townscape assessment: the development of a ... - urban … · townscape assessment: the development...

17
Urban Morphology (2007) 11(1), 25-41 © International Seminar on Urban Form, 2007 ISSN 1027-4278 Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built environment Alan Reeve Townscape and Heritage Research Unit, Joint Centre for Urban Design, Oxford Brookes University, Gipsy Lane, Headington, Oxford OX3 0BP, UK. E-mail: [email protected] Brian Goodey Joint Centre for Urban Design, Oxford Brookes University, Gypsy Lane, Headington, Oxford, OX3 OBP, UK. E-mail: [email protected] and Robert Shipley School of Planning, University of Waterloo, Waterloo, Ontario, Canada, N2L 3G1. E-mail: [email protected] Revised version received 24 July 2006 Abstract. ‘Townscape’ as an approach to understanding one aspect of quality in the built environment has had mixed fortunes over the last few decades. Nonetheless, it remains a useful category within which the question of how places work at an aesthetic level can be considered. The problem has been, however, the absence of a comprehensive and relatively objective system for recording and representing the findings from townscape evaluation in the field. This paper presents the evolution of one such approach, and illustrates how it has been applied as part of a more general impact assessment research project to determine the effectiveness of the Townscape Heritage Initiative (THI) regeneration programme in the UK. The paper concludes with a discussion of the utility of the method for future built environment monitoring and evaluation programmes. Key Words: townscape, urban design, urban quality, evaluation methods As a focus, ‘townscape’ has dominated the more popular literature on urban design in the UK since the end of the Second World War. The term’s specific use seems to date to Hissey (1889, p. 263) who recorded ‘some of the quaint townscapes (to invent another word) of our romantic, un-spoilt English towns’. Its near equivalent outside the UK is ‘urban landscape’, a term that occurs frequently in this journal, as does the term ‘townscape’ itself, which is used widely not only by architects but also by British geographers (for example, Smailes, 1955). In the immediate post-war period townscape was seen by many architects as a concept that re-emphasized character, tradition and differ- entiation in distinction to the imported and imposed uniformity of the Modern Movement (De Wolfe, 1949). Use and development of the term shared some of the implications (and confusions) of the seventeenth-century coining of the term ‘sharawadgi’ as applied to the studied beauty of irregularity (Pan, 1995). Although always criticized as a sub- theoretical set of tools for partial analysis and prescription (as well as the basis of strongly felt and elegantly argued critiques of change in British towns – a prime example being Ian Nairn’s London, 1966), the townscape approach to urban analysis dominated planning thought for a considerable period. It remains a key consideration, especially at the interface between professional and public views of existing and future environments,

Upload: buimien

Post on 05-Jun-2018

225 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Urban Morphology (2007) 11(1), 25-41 © International Seminar on Urban Form, 2007 ISSN 1027-4278

Townscape assessment: the development of a practical toolfor monitoring and assessing visual quality in the built

environmentAlan Reeve

Townscape and Heritage Research Unit, Joint Centre for Urban Design, OxfordBrookes University, Gipsy Lane, Headington, Oxford OX3 0BP, UK.

E-mail: [email protected] Goodey

Joint Centre for Urban Design, Oxford Brookes University, Gypsy Lane, Headington,Oxford, OX3 OBP, UK. E-mail: [email protected]

andRobert Shipley

School of Planning, University of Waterloo, Waterloo, Ontario, Canada, N2L 3G1.E-mail: [email protected]

Revised version received 24 July 2006

Abstract. ‘Townscape’ as an approach to understanding one aspect of qualityin the built environment has had mixed fortunes over the last few decades.Nonetheless, it remains a useful category within which the question of howplaces work at an aesthetic level can be considered. The problem has been,however, the absence of a comprehensive and relatively objective system forrecording and representing the findings from townscape evaluation in the field.This paper presents the evolution of one such approach, and illustrates how ithas been applied as part of a more general impact assessment research projectto determine the effectiveness of the Townscape Heritage Initiative (THI)regeneration programme in the UK. The paper concludes with a discussion ofthe utility of the method for future built environment monitoring and evaluationprogrammes.

Key Words: townscape, urban design, urban quality, evaluation methods

As a focus, ‘townscape’ has dominated themore popular literature on urban design in theUK since the end of the Second World War.The term’s specific use seems to date to Hissey(1889, p. 263) who recorded ‘some of thequaint townscapes (to invent another word) ofour romantic, un-spoilt English towns’. Itsnear equivalent outside the UK is ‘urbanlandscape’, a term that occurs frequently inthis journal, as does the term ‘townscape’itself, which is used widely not only byarchitects but also by British geographers (forexample, Smailes, 1955).

In the immediate post-war period townscapewas seen by many architects as a concept thatre-emphasized character, tradition and differ-entiation in distinction to the imported and

imposed uniformity of the Modern Movement(De Wolfe, 1949). Use and development ofthe term shared some of the implications (andconfusions) of the seventeenth-centurycoining of the term ‘sharawadgi’ as applied tothe studied beauty of irregularity (Pan, 1995).

Although always criticized as a sub-theoretical set of tools for partial analysis andprescription (as well as the basis of stronglyfelt and elegantly argued critiques of changein British towns – a prime example being IanNairn’s London, 1966), the townscapeapproach to urban analysis dominatedplanning thought for a considerable period. Itremains a key consideration, especially at theinterface between professional and publicviews of existing and future environments,

Page 2: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

26 Townscape assessment

particularly in the UK where it has had asignificant history in urban design, planningand conservation. Its development will alwaysbe associated with the name, writing andespecially the illustrations of Gordon Cullen(Gosling, 1996; Whistler and Reed, 1994).

The subject remains central to the processesof conservation, development and urbandesign (Tugnutt and Robertson, 1987),although interpretations differ considerably.In the early 1990s, to the English HistoricTowns Forum (1992) it appeared to consist ofquality in building detail, whilst in anAmerican survey exploring popular visualpreferences through photographs (A. NellesenAssoc., 1993) it was the quality of spaceswhich was the focus. More recent literature(Shipley et al., 2004) has suggested thattownscape can be used in conjunction withother approaches for assessing change in thebuilt environment, relating this to criteria ofevaluation and performance. That is,townscape is becoming again the focus ofinterest in specific fields of urban conservationand design as a potential means for measuringchange brought about by regenerationprogrammes such as those funded by theHeritage Lottery Fund (for example, theTownscape Heritage Initiative, THI). It isbeing used as a tool for understanding howinvestment in the physical fabric of places atthe urban scale can affect their visualperformance.

Although many researchers have exploredgeometric analysis, photographic recognitionand comparison, and spatial analysistechniques as surrogates for the comparativeanalysis of townscapes (see, for instance,Cooper, 2005, on the application of fractalanalysis in streetscape evaluation), no singletechnique is either comprehensive or generallyaccepted. The search has been for a fieldtechnique which records and / or assesses anurban or street view, and translates this into amanageable set of data which can be mappedand compared with other, similarly achieved,data. The purpose of this paper is to describeand explain the development of such atechnique over a 25-year period, and toprovide some recommendations for its

application in urban design and urbanmorphology. The paper describes a number ofprojects in which this technique has beenevolved, and concludes with a detailedaccount of its current incarnation in itsapplication to a longitudinal evaluation of theTHI programme.

Identifying environmental opportunities inTower Hamlets, London, 1978

The method employed in the THI work relieson detailed field observation and recording ofthe quality of townscape elements on aprepared and piloted pro forma and the use ofsummary scores for each view (not view-point) for the construction of score sheets andmaps which suggests the patterns andcomplexity of townscape quality within agiven area. The use of a standardized proforma permits comparison both within, andbetween areas, and also allows comparison ofindividual, or groups of, variables. Themethod was initially developed by Goodey(Goodey and Ashford, 1978) in response to arequest to identify environmentalopportunities in the London Borough ofTower Hamlets for the Tower HamletsEnvironment Trust.

The brief focused on the emerging conceptof ‘opportunity’ in the area adjacent to theCity of London and the recognition, as in theDepartment of the Environment’s Makingtowns better studies of 1973 (especially thatfor Rotherham) (Department of the Environ-ment, 1973), that visual quality was a keycomponent of urban improvement. Elementsdrawn from both the ‘townscape analysis’ and‘environmental quality measurement’ in theRotherham reports were influential in shapingthe methods described below.

Initial discussion about both the factorswhich might comprise a townscapeassessment and the methods, includingscoring, that were to be used took place at atime when Alice Coleman (1984, 1986) wasdeveloping a method for the assessment ofbuilt environmental quality in nearby areas.Her work, when published, proved both

Page 3: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 27

provocative (Campbell, 1985) and influentialin decisions on public housing renovation. Itwas particularly concerned with the presenceor absence of quality or condition markers,rather than the overall aesthetic quality oftownscapes or streetscapes.

From a range of existing analyses andstudies we determined the following initial setof variables as describing townscape quality:

General impression:character of buildings maintenance of buildingsgrouping of buildings

Detailed observation:character of boundaries maintenance of boundariesstreet furniturestreet amenitiesstreet maintenance planting: shrubs and flowers trees lawns

With the addition of perceived land-usediversity, a pro forma allowing scores for eachvariable from 0 (low) to 5 (high) was used in asample of road junction views in the largelyresidential areas of Bromley-by-Bow, eastLondon. Findings from this pilot survey werecompared with the results of an interviewsurvey with a sample of the area’s residentswhich sought their evaluation of the variablesselected.

To further refine the method, especiallywith regard to observation points, visualenvelopes and the weighting of variables, asecond pilot survey was undertaken inMiddleton Cheney, a village of then some4000 inhabitants with a wide range ofresidential settings, located in southNorthamptonshire. As the residence of one ofthis paper’s authors, this offered manyadvantages with regard to view definition andknowledge of spatial coverage. In thisrevision, traffic and pedestrian environmentsreceived increased emphasis, as did skyline

features, and evidence of historic orcommunity concerns. As a result of both pilotsurveys we were able to advise on the densityof observation points, the form of views beingassessed, the locations to which scores mightbe attributed, and the firmer definition andscoring of all variables employed. A six-pointscale was suggested, although not all scoreswould be available for all variables. Thereport outlines the use of the method as one ofa suite of techniques for assessing localenvironmental quality, several of which wouldinvolve community input or comment.

The final set of variables proposed forfuture use in the Tower Hamlets context was:

1. First impression 2. Evident functional diversity 3. Built form

characterqualitygroupingmaintenancepersonalization

4. Boundaryqualitymaintenance

5. Green spacepercentage of view public or private?qualitymaintenancewaste land

6. Streetstraffic suitabilitysurface quality

7. Pedestrian environmenttraffic suitabilitysurface qualityvandalismgraphics/advertising

8. Skyline/‘wirescape’ 9. Nostalgic reference10. Evident community concern11. Presence of selected features

children’s play areaspublic eating facilitiespublic toiletsrecreational water space

12. Work in progress

Page 4: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

28 Townscape assessment

The report concluded that ‘what we havebeen asked to consider is a technique which‘formalises subjective observations over alimited ... and biased ... set of variables’ ’. Ifthe scores generated are treated as sacrosanctor in isolation, then there is little chance thatany changes that are sensitive to either thecommunity or the environment will beaffected. Scores offer a starting point, therequested ‘base line’, from which people whocare about a place called Tower Hamlets (orparts thereof) can begin to probe further andask some questions that have not previouslybeen raised.

The report’s findings and implications wereconsidered in detail by the Trust. Anyborough-wide survey by an agency intent onattracting investment and making its mark inthe area was seen as too heavy an investmentin terms of time and money.

It was the presumed high cost of surveyingthat prevented the Tower Hamlets report beingdeveloped further in consultancy studies.Nonetheless, the techniques found a place in anumber of graduate student research studiesand, although the labour in their studies mightbe seen as voluntary, were not reported asparticularly burdensome. Several theses wereable to employ, refine and appraise the utilityof the methods. Issues of initial definition,variables to be used, density of observations,weighting and application of results were themajor themes for discussion.

Coventry environmental quality survey,1989

As a key element in the development ofCoventry’s Unitary Development Plan, theFirst Interpreters planning consultancy, basedin that city, was commissioned to undertake anenvironmental quality survey of the city’s area.The report on work undertaken in early Spring1989 details the process and results (FirstInterpreters, 1989).

The survey called for a particular emphasison green space within the urban context, andwas to represent the initial stage of anenvironmental quality enquiry which would

then be pursued with communities at the locallevel. The framing of the work implied a highvalue attached to landscape-dominated design,‘arcadian’ views and English landscape tastesas, for example, maintained by both the first(and second) editions of the ‘Essex DesignGuide’ (Essex Planning Officers’ Association,1997), although possible differences amongstethnic communities were recognized.

The pro forma for observations thatemerged from a pilot study on Coventry’sfringe, resulted in assessment of the followingvariables on a seven-point scale:

1. Built formhuman scalebuildings design detailfunctional varietypermeability

2. Green spacepresence ofvariety/species/formsseasonal visionmaintenance

3. Public spacedesign qualityedge feature qualitycleanlinessmaintenance

4. Waterpresence ofmaintenance

5. Explicit heritage6. Street furniture design7. Skyline qualityObservations (unscored)

presence/absence of enclosurenew buildtrafficparking issuesgraffitineighbouring wildlife, etc.

With the intention of providing a map basefor the whole city area, a regular grid of 1 kmsquares within the city boundary wasemployed, with road junctions or open view-points closest to the grid intersections beingused for four cardinal observations. Thisbasic grid was intensified in the more central,

Page 5: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 29

built-up, areas and was supplemented byadditional observations where areas of contrastwere hidden from initial view. In addition,special attention was paid to a nominatedrange of post-war residential schemes.

The whole survey, involving 249 sites andjust less than 1000 survey forms, was under-taken by one surveyor, and photographs weretaken of each view observed. This wasachieved in 8 working days in March 1989,during daylight hours and with the benefit offine weather throughout.

A photographic follow up in better lightrevealed the continuing process of modestenvironmental change which endorses eachview as dynamic, but seldom altering radicallyover a year. The presence of one surveyoraided consistency in the observation, inter-pretation, and supplementary notes on areaswhich formed part of the final report. Viewscores were mapped, both to provide an overallenvironmental quality map of the city andspecifically to draw on the green spacevariables. The results were used as a basis forfurther enquiry and discussion in thedevelopment of Coventry’s UnitaryDevelopment Plan.

Consequent upon the Coventry work, afurther study was undertaken in the Ancoatsarea of Manchester by First Interpreters. Thistook the method back to an area not dissimilarto parts of Tower Hamlets – inner cityindustrial and residential with a few, fading,buildings of note. Two surveyors were used ina study drawing on observations at all inter-sections, providing a high-density coverage.

After a pilot survey, the variables selectedfor Ancoats marked one significant changefrom those used in previous studies. The issueof safety brought into perspective thedifferences between the sexes in townscapeevaluation and joint observations by observersof both sexes were employed at each point.Issues of personal safety were expanded in thevariables to be assessed. This study was notpublished, but fed into the local regenerationprocess.

1990-1998

In October 1989 a further application wasemployed in an environmental audit ofBroseley, an historic industrial town adjacentto Ironbridge in Shropshire. The survey wasat a very detailed scale, with observations atall road and path intersections, and ofbetween-building views. It yielded maps ofoverall townscape quality, and especially ofvisible green space. The data and maps wereused to guide detailed analysis of public spacequality (First Interpreters, 1990).

Summarizing the qualities implicit in thesurvey, the authors noted:

... we can characterise a good view as one inwhich there are no out-of-scale buildings,there is attention to design detail, wherethere are several uses in view, and wherethere are alleys and side roads which suggesta variety of routes through the scene. InBritain we expect greenery in our towns andthat a variety of species ensure all yearcharacter; we also expect planting to be well,or appropriately, maintained. Good publicspaces have been designed or directed toachieve public functions, they haveattractive edges and are clean and wellmanaged. Water adds movement and lightto the scene. Clear evidence of history is anattraction which reminds visitors of theirroots. Street furniture is always in evidenceand can be an asset. Long-distance andskyline views help place a town in itssetting. A bad view is easy to deduce fromthe above. Single use areas of poorlydetailed buildings with little planting, aridstandard street furniture, the whole poorlymaintained, represent the town view whichgains little attention or interest.

There followed a fairly fallow period in thedevelopment of the approach in the publicarena, although again student work reflectedon the problems and opportunities. It was thepresumed cost of field survey which held themethod back, although complex combinationsof desk, community and field survey wereemployed to reach related ends, as in the 1989environmental assessment of the residentialareas of Harrow by Wootton Jeffreys (Munroand Lane, 1990). Our experience in this work

Page 6: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

30 Townscape assessment

since then has reinforced the conclusion thatsystematic townscape evaluation using ourmethods does have serious resource impli-cations if applied unthinkingly. However,assessments using the methods being describedhere can be made reasonably cheaply andefficiently to evaluate specific and clearlydefined areas, or even single views, in whichchange is occurring or is about to occur; orwhere there is a need to record the visual statusquo for future reference – such as inconservation areas.

A considered view of the methodology,developing and latent, over the 20-year periodreturns to four key factors. The first two areidentified by Findlay, Rogerson and Morris(1988, p. 96) in their discussion of quality oflife indicators:

Two main technical problems are faced byresearchers in the evaluation of quality oflife. The first relates to the identification ofthose elements or dimensions of anenvironment which are involved in people’sassessment of its quality of life. The secondis to decide how much importance to attachto each of the elements when they areaggregated to form an overall evaluation ofquality of life.

Problems also arise in the case ofenvironmental and townscape quality. First,what are the variables to be considered? Andsecond, how are they to be weighted in theproduction of a composite indicator? The thirdkey issue in the method outlined above is thedensity and location of observations and scoreallocation, a matter which was seeminglyresolved in the Coventry study. The fourthissue is the more practical one of how the dataand their presentation are to be used in theurban environmental management process.Given that a survey can reveal contour zonesof related quality, and that maps can suggestboth high and low points of visual environ-mental or townscape quality, there is thequestion of how these data can be used tobetter understand the processes involved inenvironmental change. The next studydescribed below begins to address thisquestion.

Townscape evaluation of the impact of theJubilee Line Extension (JLE) on theLondon Underground, 1999-2003

The London Transport study provided aparticularly useful focus for addressing thislast issue. By monitoring the areassurrounding the Jubilee Line Extension stationsites both before and after opening, a measureof townscape quality change could beattempted (see Willis, 1997 for an account ofthe background to the JLE). Indeed, this wasessentially the brief for this work, whichformed a part of a large-scale evaluation of theimpact of the JLE economically, socially,environmentally and in terms of other issues(Reeve and Goodey, 1999).

The JLE townscape evaluation took theform of a detailed survey of seven of theeleven new station sites and their immediatecontexts in 1999 prior to opening, and 2 yearslater once operational. The assumption wasthat the impact on the townscape would beboth in terms of built fabric (quality ofbuildings, streetscapes, etc.) and the levels ofactivity or business around each station. Thepro forma developed for this projectattempted, therefore, to record both ‘static’and ‘dynamic’ elements in the visual scene,and to assess them qualitatively before andafter the JLE was operational. The methoddeveloped earlier thinking regardingevaluation of townscape by considering thescale of, or boundary between, elements ordimensions in the visual scene. Weconsidered that the scene could be understoodat four levels of resolution: pedestrian (local);street long (neighbourhood); elevated views(districts and above); and skyline views(metropolitan and above). For the purposes ofthe JLE evaluation, levels one and two werethe primary focus, although in NorthGreenwich and Canary Wharf townscape canbe experienced at an undoubtedly elevatedlevel. Our method was chiefly concerned,however, with the public and pedestrian viewpoint rather than that available to those withaccess to high buildings, for example.

We also sketched out ‘expanded’townscape criteria to think through the

Page 7: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 31

qualities that might both be observed andmight impinge on the experience of the placeand thus affect perception – although made upof a number of different visual clues. Thesewere organized into three categories:

General affective qualityambiencedistinctivenesslegibilityhuman scalepersonal safetycleanlinessvitalityfunctional variety

Detailed morphological elementarchitectural qualitygrouping/blocksstreet pattern‘floorscape’street furnitureedge featuresplantingtreessignagevisual stopsbarriersviewing placeschanges in scalecontainmentlandmarks

Dominant content of arealisted buildingsconservation areahistorical associationsdominant buildingsdominant building heightdominant buildingfunction

The criteria for assessment are complex, inso far as they require knowledge of siteconditions beyond the immediately apparent(such as conservation area status). However,this complexity seemed to be justified if wewere to build up a clear understanding of thecontext within which visual change wasoccurring, and the relative status or

significance of the views being scored. For adedicated research project such as the JLEwork, the time and other resources involved inbuilding up this complete picture could beargued for in relation to the overall budget forthe JLE itself and on the basis of thepermanence and significance of alterations tothe streetscapes around the new stations.However, the criteria were simplified after thepilot survey, so that greater emphasis wasgiven to those qualities that could be observedin the field rather than those requiringadditional research. In addition, only themorphological elements were scored – on asix-point scale – and the others simply noted.We also developed a set of guidance notes forscoring views, so that the field worker couldhave confidence in terms of what counted ashaving a high score and the opposite. Forexample, for the criterion of cleanliness thisguidance was given:

a.Cleanliness: evidence of high level ofcleansing of all areas of public space – litter,casual obstructions such as road works ingood order, graffiti, redundant signs, posters,etc.High: a clean and appropriately ‘tidy’environment, bearing in mind this is themost frequent concern of street users.Low: long-standing litter, redundant publicworks debris, torn posters, overflowing litterbins, dumped black bags, etc.

The implication here, which graduallyemerged as we thought through the appraisalmethod, is that what we were assessing wasperformance against an explicit standard.This is critical to understanding how themethod has evolved, and how it differs fromprevious instances of townscape evaluation.More will be said on this later in the paper.

In all, over 1000 views were scored bothbefore and after the opening of the JLE,building up a clear picture of what changeshad occurred and their significance intownscape terms. Unfortunately, a proposedfurther follow-up 5 years after opening willnot be conducted, for funding reasons,although several types of townscape impacttake several years to become evident.

Page 8: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

32 Townscape assessment

The main finding, in terms of method, isthat it provides a reasonably robust andsystematic technique for capturing change withrespect to clearly specified elements in thevisual scene. Moreover, saturating an areawith views for scoring allows a potentiallyvery detailed picture of change to be built upover time, in a way that permits this to begraphically presented through isoline maps,score cards and other means. Figure 1provides an example of how isolines were usedin this study to represent areas of differenttownscape quality.

The method is also adaptable for the needsof particular places. For instance, differentcriteria can be used for determining whichviews to score, depending, for example, on thepresence of new work, or of particularlysensitive townscapes. Likewise, the elementsscored can be determined on the basis of thetype of change being monitored and evaluated.

In the case of the JLE research, the elementsconsidered covered a wide range of visualissues because the research was concernedwith the overall impact on the immediate andlocal vicinities of the station sites. In the caseof our work for the Heritage Lottery Fund(HLF), (to be considered next), the focus wason conservation quality and the impact ofinvestment and regeneration, and the methodwas adapted to take account of this.

The Heritage Lottery Fund, 1989 andbeyond

The Townscape Heritage Initiative (THI) wasestablished by the Heritage Lottery Fund in1998. The THI Evaluation is a 10-yearproject designed not just to monitor the resultsof THI schemes being carried out across theUnited Kingdom, but more importantly to

Figure 1. Map of the immediate environs of Bermondsey Jubilee Line Extension Station,London, showing isolines. The higher the number on each isoline, the higher the

townscape score and hence quality.

Page 9: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 33

measure their effectiveness. The rationale,particularly from a townscape perspective, isthat regeneration projects at the streetscape anddistrict scale not only take a considerable timeto implement, but that their effects can take aslong to materialize. The research team havethe advantage of a relatively long period toobserve change: built into the research is aprogramme of at least three visits to all 17individual THI projects (sampled from a totalof around 75 schemes at the outset of thework) to undertake field work, including atownscape scoring. The townscape methodhas particular application here since it is idealfor monitoring change, not recording staticconditions. In addition, the work has involveda baseline study to capture the quality of thetownscape in the selected schemes prior toinvestment. As with the JLE research thisprovides an unambiguous benchmark againstwhich change can be measured.

Adapting the townscape method and thecriteria for evaluation to the needs of the THIprogramme

As with the earlier research projects, theessential method employed in the THI surveyrelies on detailed field observation andrecording of the quality of townscape elementsusing a pro forma (Figure 2). The elementswithin the pro forma differ from those used forthe JLE assessment, particularly in theiremphasis on indicators of conservation quality.These are grouped under three distinct themes:streetscape quality and maintenance; privatespace in view; and heritage in view. This is toallow aggregation of scores under these heads,and analysis of change with respect to themonce the data have been collected. This relatesto a central assumption of the research, as wellas the main focus of the projects: that privateand public space, whilst both contained withina single view, are differently and separatelymanaged. Whilst the quality of elements in thestreet are normally the product of publicspending and public maintenance and design,for example by a local authority, the elements

in private space are managed by theindividuals owning or controlling that space.The further hypothesis tested in the research isthat spending on elements in the publicly-owned part of a view will have an impact onthose in the privately-owned part. The thirdtheme, heritage in view, is clearly the mostdirect indicator of change with respect toheritage provision, maintenance orconservation.

A further refinement of the method asemployed for the THI study involves the fieldworker giving an impression score to serve asa check on the aggregate total derived fromthe individually-scored elements. In mostcases the impression score and the aggregateof individual scores do not vary that much.Where they do, this is usually because there issome additional quality – such as a distantvista – that either improves the overallimpression, or gives an impression ofdilapidation, decay and visual dreariness oroppression that is not entirely captured by thescored elements. The full emotional oraffective ‘content’ of a view or scene cannotbe captured by reference to its visualcomponents alone, since this ‘content’ mustinclude other aesthetic sensations such assmell and sound; it must also invokesensations which go beyond the aesthetic –including, for example, fear.

In terms of defining criteria, as suggestedearlier, these are now based more clearly on anotion of ‘performance’ than in previousiterations of the method. In other words, thequestion is asked, ‘how well is the urbanquality delivered in the view as measuredagainst a reasonable expectation?’ What willbe observed is that the qualities listed rangebetween the cognitive and the affective, but allare capable of being scored on a five-pointscale using the criteria, and all assume thatexpectations are both definable and realistic inrelation to particular views. They are,therefore, quantifiable.

With respect to conservation in particular,the articulation of clear criteria for assessmentcan be shown in this example from Table 1:

Page 10: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

34 Townscape assessment

C. Heritage in viewC.25 Neglected historic featuresSome buildings of historic significance,either listed or at least part of the streetscapeof conservation areas, may be in such poorrepair that their future is not certain. Oftenthese structures are vacant. It will be

important to note the presence of suchbuildings.High: No visible evidence of neglectedhistoric buildings.Low: Several historic buildings whichappear to be in poor repair and may be indanger of eventual loss.

Figure 2. Blank townscape pro forma as used in the field to score different townscapequalities.

Page 11: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 35

Table 1. Summary score-card for Bodmin, Cornwall at baseline stage, before anyinvestment under the THI programme.

Score Out of % Out of 5A. Streetscape qualityA1: Pedestrian friendly 138 220 63 3.1A2: Cleanliness 161 220 73 3.7A3: Coherence 155 220 70 3.5A4: Edge-feature quality 137 220 62 3.1A5: ‘Floorscape’ quality 136 220 62 3.1A6: Legibility 162 215 75 3.8A7: Sense of threat 150 220 68 3.4A8: Personal safety: traffic 141 220 64 3.2A9: Planting: public 88 140 63 3.1A10: Vitality 124 215 58 2.9A11: Appropriate resting places 58 100 58 2.9A12: Signage 89 170 52 2.6A13: Street furniture quality 105 215 48 2.4A14: Traffic flow appropriateness 144 210 68 3.4Total 1788 2805Average 63 3.2

B. Private space in viewB15: Advertising, in keeping 94 175 53 2.7B16: Dereliction, absence of 130 210 61 3.1B17: Detailing, maintenance 133 220 60 3.0B18: Façade quality 135 215 62 3.1B19: Planting: private 81 120 67 3.4Total 573 940Average 61 3.0

C. Heritage in viewC20: Conserved elements evident 116 180 64 3.2C21: Historic reference seen 80 115 70 3.5C22: Nomenclature/place reference 58 115 50 2.5C23: Quality of conservation work 107 180 59 3.0C24: Quality of new development 84 170 49 2.5C25: Neglected historic features 91 180 5 2.5Total: 536 940Average 57 2.9

Impression score 273 440 62 3.1Aggregate score 2897 4685 62 3.1

In this case it is simply a matter of observingwhether neglected buildings of heritage value,or whose contribution to the heritage value ofthe streetscape is significant, are present and to

what degree. The THI application anddevelopment of the method have allowed onekey principle to come to the fore in ourevaluation of places: that we assess them not

Page 12: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

36 Townscape assessment

against an ideal aesthetic standard, but againstwhat they can reasonably be expected toachieve. This means that different places withdifferent townscape ‘quality’ (uniqueness,architectural merit, townscape variety, forinstance) can achieve the same score. Whatmatters is whether that score improves ordeclines over time and with respect toparticular qualities in the visual field. This isjustified in that we are monitoring change, ormovement towards a standard that satisfiesseparately all of the criteria explained in thescore sheet.

Representing the data for the THI evaluation

As suggested earlier, the townscape methodallows the data to be represented in a numberof ways. The scoring also means, of course,that this representation can show changes overtime at the micro scale if necessary. For theHLF work we have used two methods forpresenting the data: mapping and a score-card.The score-card allows scored values to bepresented at a number of levels of resolution:for example a summary can be produced of allthe elements in a project and of how thesehave changed over time (Table 2).Alternatively, a particular theme can besummarized for the whole project, andcompared over time – for example heritage inview. A third option is to show change withrespect to a particular view.

The other means of showing the data is bymapping. We have found that grouping scoresinto bands and plotting these on ‘view maps’– that is maps that show the location anddirection of views – can be a very powerfulmethod of revealing and communicating theimpact of spending on the townscape of aplace. This allows us to highlight graphicallywhere change has occurred, and to say howthese changes map against where spending hashappened at the very local level. That is, themapping can test whether there has been apositive or negative response in the townscapeto spending on a cluster or terrace of buildings,or on a public space. Indeed, the scoring ofindividual themes and elements makes it

possible to produce a whole series of mapsshowing change for particular elements. Suchmaps avoid the inherent problem of isolinerepresentation, with its assumption that thereare regions within an area under assessmentwith the same townscape score when the scorerelates to a particular view, not to ageneralized area. So, with sufficient resources,over sufficient time, an extremely clear anddetailed picture can be painted of how a placehas been changed as a consequence ofspending on the physical fabric. For example,in West Wemyss, on the east coast ofScotland, 25 miles from Edinburgh (Figure 3),there is a very clear finding that in those areaswhere buildings have been restored thetownscape improved dramatically oncebuilding work was completed.

Discussion

The paper has described the evolution of amethod for townscape evaluation over a periodof some 25 years, by a number of researchers,and through half a dozen significant researchprojects. The method has arrived at a statewhich, arguably, allows it to be used to assessthe performance of townscapes using explicitcriteria to produce numerical results. The factthat the method involves numerical scoringmeans that the data generated, whilstqualitative themselves (that is, based onempirically observed qualities), can beunambiguously analysed, described andrepresented. The method also makes noclaims to capturing, objectively, the aestheticquality, per se, of places assessed since it isnow based on explicit criteria to do with howwell a view performs in relation to particularqualities.

However, a number of issues raised by themethod should be acknowledged. The mostimportant of these is the question of theobjectivity of the scorers in the field and theircompetence to score consistently the sameview. That is, do different observers arrive atthe same score for the same view? Ourexperience suggests that they do. Although, togive greater confidence in this, the method is

Page 13: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 37

Table 2. Summary score-card for Bodmin, Cornwall at interim stage, 2 years after workcommenced.

Score Out of % Out of 5A. Streetscape qualityA1: Pedestrian friendly 125 175 71 3.6A2: Cleanliness 126 175 72 3.6A3: Coherence 130 175 74 3.7A4: Edge-feature quality 127 175 73 3.6A5: ‘Floorscape’ quality 120 175 69 3.4A6: Legibility 137 175 78 3.9A7: Sense of threat 131 175 75 3.7A8: Personal safety: traffic 122 175 70 3.5A9: Planting: public 90 135 67 3.3A10: Vitality 125 175 71 3.6A11: Appropriate resting places 48 70 69 3.4A12: Signage 64 100 64 3.2A13: Street furniture quality 50 100 50 2.5A14: Traffic flow appropriateness 117 175 67 3.3Total 1563 2225Average 70 3.5

B. Private space in viewB15: Advertising, in keeping 90 155 58 2.9B16: Dereliction, absence of 131 175 75 3.7B17: Detailing, maintenance 127 175 72 3.6B18: Façade quality 122 175 69 3.5B19: Planting: private 55 90 61 3.1Total 525 770Average 68 3.4

C. Heritage in viewC20: Conserved elements evident 86 130 66 3.3C21: Historic reference seen 39 50 78 3.9C22: Nomenclature/place reference 35 55 63 3.2C23: Quality of conservation work 60 115 52 2.6C24: Quality of new development 54 110 49 2.5C25: Neglected historic features 97 150 64 3.2Total 403 630Average 64 3.2

Impression score 242 350 69 3.5Aggregate score 2491 3625 69 3.4

always applied with two field workers andrequires both to agree a score.

In addition, to test the robustness of themethod as an instrument for evaluating the

quality of townscapes and when a number ofobservers are involved, an experiment wasundertaken by one of the authors. The testinvolved 60 third-year students from the

Page 14: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

38 Townscape assessment

Figure 3. Scores for key views, West Weymyss, Scotland (the higher the number, thebetter the view). A. Before investment under the Townscape Heritage Initiative. B. After

investment. Reproduced by kind permission of the Ordnance Survey (license no.100017958).

Page 15: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 39

School of Planning at the University ofWaterloo in Canada. The participants weregiven only the written instructions for con-ducting the survey and sent out to evaluatefour views each. The students were dividedinto groups with each group visiting one offive different local towns or city centres.

The results of their surveys were thenanalysed and a number of findings emerged.First, there was a wide variation in therecorded scores, varying by as much as 15 percent above and below the median score. Thatis, if the highest possible score is 125 and themean reported for a given view was 65, thereported scores might range from 45 to 85.When the top and bottom 10 per cent of thescores in each group were eliminated the rangewas reduced to about 8 per cent above andbelow the median (from 55 to 75). That wouldstill render the technique unreliable if used inthis way. However, the second finding wasthat the aggregated impression scores and theaggregated reported scores were very close,meaning that the observers were veryconsistent in their own individual impressionsand scoring. Finally, it was found that whilescores varied greatly among observers, theirrankings of the views were quite consistent.

The second part of the experiment involveda sub-set of the class that had undertaken thefirst surveys. There were two differences fromthe first test. First, whereas the originalexercise was a compulsory class project, thesub-group for the second exercise wasvolunteers and they were paid a small fee fortheir services. Secondly, though theparticipants were again given the writteninstructions for the survey, they were alsoinstructed to observe one view together,discuss the instructions and arrive at a scorefor each of the 25 criteria on which they allagreed. In other words they ‘calibrated’ theirscoring system. This exercise took about 40minutes. The results of analysing the scores ofthis group revealed that the deviation from themedian was between 2 per cent and 3 per centabove and below.

This experiment would seem to support theassertion that motivated users with a minimumof training in the use of the method can be

relied upon to produce consistent andcomparable scores. The training is necessary,clearly, to induct field workers with theapplication of the method, and to deal with anypossible ambiguity about the meaning ofelements, or the criteria for scoring them.

What the method does assume is visualcompetence and some degree of familiaritywith basic planning and urban design termssuch as ‘legibility’. For the THI evaluation, atleast one field worker is required to haveconservation knowledge to make informedevaluations of elements under the ‘heritage inview’ theme. In other words, the method isnot designed to elicit lay responses to thevisual scene; but the views of researchers,students and professionals with sufficientcompetence to observe change and to makesimple judgements using explicit criteria.

Finally, we considered the use ofphotographs as the evidence for scoring asopposed to direct observation. The benefit ofthis approach would be that it allows forgreater consistence in the environment inwhich the assessment is actually made. Wehave not conducted any experiment to testwhether environmental factors (temperature,weather, light, field-work ‘fatigue’) affect thescoring. However, our view is that whilst aphotograph provides a useful record of viewsscored, it is no substitute for field observation,particularly where certain qualities are reallyonly evident when directly observed. Forexample, the quality of conservation work isnot always captured in a photograph – as in thecase of the type of mortar used for re-pointing.

Conclusions

We started this paper with an assertion thatprior to our work there had been no reliabletechnique for translating townscapeobservation into a manageable set of data thatcould be used to compare visual environmentalquality over time, or between places. What wehope to have shown is that with the THIversion of the townscape survey method sucha technique now exists, and can be adapted byother researchers in this field. For others

Page 16: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

40 Townscape assessment

considering using this method in their work wewould make four observations.

First, the technique is adaptable to a rangeof different problems and at different scales inthe evaluation of environmental quality; butbecause of this the examples of elements,themes and criteria given in this paper shouldnot be applied unthinkingly or uncritically.Secondly, the key to using the method is tomake clear what counts as an indicator ofvisual quality, and what the criteria are forassessment – principally that they shouldalways refer back to identifiable physicalelements. Thirdly, it is recommended thatelements within a scorecard be groupedthematically to avoid the problem ofweighting: in this way particular themes (suchas conservation quality) can be evaluatedseparately from other issues without gettingembroiled in whether one element has moresignificance than another. Finally, we wouldstress that the main value of our method is thatit allows the researcher to accurately identifyand map areas of different visual performanceand quality, and thus can provide powerfulevidence for arguing the need to protect,improve or treat in some other way specificstreetscapes, groups of buildings or othermorphological elements within a giventownscape. As a systematic tool for observingtownscape it forces the researcher to lookclosely and assess clearly where change isoccurring, and to decide, using explicitcriteria, whether this change is desirable.

References

A. Nelessen Associates (1993) Visions fordowntown Saint Paul: results of the visualpreference survey, for Department of Planningand Economic Development, City of Saint PaulMinnesota.

Campbell, K. (1985) ‘Utopia found wanting’,Architect’s Journal 140, 31, 57.

Coleman, A. (1984) Utopia on trial – vision andreality in planned housing (Hilary Shipman,London).

Coleman, A. (1986) ‘Dangerous dreams’, Land-scape Design 163, October, 29-31.

Cooper, J. (2005) ‘Assessing urban character: theuse of fractal analysis of street edges’, Urban

Morphology 9, 95-105. Cullen, G. (1971) The concise townscape (Archi-

tectural Press, London).De Wolfe, I. (1949) ‘Townscape’, Architectural

Review 106, 355-62.Department of the Environment (1973) Making

towns better: the Rotherham study, 2 vols(HMSO, London).

English Historic Towns Forum (1992) Townscapein trouble: conservation areas – the case forchange (English Historic Towns Forum andButterworths, London).

Essex Planning Officers’ Association (1997) TheEssex design guide for residential and mixed useareas 2nd edn (Essex Planning Officers’Association, Chelmsford).

Findlay, A., Rogerson, R. and Morris, A. (1988) ‘Inwhat sense ‘indicators’ of quality of life?’, BuiltEnvironment 14, 96-106.

First Interpreters (1989) Coventry environmentalquality survey, Coventry (First Interpreters, Cityof Coventry).

Goodey, B. and Ashford, P. (1978) Identifyingenvironmental opportunities: suggested surveytechniques and applications, unpublished Reportfor the Tower Hamlets Environment Trust.

Gosling, D. (1996) Gordon Cullen: visions ofurban design (Academy Editions, London).

Hissey, J.J. (1889) A tour in a phaeton through theeastern counties (R. Bentley, London).

Munro, B. and Lane, R. (1990) ‘An environmentalassessment of the residential areas of Harrow’,The Planner 12, 15-18.

Nairn, I. (1966) Nairn’s London (Penguin,Harmondsworth).

Pan, D. (1995) ‘Decoding Sharawadgi: Taoistinfluence on the Chinese landscape garden’,Landscape Research 20, 11-18.

Reeve, A. and Goodey, B. (1999) JLE townscapeimpact study, unpublished research report, JointCentre For Urban Design, Oxford BrookesUniversity.

Shipley, R., Reeve, A., Walker, S., Grover, P. andGoodey, B. (2004) ‘Townscape HeritageInitiatives evaluation: methodology forassessing the effectiveness of Heritage LotteryFund projects in the United Kingdom’,Environment and Planning C: Government andPolicy 22, 523-42.

Smailes, A.E. (1955) ‘Some reflections on thegeographical description and analysis oftownscapes’, Transactions and Papers of theInstitute of British Geographers 21, 99-115.

Tugnutt, A. and Robertson, M. (1987) Makingtownscape: a contextual approach to building in

Page 17: Townscape assessment: the development of a ... - Urban … · Townscape assessment: the development of a practical tool for monitoring and assessing visual quality in the built

Townscape assessment 41

an urban setting (Mitchell, London). Whistler, W. and Reed, D. (1994) ‘Townscape as a

philosophy of urban design’, Urban Design

Quarterly 52, 15-30.Willis, J. (1997) Extending the Jubilee Line: the

planning story (London Transport, London).

Density inside out

The Departments of Geography and Architecture atthe University of Edinburgh are organizing aconference on urban density. It will be held inEdinburgh, UK, 6-8 June 2007.

To think of the city is to invoke the question ofdensity. Urban density has been celebrated,cultivated, worried about, managed and shunned.For some, density is what makes the city full ofpromise; for others it is what determines itsproblems. Derived from the physical scienceformula for the ratio of mass to volume of inertmaterials, in urban applications density hasoperated as a seemingly objective measure of theratio of people or activity to area. As a diagnostictool, density has been employed in fields rangingfrom the pragmatic science of urban planning, tothe art of urban design. But the city is no mereinert material. The conference offers anopportunity to reconsider the relationship between

conceptions of density and how technology,infrastructure and buildings are organized. Thefollowing are some of the suggested themes:

1. Metaphorical and ideological fortunes of‘density’

2. Density/intensity3. Measuring density4. The history of density in urban planning,

design and architecture5. Typologies of density6. Cultures of congestion7. Crowding, proxemics and territoriality8. Prosthetic, dispersed and networked densitiesThe conference organizers are: Jane M. Jacobs

([email protected]), Geography, University ofEdinburgh; Stephen Cairns ([email protected]),Architecture, University of Edinburgh; IgnazStrebel ([email protected]), Geography,University of Edinburgh.

A suburban world?

An international conference on global decentral-ization and the new metropolis will be held atReston, Virginia, USA from 6 to 8 April 2008.

A new era has begun in which areas outsideworldwide core cities are forming differentsettlement systems. In Europe, even tight buildingcontrols have not curbed the appetite of householdsto move out of the central cities. China and India,with fast-growing middle classes, are witnessing astartling expansion of their own urban areas withrapid automobile-dependent growth on the fringe.Mexico and Indonesia are beginning to adoptAmerican-style suburban development patterns. Asurban form changes, new research and policy haveto be created to address the challenges of sprawl,congestion and affordable housing. This confer-ence will explore these and other issues in an

American as well as an international context.Themes to be explored include:

1. Immigration and the demographics of thesuburbs

2. Segregation by lifestyle and economics;including gated communities

3. Metropolitan form4. Energy consumption and sustainable

environments5. Governance and regulation6. Suburban decline7. Urban form at the neighbourhood level8. New suburbs and new towns

Conference website: http://www.mi.vt.edu/ Further information may be obtained from:[email protected]