Towards a Theory of Sustainability andUrban Quality

A New Method for Typological Urban Classification

Johan Rådberg

Planners and politicians are searching for action plans to increase the environmental quality and sustainability inthe large cities. But the search is hampered by the lack of theory. The confusion over the issues of quality andsustainability stems from the fact that theories are formulated on a general and abstract level. What is needed is atheoretical framework to analyse the empirical observations on the existing urban environments. We need a sys-tematic descriptive classification of the urban structure on the microlevel.

The paper outlines a methodology of classification of Swedish urban types. The classification is based on pa-rameters that can be obtained from central data bases. The registration of objects will be implemented in a Geo-graphical Informations System (GIS). The result is (a) reliable descriptions (maps, data bases) of the existing ur-ban structure in a given city and, (b) in the long term, new knowledge of the relationship between urban form andenvironmental capacity.

Keywords: GIS-system, sustainability index, urban density, urban quality, typology.

INTRODUCTION — THE LACK OFTHEORY

Urban planning is in a crisis. Planners and poli-ticians all over the world are aware of the urgentneed for action plans to increase the environ-mental quality and sustainability in the largecities. The campaign for improved urban affectsa broad field of action: influencing the attitudesand lifestyles of the inhabitants as well as devel-oping alternative transportation systems, waterprocessing systems, waste processing systems, aswell as changing the urban structure, the landuse pattern etc. It is clearly a formidable task,both in terms of time and money needed. Butthe greatest obstacle is probably the lack of anadequate theory.

Everybody will agree on the need to plan forthe sustainable city. (See, for example, GreenPaper on the Urban Environment, 1990). Whatwould the ideal sustainable city look like? Thereis no definite answer today. On one hand, wehave the proponents of the compact city. On theother, we have the proponents of the the greencity. The first implies a strategy of concentra-tion and increasing urban density (in order tominimize the need for transportation). The sec-ond implies a strategy of deconcentration andspreading out, using the unbuilt land for local

water infiltration and cultivation. So we are con-fronted with two strategies — seemingly irrec-oncilable — for the sustainable city. This maybe called “the density paradox”.

The issue of urban quality and urban form isno less confused. The leading theorists of urbanplanning and design during the twentieth cen-tury (for example, Le Corbusier, Raymond Un-win, Lewis Mumford, Jane Jacobs, Kevin Lynchand Christopher Alexander) argue in favour ofdifferent ideal solutions: high-density develop-ments or low or medium density developments.They argue for different urban models: someadvocate high-rise buildings and large openspaces, others argue for traditional grid-ironplans, streets and compact blocks, still others ar-gue for small-scale garden suburbs. There isevidently no consensus on the question of ur-ban quality and urban form.

To sum up: there is a great confusion in thefields of quality and sustainability in urbanplanning. Much of this confusion stems fromthe fact that the theories are formulated on avery general and abstract level. We need empiri-cal facts, observations. Above all, we need atheoretical framework for these empirical ob-servations. We need a systematic descriptiveclassification of the urban structure on the mi-crolevel in order to be able to process the ac-

2 IAPS 14 Book of Proceedings

cumulated information on existing urban envi-ronments.

Sustainability and the density paradoxLet us first look closer at the density paradox;the opposing strategies of the compact city andthe green city. What are the theories behindthese strategies?

The obvious merit of a compact high-densitycity would be the short distances from the cen-tre to the periphery. Short distances would (atleast in theory) mean less transport of goodsand people and this in turn would mean a greatsaving in terms of time and energy, as well asless pollution from car traffic.

The theory of the green city is the following:Many of the environmental problems of todayare caused by the lack of local recycling. Thesolid and fluid waste from the cities should notbe flushed into the sea, burnt or dumped on acity waste dump. It should rather be recycled; inparticular the organic waste ought to be broughtback to fertile soil. This means we must providemore land, unbuilt land, for local infiltrationand recycling of waste water and also forplanting and cultivating. The resulting strategyseems to point towards a less dense, more spreadout, city structure than today.

Both theories relate to general theories, re-lated to abstract models of cities. The compactcity or the green city are both stereotypes, theydon’t exist. In reality, cities are not homoge-nous. The urban density is not uniform. Themodern city is an aggregate of small fragments,districts, blocks which differ from each other, interms of urban density, in types of buildings,urban quality etc. We can identify some recur-rent types of developments: the urban core,early twentieth century villa suburbs, the post-war residential enclaves, the industrial enclavesand institutions.

The choice between the compact city or thegreen city is a fictitious choice, between twostereotypes. In reality, we have to consider citesas complex aggregates of urban elements on themicrolevel.

A general policy for urban sustainabilityshould consist both of strategies for reducingcar traffic and strategies for increasing local re-cycling and cultivation. The strategies shouldfocus more on the local level — the micro-scale— and less on the macro-scale. One reason isthe obvious fact that the existing urban struc-tures are very permanent; it is very hard tochange the macrolevel. However, on the mi-crolevel, — the urban block level — strategicadjustments (rebuilding, restructuring of theground) are much easier to implement.

The problem of quality in the urban envi-

ronment is also primarily related to the mi-crolevel, the individual districts or urban blocksin the city. In every city, there are more attrac-tive and less attractive residential districts. Weknow (and recent research has furnished proof)that urban form is an important factor.

In many industrialized countries, the post-war large scale housing estates are the most af-fected by problems (like hard-to-let apartments,negative social segregation, high levels of van-dalism etc). Why are these areas so problem-ridden? Recent research suggests that the degreeof vandalism and other social ills is correlated tocertain features of the urban design: the numberof storeys, the number of flats in each building,the relation of the buildings and the entrancesto the streets and the surrounding open spaces.(See, for example, Oscar Newman, 1973; or Al-ice Coleman, 1985).The key to understandingthe social ills in the large-scale post-war housingestates is the study of the urban form (urban de-sign) on the microlevel, the urban block.

The problem of urban sustainability, qualityand urban form should, therefore, be focusedon the microlevel, rather than on the mac-rolevel. The statistical averages for a whole cityor a whole region are not very informative andmay even be misleading. We need a theoreticalframework for a study of the small elements inthe urban structure, the urban blocks, or groupsof buildings.

TOWARDS A THEORY OFTYPOLOGICAL CLASSIFICATION

This project proposes that a new micro-level ty-pological classification is the sort of theoreticalframework that we need. The project is abouthow to develop such a typology. Before wemove on to the practical methodology, weought to comment on the meaning of typology.

Typology is by no means a novelty as such;it has been a part of the common practice in ar-chitectural analysis and design for a long time.However, we have relied almost entirely onbuilding typologies and these typologies havebeen based on the functional content (see forexample the works of architectural historianNikolaus Pevsner). According to functional ty-pology schools, hospitals, museums, residentialbuildings are studied as separate categories ofbuildings. Although functional typology maybe quite productive for the design of buildings,it is much less useful for urban descriptions andurban analysis. Functional typology does nottake into account the immediate surroundingsof the building, the context. Nor does it takeinto account that functions often change overtime, while urban forms are very stable.

A typo-morphological urban analysis (as

Johan Rådberg, Sustainability and Urban Quality 3

opposed to the functional typology) means thatbuildings are studied in context, together withthe surrounding public and private spaces. Theobject of such analysis may be a group ofbuildings and open space (an urban block), thebuilding lots, or the street-pattern.

Typo-morphological urban research can betraced back to the 1950s and 1960s. At thistime, Italian architects were systematicallystudying building types in historical city centresin Italy. At the same time, British geographerM.R.G. Conzen studied the historical evolutionof British medieval cities, from a somewhat dif-ferent standpoint. Conzen developed a typo-logical (or typo-morphological) study of theurban forms, the street-net, the urban blocksand the individual plots. Similar research hasbeen done in a number of European cities, andin San Francisco (see Moudon, 1986; Lawrence,1986). These historical studies demonstrate thaturban elements (the urban block, the street pat-tern) are very stable over long historical peri-ods. They also demonstrate that functions canchange to a surprising extent within a traditionalurban pattern, without breaking the pattern.

Typomorphological urban research is stillonly in the early stages; in the future it can beexpected to develop along several lines (see alsoMoudon, 1987):1 As a tool for description of the existing ur-

ban structure. The classification into urbantypes will provide a basis for description ofthe existing urban structure in a specific city,in terms of characteristic urban typologies.

2 As a tool for analysis. It can provide deeperinsight into the “sustainability” of the dif-ferent urban types: by gathering basic envi-ronmental data (for example annual energyflow per capita, consumption of water, pro-duction of waste, recycling capacity, capacityfor local cultivation, potential of recyclingbuilding materials etc.). It will also be possi-ble to evaluate the relative attractiveness ofthe different urban types, by gathering so-cio-economic data (for example annualturnover of inhabitants, average income, so-cial interventions, signs of vandalism etc.).

3 As a tool for planning and design. In pro-viding a deeper understanding of the urbantypes, of sustainability and quality, a betterdescription of the existing built environment,it will pave the way for a better planningpractice, on both the micro- and the mac-rolevel.

A TYPOLOGY FOR SWEDISHURBAN BLOCKS

The basic methodology for typological classifi-cation should be essentially the same regardless

of the country. The proper starting point shouldalways be the history of urban development. Wecan expect to find specific regional and nationalcharacteristics in Swedish urban types, but —this is perhaps even more striking — we alsofind that Swedish urban types are very similar tothe urban types of other European countries,for example Germany, Austria, Denmark, Nor-way and Finland. The historical periods haveproduced similar urban types over a great partof Europe.

In the case of Sweden, the following histori-cal periods can be identified:1 The preindustrial period (before 1850)2 The Haussmann period (1870–1905)3 The period of Garden suburbs (1905–1930)4 The functionalist period (1930–)

The preindustrial period (before 1850)The characteristic urban type was the traditionalSwedish wooden town, with timbered houses ofone or two storeys. The streets are narrow: 6–10metres. The dwelling-houses are as a rule placedalong the streets, wooden fences protecting theprivate courtyards and small planted gardensfrom the streets.

The Haussmann period (1870–1905)High density blocks of flats in 5–6 storeys, withbuildings in the back. The transformation ofinner Stockholm (and other large Swedish cit-ies) in the late nineteenth century followed thecontinental urban pattern (Paris, Berlin, Vi-enna). The Building Code of 1874 set the rules.Apartment buildings of 5–6 storeys liningbroad streets (18 m). Urban density was veryhigh (e=2,5–3.0). Public parks were laid out inorder to compensate for the loss of private gar-dens.

Around 1900, low density suburban villadistricts for the affluent were laid out along newelectric tramways from the city centre.

The period of Garden suburbs (1905–1930)Reformed inner city blocks of 4–6 storeys withlarge open courtyards. Following new buildingregulations that were introduced during1910–30 back buildings were no longer per-mitted and new inner city blocks were designedwith large, planted courtyards. The urban den-sity was accordingly reduced.

New suburban developments were laid outfollowing the pattern of English and Germangarden suburbs. Mixed developments (detachedhouses, semi-detached houses, row houses) withlow or moderate urban density (e= 0,2–0,3)

The functionalist period (1930–)The functionalist ideal of open blocks of flats in

4 IAPS 14 Book of Proceedings

green space made its breakthrough in Swedishurban planning as early as 1930.

In Sweden, the most common type of blockof flats during the period 1930–50 was thethree storey “walk-up” (or lamella house). Thetraditional separation of private and publicspace was abandoned.

In the period 1955–1975, high rise blocksof flats became more common, with pointblocks (6–10 storeys) or slab blocks (6–10 sto-reys). Traffic separation was introduced, sepa-rate roads for walking and biking were builtapart from the regular network of roads formotor traffic.

After 1975, the era of high-rise blocks offlats came to an end; instead two-storey blocksof flats or terraced houses (in Sweden called“row-houses”) increased their share of the to-tal. The traditional closed block of 4–6 storeysmade a come-back in central areas of the largecities.

A preliminary typology would thus consistof eight basic types or classes. In each of thesebasic classes or types we can further distinguishbetween sub-types or variants.

a) above c) tob right

b) above

d) top right

Figure 1 a—d. Some examples of Swedishurban types. a) Preindustrial low-risetraditional blocks, b) High density in-ner city blocks of flats in 4—6 sto-ries.c) Suburban villas, large de-tached houses on large plots. d) Gar-den suburb, semidetached housesand 3-4 family houses on small plots.

1 High density inner city blocks of flats in 4–6storeys.

2 Inner city blocks of flats with open plantedcourtyards.

3 High rise point blocks or slab blocks, 6–12storeys.

4 Medium density blocks of flats, 3–4 storeys.5 Low-rise, pre-industrial districts in city cores.

Johan Rådberg, Sustainability and Urban Quality 5

6 Garden suburbs, mixed developments withterraced or semidetached houses.

7 Detached houses (“bungalows”) on smallindividual plots.

8 Detached houses or villas on larger plots.These basic types are illustrated by Figure 1 a–hon these pages.

On modernist and traditional urban patternsThe eight basic urban types differ from each

other in many aspects. They were built duringdifferent historical periods; the architecture,building materials and techniques are different.Some have a quite central location, others haveperipheral locations. Some are rented, othersowner-occupied. But the primary interest in thisproject is in the morphological characteristics,that is, the urban form and how the form affectsthe quality and sustainability.

e) Photo: Anders Friberg

f) Photo: Anders Friberg

g)

h)

Figure 1 e—h. Swedish urban types. e) Inner city blocks of flats with planted courtyards f) 3-4 storey blocks of flats, “walk-ups”. g) Small detached houses on small plots,h) high-rise developments, point blocks or slab blocks 1950s—70s.

As a first step, we must distinguish between tworadically different urban patterns: the traditionalurban pattern and the modernistic urban patternwhich has dominated the post-war period.

The traditional urban pattern is made up of

small-scale elements, individual building plots,that are aggregated into blocks, along streets.The modernist urban pattern is made up oflarge elements, large buildings with surroundingunbuilt space.

6 IAPS 14 Book of Proceedings

Figure 2 Idealized figures showing traditional urban pattern (left) and modernistic urbanpattern (right). The traditional pattern is made up of small-scale units (individualplots); the modernistic pattern is one large-scale unit (a housing estate). The openspace in the traditional pattern is divided into public space (streets) and privatespace (gardens). In the modernistic pattern the open space is unified, continuousand there is no boundary between public and private space.

The basic urban pattern — traditional or mod-ernistic — is very important. It influences theway that a certain area can adapt to changingdemands and changes in functions and life-styles. The traditional urban pattern can easilyand gradually adapt to external changes. Thebuildings on individual plots can be enlarged,transformed internally or even be pulled downto be replaced by new buildings. In the mod-ernistic urban pattern, gradual adaption is notfeasible. Rebuilding often affects the entirehousing estate and is often quite radical.

On classification of urban densityThe other parameters that are used for classifi-cation of the urban types in the project are fol-lowing:a residential density, (e= ratio of total residen-

tial area to area of land)b building height, (average number of storeys)

(n)c percentage of built-up area. (v)

The mathematical formula is as follows: (e) =(v) x (n)

These parameters can also be analysed in adiagram (Figure 3). If we register a number ofdifferent urban blocks and place each block asa dot on the graph, (according to their urbandensity and number of storeys), we find that theindividual observations of blocks (dots in thediagram) tend to cluster into a larger bubble.(Rådberg, 1988).

Johan Rådberg, Sustainability and Urban Quality 7

Figure 3. Urban density, building height and percentage of built-up area in eight Swedishurban types. (1) High density inner city blocks, 4–6 storeys. (2) Blocks with plantedinner courtyards, 4–6 storeys. (3) High rise developments, point blocks or slabblocks 6–12 storeys. (4) 3–4 storey “walk-ups” (lamellas). (5). Preindustrial low risetraditional blocks. (6). Garden suburbs, mixed developments. (7). Small one-familyhouses (bungalows) on small individual plots. (8). Villas on larger plots.

Project objectivesThe general objective of the project is to de-velop a typological urban classification system,as a basis for research, for descriptions (map-ping) of the existing urban environment and asa basis for urban planning and design.

The classification system is primarily basedon the Swedish urban environment, but thestructure of the typology and the basic pa-rameters should be adaptable to the urban typesof other countries in Europe.

THE PROJECT IS EXPECTED TORESULT IN

1 A new method for registration and analysisof the urban environment in a given city, onthe basis of a typological classification sys-

tem.2 Descriptions of the existing urban structure

in a given city, for example digitalized mapsand/or databases.

3 New knowledge of the relationship betweenurban form and environmental characteris-tics, summarized in a sustainability index.

4 New knowledge of the relationship betweenurban form and quality (in terms of long-term attractiveness), summarized in an indexof urban quality.

5 A method for identification and registrationof historically and culturally valuable areasin given cities

Project methodologyThe project will proceed in five consecutivesteps or project-parts. The first project-part,

8 IAPS 14 Book of Proceedings

(Step 1) concerns the formulation of the urbanclassification system. The basic parameters willbe specified. Tests will be made on which pa-rameters could easily be obtained from centraldata bases. The reliability of the given parame-ters will be studied. The classification systemand the parameters will then be modified andtested in small-scale studies in parts of a selectedcity.

The registration of objects will be imple-mented in a Geographical Informations System(GIS). The primary source for information onurban elements (building lots, residential andother buildings) in Sweden is the Central Insti-tute for Information on Real Estate (CFD) inGävle. This centre has updated digitalized in-formation on about 3 million units, almostevery building in the urban areas of the coun-try. This central data base comprises about 100different parameters; the number of parametersis continuously enlarged. The central parame-ters in this project are (a) the urban form (givenas coordinates), (b) the total area, (c) number ofstoreys (d) the built-up area in relation to openspace (e) the ratable value. The CFD data base isthe single most important source of buildingdata in Sweden, but there are several othersources. The city planning offices often havedigitalized maps with information of actualplans, regulations, roads, infrastructure etc.

A tailored GIS program will be designed as atool for registration and classification of the ur-ban elements. (See Aglialas, D & C. Harlow,1990.) The registration and classification proc-ess will be semi-automatic, which means that theoperator can intervene manually when a specificobject cannot be identified according to thedetermined geometric conditions.

After the the first step in the project, the se-lected parameters and geometric conditions willbe evaluated and modified if necessary.

Step 2 will be a full-scale study of a selectedcity. The urban elements (blocks) will be classi-fied, with the help of the GIS. The results will beillustrated in a detailed digitalized map showingthe urban types, their extent and geographicallocation.

The elements (blocks) are also registered in adata base. This means that additional informa-tion related to each object (block) can be

fetched from other data bases, using a geo-graphical code as a link. Such additional infor-mation include: the number of households ineach object, age distribution, income distribu-tion, taxation, percentages of empty flats, socialcare, car ownership, the annual per capita con-sumption of electricity, production of wastewater, an so on. The range of information avail-able from other data bases is almost unlimited.

In the next stage of the project (Step 3), atypological catalogue for urban research will becreated. This typological catalogue should con-sist of at least 300–500 objects (urban blocks)from different cities (large and medium sized)all over the country. The objects are selected bystatistical methods to get a proportionality to thetotal population in each of the typologicalgroups. The catalogue can be used as a refer-ence selection for specialized studies of eco-logical problems. Such studies could be thestudy of the flow of material and energy, orevaluations of the opportunities or constraintsfor new ecological innovations in existing resi-dential areas with different urban form. In thenext step (Step 4), the accumulated informationrelated to the objects in the research catalogue,(such as specific energy consumption, wasteproduction, possibilities for local infiltrationand cultivation etc.) will make it possible toconstruct a sustainability index. This indexshould be tailored as a tool for comparisonsbetween existing areas.

In a similar process, the project will try todevelop a quality index, (containing a mix ofparameters based on social data and urbanform). This index would also be used for com-parisons between existing areas.

Finally, in Step 5, the project will developmethods to supplement the basic typologicalmap with complementary information con-cerning culturally and historically valuable dis-tricts or objects, or valuable landscape elementsetc.

REFERENCES (SELECTIVE LIST ONLY):Aglialas, D. & C. Harlow, 1990. “Computerational Image Interpretation Models: An Overview and a Perspective”

in Photogrammetric Engineering and Remote Sensing, Vol. 56, pp. 871–886.

Coleman, Alice, 1985. Utopia on Trial. London.

Conzen, M.R.G., 1968. “The Use of Town Plans in the Study of Urban History” in H J Dioz, The Study of Ur-ban History. New York.

Johan Rådberg, Sustainability and Urban Quality 9

Green Paper on the Urban Environment. 1990 EC. Luxembourg.

Lawrence, Roderick, 1987. Housing, Dwellings and Homes, Design Theory Research and Practice. Chichester.

Lynch, Kevin 1981. A Theory of Good City Form. Harvard.

Marling, Gitte, Mary-Ann Knudstrup & Nina Værum, 1995. Indicators on Urban Environment. Aalborg.

Moudon, Anne Vernez, 1986. Built for Change. Cambridge, USA.

Moudon Anne Vernez, 1987. “The Research Component of Typomorphological Studies”. Paper presented at theAIA/ACSA Research Conference, Boston, November 1987.

Newman, Oscar 1973. Defensible Space, Crime Prevention through Urban Design. New York.

Rådberg, Johan, 1988. Doctrine and Density in Swedish Urban Development 1875 — 1975. Stockholm:Byggforskningsrådet R11:1988.

The International Institute for the Urban Environment, 1994. The European Sustainability Index. Project report.Delft.