the structure of pattern languages

Upload: dana-alhasan

Post on 10-Apr-2018

222 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/8/2019 The Structure of Pattern Languages

    1/14

    ^ u fii! \ \ //

    NikosA. SalingarosThe structure ofpattern languages

    Author's addressDivision of MathematicsUniversity o f Texas at SanAntonioSan AntonioTX7 8 2 4 9US A

    Pattern languages help us tackle the com plexity ofa variety of systemsranging from com puter software, to buildings and cities. Each'pattern'represents a rule governing one wo rkin g piece of a complex system ,and the application of pattern languages can be done systematically.Design that wishes to connect to human beings needs the informationcontained in a patte rn language. This paper describes how to va lidateexisting patte rn languages, how t o develop them, and how theyevolve. The connective geom etry of urban interfaces is derived fro mthe arc hitectural patterns of Christopher Alexander.

    We observe the world arou nd us and learn itsstructure by abstractin g cause and effect, and bydocumenting recurring solut ions obtained un derdifferent cond itions. Such empirical rules ,repre senti ng reg ularities of behaviour, are called'pattern s'. Visual patterns are the simplestexpression of th e p atte rn concept (Salingaros, 1999).Many patterns are hard-wired into our min d: weinhe rit actions and reactions that guarantee oursurvival. Other pa ttern s have to be learned, and forman artificial extension of the hum an min d. Theability to observe pattern s gives us the hu m anadvantage of bo th adap ting to, and changing ourenviron ment. Of course, the complexity enveloping apa tte rn in each specific setting has to be partiallycleared so as to get at its basic m echa nism .The language of a group of patterns forms theground work for any discipline. Learned patte rnlanguages - not intrinsic to the hu ma n mi nd - werecarefully preserved in t he past. Many patte rns ofhu ma n relations are codified into religions, myths,and literary epics. A collective intelligence developsfrom poolin g discoveries accumulated overgene ration s. This process is entirely gen eral. Thesciences rely on math em atics for th e ability toorganize data and explain phen om ena by means ofregularitie s, or logical pattern s (Steen, 1988).Breakthroughs occur when patterns in one area linkto patterns in oth er areas.This paper discusses the language th at link spat terns. A patte rn language contains usefulconnective information that helps both to validatethe pattern s, and to apply them. We are going to cast

    the structur e of a patt ern language in terms of theproperties of patt ern co mbin ations. Such anapproach reveals the orderin g of patterns in space,time, and human dimensions. I will assume readershave a min imal familiarity with th e architecturalpatterns of Christopher Alexander as published i n APattern Language (Alexander, Ishikawa et al., 1977).1Although introduced in to architecture mo re tha n 20years ago, thei r tr ue significance has beenappreciated by only a few practitio ners. Patterns ar ea powerful to ol for control ling complex processes,but because of misund erstandin gs, they have notplayed a wide role in a rch itectu ral design. Instead,patterns have found unexp ected success in com puterscience.

    The audien ce for this paper is anyone interested inconnecting thei r designs to hu m an beings. We willshow that this cannot be done withoutincorpo rating patte rns. After describing in generalterms what pattern s are, and the ways they cancombine, I will discuss the relationship betweenpatte rns an d science. Graph t heo ry visuallyillustrates some key aspects of pattern lang uages:how patterns co mbin e to form higher-level pa tternscontaining new information; how linked patternsexist on different levels; how to find p attern s in anew language; and ho w a patte rn language isvalidated thro ug h its connective structu reindepend ently of each individual pattern's validity.Amajor concern is how a pattern language is damagedthro ugh the imp osition of arbitrary stylistic rulesand anti-patterns, which are often mistaken forpat terns. All too often, peo ple have tried to chan ge a

    theory arq V0I4 no 2 2000 149

  • 8/8/2019 The Structure of Pattern Languages

    2/14

    0 arq V 0 I 4 no 2 2 0 0 0 t h e o r y

    1 Small parking lots(#103). One of the 253design 'patterns ' th atrecur in archi tecturea Each of theAlexandrine patternsopens wi th aphotographsummariz ing thepattern subject. Noneis ident i f ied: th is on eil lustrates the Fellows'(faculty) car park at St.Catherine's College.Cambridgeb The i l lustrat ionencapsulat ing the tex tsect ion ent i t led'Vastpark ing lots wreck th eland for the pe ople'c The i l lustrat ionconcluding thediscussion sectiond The prescript ion andthe patternThe il lustrations inthese Figures and Figs.2-14 and 21 are from APATTERN LANGUAGE byChristopher Alexander,copy r igh t -1977 byChristopher A lexander.Used by permission ofOxford Univers i tyPress, Inc. (Availablefrom Oxford Un ivers i tyPress, 1171 pp .Hardcover, GBP40 or$40. ISBN 019 5019199www.oup.com)

    Make parking lots small, serving no more than five toseven cars, each lot surrounded by garden walls, hedges,fences, slopes, and trees, so that from outside the cars arealmost invisible. Space these small lots so that they are atleast 100 feet apart.five to seven cars

    i d

    society by changing its architectural patte rnlanguage. An application to the geometry of u rbaninterfaces is given from the patterns approach.W h a t is a pattern?In A Pattern Language, Alexander and his colleaguespublished 253 solutions or design 'patte rns' thatrecur in a rchitecture, such as the need for SMALLPARKING LOTS (#103) [Fig. 1a], or SIX-FOOT BALCONY -the minimum depth that makes it useful - (#167)[Fig. 2] (Alexander, Ishikawa et al., 1977). They arguedthat built designs violating the derived patterns werenoticeably less successful than those that followedthe m. The Alexandrine format fixing a patternconsists of a statement summarizing the philosophyabou t a specific topic (i.e., for SMALL PARKING LOTS):'Vast parking lots wreck the land for people' [Fig. ib[.

    They follow the patt ern statem ent by anexplanation th at suppo rts the pattern : statisticaldata; a scientific analysis; discovering thesimultaneous occurrence of this pattern in totallydifferent cul tures; psychological, structura l, orcultu ral reasons; and so on. For example, thediscussion following the above patt ern includes:'... the fabric of society is threatened by the mereexistence of cars, if areas for parked cars take upmore than 9 or 10% of the land in a community....tiny parking lots are far better for the environmentthan t he large ones, even when their to tal areas arethe same ... Large parkin g lots, suited for the cars,have all the wrong properties for people' [Fig. ic].

    A patter n ends with some sort of prescription inpractical terms, to help incorporate the pat tern intoan actual design. For example: 'Make parking lotssmall, serving no more than five to seven cars, eachlot surround ed by garden walls, hedges, fences,slopes, and trees, so that from outside the cars arealmost invisible ... [Fig. id]'.

    Many criticisms of Alexander's Pattern Languageare valid to some degree - tha t it reflects thephilosophy of the 1960s, that it is too radical and noteasily incorporated into contemporary design andplanning , tha t it ignores almost all of what isconsidered impor tan t architecture in the twentiethcentury - but these are trivial compared with theimp ort ant message it offers. This paper will attem ptto show that any design that ignores patterns cannever hope to connect to human beings.Combining Alexandrine patternsYou can combine design patterns in an infinitenumber of ways. However, the connective rules - i.e.,the language - were only briefly sketched out. Toobtain an understanding of the relationshipbetween patterns, you have to go back to Alexander'searlier work (Alexander, 1964; Alexander, 1965).Other than Chapter 16 of The Timeless Way of Building(Alexander, 1979), Alexander himself has not dwelton the synthesis between patterns. Any perceivedweakness of patte rns could he in individual patterns,but it is more likely the result of not unders tandingtheir combinatorial language. Although designpatterns written in Alexandrine form allude to theirconnectivity to other patte rns (in the prelude and

    Nicos A. Salingaros The structure of pattern languages

  • 8/8/2019 The Structure of Pattern Languages

    3/14

    theory arq V0I4 no 2 2000 151

    2 Six Foot Balcony(#167). Another ofthe patterns - seealso 21a Encapsulatingillustrationb The pattern. Theaccompanyingprescriptive noterecommends that, ifpossible, part of thebalcony should berecessed into thebuilding so that it isnot cantilevered outand separated fromthe building by asimple line and thatit should be partiallyenclosed

    six feet deep

    2b

    postscript), it is difficult to visualize those with ou t aconnective map . Even architects who use p attern stend to be unaware of how patterns link to eachother, so the result ing d esign frequently lacks large-scale coheren ce.In an entirely unanticipated development, thePattern Language format has found a basicapplication in computer pro gram min g. Anyprog ramm ing solution th at reappears in separateinstances may be identified as a 'patt ern', and besubsequ ently re-used as a un it. Patterns are nowrecognized as a powerful th eoretica l framework inwhich to assemble complex computer program s(Coplien and S chmidt, 1995; Gabriel, 1996; Gamma,Helm et al., 1995). The proponents of softwarepatt erns believe tha t patt erns can help to solve awide range of practical problems t hat w ouldotherwise be too cumbersom e or time-consuming.To give readers a bette r sense of wh at is mean t bypatterns con necting to each other, we list someexamples of coup ling. One pattern co ntains or generalizes anoth ersmaller-scale pattern. Two pattern s are comp lementary and one needsthe oth er for completeness. Two patter ns solve different p roblem s that overlapand coexist on t he sam e level. Two pattern s solve the same problem inalternative, equally valid ways. Distinct patterns share a similar structure, thu simplying a higher-level con nection .

    With connective rule s, two different aspects of apatte rn come into play. On one hand, a pattern'sintern al compon ents will determ ine its inclusioninto a larger pattern. On the o ther han d, it is theinterface tha t determ ines overlap, or connection on

    th e same level. Two pattern s on t he same level mayeither compete, loosely coexist, or necessarilycomplement each other.One criticism of Alexandrine pat tern s arises fromtheir clash with existing economic practice andconstruction process. The Pattern Language extendsfrom the scale of surface detail, to th e scale of a largecity, an d covers Alexander's ideas on how to bestimplement a more hum an buil t environment(Alexander, Ishikawa et al., 1977). Some of the u rba npattern s flatly contradict land speculation and theerection of mega towers, while the building pattern smake obvious the need for more structu ral qualityth an today's contra ctors are used to providing . Bothof these points threaten a profit source in th econstruc tion industry. While itis not yet clear how toreconcile those differences, Alexander's critics find inthis a n excuse to dismiss all of the Pattern Languageas impractica l and unrealis tic (Dovey, 1990). That isvery short-sighted.A more serious concern comes from practition erswh o attem pt to apply Alexandrine patterns to shapethe b uilt en vironm ent. The Pattern Language is not ,and was never claimed to be a design metho d an d itis always a struggle to integ rate pa ttern s int o anactu al design project. Architects, however,desperately need a self-contained design method,and, no t finding it in Alexander's theories, will ad optwhatever design method is currently in fashion. Thetools tha t Alexander is proposing are therebybypassed, appe aring useful only in retrospectiveanalysis, which also explains t he Pa ttern Language'srelative lack of impact. Design is treme ndo usly ha rdwork, and I would like to help show how to utilizepatter ns in practice.A set of connected patt erns provides a framew ork

    The structure of pattern languages Nikos A. Salingaros

  • 8/8/2019 The Structure of Pattern Languages

    4/14

    arq V0I4 no 2 2000 theory

    up on w hich any design can be anchored. Thepat tern s do not determine the design. By imposingconstraints, they el iminate a large num ber ofpossibilities whil e still allowing an infinite nu m berof possible designs. The narro win g of possibilities is,after all, an ess ential p art of a practical desig nme tho d. In this case, the rem ainin g choices areprecisely those that connect to hum an beings ei thervisually, emotionally, functionally, or by facilitatingthe ir in teractio ns a nd activities. People havefunda menta l physical and emotional needs thatshould be sat isfied by the bui l t environment , thou ghmo st of th em a re neglected nowadays. Architecturaldesign tha t accomm odates - or, better still, enhances- a framework of Alexandrine patte rns will be felt asmo re 'natura l ' than one which doesn't .Theconnective geometry of urban interfacesIn a l i v ing ci ty , bo un da r i e s de f ine and conn ec td i ff e re n t r e g i o n s , a n d e n c o u r a g e m a n y h u m a nprocesses t ha t mak e the c it y success fu l. W he t he rthese fun c t ions t ake p l ace is l a rge ly a conse quenc e o ft h e g e o m e t r y o f t h e u r b a n b o u n d a r i e s : it h a s t o b eb o t h c r i n k ly a n d p e r m e a b l e . ( In m a t h e m a t i c a lt e r m s , i t i s acc ura te to ca l l suc h a l ine a ' f rac ta l ', s incei t i s ne i th e r co n t in uo us , nor pe r fec t ly smo oth . ) Thenee ded in for ma t ion fo r t h i s a l r eady exi s ts i n seve ralAlexandr ine pa t t e rns , which combine to g ive ade f in i t e u rb an geo me t ry ve ry d i f fe ren t f rom tha tf o u n d i n c o n t e m p o r a r y c it ie s .

    In p rac t i ce , i t i s ve ry cum ber som e to wo rk f rom acom ple t e ca t a logue o f d iscove red pa t t e rn s t o c rea t e apr od uc t . A simp l i f ied co nnect iv e li s t can dras t ica l lyimprove the u t i l i t y o f any pa t t e rn l anguage . Apro ced ure fo r gen e ra t i ng such a m ap i s based on th econ cep tu a l ' ch un kin g ' o f i n fo rma t ion (Mil le r, 1956) .The goa l is t o c lus t e r pa t t e rn s i n to g rou ps o f abo u tf ive or fewer o n ea ch level of scale . Sup pose on en e e d s t o d e s i g n s o m e t h i n g u s i n g a v ai la b l e p a t t e r n s ;p i ck those t h a t a re mo s t r e l evan t t o t he p rob lem a th a n d , t h e n c h o o s e n o t m o r e t h a n a b o u t a d o z e nr e l a t e d p a t t e r n s f r o m a n e x i s ti n g p a t t e r n s c a t a l o g u e .Ident i fy a ver t ica l dimension (e .g. t ime, space , org roup s i ze ) appropr i a t e t o t he p rocess t ha t gene ra t e st h e e n d p r o d u c t , a n d s t u d y h o w t h e g e n e r a ti v eproc ess develo ps as on e moves up t he levels of scale .

    Once you a ssemb le a g r ou p of pa t t e rn s f rom apa t t e r ns ca t a log ue , you can go back and d eve lopo th e r s fo r r e l a t ed p rocesses , wh ich wi l l i nc lu depa t t e rns l e f t ou t i n t he i n i t i a l round . Pa t t e rn g roupsfor d if fe ren t r e su l t s shou ld be sepa ra t e , and n o tconfuse each o th e r ' s c l a r it y . In t he ca se o f u rb aninterfaces, severa l pat terns are di rec t ly re levant . Ih a v e l i st e d t h e m h e r e , n u m b e r e d a s i n A PatternLanguage (Alexander, Ishikawa et al . , 1977).13 . SUBCULTURE BOUNDARY15 . NEIGHBOURHOOD BOUNDARY42. INDUSTRIAL RIBBON53 . MAIN GATEWAYS108. CONNECTED BUILDINGS119. ARCADES121. PATH SHAPE122. BUILDING FRONTS

    124. ACTIVITY POCKETS160. BUILDING EDGE165. OPENING TO THE STREET166. GALLERY SURROUNDThese doz en pat te rns serve as an emp ir ica lfoun da t ion fo r a geom et ry o f u r ba n in t e rfaces .Reversing the order of the patternsAlexander numbered the pattern s according todecreasing size, yet I will reverse the order i n theabove list for our discussion. GALLERY SURROUND[Fig. 3] proposes th at people sho uld be able to walkthro ug h a connecting zone such as a balcony to feelconnected to the outside world. OPENING TO THESTREET [Fig. 4] is the corollary: people on a pavementsho uld feel connected to function s ins ide a build ing,mad e possible by direct openings. BUILDING EDGE[Fig. 5] shou ld be such as to encou rage life, creatingpedestrian nodes and the necessarily crinkly,crenellated geometry that they require. ACTIVTrYPOCKETS [Fig. 6] reveal tha t any pub lic space issuccessful only if its edge con tains andaccommodates successful pedestrian nodes.

    galleries adjacent indoor rooms

    balconies

    open, without glass

    i) M '- straddling the pathpedestrianpftttl

    crenelation

    depth along the edge

    shelter

    Nicos A.Salingaros Thestructure of pattern languages

  • 8/8/2019 The Structure of Pattern Languages

    5/14

    theory arq V0I4 no 2 2000 15 3

    BUILDING FRONTS [Fig. 7] define th e life at the b uiltedge of a street, while u nifo rm set-backs 'alm ostalways destroy the value of the ope n areas betw eenthe buildings'. PATH SHAPE [Fig. 8] requirespedestrian nodes along a path, and these will deformany straight edges into a more fractal form . ARCADES[Fig. 9] connect the inside of buildings with the worldoutside via an intermediate partially-enclosed space;withou t them, the t ransi t ion is too abrup t .CONNECTED BUILDINGS [Fig. 10] create both a

    bou ndar y and a pat h alo ng it , which is destroyed byhaving interm ediate space between the bui ldings.MAIN GATEWAYS [Fig. 11] give significance - bydefining access - to wha t would otherwise be auseless space between buildings. INDUSTRIALRIBBON [Fig. 12] functio ns as one possible way tocreate a wide bo un dar y for separat ing regionscontaining ot he r types of building s. Finally, the twopa tte rn s NEIGHBOURHOOD BOUNDARY [Fig. 13] andSUBCULTURE BOUNDARY [Fig. 14] stress th e ne ces sit y

    paths

    .pocketsof activity connections

    gateway

    no setbacks

    slight angles

    boundary

    ribbons 200 to500 wide

    bulge in the middleroad

    narrow ends

    continuous arcades1 i 111 i 11 nTflYf meeting places

    3-14 a dozen patternswhich serve as anempir ical foundat ionfora geometry ofurban interfaces

    3 Gallery Surround(#166) form s aconnect ing zoneproviding a feeling ofconnect ion to theouts ide world

    4 O pening to the Street(#165) provides aconnect ion betweenthe pavements andthe functions insidebuildings

    5 Building Edge (#160)should encourage l ife,creating space forpedestrian nodes

    6 Activity Pockets(#124). Public spaceonty work s if i ts edgesconta in successfulpedestrian nodes

    7 Building Fronts (#122)define the l ife at thebui l t edge of thestreet

    8 Path Shape (#121)requires pedestriannodes along a path ,deforming straightedges

    9 Arcades (#119)provide a t ransi t ionbetween bui ld inginter iors and theworld outs ide

    10 Connected Buildings(#108) create both a

    boundary and a pathalong it

    11 Main Gateways (#53)give significance t ootherw ise uselessspace

    12 Industrial R ibbon{#42) forms a way o fcreat ing a wideboundary forseparating groups ofother types ofbui ld ings

    13 N eighbourhoodBoundary (#15)stresses thenecessity forcontainment in aliving city

    14 Subculture Boundary(#13) is sim ilar

    gateways

    restricted access

    200 "feet of land

    man-madeboundaries meeting places

    ' natu ral boundaries

    The structure of pattern languages Nikos A. Salingaros

  • 8/8/2019 The Structure of Pattern Languages

    6/14

    arq V0I4 no 2 2000 theo ry

    of conta inm ent in a living city, and show how onezone can destroy an adjoinin g zone if theappropriate boundaries are absent. Together, theabove pat tern s com bine to create the pictu re of aliving city tha t depen ds in large part o n itsconvoluted, permeable interfaces. The informationgathe red by Alexander and his colleagues in pu ttin gtoget her th e Pattern Language offers a conception ofthe u rba n fabric as a highly connected str ucture ,whose subdivisions are defined by complexboundar ies .Some critics may wish to dismiss th e first gro up ofpatt ern s as relevant only to a pedest rian city, whichin their est imation, no longer exists. Quite theoppo site is true. The discussion of this pape r makesit clear that, since hu m an beings are anatomicallygeared for walking as their p rincipal mo de oftranspo rt , these pat ter ns are t imeless and relevanteven if th eir d om ain is restricte d in today's car-dom inate d u rba n landscape. They still applywherever we walk, wh eth er it be in park ing lots,along storefronts, sub urban pavements, or indo orsho pp ing m alls. Decades of suppression by pat tern sfor the automobile network has erased mostpede strian p atte rns (Newman and Kenworthy, 1999).Whenever there is an archi tectural opportuni ty,however, these pattern s re-emerge spontaneously tocreate a living interface.Validation of the patternsAlexander presents the P attern Language as apractical tool, and orders the patterns in roughlydecreasing size. That is the correct ordering w henone is using the m for design, since decisions on thelargest scale have to be ma de first. Nevertheless, tha tpresupposes that the pat terns are understood to betru e in a fund ame ntal sense. The prob lem is thatmainstream architecture never entirely acceptedAlexandrine pat terns; i t was the m ore sensitive andspiritual fringe movements that did. Tovalidate theabove patt erns , they have to be read in th e opp ositeorder: small to large. The hu ma n m ind can com binethe smaller pat terns in to groups; the larger pat ternsutilize these grou ping s and also generate newpropert ies th at are not present in the comp onentpat tern s. The min d is capable of validating thepat tern s subconsciously when we read the pat ternsin a n evolving (small-to-large) order .Even now, mo re th an 20 years after its publicati on,the fundamental significance of the PatternLanguage is hardly appreciated. Many people stillthin k of i t as a catalogue of personal preferences,which is a total misconception (Dovey, 1990). Eventhos e who realize th at each pat tern is establishedeith er th ou gh em pirical observation, or by scientificreas on ing , often fail to see its inevitability. Irecomm end, tho ugh, that you photocopy therelevant patterns from A Pattern Language (Alexander,Ishikawa et al., 1977), and staple them together in th ereversed order. Reading th em wit hou t th edistraction s of all oth er patter ns helps to connectthe m in the reader 's mind , and the n aturalprog ression sm all to large reveals th e con nectionsbetw een successively larger scales. Doing this leads to

    the conclusion th at th e type of urban boundarydescribed is no t simp ly ou r suggestion, but isnecessary for a living city .Quite sepa rate from th e intern al validation offeredby their abi l ity to com bine, what demonstrates thepat terns ' inevi tabi l i ty is their connection tofundamental pat terns of human behaviour andmovem ent . Many h u m an funct ions and interactionsare facilitated by the p ropo sed u rban geometry, andwe could graphi cally li nk behavioural patterns tothese architec tura l pat ter ns directly. In mostinstances, this conn ect io n is revealed as an i ntu itio nthat the patterns for urban boundaries 'feel right '.Alexander based m uc h of the validation for thePattern Language on th is intuitive assessment(Chapter 15 ofThe Timeless Way ofBuilding [Alexander,1979]). wh ich was dis missed as unscientific. Butagraphic and theoret ical basis underl ies this.The smaller t he scale o n wh ich a pattern acts, themore imm ediately i t connects to hum an beings.Architectural pa tte rns o n the hum an range of scaleslc m - i m create a visceral response because we canexperience th em wi th m os t of our senses. Largerpat terns t hat canno t be touched or fel t requiresynthesis and rec ogn it ion ; they become moreintellectual. People wh o have not experienced themin person (in som e regio n of the world where theystill exist) can rarely im agi ne th eir emotionalimpact. This is the reason why the sequence small-to-large works in a val ida tion process: it brings in thestrongest personal conn ect ion at the beginning, andsuccessive pa tte rns bu ild u p on an intuitivelyaccepted base.Patterns and scienceIn the remain der of this paper, I will discuss pat tern sin very general term s, with the intent ion ofdemo nstrat ing th eir inevi tabi l i ty.A pattern is adiscovered sol utio n t ha t ha s been tested for somet ime, and un der varying co ndit ions. Forarchi tectural and urb an p at tern s, the t ime-framecan be several mill enn ia. A pa tter n is not usuallyinvented, so creativity is subordinated here toscientific inq uiry a nd observ ation. Although you canfind novel ways to com bin e and relate patterns,creativity is reserved for the products arising from anapplication of the pattern language, not the process.Since patt erns are derived em pirically fromobservations, they differ from scientific theory,which derives solu tion s start ing from firstprinciples. Nevertheless, discovered pattern s providea phenom enological fo und at ion out of whichscientific theories can grow. Once established, thosetheories explain why some pat tern s work.Sometimes, a pa tte rn m ay arise as an informedconjecture. It has to survive t he intense criticism andscrutiny that are pa rt of the scientific metho d ofvalidation. Alt hou gh p at te rn s are pre-scientific, theyare in fact mu ch b road er t ha n science. A pattern maybe the interse ction of separa te scientificmechan isms. Many pa tte rn s do not yet have ascientific exp lanat ion ; for oth ers tha t do, theexplanat ions may be bu lky an d convolutedcompared with the simplici ty of the pat tern itself.

    Nicos A. Salingaros The structure of pattern languages

  • 8/8/2019 The Structure of Pattern Languages

    7/14

    theory arq V0I4 no2 2000 155

    Medicine, pharmacology, and psychology are basedat least partially on patte rn languages, while theirphenom enological foundation is slowly beingreplaced by a biological/chemical basis.Morphological and scaling rules that apply broadlyacross many different disciplines (West and Deering,1995) are pat tern s tha t are useful indep end ently ofthe particular mechanisms that generate theobserved phenomena.Unfortunately, architecture as a disciplinecurrently has no means of validating anarchitectural patte rn, so the basic mechanism forpattern formation doesn't exist. Architects wh o arenot also trained in the scientific me thod w ill notdistinguish between a design metho d or proceduretha t gives successful results and one t ha t fails; th evalidation process that should follow any proposedsolution does not form part of architecturaleduca tion (Stringer, 1975). The reasons why som ebuild ings fail - in the sense of being un plea sant and

    difficult to use - are never seriously ex amin ed.Consequently, design mistakes tend to be repe atedindefinitely.A philosophical reversal presents an even moreserious imp edime nt to the use of architecturalpattern s. Architecture has changed in th is c enturyfrom being a trade serving hum anity withcomfortable and useful structures, to an art thatserves primaril y as a vehicle for self-expression forthe architect. In the curren t architectural paradigm,the em otional and physical comfort of the user areof only mino r impo rtanc e. Architects resist using th ePatter n Language because they errone ously believe it

    hin ders artistic freedom. Declaring tha t they wish toexpress their creativity freely, they nevertheless forcethemselves to work withi n irrelev ant stylisticconstraints. Contemporary architecture has becomeself-referential, validated only by how well itconforms to some curren tly accepted style, and no tby any objective extern al or scientific c riteria(Stringer, 1975).The nature of a pattern languageIn practice, pa tte rn languages arise from two verydifferent need s: (a) as a way of und erstand ing, andpossibly controlling , a complex system; (b) asnecessary design tools wit h which to bu ildsomething that is functionally and structurallycoherent. To visualize patterns and theirinterconnections, we use a graph representation .Patterns may be identified with n odes in a graph,and the graph is connected by edges of differentlengths [Fig. 15]. A pattern is an encapsulation offorces; a general solu tion to a proble m. The'language' combines the nodes into a norganizational framework. A loose collection ofpatt ern s is not a system, because it lacks con nectio ns.The rules by which t he patt erns (nodes) connectare just as important as the pattern s themselves.Words with out connection rules can not make up alanguage. A coherent comb ination of pattern s willform a new, higher-level patte rn th at possessesadditional properties [Fig. 16]. Not only does eachoriginal pattern work in comb ination as well as it

    did individually, bu t the whole c ontainsorganizational information that is not present inany of its constituent patter ns. A higher-level p atte rncann ot be predicted from the lower-level patte rnsalone. Sticking patterns to gether withou t p rop erordering will no t provide a n overall coherence. Eachcomp onent m ight work individually, but the wholedoes no t work, precisely because it is not a whole.A pattern language is more than just a patternscatalogue. Individual pat tern s are easier to describeth an the ir languag e, yet a catalogue is onlyadictionary. It does not give a script; it has no rul es forflow, internal connections, or ordered substructu res.A patterns catalogue lacks the essential validationthat comes from recognizing the com binatorialproperties in t he languag e. Some patterns willrequire other complem entary patterns forcompleteness, and the allowed comb inations are

    15 Individual patternsgroup to form sixhigher-level patternshaving additionalproperties

    16 Further connectionsorganize thepatterns in Figure 1into a pattern onthenext higher

    level. Newproperties ofthe wholecorrespond tonew symmetries

    The structure of patter n languages Nikos A.Salingaros

  • 8/8/2019 The Structure of Pattern Languages

    8/14

    arq V0I4 no 2 2000 theory

    usually infinite. A languag e tells you which of th emcan be comb ined, and in what man ner, in order tocreate a higher-level patte rn . Drawing an analogywith biological systems, the system works because ofthe co nnectio ns betw een subsystems (Passioura,1979)-Hierarchical c onnections across scalesEvery complex system has a hierarchical structu re;i.e., different processes are occu rrin g on differentscales or levels. Conn ections exist both on the samelevels, an d across level (M esarovic, Macko et al., 1970).The same is tru e for a patte rn lan guage. The'language' generates a connective network by whichthe o rderin g of nodes o n on e level creates nodes at ahig her level. This process goes on all the way up, andall th e way down in levels [Fig. 17]. The cohesiveframework p rovided by the language enables theupw ard tra nsitio n to all the higher levels.We ca nbet ter und erstan d a language i f i t has organizat ionat different levels, becau se each level is shielded fromthe complexity in all the ot her levels.A pat tern language does n ot have a strictlymodular rule st ructure - as would b e the case if thelanguage were defined by only a few basic un its - bu t

    adds new r ules as the scales grow. Higher levels in asystem are dep end ent on all lower levels, bu t n otvice-versa (Passioura, 1979). Even thoug hdisconn ected lower-level patte rns can work wit ho utnecessarily form ing a higher-level patter n, such asystem is no t cohesive, because it exists on on ly on elevel. Each level in a complex hierarch ical system issup port ed by the pro perties of the next-lower level.The comb ination of patterns acting on a smallerlevel of scale acquires new and unexp ectedproperties not present in the constituent pattern s,and these are expressed in a higher-level pattern [Fig.18]. Patterns on higher levels are therefore necessarybecause they incorpo rate new in formation.

    Many failures in describing a complex system aredu e to no t allowing for enoug h levels.A gap betweenlevels disconnects the p atte rn langu age, since thepat tern s on different levels are then to o far apa rt tobe relate d [Fig. 19]. We tend to fall into th is tra pbecause of non-hierarchical think ing. Some urb anpatt erns w ork on th e scale of 100m and c onta inarchitectural p atterns that work on the scale of 1 m,bu t what about the patterns on all the in termed iatescales? An even mo re serious problem is thewidespread associatio n of imp ortan ce with size in

    17 Hierarchicalconnections showhow patterns onhigher levelsdependon those on lowerlevels18 Patterns on one levelcombine to helpdefine a new patternon a higher level19 Two groups ofpatterns are too farapart in scale toconnect effectively

    Nicos A. Salingaros The structure of pattern languages

  • 8/8/2019 The Structure of Pattern Languages

    9/14

    th eo ry arq V0I4 no 2 2000 157

    our cultu re. Working within t hat mindset, itis veryeasy to conc entra te only on the large-scale patter ns(or anti-patte rns), and ig nore th ose on lower levels.That makes it impo ssible to validate patt ernsthro ugh their vertical connections, which areillustrated in Figures 17 and 18.One of the prin cipal metho ds of validating apatter n language is that every patter n be connectedvertically to patter ns on b oth higher and lowerlevels. Damage to a patt ern language can beund ersto od visually, by crossing out any singlepatter n in Figure 17. This will remove t hecoordination of all the linked pattern s below it;moreover, if a vertical relation is one of inclusion,the n obviously those patt ern s below are alsoeliminated. In addition, all linked patterns above thecrossed patt ern are automatically eliminated.Therefore, removing one pat tern w ithou tund erstan ding its connections damages a significantportio n of the p attern language because it alsoremoves at least one vertical chain of patter ns.It is necessary to address a misun derstandin g thatidentifies any multi-level struc ture with an inv ertedtree-like hierarc hical orde ring. In a tree, everythingis ordered from a single nod e above, and no des o nthe sam e level do not link directly. Althou gh som eauth ors use this terminology, tha t is not what ismeant here. Figure 17 shows that th e hierarchy wepropose for pattern languages is not an inverted tree,because it has multip le tops and ho rizontalconn ection s; i.e., several times mor e con nectionsthan a t ree has. A hierarchical inverted tree struc tureis too restrictive, since all com mu nica tion has to passthro ugh higher-level nodes. Inverted tree-likehierarchie s are associated wit h systems that ex erttop-down control (Alexander, 1965).Finding patterns for new disciplinesA new discipline needs to abstract its patte rns as theyappear. It is bui ldin g its own foun dation and logicalskeleton, upo n which future growth can besup port ed. Knowing its basic patt erns early on willspeed up the language's development, and guide it inthe right direction. You may obtain insight into anew field lacking a patte rn lan guage by studyingpatterns from established disciplines. A universalhigh-level structu re is inheren t in all pat ternlanguages. The solu tion space, which is distinct fromthe pa ram eter space, is rarely one-dimension al,which means th at knowing what doesn't workcannot give what w orks simply by doing theoppo site. There may be an infinity of differentopposites. One needs to exhaust th e solution spaceby identifying many neigh bourin g anti-patternsbefore zeroing in on the pat tern itself.Here we need to warn against the destructivetendency in our times of ud ging patternsprematurely using strict criteria such as efficiency,cost reduction, and streamlining. It is not that theseare inapp ropriate criteria, but ra ther tha t they tendto ignore the linkage between patte rns. In o therwords, pattern s in a pattern language depend oneach other i n a complex manner, and a hasty cullingof what are erroneously deemed 'superfluous'

    pattern s may damage the cohesion of the lang uage .Many fundam ental p atterns have be en d iscarded inthe false interest of economy, w ith ou t realizin g th atthey are essential to a system's co he ren ce a nd overallperformance. The long-term con sequen ces of thisare negative, and significant. You may a t tempt tostream line a process after its compl exity is wellunderstood, but not before. Promising new p at terns,and time-hono ured old ones, have be en ruthl esslyscrapped by short-sighted thin kin g, bo rn e out of thebelief that complex systems have to con form to so mesort of 'minimalist design'. This com es from asuperficial underst andin g of how a system work s.The most elegant complex systems are nea rly (butno t perfectly) ordered. Having to ac com mo da tepatter ns on the smaller and inte rm edi ate scales -indeed, actually growing out of th em - t he larger-scale pattern s canno t be perfect in th e sense of beingpur e or too simple. Good design avoids unn ecessa rycomplication. It is balanced between arising out ofloosely organized small-scale pat ter ns , whic h couldlead to somewhat rand om forms o r processes, andpat terns which might pay too mu ch at ten t ion to thelarge scale. Going too far in eith er e xt rem e d amag esth e coherence (and therefore the efficiency) of t hesystem.The general ideas offered h ere pro ve u seful inextending urb an patterns to the electro nic city. Thenotion of an 'intelligent environment' defines theurba n connect ivi ty of the new mil l enn ium . On top ofthe existing path structure governed by Alexandrinepatterns (Salingaros, 1998), we need to develop ru lesfor electronic connectivity (Droege, 1997; Graha mand Marvin, 1996). To define a cohe rent , w orkingurban fabric, the pattern language of electronicconnections (which is only now be in g developed)mu st tie in seamlessly to the lang uag e for physicalconnections. Already, some auth or s mislead inglydeclare that th e city is made redun dan t by e lec tronicconnectivity. Such opinions ign ore new observedpatterns , which correlate electronic no des tophysical nodes in the pede strian ur ba n fabric. Thetwo patt ern languages will mo st likely com ple me ntand reinforce each other.Consistency and connectivityOf the two criteria: (a) interna l consistency, and (b)external connectivity, the second is by far th e mo reimportant . A system's complexity - the extent ofwhich may not be known for som e tim e, if ever - canprevent a new patter n language from having asmooth internal structure. It is essential, however,tha t any patte rn language link to existing languagesat its boun darie s [Fig. 20]. For exam ple, a bui ld ingthat is internally inconsistent wou ld be u nusab le.Once a buildin g has achieved a m in im u m degree ofinter nal consistency, however, ex ter na l connectivitywith other pat terns becomes m or e im por tant . Thepoint is to avoid the isolation of path olog icalsystems, which then survive because they are no tsubject to interactive checks an d balanc es.It is possible to define a set of ant i -pat terns that'clean up' complexity by im po sing rigid, on e-dimensional ideas. Such a langu age could i tself be

    The structu re of patter n languages Nikos A. Salingaros

  • 8/8/2019 The Structure of Pattern Languages

    10/14

    arq vol 4 no 2 2000 theory

    perfectly consis tent internally, hu t it can no t coexistwith ot her pa t tern languages that respectcomplexity. The best example comes fromgovern ment . Fascism and total i tarianism clean upthe messiness of hu m an society, bu t clash wi th ou rmost deeply-held pat ter ns of hu ma n values. In th esame way, any organizat ional pat tern language th atattem pts to create a positive work env iron me nt willnecessarily connec t wi th and provide a tran sitio n toAlexander's archi tectural pat tern language, wh ichdete rmin es b uil t form on all levels of scale(Alexander, Ishikaw a et al., 1977).The archi tectural p at tern SIX-FOOT BALCONY[Fig. 2] helps to illustra te connectivity (Alexander,Ishikawa etal., 1977). Many social pat tern s of familylife, such as si t t ing arou nd a table; eating a meal;child ren playing with toys on the floor; grow ingplants in large pots; outdoor cooking on a charcoalgrill; and so on, can occur on a balcony only if it is atleast 6ft (2m) dee p [Fig. 21]. When a balcony is ma detoo narro w so as to follow some arbi trary designcano n or sim ply to be cheap (which satisfiesintern ally co nsisten t criteria), it fails to conn ect tothe above social pat tern s. Connection here meansaccommodation and inclusion among pat ternsbelon ging to two different languages. Mathem aticalisolation, as in Figure 20, guarantees the physicalisolation of the balcony from potential users.We don 't appreciate how completely archi tecturalpat terns connect to social pat terns; the former makeup a significant p art of the tra dition al culture in anysociety. Losing th em irreparab ly damages t he way asociety functions, because architectural patternshelp to define all the higher-level social pa tter ns[Fig. 22]. Especially am on g the rura l poor, traditio n isthe only way of safeguarding their cul ture. Traditio nembo dies solu tion s evolved over countlessgenerat ions, so design pat terns are connected w ithand have beco me p art of a way of life. This po int h asbee n stressed by Alexand er (Alexander, 1979), and isvery eloq uen tly a rgu ed by Hassan Fathy (Fathy, 1973)(pp. 24-27). Sensitive architects pay attent ion so t ha ttheir designs accomm odate and nu rtu re socialpa t te rns .Sometimes, a pat tern m ight have an un wante dsecondary characte ristic; the same way an in heri tedtrait in a n org anism may be essential for survival, buthave a mil dly neg ative side-effect. The sam e pa tt er n isexpressed as two different features. Attempting toremove the secondary, unwa nted feature (forexample, getting rid of every architectural elem entor social pat ter n th at 'spoils' an overall perfectsymmetry) wit ho ut realizin g what it connects to candestroy the en t ire language. By condemningsecondary features of hu m an pat terns because theyare not co nsisten t with arbitrar y ideas of style, orbecause of som e anti-social aversion, architects havesucceeded in el iminat in g t radi t ional p at ternlanguages aroun d the world.Stylistic rules and the replication of virusesDuring a t ime of crisis, or in the desire to be totallyinnovative, established disciplines sometimeswillingly replace th eir p att ern languages by stylistic

    ru le s . Those are entirely arbitrary , however, comingeit he r from fashion or dogma (someone in au thorityprono unces a rule that is never questioned), or theyrefer to a very specific situati on t ha t does no t applybroa dly. Stylistic rules are incom patib le wi thcomp lex patterns such as the one shown in Figure22. The mechanism by which stylistic rules propagatebears essential similarities to the replication ofviruses. A stylistic rule is usually given as a template,an d pro pon ents are required to replicate it in theenv iron men t. Its success is measured not b y how wellit serves any hum an activity, bu t rat her by how manycopies are produced.Stylistic rules frequently have no connection toh u m an needs: they are just images with a superficialsymb olic content. While some are benign, many arepathological . An information code for built form -

    / V^ Six feet

    Six feet deep.NicosA. Salingaros The structure of pattern languages

  • 8/8/2019 The Structure of Pattern Languages

    11/14

    theory arq V0I4 no 2 2000 159

    for example, 'flat, smoo th, con tinu ou s walls at streetlevel' - enters the mind of a designer either th roug hteaching, or from seeing bu ilt exam ples. Otherwiseintelligen t peop le are easily seduced by simplisticideas in a design me tho d, wh ich is easy to applybecause it eliminates or suppresses naturalcomplexity. That individual then becomes an agentfor replicating the virus. Every time this code isreplicated, it destroys hum an connections in thatregion of the city; the resu lt is obvious because th isparticular virus undoes all the patterns forconnective urba n interfaces discussed previously.By contrast, a patte rn is not dictated or forced, butarises out of use, an d is accepted o n its benefits. Itfacilitates hum an life and intera ction s, and has tocontin ually stan d u p to tests of its efficacy in thisrespect. An essential difference is tha t, because of its

    I

    20 The enclosed patte rncandidates areinternally consistentbut fundamental lyflawed, because theyfail to connect toexternal patterns

    21 Six-Foot Balco nypattern shows howmany social patternsof family l i fe canoccur wi thin such aspace

    22 Architecturalpatterns that pai rwith social patterns(solid) furthercombine to create asocio-archi tecturalpattern on a higherlevel

    underlying forces, no architectural p atte rn can bereprese nted as a simple visual image. A pattern solvesa complex problem ; it is not a template to bemindlessly cop ied. It is far easier to repr od uce avisual template tha n to solve a fund ame ntal designproblem, however, because the former requi res noreasoned thou ght ; only intuitive mat chin g. Theintellect does no t need to work, and the designer canwithdraw from the responsibility of mak ing difficultdecisions abou t the complex interaction s betw eenbuilt forms and h um an activities. Partly as a result ofthis shift, architectural design is now heavilyoriented towards visual templates defined by d esignstyle.Many stylistic rules are anti-p attern s: they areneith er accidental, nor the simple preferences of anindividual. They intentionally do the opp osite ofsome trad itio nal p att ern for the sake of novelty. Bymasquerading as 'new' patterns, they misuse apatte rn language's natu ral process of rep air todestroy it. Patterns work via cooperatio n to build upcomplex wholes th at coexist and co mpete in somedynamic balance. By contrast, stylistic rules tend tobe rigid and unacco mm odatin g. Their repli catio n inman y cases fixes the geom etry of built fo rm so as toexclude hu ma n pat terns. Any single stylistic ru le iscapable of suppressing an entire chain of link edpat tern s on m any different scales [Fig. 17]. Adestructive stylistic rule, like a virus, is aninform ational code that dissolves the com plexity ofliving systems.Today's architects are train ed to use a lim itedvocabulary of simple forms, materials, and surfaces.Their possible combinations are insufficient to evenapproa ch th e struc ture of a language. This replacesan accumulated l i terature of pat terns corresp ondin gto words, sentences, paragr aphs, chap ters, and book sthat encapsulates meaning from h um an experienceand life. Few people realize the enormousconsequences on society of adoptin g a p arti cul ardesign vocabulary. Decisions concerningarchitectural style affect the surro un din g cu lture ;contrary to what is widely proclaimed, one person'svisions are no t restricted to a building as a single artwork. A single visual temp late can eventually destroya cu ltu re jus t as effectively as a deadly virus .Evolution and repair of pattern languagesValidated pattern s are more or less per ma nen t, yetthere exists a process of repair and repla cem ent. Nowand the n, we may play devil's advocate an d ig no re o ldsolutio ns to see new, innovative ones in an olddiscipline.A new pat tern is superior if it increasesthe connectivity with the majority of establishedpat terns compared with the old pat tern it isreplacing. It could have a broade r con text, orsupersede several older pattern s, thu s tigh ten ing th elangu age. This is a process whose goal is tostrengthen an existing pattern language by repairand evolution, to preserve accum ulated w isdom bykeeping it relevant to changing n eeds.Much less frequently, a parad igm s hift o ccurs tomake an entire pa ttern language irrelevan t: e.g.,horse-drawn vehicles are replaced by auto mo bile s.

    The structure of pattern languages Nikos A. Salingaros

  • 8/8/2019 The Structure of Pattern Languages

    12/14

    arq V0I4 r>0 2 2000 theory

    That does not inval idate the pat ter n languageshowing how to create the former; i t just m akes t hatend p rodu ct less desirable. While the technolog y andmaterials changed, however, man y pat t erns weresaved almost int act in going from carriages to cars. Ingeneral , the adop tion of innovat ion is great lyfacil i tated by minim izing the percept ion of change;and consequently the nu mb er of pat ter ns tha t needto be replaced. It is wasteful to thro w ou t a rep osito ryof pat tern s, some of wh ich may have beenestablished over millennia.The introd uct ion of a new pat t ern langu age needno t displace an older o ne entirely. Coexistence ofcomp eting or comp lementary pat te rns is oftendesirable and even necessary, especially if the newpatterns occupy different positions in the hierarchy(by actin g on different scales). If pro per ly co nne cted ,they will lead to a riche r and m ore stable comp lexsystem. Pat terns for the automobile t rans port at ionnetw ork were falsely believed to be threa ten ed bypat te rns for pedestrian and m ass-transi t networks.On the basis of this misundersta nding , urb anplann ers and car manu facturers simply suppressedth e latter (Newman and Kenworthy, 1999). Nowadays,we are beginn ing to understand that a balancedcoexistence of all three languages - describin gpedestrian, automob ile, and mass-transi t movem ent ,respectively - is a necessary prerequisite for acomprehensive t ransp ortat io n system (Salingaros,1998).

    A few pat terns mig ht work equal ly well ondifferent levels, thoug h m ost pat tern s ' con textestablishes the ir place in a parti cular scale of thepat t ern langu age. Some pat terns can be moved up ordown vert ical ly with in a language. Such a pro pertyleads to economy in a pat te rn language th rou gh self-simil ar scaling, whic h means th at one scale looks t hesame as anoth er scale wh en m agnified. A pa t t e rnlangu age th at develops coherence over time may alsodevel op a degree of self-similar scaling as a res ult ofth e con nection s across levels.As the ensemble ofpat tern s evolves a cooperative struct ure , driven bythe al ig nme nt of pat terns (or ant i -pat terns) ondifferent levels, it creates unexpected similarities.Thus, each level of a cohere nt str uct ure expresses aproper ty tha t is characteristic of the wh ole.The importance of detailA language requires patterns on as many levels as ittakes to conn ect to nat ur al processes. Every level isimpo rtant by itself. In any complex system, detai l ispa rt of th e lower scales in a hierarchy. If these ar eunco nnec ted, or missing, then th e system is no tcoherent , and cannot work (Mesarovic, Macko et al.,1970). Neglecting a patt ern because it is on a lowerlevel handicaps t he entire stru cture. Itis no t alwaysobvio us wh at th e lowest level of a system is u p o nwh ich all the h igher levels depend. Detail tha t is par tof a scaling hierarchy will be connected to all high erlevels of complexity, an d is no t just 'added on'.Physical forms have stru ctu ral features on differentscales as a result of inter nal a nd ex tern al forces. Fromthe m icroscopic to the macroscopic throu gh allintermediate scales, different levels of scale coopera te.

    In th e design of build ings , the re are several scales -correspondin g to th e h um an range of scales, 1cm to1m - th at are difficult to justify purely on structu ralgrounds. Yet, to define a conn ected hierarch y ofscales, thos e scales have to be presen t in the stru ctur e(Salingaros, 2000). Therefore, either the designshould allow the emergence of structure andsubdivisions on th ose scales, or substr uctu re has tobe intention ally generated on those scales. This needcreates tradition al 'orn am ent' and all the pattern stha t gener ate it (Alexander, Ishikawa et al., 1977;Salingaros, 1999). The appropriate orn ament isessential for a large form to be coheren t (Salingaros,2000). An analysis of structural coherence arisingfrom a linked hie rarchy of scales reveals th enecessity for orna me nt, t hou gh nowadays,orn am ent is discordant because it is unrelatedto the larger form.Detail is a separate ques tion. The smallestperceivable detail at arm's le ngth goes down to0.25mm, which relates to a visual system such as atextile or a comp uter display. While such detail isavailable in richly-textured materia ls, itis usuallythe scales between texture and orn am ent(im m-i cm ) that are missing from contemporarybuildings. Our minimalist design tradition removesthe interm ediate an d smaller scales from b uiltform. After half a centur y of train ing in th is idiom,we tend to forget th at t he best-loved architectu re(Modernist included ) works especially well on th esescales. People need to conne ct to struc ture o n everyscale.ConclusionPattern languages encapsulate hu ma n experience,and help us cope with complexity in ourenviron men t. They apply to everything fromcomp uter prog rams, to buildings, to organizations,to cities. A civilization's pa tte rn langu ages are oftensynonymous with its technical and cultural heritage.New spheres of hu m an endeavour develop their ownpatte rn language, which mu st link to existingpattern languages in related fields. Individualpattern s are validated empirically over t ime. Thelanguag e itself will be on th e rig ht track if it evolves aconnective structu re th at incorpo rates scaling andhierarchy. Architecture and u rban design in th etwe ntieth cen tury rely on a set of stylistic rules tha tfail to connect to patte rns of hum an life. People havebeen tau gh t by schools, critics, television, andmagazin es to prefer abstract visual forms, and toignore the fact th at env ironm ents generated by suchtemplates cannot accommodate their ownbehavioural patter ns. An example of this wastraced to a fundamental misunderstanding abouturb an geometry. It is believed that the removalof urban interfaces would help to create thecontem porary city, bu t it has seriously damagedit instead.

    This paper argued t hat p atterns provide anecessary foundation for any design solution toconnect with h um an beings. Contradicting themdisconnects the built form from peo ple. Thisconclusion has profound consequences forNicos A. S alingaros The structure of pattern languages

  • 8/8/2019 The Structure of Pattern Languages

    13/14

  • 8/8/2019 The Structure of Pattern Languages

    14/14

    After the CityLars LerupAn architect's view of the newmetropo litan consciousness and thesuburban metropolis as the futurefrontier.216 pp., 52 illus. $24.95

    The CIAMDiscourseon Urbanism,1928-1960Eric MumfordThe first complete history of a m ovementwhose mission was to revolutionizearchitecture and create an agenda formodern urbanism.408 pp., 107 illus. $45The Unknown CityContesting Architectureand Social Spaceedited by lain Borden, Joe Ken,and Jane Rendell with Alicia PivaroEssays on architecture as narrative andurban space as experience and the newgeographies they create.600 pp., 150 illus. $50

    TheProvisional CityLos Angeles Stories ofArchitecture and UrbanismDana CuffHow the modern city is formed throughan en dless cycle o f upheavals,demolit ion, and debate.400 pp., 188 illus. $40The ArchitecturalDrawings ofAntonio daSangallo theYounger andHis CircleVolume II: Churches, Villas,the Pantheon, Tombs,and Ancient InscriptionsChristoph L Frommel, general e ditorNicholas Adams, editorThe second volume of a trilogy containingthe com plete drawings of the famoussixteenth-century Italian workshop.512 pp., 509 illus. $95

    To order call 800-356-0343(US & Canada) or 617-625-8569.Prices subject to change without notice.

    Perspecta 3 1The Yale Architectural Journal"Reading Structures"edited by Carolyn Ann Fougand Sharon L JoyceExamines the relationship betweenarchitectural design and structuralengineering through the processes andproducts of both disciplines.160 pp., $20.00 paper

    nnw inEccentric SpacesRobert Harbison"[MJakes me want to rush out in everydirection at once and reexamine all I haveever seen. You can hardly ask a book to domore than tha t." New York Times192 pp. $15 paper

    The Springboardin the PondAn Intimate Historyof the Swimming PoolThomas A. P. van Leeuwen"A bracing intellectual plunge."Los Angeles Times330 pp., 244 illus., 26 color $35 paperArchitecture Theorysince 1968edited by K. Michael Hays"This collection insistently raises importantquestions and helps us elucidate problemsthat might not have otherwise occurred to us.' Architecture824 pp., 70 illus. $30 paper

    Architectureand ModernityA CritiqueHilde Heynen"[A] very helpful syn thetic overview of theprincipal positions in critical theory'sarguments over mod ernity." South Carolina Review280 pp., 103 illus. $22 paperLe Corbusier,the Noble SavageToward an Archaeologyof ModernismAdolf Max Vogt"Scholarly and w itty." Building Design381 pp., 255 illus. $25 paper

    http://mitpress.mit.eduThe MIT Press