museums, university,...) most of which are built in the typical "Ringstraßen-Stil", a style of architecture which makes this narrow ring a special ring-like district in the city of Vienna.

Fig. 8: Map of Vienna.

In Vienna there are two strong circles around the city (see figure 8). The inner circle borders the old city, which is the first district. (Here the word "district" signifies the bureaucratic division of Vienna in 23 districts. In most cases these districts form also "districts" in the mental image of the city - especially as it is very common for Viennese people to give directions according to these districts. The reason is, that these districts are not purely bureaucratic in origin but that most of them were villages, that were integrated into the city as the city grew beyond its former limits.) Around this area there is a circle of districts bordered by the so-called "Gürtel" (= belt). The districts inside the Gürtel are arranged linearly along the Ringstraße so there is a clear directional quality in this pattern of districts too. Outside the Gürtel the districts starting with 10 and going up to 23 are arranged - wrapped around the city like a shell. A third ring is formed by larger streets surrounding the city near its border.

The major roads that once lead to the villages, which later became the districts, are still recognizable in the names of major streets leading out of the center. The names of those streets act as reminders of the districts (villages) they are leading to. Examples are the Margaretenstraße or the Ottakringerstraße. Similarly other major roads lead out of town to destinations farther away - like the Triesterstraße that goes into the direction of Italy to the city of Trieste - the former Adria-harbor of Austria.

Districts are larger areas in a city which often consist of several other city elements. Most districts have a well defined boundary consisting of a path or an edge. Buildings in a district show common character or functionality. In grown cities the pattern of districts normally has historic reasons.

3.2.5. Landmarks

Landmarks are of special importance when giving directions to other people since they are so easily recognizable. Path descriptions often rely on landmarks. In route descriptions it is very common to describe a series of landmarks. Encountering one of them triggers attention for the next in the series. This sequence gives the traveler a feeling of progress and allows to estimate distances in the description. Landmarks often occur as starting or ending points of major paths or in the center of districts.

Landmarks are easier to recognize if they are prominent against the background, if they have a clear form and are prominent in location. Prominence makes a landmark visible from a greater variety of places and directions.

Lynch mentions that a high landmark can be perceived as being bottomless. This means that when standing in front of a tall house people are sometimes not immediately able to associate this building with the skyscraper which can be seen from far away from several points in the city.

Typically landmarks vary widely in scale - they may be mountains, the sun, a golden doorknob that blinks in the morning light or the tower of a church. The concept of scale in use here is the scale as defined by Montello (see section 2.2.3.)

When occurring as a pair landmarks can give directional information. A famous example is the Duomo in Florence with its campanile (bell tower). The relative position of these two landmarks is a strong indicator for direction. Another example of paired landmarks is the World Trade Center in New York City. However as the World Trade Center consists of two seemingly identical towers it does not give as clear directional information as the Duomo in Florence.

3.2.6. Combinations of city elements

The conjunction of several strong elements in a city has not only additional but even synergistic effects for the city image. A typical example is the Viennese Ringstraße mentioned above. A combination of edge and path in the same objects - forms a much stronger element than the path or the edge in isolation. An example for such a combination is the "Great Wall" in China.

The discussion further described the city elements as clear cut entities. This is not always the case. For example the image of a node can be abstract or clear. In some cases people can mentally "see" the element - in other places it is mainly an abstract referent like "a gas station", a big yellow house and so forth.

The connections between several elements in the city image can be either a loose one, or the elements have a positional relation - a relation which Lynch calls the structural quality. Structural quality is based on the multimedia representation of the environment in the city image and is a source for distortions of the city image.

In the case of two cities growing till they form one bigger city there is a very strong and fixed internal structure in each of those cities. New structures building between the cities will be rather flexible in the beginning and slowly get static as the number of connections in the image increases.

Designing for a clear city image

Learning a city image is a long and complicated process. How can the design of a city aid travelers in learning the city image? Is there a certain set of guidelines that tells how paths shall be designed and cities laid out to give the observer a sense of the whole structure - to help building an image?

Indeed there is such a set of "guidelines". The most important rule is to heighten the visual identity of city elements and to structure the environment in a clear way. Above all paths help getting order into chaos. They act as keylines that should show a singular quality. A visual hierarchy of paths and edges has to be provided. Paths and edges should exhibit a sense of directionality.

Breaks in transportation should be made special places so they are perceived as strong nodes in the image. The entering and leaving of the node must be clearly communicated to the traveler. "Where physical homogeneity coincides with use and status the effect is unmistakable" according to Lynch [LYNC60, p.104].

Lynch summarizes these and other rules in a set of 10 guidelines:

1. "Singularity or figure-background clarity and sharpness of boundaries". The contrast may be to the immediate surroundings or to the observer's experience. This guideline aims at the qualities of city elements which make them identifiable.
2. "Form simplicity". City elements should strive for clarity and simplicity of visible form in the geometrical sense as these forms are more easily incorporated into city image. Lynch points out that there is evidence, that observers tend to distort complex forms to simpler ones.
3. "Continuity" in edges or surfaces, nearness of parts, repetition of rhythmic intervals. These qualities ease the perception of a "complex physical reality as one or as interrelated".
4. "Dominance of one element in an ensemble" translates to a cluster of elements grouped around one major element. This quality, like continuity, allows simplification of the image. Lynch calls this process "abstraction".
5. "Clarity of joint" means high visibility of joints and seams, for example at major intersections. Such joints are the "strategic moments of structure and should be highly perceptible".
6. "Directional differentiation" means that city elements should exhibit directional qualities. These qualities differentiate one end from the other and are very useful in structuring on a larger scale.
7. "Visual scope" are qualities of the environment which influence the range of vision - be it actually or symbolically. Examples for such constellations are transparencies, overlaps, vistas, panoramas and several others. According to Lynch these qualities help grasping of larger complex wholes by "increasing the efficiency of vision".
8. "Motion awareness" argues that actual and potential motion shall be made explicit to the observer and that the quality of motion shall be made "perceivable" to the observer by visual or kinesthetic means.
9. "Time series", for instance sequences of landmarks, are series of elements which are sensed over time. An example is a series of landmarks to be encountered on a route.
10. "Names and meanings" are non-physical characteristics which may enhance the imaginability of elements. They sometimes give clues about the location (like in "North Station") or trigger historical, functional or economical associations. Examples are the Vienna districts which still hint at the names of the former villages, or the "Street of the Weavers", examples which were discussed in more detail above. Such non-physical characteristics facilitate structuring of elements in the environment.

Conclusion

Lynch describes five elements of the city environment and calls them path, edge, node, district and landmark. These elements are the constituents of the "image of the city", which is an early model of spatial memory for cities. The most important elements for learning the city environment are the path and the landmark. City elements seldom occur in isolation. The coexistence of several city elements in a location can heighten the overall significance of that place enormously. Lynch defines a set of guidelines useful for designing city environments that are easier to navigate and easier to learn. The most important of these is to provide singularity of figure-background clarity. That means that locations in the city need unique appearance to be easily recognizable.

3.3. Communicative and social aspects of the environment

The landscape plays a social role as well. The named environment, familiar to all, furnishes material for common memories and symbols which bind the group together and allow them to communicate with one another. The landscape serves as a vast mnemonic system for retention of group history and ideals. [LYNC60, p.126]

Space as an ordering principle

Space is a strong ordering and organizing principle and therefore is used by many people as a mnemonic system. This use was known already by the Greek as "mnemotechnics" or the Art of Memory, and used primarily to memorize long pieces of text. Speakers mentally created house and walked though it. Putting various objects like pictures or statues at distinguished places in the house structured that space. At those objects pieces of text were mentally attached. When giving the speech the speaker mentally walked the house and thus was able to recall the pieces of text when looking at the objects. The first written proof of this technique stems from Cicero in his "De Oratore". The Art of Memory and mnemotechnics was further developed till the renaissance [YATE66].

3.3.1. The environment as information carrier

3.3.1.1. Knowledge in the world and Writing on the world

Affordances

Objects in space give information about their usage and their use. Many objects are formed in a way that it is evident what they are meant for. This information has been called "affordance".

What we see as the behavior of the system, object or event is that which is afforded or permitted by the system, object or event. When the affordances of an object are perceptually obvious, it is easy for us to know how to interact with it. Conversely, when the affordances are less obvious or ambiguous, it is easy for us to make mistakes when trying to interact with the object. [PREE94, p.81]

An example is a door that can be opened into one direction only. It is possible to design a door handle to clearly show if the door has to be pushed or pulled to be opened [NORM88] - the handle either "affords" pulling or pushing. The affordance in this example comes from physiological properties of the human arm and hand. Affordances can also stem from size and form qualities. Certain keys will not fit into certain keyholes because of their size and form for example.

Such affordances are knowledge about objects that is present in the environment. Norman calls this "knowledge in the world" [NORM88]. This thesis sees the concept of "knowledge in the world" as even more general - the age and functionality of houses and every object in the environment is knowledge in the environment, that is presented by the objects themselves. Also this information is "knowledge in the world."

Writing on the world

Writing on the world is a concept to enrich the environment with the sort of information described above. Initially this term was coined by Jay Bolter [BOLT92] as a concept to enrich virtual environments (see section 5.1.5.) but it is equally valid in real environments. Writing on the world wants to point out that the term "writing" is not restricted to text but that all information in the environment constitutes "writing" [BOLT91] - be it pictorial writing, textual writing or "writing" using affordances - that is: writing using the design of objects.

The writing is "on the world" as every surface or even every object can act as information carrier. The writing therefore is not simply on a surface but really on the world. Writing on the world is not only a different way to look at information contained in the environment: it advocates enriching environments and spaces with information as this information richness makes navigating and using environments easier. An example of this writing on the world was described already in section 3.1.: The Sign of the Awl is a typical form of pictorial writing on the world. As an abstract representation of information it requires the user to possess reading skills but this is equally true for every other form of writing and reading.

Such a writing can be found in many cities. An example are the old gracht houses of Amsterdam. On those houses it was common to place a gable stone featuring a graphic description of the "functionality" or "contents of the house" (see figure 9a). The system with the gable stones was so effective for navigation, that it was not necessary to have street numbering in the medieval city. When the French invaded Amsterdam they rapidly got lost in this maze of buildings with strange symbols on each house because they did not possess the necessary reading skills.

Fig. 9a: Writing on the world (gable stone).

Fig. 9b: Writing on the world (guild sign).

Similar concepts were used in many other medieval cities: "Zunft-zeichen" (guild-signs) point out shops belonging to a guild, for instance the guild of the shoemakers (see figure 9b). Today pictorial writing becomes important again in advertising, in the signs of companies, the look of restaurants, subway stops and so forth.

3.3.1.2. Unmistakable form

Looking at architecture this unmistakable form is a typical quality of landmarks (see section 3.2.5.). Landmarks can be seen as an extreme example of "writing on the world". They give information about their location. Landmarks must have an identity and must not be confusable with other objects in the vicinity.

Unmistakable form in architecture relies not only on the form of the object itself but on its surroundings: There may be several similar houses in a city - but if only one of them is standing on a hill such that it can be seen from far away this locational aspects make the house special and therefore unmistakable. Typical landmarks have a distinct look which separates them from every other object in the environment. A good example is the "Atomium" in Brussels (figure 10).

Fig. 10: Unmistakable form in architecture.

3.3.1.3. Navigational infrastructure

• infrastructure serving as information carrier.
• infrastructure serving as transportation.

This type of information or navigation infrastructure gives very localized information. There are other types too but these are present mainly at special places in the environment. Examples are information kiosks, tourist information offices, street maps provided in subway stops and so forth. Travelers in an environment can take it for granted that such information infrastructure is present in major navigational nodes in the environment - like train stations, airports and major subway crossings.

People know that when they reach an important node in their environmental image of - say - a continent that there will be infrastructure available to help in switching from the geographical-scale to the environmental-scale image. They rely on the second type of navigational infrastructure to be present at such places.

Transportational infrastructure

Transportational infrastructure allows people to reach a destination without having to learn the environmental image of this area. They want to reach a destination and if they need to reach it only once they do not even want to learn the environmental image. A typical example for such infrastructure is a taxi. It is a navigational tool that transports people to certain places, like a hotel, without demanding any local knowledge from the user.

Other examples of such infrastructure require local knowledge: subways, tramways, buses and so forth. The necessary navigational knowledge to use them in order to reach a destination is limited however and it is very common that information obtained from the first type of navigational infrastructure is sufficient to reach a certain place using transportational infrastructure.

3.3.1.4. Change in the city environment

Change in a city tends to leave the overall city image intact. This is possible because the image does not rely only on single objects and locations, but is more a network of places and because a location is not identifiable by only one quality but by several (see section 3.2.6.).

For example if St. Stephen's cathedral in Vienna would vanish, one of the most important landmarks of Vienna were gone. The city would loose its clearly marked center and much of its prestige. Navigation still would be possible because of the synergistic effect of many strong elements in the same place. There still would be one of the most important subway stops and many important meeting-places and shopping areas. After a long time people still would speak of the "Stephansplatz" and sell postcards with pictures of St. Stephen's cathedral there.

The city image therefore can live well with change if this change is not too fast and not too global. Normally change does not involve landmarks as in the example above, because these are culturally important to a city and therefore preserved. Changes in ordinary houses do not influence the character of an area as strongly and a large number of houses has to be changed in an area before it looses its character.

In an "old city" almost all old houses would have to be replaced to loose the character of an "old city" because otherwise the narrow streets never could be opened. Established city images tend to be robust. This robustness can be enhanced with a sensible naming scheme.

Catastrophic events

Greater changes in a city can happen as well but these changes happen seldom and are caused mainly by catastrophes. An example is the center of Salzburg where almost the entire center was demolished and rebuilt using much broader streets and larger houses. This changed the character of parts of the city enormously - what was an medieval city became a "modern" baroque city. Still the center of Salzburg today looks "old" to us because what was considered a "broad" street several centuries ago is considered a small lane today. Another example of such a restructuring is London after the great fire of 1666 (described in [LYNC72]).

As such changes in a city changes the character of an entire area this area often is seen as one element of the city image - it becomes a district or at least a node. Often the boundaries of such an area can be seen in the environment a long time after the gradual change in the environment has already leveled the difference between the new district and the surrounding areas. The reason for this is that gradual changes are localized - in a circular building structure consisting of several houses the demolition and rebuilding of only a few houses cannot destroy the circular structure. An example for such a structure is a place in Rome where a theater was demolished several centuries ago. Still the principal structure of the former theater shows in the building structure even today (see figure 11 on the next page).

Changes in the environment can lead to navigational problems because the city image suddenly is wrong. In real cities the city elements occur in combination and real cities tend to change slowly. The environmental image of well designed, grown environments therefore tends to be rather stable.

Fig. 11: A former theater in Rome influences the building structure [LYNC72].

3.3.2. Communicating about the environment

Those and other aspects of describing space textually and verbally have been pointed out already in section 2.2.3. which looked at descriptions of space and routes. After describing the elements of the city image it should be evident that communication about routes relies mainly on describing the qualities of landmarks and objects encountered. The richness of the environment allows users to check their position easily and to realign themselves according to the description using knowledge in the world or navigational infrastructure.

It was also pointed out already that people tend to ask other people for the way and that navigation thus becomes an iterative approach to the final destination.

Communicating about the environment is another way of retrieving navigational information in the environment. As cities never are empty places this "asking for the way" is a common process in city navigation. It frequently occurs to reassure oneself that one is still on the right path. Descriptions of the environment depend on well identifiable features in the environment - for instance landmarks.

3.3.3. Social aspects of space

Although social aspects of space are not the main focus of this thesis they should be considered as being an additional - sixth - element of the city image which may well be at least as strong a structuring principle for environments as are paths and landmarks.

3.3.3.1. Giving meaning to place

A typical example for a place is a meeting place. Such places are designed to stimulate social activities by making them comfortable and providing recreational facilities and so forth. Such places, like a pub or a place where a bulletin board can be found, are always places where people go to meet other people. They often are providers of informational infrastructure too. They sometimes have a very distinct look and feel and since it is a place that many people frequently visit it often is a node and landmark and it certainly structures the vicinity.

Social activity can be catalyzed by spatial constraints [ERIC93]. An example is a red traffic light at a pedestrian crossing which will make people halt in their walking. While waiting most people examine the surroundings more closely. A object or information carrier or piece of art in such a location is bound to be noticed and stimulates communication and thereby social action.

A piece of art can distinguish a place not only at a pedestrian crossing. Erickson describes how a bronze sculpture in a Seattle railway station made this station a special place where people even interact with the piece of art and each other though the medium of the sculpture. Residents there sometimes even make detours and (...) go by the sculpture to see what's happening. [ERIC93, p.399]. Such objects not only encourage interaction with themselves but they are able to catalyze interaction between people. The environment thus becomes a catalyst for social interaction, and for interaction with the environment. An other example for such a piece of art is a sculpture found in an airport (see figure 12).

Fig. 12 - Public art in an airport.

This "giving meaning to place" is also a concept used in the art of mnemotechnics which was described at the beginning of this chapter. There this term has no social implications however.

This section describes the concept of "giving meaning to place" according to which social interaction with a space heightens the importance of a location. This principle could be seen as a sixth element of the environmental image.

3.3.3.2. Cultural habits, taboos and private space

An other example are doors. This particularly interesting example has been studied by Maria Catedra:

The door as a motive can be seen as a marker for a physical limit, an abstract concept, an image for a social entrance, a metaphor for time, a symbol for spiritual passage or cultural domination, or ideology about women. In some examples the existence or absence of a door is the point of interest. The door becomes a highly-charged, condensed, and thick idea. [CATE91, p.53]

The door thus is a catalyst and restrainer of social interaction and therefore one of the major building blocks to structure (social) spaces. Alexander points out that:

The movement between rooms is as important as the rooms themselves; and its arrangement has as much effect on social interaction in the rooms, as the interiors of the rooms. [ALEX72, p.628]

The concept of the door therefore becomes a fitting example of cultural differences in social constraints may structure space and vice versa:

But while the predicative notion is shared by everybody, the expression drastically changes according to cultural norms. A piece of cloth may close more effectively than a door closed with seven keys; for example, the purdah: a thin curtain, sometimes veil which in India divides the part of the house where woman are secluded. Cultural concepts may form limits as impenetrable as stone walls. [CATE91, p.55]

The door is an important concept in religious sites where they again act as strong catalysts for social action. In the religious context the door is an example of a object that is used metaphorically and thus gets yet another, slightly different meaning:

What is a church or temple? It is a place of worship, spirit, contemplation, of course. But above all, from a human point of view, it is a gateway. A person comes into the world through the church. He leaves it though the church. And, at each of the important thresholds of his life, he once again steps through the church. [ALEX72, p.332]

As these social aspects of spaces and objects give such a fitting structuring principle Alexander recommends to make use of this principle to structure space and thus make people more comfortable using those spaces. For example:

In each community and neighborhood, identify some sacred site as consecrated ground, and form a series of nested precincts, each marked by a gateway, each one progressively more private, and more sacred than the last, the innermost a final sanctum that can only be reached by passing through all of the outer ones. [ALEX72, p.334].

Private space

This religious structure is similar to the structure of the private space which is structured into nested and shielded areas of progressive privacy as described by the "onion model" of privacy. This model distinguishes 5 spheres (levels) of privacy [FLEI94]:

• The secret sphere
• Intimate sphere
• The Trust Sphere
• The Social Sphere
• The Public Sphere

The important consequence of those thoughts is that the structure of a space follows social spaces and that no space used by more than one person can fulfill both social, work and navigational requirements if the design of this space does not consider social issues - be it the size of the ideal work group, the flow of information and material, the need for common ground or private space and many other issues.

No building ever feels right to the people in it unless the physical spaces (defined by columns, walls, and ceilings) are congruent with social spaces (defined by activities and human groups). [ALEX72, p.981]

These issues are both valid for smaller, closed spaces like single buildings as they are valid for large (open) spaces like cities and landscapes.

Conclusion

This section shows the importance of communicative and social aspects of space. There are several concepts to enricht an environment. A richer environment is easier to navigate because the environment provides information useful for navigating and for using objects in the environment. Affordances are a typical example of such knowledge in the world. They show how to use objects. Writing on the world is a concept to use the world as a surface to write on. Examples of this are gable stones or guild signs. These pictorial writings enhance the environment with navigational information. Social issues were described in this section as well. Social space follows real space. A real space cannot be really useful if it is not congruent with social space.

3.4. Summary of the city environment

• Structure-giving objects and functions.
• Information providers and information carriers.
• Transportation.

3.4.1. Structural objects and functionality

At this level of detail identifiable objects in the city are:

• Major navigational paths and edges. Paths and edges seldom show explicit directionality at this level of detail. Paths show an identifiable starting point and an ending point also in this level of detail.
• Crossings of paths and connections of paths should be considered important nodes in the overall city structure.
• Districts are areas separated by major paths. Districts of this size often are bureaucratic or historic in origin and not necessarily show unique character.

The next level of detail shows single buildings. This level is the level of landmarks and architectonic properties of single buildings. Specialized buildings are identifiable in this level - like subway stops, churches, and other functionality like a railway station.

At this level of detail it gets interesting to look at the structure of single objects and at operations supported by those objects:

• Paths connect two locations in the city. Therefore they have a starting point and an ending point - both of which are distinct locations in the city. Along the path there may be additional distinct locations - like crossings with other paths, squares or buildings. The path is a continuous element - it is possible to stop on a path at every point.
• Lines connect two locations in the city. Contrary to the path the line is a discontinuous element. Lines can have several distinct stops but it is not possible to stop at every point on a line. Lines are used by public transport like subways or buses.
• Intersections of paths sometimes are squares. Squares often are major elements in the city. They often contain stops of public transport systems and thus are connected to lines. They sometimes contain landmarks.
• Building: A building is a container. Buildings have a unique address. As every container buildings can provide open access, restricted access or no access. Buildings consist of walls which - on the outside - are called facades.
• A facade may contain signs and writings which provides information about the building. The facade can contain pathways into the house or into substructures of the house. Doors may be open or closed (see below). A facade often contains signs and writings. The facade often gives information about the age and state of the building. Thus the facade carries information about the building and its contents.
• Some buildings are landmarks: They can be seen from far off. In case of paired landmarks they provide directional cues. In addition to the properties of the building the landmark gives information about absolute position inside the city structure. When its approximate size is known it can give distance cues. Many landmarks provide no access - they can be seen as special as they do not provide the container functionality. Examples for such landmarks are monuments.
• Some buildings provide functionality to everybody - examples of these buildings are railway stations, churches and so forth. These buildings often are starting or ending points of paths and lines. This type of building is often found in a square.

Structures inside buildings

On the outside buildings consist of a facade and openings. The inside of buildings is highly structured.

• Similar to the path in the city there are hallways (flat paths) and stairs (paths leading up or down). Hallways and stairs have walls which may be the inside of the facade or separations from other elements of the building.
• Similar to lines there are elevators. Elevators connect distinct locations. These locations often are hallways or lobbies. Some elevators provide access to transportation systems located or parked below the building. Elevators can provide restricted access to transportation facilities on top of buildings - like helicopters.
• Lobbies have a similar function as squares in the city. They commonly connect hallways, stairs, and elevators. They sometimes are locations of landmark status inside the building.
• Rooms are locations inside buildings. They consist of walls. The walls may contain doors or windows. While the other elements in the building provide mostly public access, as long as the building provides public access, rooms again may show all three kinds of access: public access, restricted access and no access.
• Doors connect buildings to the city if the door is in the facade or to other elements of the building - like hallways, lobbies or other rooms. Doors can be opened or closed. Closed doors may be locked or unlocked. The state of the door gives a hint on the accessibility of the room reachable through the door. Some doors can be opened only from one side.
• Windows are similar to doors. Most windows are located in facades. They are not used to access rooms but can be used as rapid connection from a room to the path or square in front of the building. They provide visual access to the outside of the building and some windows provide visual access to parts of the inside of the building (shop windows). Windows can be open or closed. Windows normally can be opened only from the inside of the house. Some windows cannot be opened at all. Windows sometimes provide direct visual access from one building to the facade of another building.
• Walls separate rooms, hallways and other elements of the house. An outside wall may be the inside of a facade. In these walls windows can be located. Some walls contain doors. Walls may contain information providers like pictures, pinboards, shelves and so forth.

3.4.2. Information providers

• Information in the world.
• Information in other people in the environment.
• Information in the head of the traveler herself.

• Facades provide information about the contents of a building as was described above. Facades can contain doors and signs.
• Signs are a general form of an information provider. Signs can be read. They can be attached to facades or walls or can be located on their own in the environment. Some signs are well visible all day, others are unreadable in the dark. Signs can change their appearance repeatedly - these signs provide rapidly changing information, like departures in a train station. Signs are normally as accessible as the location they are contained in. An exception are signs in a room, which is visually accessible from the outside of that room. Signs cannot be removed.
• A special form of signs is the traffic sign. It is an information provider used to provide navigational information to cars, taxis and buses.
• Sometimes information is bundled in a kiosk. Kiosks are public access special purpose rooms that often contain many signs. They can contain a person whose task is to provide information. Kiosks take the form of tourist information offices and so forth. Kiosks often provide information that can be removed, for example maps or guidebooks.

3.4.3. Transportation

• Walking is a slow form of transportation for rather short distances. Walking can use paths, squares and all open access house structures.
• Cars use paths and squares in the city. Cars typically are restricted access vehicles. When leaving the car to continue transport by walking or public transport an available free space has to be found. Sometimes parking lots are provided near major stops of public transport systems or in buildings. Using a car requires complete navigational information about the location to be reached or communication devices or information providers to provide this information during the ride (traffic sign).
• Taxis are a special form of cars. For a fee they provide public access and can navigate using incomplete information. Taxis are path based transports - therefore they ideally can be entered and left at every point on the path. Taxis do not require a parking lot when the transport is continued using another transport system.
• A bus is a special form of a car. For a smaller fee than the taxi they provide public access. Buses navigate along lines which coincide with paths in the city. As line-based transport entering and leaving the bus is possible only at predefined stops. Contrary to other forms of line-based transport buses are able to navigate around obstacles within the restrictions imposed by the path.
• Also subways and railways are line-based transports They navigate along lines which do not coincide with paths but which sometimes intersect paths. They provide public access and are not able to navigate around obstacles. Stops of these transports are often located near major locations (squares) or landmarks in the city. As they use lines that are independent from the overall city structure they sometimes define a separated space overlaid over the city structure. In the case of a subway system which travels without visible connection to the city features this separate space is sometimes perceived as another level of reality in the city.

3.4.4. Communication

• The telephone provides information and communication with other locations in the city and beyond without delay. Telephones can provide navigational information during travel.
• Mail provides slow communication or transport of objects in the city.

3.4.5. Limits of this environment

This property limits transport and information provision in the city. For example it is not possible for a room or house to be placed in a way that it is adjacent to two houses located on opposite sides of the city.

The only object in the city that can overcome this limit is the communication provider. Information using the telephone can travel at speeds which allow instantaneous information travel. The subspace of information therefore can overcome the Euclidean limitations of the city structure. Information is essentially immaterial at the current state of technology. This means that no transportation system in the city structure can overcome this limit.

Line-based transport systems like the subway seemingly manage to distort the perceived geometry of the city as they provide more direct and therefore shorter connections through the city structure than do paths.

Cities are real world constructs and are therefore limited by real world physics. Virtual environments can define spaces which are independent from these restrictions. Those spaces therefore can overcome the inherent limits of Euclidean space.

3.5. Conclusion

Lynch's "image of the city" is a mental model consisting of five elements, the most important of which are the path and the landmark. Landmarks rely on "unmistakable form" to be recognizable. City elements seldom occur in isolation. Combination of city elements makes navigation easier and leads to a city image which is stable even when parts of the city are changing. Lynch gives a set of guidelines for city-design which help designing an easy to navigate city. The most important of these guidelines is the "clarity of figure-background". It essentially means that all city elements should show typical, or even "unmistakable" form. Social aspects of space can be seen as a sixth city element. Real spaces that are not congruent with social spaces can never "feel right". The environment serves also as information carrier. Objects in the environment often provide information about their use and about other objects. They can be seen as informational and navigational infrastructure in the environment.

The last section identified objects and operations that are typical for the city environment and listed them in a detailed ontology. City objects can be classified into structural objects, information or communication providers and transportation. While the city structure essentially is an Euclidean space the subspace of transportation system can form a topological space inside the city structure. However even this topological space is not entirely independent from the overall Euclidean structure of the city. Only transportation of immaterial objects (information) can overcome the limitations of Euclidean space in a real city.