University of Copenhagen

Staged, yet unstaged Sociological inquiries into space and micro-interaction Liebst, Lasse Suonperä

Publication date: 2014

Document version Other version

Document license: Other

Citation for published version (APA): Liebst, L. S. (2014). Staged, yet unstaged: Sociological inquiries into space and micro-interaction. Department of Sociology,University of Copenhagen.

Download date: 03. okt.. 2021 LASSE SUONPERÄ LIEBST SUONPERÄ LASSE Department of Sociology faculty of social sciences Univeristy of Copenhagen · Denmark university of copenhagen PhD Thesis 2014 · ISBN 978-87-7611-771-9

Lasse Suonperä Liebst Staged, yet Unstaged Sociological Inquiries into Space and Micro-Interaction

Staged, into yet Space Unstaged · Sociological Inquiries and Micro-Interaction Staged, yet Unstaged Sociological Inquiries into Space and Micro-Interaction Lasse Suonperä Liebst · PhD Thesis · 2014

Staged, yet Unstaged

Staged, yet Unstaged Sociological Inquiries into Space and Micro-Interaction

Lasse Suonperä Liebst

PHD THESIS DEPARTMENT OF SOCIOLOGY UNIVERSITY OF COPENHAGEN

Staged, yet Unstaged. Sociological Inquiries into Space and Micro-Interaction © Lasse Suonperä Liebst

Ph.d.-afhandling Sociologisk Institut Det samfundsvidenskabelige fakultet Københavns Universitet

Opsætning: Lasse Suonperä Liebst Sats: Garamond Omslag: SL Grafik Forside: Flyfoto af Roskilde Festival 2011 Tryk: SL Grafik, Frederiksberg C., Danmark (www.slgrafik.dk) ISBN: 978-87-7611-771-9

Indleveret til bedømmelse: 11. april 2014 Offentligt forsvar: 4. september 2014

Vedlejder: Lektor Inge Kryger Pedersen, Københavns Universitet

Bedømmelsesudvalg: Professor Peter Gundelach, Københavns Universitet Professor Mats Franzén, Uppsala Universitet Lektor Henrik Harder, Aalborg Universitet

Indhold

Forord viii Introduktion 1 Projektets tilblivelse og tilgang 5 Rummets fysik 5 Rum og mikro-interaktion 9 Andre broer 15 Scener og ikke-scener 18 Undersøgelsesformål 21 Durkheimiansk rumsociologi: Bidrag til en rekonstruktion 22 Den sociale morfologi 24 Rum, densitet og følelsesmæssig smitte 30 Balkanisering eller akkumulering 33 Den metodiske tærskel 38 At måle mikro-rummet 41 Space syntax 45 Rumlig netværksanalyse – eller at gætte den goffmanske restaurant 50 Distancer og tilgængeligheder 53 Netværkshandlen 57 Ikke-diskursiv og ikke-lokal 64 At måle sig med rumstatistikken 68 Kontekst og kvantificering 70 Skitse af artikelbaserede kapitler 76 Bibliografi 79 Kapitel 1: Space Syntax Theory and Durkheim’s Social Morphology 92 Introduction 92 Revisiting Durkheim’s Social Morphology 94 Explaining Durkheim’s Non-Explanation 97 A Sociology of the Elementary Cell 99 Space Syntax and the Shortcomings of the Macro Wing 104 The Micro-Ritual Logic of Space 109 Conclusion 114 References 115 Kapitel 2: Ambiance of the Machine 118 Introduction 118 Path and Method 120 Physiognomy and Morphology 123 Quality and Quantity 126 Emotion and Affect 130 Conclusion 134 References 134

Kapitel 3: Phenomenology of the Movement Economy 137 Introduction 137 The Movement Economy Theory 138 Economic and Phenomenological Attractions 139 Statistical Hypothesis 140 Data and Methodology 141 Spatial Data 141 Measuring the Movement Economy 143 Survey and Sampling 143 Analytic Strategy 144 Measures and Definitions 144 Results 145 Movement Economic Analysis 145 Multilevel Estimation 147 Discussion and Concluding Remarks 149 References 150 Kapitel 4: Dissecting Collective Effervescence 153 Introduction 153 Theoretical Frameworks and Expectations 156 Interactional Effect 157 Selection Effect 158 Environmental Effect 159 Method and data 161 Dependent Variable 163 Interactional Measures 163 Spatio-Environmental Measures 164 Selection Measures 165 Control Variables 166 Results 166 Discussion and Conclusions 171 References 172 Kapitel 5: Hot Spots of Festival Co-Creativity 177 Introduction 177 Theoretical Framework and Expectations 180 Data and Method 182 Statistical Techniques: Quadrant Analysis and Spatial Regression 184 Measures and Definitions 184 Results 187 Quadrant Analysis: Exploring the Co-Creative Hot Spots 187 Movement Analysis: Measuring Movement Density by Proxy 187 Logistic Analysis: Explaining the Co-Creative Hot Spots 188 Discussion and Conclusions 190 References 193

Konkluderende bemærkninger og resumé 197 Mikro-rumligt inviterede kropsteknikker 197 Begrænsninger og implikationer 202 Dansk resumé 206 English Summary 210 Bibliografi 213 Appendiks 215 #1: Nørrebro-spørgeskema 215 #2: Festival-spørgeskema 216 #3: Segment angular analysis af Roskilde Festival 227 #4: Visibility graph analysis af Roskilde Festival 228

vii

Forord En lang række personer har ydet mig uvurderlig hjælp i forbindelse udarbejdelsen af nærværende afhandling. Jeg skylder jer alle min uforbehold- ne tak. Dette gælder i første omgang min fremragende og engagerede vejleder ved Sociologisk Institut, lektor Inge Kryger Pedersen, der igennem årene har udgjort afhandlingens sociologiske og solidariske substratum. Under mit otte måneders udenlandsophold ved The Bartlett School of Architecture, University College London, har jeg været så privilegeret at have lektor Sam Griffiths som vejleder, der på afgørende punkter fastholdt mig på det socialmorfologiske spor. Professor Bill Hillier fra The Bartlett skal ligeledes takkes for at afsætte tid til at blive interviewet om space syntax-tilgangens ophav og fremtid. Opholdet ved The Bartlett blev endvidere beriget af mange og lange faglige snakke med Wafa Al-Ghatam, der også har været til uvurderligt teknisk assistance i forbindelse med afhandlingens anvendelse af GIS- og space syntax-teknikker. Opholdet ved The Bartlett blev gjort praktisk muligt med hjælp fra en række generøse fondsbidrag: Augustinus Fon- det, Oticon Fondet, Dansk Tennis Fond, Ove Arup Foundation (for- midlet af Anglo Danish Society) samt Torben og Alice Frimodts Fond. Professor Lars Marcus har gennem en årrække været en vigtig rumte- oretisk sparringspartner i forbindelse med mine utallige ugelange ophold ved forskningsgruppen Spatial Analysis and Design, KTH Royal Institute of Technology. I forbindelse med disse ophold har Ann Legeby endvidere tålmodigt introduceret mig til de tekniske aspekter af place space-red- skabet, hvilket har været en afgørende betingelse for, at afhandlingens metodiske drejning kunne realiseres. I tiden før og under ph.d.-forløbet har jeg været tilknyttet forskningsgruppen Kultursociologi ved Sociologisk Institut, hvor diskussionerne med Henning Bech, Charlotte Bloch, Allan Madsen, Bjørn Schiermer, Anna Sofie Bach, Anette Stenslund og Kristine Samson har sat sig betyd- ningsfulde aftryk på afhandlingen. Særskilt skal der lyde en tak til forsk- ningsgruppens koordinator, Marie Bruvik Heinskov, for at insistere på den fænomenologiske sensitivitet, de vilde diskussioner og den altid hjer- tevarme støtte. Den faglige sparring med mine gode ph.d.-kollegaer ved Sociologisk Institut og SFI har været uvurderlig. Dette gælder ikke mindst i forbindelse med afhandlingens statistiske undersøgelser, hvor Kristian Karlson har været mere end behjælpelig, og uden hvis hjælp jeg ikke var kommet i mål. At dette er sket skyldes også Anders Trolle og Stefan Andrade, der begge har været til stor statistisk hjælp. Ydermere har Stefan været en værdsat socialmorfologisk allieret. Hakan Kalkan skal også takkes for mange lærerige arbejdsaftener og for at påminde den vordende socialsta-

viii

tistiker om relevansen af etnografi. På Sociologisk Institut har Poul Po- der været en givtig IR-teoretisk sparringsparter og Heine Andersen og Cecilie Juul Jørgensen har givet mig skarp kritik på et evalueringsseminar. I forbindelse med dataindsamlingen på Roskilde Festival har flere personer spillet en afgørende rolle. Laurine Albris inspirerede mig til at arbejde med festivalen, var medorganisator på dataindsamlingen og for- midle kontakten til festivalen. Søren Mariegård og Signe Brink Pedersen fra Roskilde Festival skal takkes særskilt for at have støttet projektet med 11 Roskilde billetter til vores seje tværfaglige team af interviewere, som der også skal lyde en tak til. Og en lige så stor tak skal der lyde til de studerende på faget Rummet er maskinen, som jeg udbød på Sociologisk Institut, foråret 2012, og som i undervisningen bidrog til den kollektive indsamling af det Nørrebrodatasæt, der først indgik i deres eksamensop- gaver og nu i denne afhandling. Gode venner og bofæller har, også selv om de måske ikke er klar over det, sat deres præg på afhandlingen: Nis Bjarnhof ved at insistere på tænkningen; Jon Ploug Jørgensen ved at insistere på at εν οίνω αλήθεια; Mikkel Dehlholm ved at insistere på det ikke-puritanske; Sam Wilkinson ved at insistere på gæstfriheden; Silas Boje Sørensen og Jonas Aller Kjeldsen ved at insistere på festivalen; kollektivet Rabarberpaladset ved at insistere på et hjerteligt og alternativt rum; Christian Hjortkjær ved at insistere på det ufattelige; Christina Ytzen ved at insistere på at insistere. Ydermere skal Ytzen og Hjortkjær også takkes særskilt for deres grund- dige kommentering af afhandlingens manus. Uden min vidunderlige mor, far og lillebror Rasmus var intet blevet til noget: Jeg føler mig så heldig, at jeg har en familie, hvor kærlighed og grundighed dyrkes med så let perfektion. Endeligt har min kæreste, Line Vistisen, været min kærlige støtte gennem hele forløbet. Du er afhandlingens sine qua none. Jeg ved ikke, hvor mange gange jeg har nydt din sociologiske faglighed, når vi har udviklet argumenter på løbeturene rundt om i Hyde Park; jeg ved ikke, hvor mange gange du har samlet mig sammen, når det hele virkede dumt; jeg ved ikke, hvor mange gange du havde ret, når jeg var stædig; en ting ved jeg dog: Jeg elsker dig.

Introduktion

Space does not direct events, but it does shape the possibility (Hillier 1996: 155).

DENNE ARTIKELBASEREDE PH.D.-AFHANDLING undersøger det byggede rum som en fysisk mikro-kontekst for sociale mikro-interaktionsprocesser. Afhandlingen er sociologisk, men et sådan socialfysisk ærinde befinder sig på kanten af, hvis ikke ligefrem på kant med, en traditionel definition af sociologiens emnefelt. Dette kan indledningsvis forklares med reference til den tyske sociolog Max Webers’ (1978: 22) klassiske forstå- else af sociologien som en videnskab, der foretager en fortolkende forklaring af former for social handlen, hvor individer relaterer sig menings- fyldt til andre individers adfærd. En sådan meningscentreret definition af sociologien kommer også til udtryk, når Weber definerer social handlen og sociologien negativt med reference til den fysisk-rumlige hændelse, at to cyklister kolliderer: “Social action does not occur when two cyclists, for example, collide unintentionally; however, it does occur when they try to avoid the collision or sock one another afterwards or negotiate to settle the matter peacefully.” (Weber 1978: 1375). Givet fraværet af mening skal en sådan hændelse snarere sammenlignes med en begivenhed i naturen; det lade sig være et jordskælv eller en stormflod, der har en eksistens og effekt hinsidiges meningsfortolkning og sociologisk forklaring. Afhandlingens kerneargument kan spidsformuleres som, at en sociologisk undersøgelse af forholdet mellem rum og mikro-interaktion kan og skal forholde sig analytisk til de fysisk-rumlige mekanismer, der øger sandsynligheden for et sådan sammenstød mellem to cyklister. Det forhold, at vi har at gøre med en hændelse uden mening eller intention, skal ikke forveksles med, at vi står over for et fænomen, der unddrager sig sociologisk forklaring og relevans. Det teoretiske og metodiske begrebsapparat, som afhandlingen udvikler, muliggør netop en sådan rum-fysisk forklaring, hvis rumsociologiske intuition jeg her indledningsvis vil for- mulere på følgende måde: Rummet determinerer ikke cyklisternes sam- menstød, men øger sandsynligheden for denne hændelse ved at fortætte antallet af cyklister i bestemte gaderum. En sådan fortætning af cyklistbe- vægelser hænger tæt sammen med, hvor centralt de enkelte gaderum er lokaliserede i det omliggende gadenetværk. Gadenettets netværksmæssige egenskaber og effekter er et nøglefokus for afhandlingen: Høj netværks- centralitet resulterer i en højere bevægelsesaktivitet, herunder strømme af cykler, der således har en højere sandsynlighed for at støde sammen i de

gaderum, hvor det rumlige netværk har fortættet deres antal (se Hillier 1996: 155; Geyer et al. 2006).1 Med dette rumsociologiske argument befinder afhandlingen sig i visse henseender tæt på sociologiens pågående “rumlige vending”, der er karakteriseret ved en udtalt interesse for, hvordan det fysiske rum ikke blot er en container for et socialt meningsindhold, men også gør en forskel ved sociale interaktioner (Warf og Arias 2008; Bjerre og Fabian 2010; Urry 2007; Farias og Bender 2012; Graham og Marvin 2001; Thrift 2008; Amin og Thrift 2002; Law og Mol 2001; Gieryn 2000; 2002). Den rumlige vendings betoning af, at rummets fysik må indtænkes som en aktiv social medspiller, afspejler mere generelt sociologiens øgede interesse for materielle objekter. Ekstraordinært indflydelsesrig er her den franske sociolog Bruno Latours (1996; 2005) aktør-netværksteori (ANT) og dennes argument om, at humane og non-humane agenter har en symmetrisk type og grad af handlingspotentiale. I dette perspektiv skal sociologiens grænse ikke, som det var tilfældet hos Weber, trækkes ved den socialt meningsfulde handlen; sociologien skal også inkludere non-humane ting og non-sociale former for aktivitet. Hvad enten det fysiske rum begrebs- liggøres som et “non-representational aspect” (Thrift 2008), et “mobility system” (Urry 2007) eller en “networked intrastructure” (Graham og Marvin 2001) står den nyere rumligt orienterede sociologi direkte på skuldrene af Latours non-human agenter. Arbejder man sociologisk med det fysisk-materielle rum, vil det i dag ofte underforstås, at man tager afsæt i et ANT-perspektiv. Dette begræn- ser den rumsociologiske fantasi fra at engagere sig i andre teoretiske perspektiver: Latours non-human agenter er ikke afhandlings teoretiske udgangspunkt. Afhandlingen argumenterer for, at det fysiske-rumlige objekt også lader sig studere af en anden vej end den, som Latour anviser, og som nutidens rumligt vendte sociologi primært har fulgt. Afhandlin- gen udvikler således et rumsociologisk perspektiv, der trækker på den franske sociologis grundlægger, Émile Durkheim, hvilket burde være utænkeligt set fra ANT-traditionens perspektiv. I Latours udlægning er ANT en antidurkheimiansk position, hvilket ikke mindst tilskrives det forhold, at Durkheims teorier angiveligt er et arketypisk udtryk for socio-

1 Når afhandlingen taler om ’rum’ refereres der, med mindre andet er anført, til den måde, hvorpå byggede rum former et fysisk netværk af relationer (f.eks. de netværks- strukturelle egenskaber ved byens gadenet). Formuleret negativt, er det aspekt, som afhandlingen fremdrager som afgørende ved rummet, ikke dets æstetiske overflader, symbolsk tilskrevne mening eller den sociale brug, der finder sted i rummet. Netværket er nøglen. Afhandlingens netværksteoretiske forståelse af rum trækker, som det vil blive uddybet teoretiske og metodisk nedenfor, i høj grad på den såkaldte space syntax- tilgang (Hillier og Hanson 1984; Hillier 1996).

logiens ontologiske eksklusion af de non-humane tings sociale relevans (Blok og Jensen 2011: 106; Urry 2000: 18; Oppenheim 2008). I denne kritiske fremstilling af den durkheimianske tradition er tingslighed noget som den ontologisk altoverskyggende sociale orden kan tilskrives af epistemologiske grunde, sådan som det kommer til udtryk i Durkheims (1982) diktum om at betragte “sociale fakta som ting”. Latour ønsker derimod at puste ontologisk liv i de non-humane ting og formulerer så- ledes sin symmetriske ontologi som en omvending af Durkheims epistemologiske diktum: “treat things as social facts.” (Latour 1996: 240)2 Afhandlingen argumenterer for, at dette opgør med den durkheimianske tradition er problematisk i mindst to henseender. For det første er Latours afvisning af Durkheim kun mulig, fordi han negligerer de mate- rialitetsstudier som den durkheimianske tradition har udført under navnet den sociale morfologi. Den durkheimianske tradition befinder sig, som fremhævet af Lindemann (2011), tættere på et begreb om non-human agency end, hvad kritikerne giver udtryk for. Den sociale morfologi placerer spørgsmålet om rum og materialitet i centrum af den durkheimianske tradition, der kontinuerligt har argumenteret for, at alt socialt liv hviler på, og aktivt præges af, et “materielt substratum” (Durkheim 1994; Mauss 1979; 2003; Halbwachs 1960; Goffman 1956; 1963, 1971, 1974; Collins 1988; 2004). For det andet er denne teorihistoriske negligering af den sociale morfologi uheldig, idet den durkheimianske rumsociologi, sammenholdt med ANT-tilgangen, har analytisk øje for andre følelsesmæssigt-relationelle aspekter af det sociale livs samspil med rummet. Dette vil jeg forklare med henvisning til Krarup og Bloks (2011) understregning af, at Latours fremlæggelse af den symmetriske ontologi som en teori om “interobjektivitet” implicerer en afvisning af en sociologisk teori om intersubjektivitet. I ANTs perspektiv handler interaktionsprocesser ikke intersubjek- tivistisk om, hvordan to eller flere individer deler et følelsesmæssigt- relationelt bånd, men interobjektivistisk om, hvordan interaktioner er muliggjort af de non-humane tings mellemkomst.

2 Et andet, empirisk begrundet, argument imod at genlæse Durkheims rummorfologi finder vi hos Castells (1996), der argumenterer for, at en sådan morfologisk forståelse af ’space of places’ bør afløses af et begreb om ’space of flows’, der er bedre egnet til at indfange netværkssamfundets post-stedslige og telemedierede karakter: “Networks constitute the new social morphology of our societies (...).” (Castells 1996: 500). Et sådan argument for det sociales tiltagende stedsløshed er imidlertid blevet kritiseret af eksempelvis Sampson (2012) og Boden & Molotch (1994), og afhandlingen befinder sig teoretisk og empirisk på linje med disse kritikere: De socialt definerende aspekter af det senmoderne liv er stadig knyttede til stedsligt forankrede ansigt-til-ansigt udvekslinger.

Vandenberghe (2002) har, efter min vurdering, korrekt kritiseret ANT’s teori om interobjektivitet for at underbelyse det genuint humane ved sociale interaktionsprocesser som et begreb om intersubjektivitet netop indfanger (se også Jerolmack and Tavory 2014). Mennesker er, i modsætning til non-humane ting, forbundne af sociale og følelsesmæssi- ge bånd. Hvad der taler for den durkheimianske traditions rumsociologiske relevans er, at socialmorfologien netop interesserer sig for, hvordan det sociales intersubjektive processer formes af disse udvekslingers morfologiske former (Emierbayer 1996; Collins 2004). Den sociale morfologi inkluderer en teori om intersubjektivitet, hvis følelsesmæssigt-relationelle processer således både studeres i egen ret og som formet i interaktionernes samspil med det rumligt-materielle grundlag (Durkheim 1994). For så vidt, at den forskningsmæssige interesse, som det er tilfældet i denne afhandling, specifikt angår samspillet mellem intersubjektive mikro-interaktioner og det fysiske rum, så udgør socialmorfologien et oplagt, men overset, rumsociologisk udgangspunkt. Afhandlingen vil dog ikke fortabe sig i teoretiske diskussioner og modstillinger af ANT og den sociale morfologi. En sådan teoretisk opta- gethed ville potentielt aflede opmærksomheden fra den mangel i den rumsociologisk litteratur, der tydeligst forhindrer rumsociologien i at bedrive tidssvarende analyser af det fysiske rum: Nemlig manglen på dedikerede rum-metodiske værktøjer. Denne påfaldende metodologiske mangel angår mest udtalt kvantitative metoder til at analysere rummets mikro-skalaegenskaber og effekter: dvs. det rumlige skala-niveau, hvor mikro-interaktionens ansigt-til-ansigt udvekslinger konkret finder sted (Logan 2012; Hillier og Raford 2010). Mangel på mikro-kvantitative værktøjer står i kontrast til det eksisterende arsenal af kvantitative metoder, der er udviklet i slipstrømmen på humangeografiens “kvantitative revolution”, men som desværre kun sjældent operationaliserer rummet med en sådan mikro-skala specificitet (Haggett et al. 1965; Morrill 1970; Goodchild og Janelle 2004; Logan et al. 2010; Porter og Howell 2012). Der kan identificeres ansatser til udvikling af kvalitative metoder til at studerede sådanne mikro-rumlige processer (Urry 2007: 39-42; Gieryn 2000; Harper 2012). Men kvalitative metoder kan, som vi skal se, ikke stå metodologisk alene, når vi specifikt har at gøre med rumsociologiske analyser af det fysiske rum. Særligt hvis rummet, som afhandlingen slår til lyd for, begrebsliggøres som fysiske net- værk af relationer, vil disses ikke-diskursive og ikke-lokale egenskaber kun vanskeligt kunne gøres til genstand for kvalitativ, metodisk analyse. Dette kalder på en mikro-metodologisk kvantificering af rummet, og det er således afhandlingens centrale metodologiske ambition at udvikle og

afprøve et rumsociologisk design, der kvantificerer mikro-rummet som en socialmorfologisk konteksteffekt for mikro-interaktionsprocesser. Konkret vil dette ske i tre empirisk-kvantitative studier, der hver især undersøger, hvordan og hvorvidt det morfologiske mikro-rum er med til at forme den intersubjektivt følte intensitet af stedsligt forankrede mikro- interaktioner.

Projektets tilblivelse og tilgang Afhandlingens ovenfor skitserede fokus var ikke givet, da nærværende forskningsprojektet påbegyndtes. Langt fra, faktisk. Stort set intet teoretisk, metodologisk eller empirisk valg i forbindelse med udarbejdelsen af afhandlingen er blevet realiseret ud fra en forudgående forskningsplan. Den idé jeg havde skitseret i min ph.d.-ansøgning kan således ikke gen- findes i afhandlingen færdige produkt. Projektforslaget lagde op til, at nærværende forskningsprojekt skulle have været en genaktualisering af den såkaldte urbanitets-tolerance-hypotese, der foreslår, at den urbane levevis udvikler byboens tolerance for socialt skæve elementer (Wirth 1938; Stouffer 1955; Wilson 1985). Løftestangen for genaktualiseringen af denne hypotese var et argument om, at en sådan toleranceproces skal ses i lyset af gadens “rumligt-materielle præmisser” uden at det dog spe- cificeredes nærmere, hvordan dette kunne realiseres metodisk konkret. To forklaringer på denne metodologiske uklarhed byder sig an: Enten var der simpelthen tale om brist fra min side eller også kan manglen fortolkes som symptom på den rumsociologiske litteraturs generelle mangel på værktøjer til at analysere det fysiske mikro-rum. Vurderet i retrospekt holder jeg på den sidstnævnte fortolkning og afhandlingen kan et langt stykke af vejen læses som et forsøg på at adressere den metodologiske mangel, som mit projektforslag arvede fra litteraturen. Vi befinder os i dag i den paradoksale situation, at et stadigt stigende antal sociologer beskæftiger sig med det sociales rumlighed, mens spørgsmålet om, hvordan rummets egenskaber kan gøres til genstand for systematisk metodisk analyse, ofte lades stå ubesvaret hen. Hvordan og hvorfor dette er tilfældet vender jeg tilbage til nedenfor, men lad mig først afklare, hvordan vejen fra projektforslag til afhandling blev formet af dette metodespørgsmål.

Rummets fysik Det var ikke den rumsociologiske litteratur, men tilfældet, der introduce- rede mig for en stringent metodologisk måde at studere det fysiske mikro-rum. I projektets opstartsfase var jeg på familiebesøg i Helsingfors

og besøgte i denne forbindelse byens velassorterede boghandel. Målrettet finkæmmede jeg sektionerne med socialvidenskab, men det var først på vej ud, under sektionen for populærvidenskab, at jeg snublede over noget fængende: Nemlig fysikeren Philip Balls’ (2004) storsællert, Critical Mass: How One Thing Leads to Another, der underkaster en lang række sociale fænomener en spidsfindig undersøgelse ud fra et socialfysisk og kompleksitets-teoretiske perspektiv. Mest fascinerende er bogens argument om, at fysisk-biologiske systemer og storbyens fysiske-materielle organisering begge kan forklares som selvorganiserende systemer med emergente egenskaber: Dvs. systemer, hvor summen er mere end delene og som ofte udvikler sig ikke-lineært og utilsigtet snarere end lineært og tilsigtet af individuelle handlinger. Tænk her på, hvordan den V-formation, som en flok fugle ofte flyver over himlen i, er lige så lidt tilsigtet af den individuelle fugl som net- værksegenskaberne ved storbyens gadenet er planlagt af politikere og planlæggere. Et sådan emergensteoretisk argument findes i en filosofisk variant inden for socialvidenskaben, der også udpeger storbyens fysisk- rumlige organisering som et emergent fænomen par excellence (Deleuze 1994; DeLanda 1997; Urry 2005). Med baggrund indenfor fysik er Balls fokus dog ikke et sådan filosofisk argument, men derimod de metodiske værktøjer, der inden for de seneste årtier er blevet udviklet og anvendt empirisk til kvantificering af storbyens emergente egenskaber. Her fremhæver Ball særskilt den såkaldte space syntax-tilgang, udviklet af de to engelske analytiske arkitekter, Bill Hillier og Julienne Hanson (1984), til kvantificering af netværksmæssige egenskaber ved storbyens fysiske rum. Under anvendelse af deres space syntax-metode har Hillier, Hanson og deres kollegaer ved UCL Bartlett School of Architecture på- vist, hvordan den gadenetværks-konfiguration, der emergerer som et utilsigtet produkt af planlægningsmæssige beslutninger, har en afgørende betydning for, hvor bevægelsesintensivt gadenetværkets enkelte gader er benyttet. Denne bevægelsesøkonomiske forskel har dernæst en række sociale effekter på byens stedsligt forankrede aktiviteter. Ovenfor nævnte jeg, hvordan en sådan rumligt formet bevægelse antageligvis er korreleret positivt med sandsynligheden for sammenstød mellem cyklister. Et empirisk velunderbygget eksempel på en sådan rum-bevægelsesproces er, hvordan butikker har en stærk tendens til at klynge sig sammen i gader, hvor gadenetværket i forvejen har fortættet megen bevægelse og således tilbyder butikslivet en økonomisk fordelagtig kundebasis (Hillier et al. 1993; Hillier 1996). For mig som fænomenologisk orienteret kultursociolog, uden nævne- værdig kvantitativ interesse eller kundskab ved ph.d.-projektet start,

fremstod space syntaxen skræmmende statistisk og kausalt. En ting var dog svær at komme uden om: Her var et stringent bud på, hvordan gadens rumligt-materielle præmisser kunne operationaliseres metodisk. Men var space syntaxen alligevel for kvantitativt vanskeligt at indarbejde i min sociologisk orienterede afhandling? Her stod jeg dog overfor end- nu en overraskelse, der yderligere besejlede afhandlingens metodiske kurs. I forbindelse med afdækningen af space syntax-litteraturen læste jeg tilgangens grundlæggende bog, The Social Logic of Space af Hillier og Han- son (1984), der med god grund viste sig at have en så sociologisk klingende titel. Helt fra bogens indledende afsnit argumenterer Hillier og Hanson for, at Durkheims social morfologi er det sted i den rumteoretiske litteratur, der først og bedst har forholdt sig til, hvordan rummet har en fysisk iboende social logik (Hillier og Hanson 1984: 18; se også Hillier et al. 1976). Hillier og Hansons udgangspunkt i Durkheim sociologi virkede ved første øjekast som en sær reference i lyset af den rumlige vendings selvfortælling. Den rumlige vending er, som dette slogan antyder, vendt mod fremtiden, i en venden-sig-bort fra den fortidige sociologis mangelfulde begreb om rum. Denne opfattelse nærer sig, som anført ovenfor, af Latours forkastelse af Durkheim, der fortegnes som arketypisk blind over for det sociales spatio-materielle egenskaber. Det er ikke første gang, at sociologien har glemt sit rumteoretiske fundament; og det i en sådan grad, at jeg, som ellers er teorihistorisk interesseret sociolog, skulle belæres af en arkitekturbog fra 1980’erne om Durkheims begreb om morfologiske rum! To årtier før nutidens rumlige vending, dedikerede den danske sociolog Jens Tonboe (1993) således sin fremragende doktorafhandling om rummets sociologi til at tilbagevise den påstand, at rummet er et nyt sociologisk tema (se også Smith 1999). Durkheims rummorfologiske bidrag er et væsentligt bevis mod en sådan påstand og hvis årstallene blev opdateret, kunne Tonboes indledende ord i disputatsen have været skrevet i dag:

Det er stadig almindeligt i nyere by-teoretiske arbejder at møde den påstand, at rummet eller ‘de materielle omstændigheder’ er en overset dimension i sociologien, som først den seneste, mest avan- cerede generation af teoretikere er begyndt at tage alvorligt. Denne afhandling er et forsøg på at modbevise denne påstand. De fleste af de følgende sider er viet til gennemgang, fortolkning og kritik af det, man ikke troede fandtes, nemlig mere eller mindre indgående diskussioner af rummets sociale baggrund og betydning i den klassiske sociologi og kulturgeografi fra tiden før 1970, dvs. før temaet blev taget op af den ’nye’ generation (…). Det har vist sig, at em-

net snarere har været glemt eller endda gemt bort i en periode, og at det nok så meget er dette forhold, som at det generelt er blevet overset, der kræver en forklaring (Tonboe 1993: ix, original kursivering).

På det tidlige tidspunkt i projektet, hvor Hillier og Hanson fik mig an- sporet til at genlæse Durkheim, var afhandlingens tematiske fokus stadig urbanitet-tolerance-hypotesen. Men jeg havde nu fået en klar idé om, at hypotesens explanans, dvs. gadens rumligt-materielle omstændigheder, kunne operationaliseres ved hjælp af space syntax-metoden. I Stouffers (1955) klassiske udlægning af hypotesen henvistes til hvordan, at byboer “(...) does rub shoulders with more people who have ideas different from his own, and he learns to live and let live.” (Stouffer 1955: 221f). Hvis space syntax-metoden vitterligt kan forudsige, hvor byens bevægelses- strømme fortættes, ville det principielt være muligt at operationalisere denne grad af “skulder-gnubben” med metoden. En test af dette ville imidlertid være betinget af eksistensen af georefererede surveydata om storbyens toleranceforhold med en ekstrem høj rumlig opløsning. Space syntax-tilgangen er metodisk orienteret mod det byrumlige netværks mindste mikro-skalaegenskaber, mens de eksisterende tolerancedatasæt har en langt mere grovkornet rumlig opløsning, der kun tillader en skelnen mellem regionale grader af land og by. De eksisterende data og min foretrukne metode befandt sig på uforenelige rumlige skalaer. Første år var nu oprandt og jeg var metodologisk fastlåst. Durkheim og Hillier var unægtelig gode følgesvende, men det hjalp mig ikke empirisk videre. Der iværksattes en kvalitativ nødplan og jeg begyndte i stedet at undersøge den urbane tolerance i bydelen Nørrebro ved hjælp af etnografiske metoder. Men når man bevæger sig ud i feltet er det altid med en vis sandsynlighed for, at man opdager, at det er noget andet end det forventede, der påkalder sig etnografisk opmærksomhed (Hastrup 2010). Fra mit kontor på Sociologisk Institut går den hurtigste vej til Nørrebro over Dronning Louises Bro. I de dage hvor jeg passerede denne bro, der krydser en af søerne i Københavns morfologiske hjerte, udspillede sig en påfaldende rumsociologisk begivenhed. Hver dag, og særligt weekendaf- tenerne, samledes her hundrede af yngre københavnere. De medbragte øl og anlæg, og der herskede en herlig urban atmosfære. Her brusede byen, blev der nu noteret i feltnoterne og som det fremgår på billedet nedenfor:

Figur 1. Eftermiddag på Dronning Louises Bro, maj 2012. Eget fotografi.

Min kultursociologiske interesse var vakt. Dronning Louises Bro havde ikke før i byens historie lagt rum til en sådan urban forsamling og fejring. Så hvad var der sket – og spillede det fysisk-materielle mikro- rum, sådan som jeg nu i lyset af space syntaxen forventede det, en rolle herfor? Det etnografiske arbejde jeg foretog på Dronning Louises Bro var på mange måder det, der gav afhandlingen dens definerende fokus, og hver af afhandlingens artikelbaserede kapitler er – helt eller delvist – afledt af tanker, hypoteser og observationer, jeg har gjort mig på Dron- ning Louises Bro. Det etnografiske arbejde på Dronning Louises Bro betød således, at det bysociologiske spørgsmål om urban tolerance blev afløst af et rum- og mikro-sociologisk spørgsmål om, hvordan det fysiske mikro-rums egenskaber bidrager til intersubjektive oplevelser af følel- sesmæssig intensitet à la dem, der brusede på Dronning Louises Bro. Min fornemmelse var og er, at broen udgør en eksemplarisk illustration af, hvordan det fysiske rum sommetider kan være med til at forme den følelsesmæssige intensitet af stedslige mikro-interaktionsprocesser.

Rum og mikro-interaktion Givet afhandlingens fokus på rumfølelses-krydsfeltet og min nyvakte interesse for Durkheims rummorfologi begyndte jeg i stigende grad at orientere mig mod den mikro-sociologiske tradition, som Erving Goff- man grundlagde med inspiration fra særligt Durkheims (1995) sene ho- vedværk, The Elementary Forms of Religious Life. Hvad der taler for denne inspiration er, hvordan den mikro-durkheimianske tradition udpeger følelsesprocesser som hjertet i det sociale liv. Det er som konsekvens af følelser af “pinlighed” (Goffman 1956) og “skam” (Scheff 1990), at det sociale overhovedet har en eksistens sui generis, der kan virke moralsk “eksternt” og “tvingende” på individer, sådan som Durkheim (1982) formulerer det. Omvendt stod det også klart, at den opstemte mikro-

interaktion, der udspiller sig på Dronning Louises Bro ikke primært drejer sig om sådanne goffmanske følelser af ekstern tvang. Broen emmer snarere af følelser af at være kollektivt opløftet og motiveret til at tage del i denne følelsesmæssigt opladende situation. Indenfor den mikro-durkheimianske tradition er en sådan nydelses- mæssig, frem for tvangsmæssig, side af det sociale bånd mest systematisk blevet fremanalyseret mikro-sociologisk af Randall Collins (1993; 2004) i forbindelse med hans formulering af teorien om interaktionsritualer (IR- teorien).3 Den følelsessociologiske motor i IR-teorien er Collins’ antagelse om, at individer mest grundlæggende er motiverede til at maksimere oplevelsen af “emotionel energy” (EE), forstået som den subjektive side af den kollektive opbrusen, (“collective effervesence”), som Durkheim (1995) udpeger, som det centrale følelsesprodukt af socio-rituel aktivitet. Det var en sådan følelsestilstand af kollektiv opbrusen, som havde fanget min sociologiske opmærksomhed, da jeg gik over Dronning Louises Bro med et andet analytisk ærinde for øje. Collins’ neo-durkheimianske for- slag om, at individer opererer følelsesmæssigt energi-maksimerende på interaktionsritual-”markeder”, gav mig et mikro-sociologisk praj om, hvordan Dronning Louises Bros socialitet kan forklares og ethvert af afhandlingens argumenter er udviklet i eksplicit eller implicit dialog med IR-teorien. At den mikro-durkheimianske tradition tildeles en så central rolle i afhandlingen skyldes ydermere det forhold, at denne tradition eksplicit vedkender sig Durkheims socialmorfologiske indsigt, at sociale symboler og følelser formes af det sociales rumligt-materielle substratum. Kontinu- iteten til trods repræsenterer den mikro-durkheimianske tradition dog også sin egen situationelle læsning af Durkheims sociale morfologi. IR- teorien implicerer ifølge Collins: “(…) a return to an older Durkheimian formulation in which social morphology shapes social symbols. Current IR theory differs from the classic version mainly in giving a radically micro-situational slant, stressing that the social morphology that counts is the patterns of micro-sociological interaction in local situations.” (Collins 2004: 32).

3 Andre durkheimianske mikro-sociologer har også betonet et nydelsesaspekt ved ritualet. Det måske tydeligste eksempel herpå er Maffesolis (1995) neostamme-teori. At afhandlingen tager afsæt i Collins fremfor Maffesoli hænger sammen med, at Maffesoli repræsenterer et forsøg på at re-orientere ritualteorien i en fænomenologisk og således deskriptiv og ikke-forklaringsorienteret retning (Evans 1997). I kontrast hertil, og i samklang med Collins’ (1989) sociologiske IR-teori og epistemologi, argumenterer afhandlingen i stedet for relevansen af at etablere og teste generaliserbarheden af sociologiske årsagsforklaringer – også når det gælder et sådant typisk kultursociologisk spørgsmål om følelsesmæssig nydelse. Dette udgangspunkt vender jeg tilbage til senere.

Den tætte kobling mellem situationens symbolske orden og materielle grundlag kommer også klassisk til udtryk hos Goffman (1956), som Col- lins i øvrigt anser som grundlægger af den nutidige IR-tradition. Dette kommer måske klarest til udtryk i forbindelse med Goffmans dramaturgiske begrebspar om “frontscener” og “bagscener”, hvor den faktiske og fysiske opdeling af rummet i to scener er med til at opretholde disse sce- ners forskelligartede symbolske definition af korrekt adfærd. Uden for- ankring i rummets fysiske egenskaber ville scenernes symbolske koder ikke være socialt virksomme (Goffman 1974: 247ff; Scollon og Scollon 2003: 55ff). Det er således Goffmans fortjeneste at have introduceret de fysisk-rumlige egenskaber til den mikro-sociologiske analyse af sociale interaktioner, sådan som allerede Giddens (1984: 199ff) anerkendte,4 og som Turner (2010) for nyligt har opsummeret på følgende måde: “Er- ving Goffman was the first sociologist to recognize fully the significance of the configuration of space and the props in space for structuring the flow of interaction in focused and unfocused encounters.” (Turner 2010: 11). Den goffmanske mikro-sociologi implicerer altså en understregningen af, at rummets materielt-rumlige egenskaber ikke blot er en simpel funktion af socio-symbolske konstruktioner. Forholdet mellem socialt rum og symbol forstås mere symmetrisk: At situationer “defineres som virkelige” har betydelige sociale konsekvenser. Men som Goffman (1974: 1) samtidig understreger, er en sådan symbolsk konstruktion af virkeligheden kun mulig at opretholde for så vidt at den er “forankret” i den rumligt- materielle, “reelle virkelighed.” (Goffman 1974: 251). At det sociale er en scene er, ifølge Goffman, ikke ensbetydende med, at individernes symbolske definition af virkeligheden har forrang, som eksempelvis Blumer

4 Det skal tilføjes, at Giddens, i forlængelse af Goffman, også selv har ydet et originalt bidrag til rumsociologien, som særligt Löw (2007; 2008) har videreført. Efter min vurdering, må Giddens dog vurderes som et problematisk rumsociologisk udgangspunkt, idet han afviser, at rummet skal studeres på dets egne ontologiske og metodologiske præmisser. Hans antimorfologiske argument er, at det bør afvises, at “(…) space has its own intrinsic nature, a proposal that is logically questionable and empirically unfruitful. Space is not an empty dimension along which social groupings become structured, but has to be considered in terms of its involvement in the constitution of systems of interaction. The same point made in relation to history applies to (human) geography: there are no logical or methodological differences between human geography and sociology!” (Giddens 1984: 368).

(1986) og Garfinkel (1967) giver udtryk for.5 Dette udtrykkes på fremragende vis i indledningen af Goffmans Frame Analysis:

Defining situations as real certainly has consequences, but these may contribute very marginally to the events in progress; in some cases only a slight embarrassment fits across the scene in mild concern for those who tried to define the situation wrongly. All the world is not a stage – certainly the theater isn’t entirely. (Whet- her you organize a theater or an aircraft factory, you need to find places for cars to park and coats to be checked, and these had better be real places, which, incidentally, had better carry real insur- ance against theft.) (Goffman 1974: 1).

Det er netop denne socialmorfologiske Goffman, som Collins (bl.a. med henvisning til ovenstående passus) forholder sig til og som lægger til grund for sin senere udvikling af IR-teorien:

We are back at Durkheim’s vision of a real material world in which human bodies come together, rituals are carried out, and collective mental representations are thereby created. Goffman adds layer on layer of how these ceremonies and mental definitions can play off each other, but the material world of human bodies is still the basic. When a fire breaks out in the theatre, all other games are off (Collins 1994: 281).

Feltarbejdet på Dronning Louises Bro havde netop fremdraget en sådan morfologisk dynamik, hvor ritualet er præget af den kropslige tilstedevæ- relse af individer i et faktisk, materielt rum. Næste skridt var at validere denne rummorfologiske betragtning med space syntax-metoden. En space syntax-analyse af det københavnske gadenet viser, at broen er en af de mest netværksmæssigt integrerede gader i gadenetværket, hvorfor forventningen er, at der sandsynligvis er tale om et gadeforløb med et meget betydeligt bevægelses- og livlighedspotentiale (Hillier 1996; Hillier et al. 1993). Dette rummorfologiske ræsonnement er tæt på cykelkollisi- ons-eksemplet fra foroven: Gadenettet fortætter bevægelsestrømme i bestemte gadeforløb, hvilket ikke blot øger sandsynligheden for cykel-

5 Latour (1996) antager fejlagtigt, at Goffman skal forstås som en symbolsk, ikke- materiel interaktionist. Som Jensen (2010) fremhæver, står Goffmans mikro-sociologi dog ikke i modsætning til ANT, men må snarere betegnes som en materielt orienteret sociologi, der ’predates’ ANT (Jensen 2010: 337). Latour burde i stedet have refereret til eksempelvis Smith og Bugnis (2006) symbolsk interaktionistiske tilgang til arkitektur, der netop opløser det arkitektoniske objekt i sociale symboler.

sammenstød, men også for, at de bevægelsesintensive gader har en livlig mikro-rituel aktivitet. Indfinder man sig på Dronning Louises Bro kon- stateres det let empirisk, at der er tale om et ekstraordinært bevægelsesin- tensivt gadeforløb (bemærk, at kapitel 3 underbygger denne kvalitative betragtning statistisk-kvantitativt). Resultatet af space syntax-analysen af Dronning Louises Bros placering i det københavnske gadenetværk er illustreret nedenfor (det vejstykke som broen er del af, er angivet som den linje, der går fra øverste venstre hjørne til nederste højre hjørne):

Figur 2. Space syntax-analyse af Dronning Louises Bro. Farveskalaen illustrerer, hvor rumligt integreret vejene er placerede i gade-netværket. Skalaen går fra høj (røde nuancer) henover mellem (gulgrønne nuancer) til lav (blå nuancer) grad af netværksmæssig integration.

Rumligt genereret bevægelse er imidlertid ikke altid socialt tiltrækken- de og rituelt stimulerende. Indtil for nyligt havde den netværksmæssige tilgængelighed være broens sociale død. Dronning Louises Bro havde i mands minde været et larmende inferno af biler og motorstøj, som ingen havde lyst til at gøre ophold i. I 2008 blev trafikken på broen omlagt som del af Københavns Kommunes ambitiøs byplan for indre Nørrebro. Broen blev spærret for gennemkørende biler og hobetallet af cykler og fodgængere fik bedre færdselsforhold. Og netop da opstod det urbane interaktionsritual på broen, men hvorfor egentligt? Det var næppe de opstillede bænkes effekt, sådan som Whytes (1988: 110) ville forvente det, idet de først blev opstillede efter ritualets påbegyndelse ligesom, at den billedskønne udsigt fra broen næppe er en tilstrækkelig forklaring, sådan som den fænomenologiske forventning ville være (Böhme 1993; Albertsen 1999; Bech 2005): Der er mange andre bilfri rum i Køben-

havn, der har en billedskøn udsigt, men som ikke er steder, for en sådan mikro-rituel aktivitet. Kombinationen af space syntax-analysen af gadenettet og Collins’ IR- teori peger i stedet på, at broens urbane ritual snarere må forklares i lyset af broens ekstraordinært velintegrerede indlejring i gadenettet, der giver dette vejforløb et højt bevægelses- og livlighedspotentiale og – som en afledt effekt heraf – gør dette til et byrum med en socialt medrivende atmosfære; dvs. et bymæssigt sted, hvor tingene sker, sådan om Collins formulerer det nedenfor:

At a higher intensity are situations with a buzz of excitement: being on a busy street in a city, in a crowded restaurant or bar. There is a palpable difference between being in an establishment where there are lots of people and one that is nearly empty. Unfocused crowds generate more tacit interaction that very sparse assemblies, and thus give a sense of social atmosphere. Even though there is no explicit interaction or focus of attention in such places, there is a form of social attraction to being there. Being in a crowd gives some sense of being in ‘where the action is,’ even if you personally are not part of any well-defined action; the lure of the ‘bright lights of the city’ is not so much a visual illumination but a minimal excitement of being within a mass of human bodies (Collins 2004: 82, min kursivering).

Afhandlingens neo-durkheimianske hovedbidrag er at foreslå en rummorfologisk radikaliseret version af Collins’ IR-teoretiske forklaring af sådanne følelsesintensiteter. Det er ikke kun individernes EE-søgende handlen, der har kapaciteten til at bevæge individer ind i hinandens kropslige nærhed, hvilket bevæger dem følelsesmæssigt. Sociale forsamlinger af individer, grader af densitet og hermed følelsesintensiteten af situationer, må også forstås som et rumkausalt produkt af netværkshandlen

(“network agency”) (Hillier 2005).6 Samme pointe formuleret kontrafak- tisk: Havde Dronning Louises Bro ikke haft den netværks-centrale placering, der understøtter gadeforløbets høje bevægelsesdensitet, ville det urbane ritual, der i dag udspiller sig her, med mindre sandsynlighed have fundet sted. Udtynder rummets kræfter densiteten, taber interaktionsri- tualet ofte sit intensive moment. Den EE-søgende trang til at nyde the “bright lights of the city” ville lede ritualdeltagerne til andre, mere bevæ- geles-livlige byrum, hvor gadenetværkets handlen har øget den sociale densitet og rituelle intensitet. Afhandlingens centrale påstand, som vil blive diskuteret teoretisk og testet kvantitativt-metodisk, er altså: at mikro- rummet er med til at forme den følelsesmæssige intensitet af stedslige mikro-interaktionsprocesser.

Andre broer Ovenstående space syntax-inspirerede fortolkning af de rumlige kræfter, der udfolder sig på Dronning Louises Bro, minder, ved første øjekast, om de klassiske planlægningsteoretiske argumenter, som vi finder hos eksempelvis Gehl (2003; 2010) og Appelyard (1981). De foreslår, at fjer- nelsen af biler ofte transformerer gaderummet til en bymæssig, social attraktion. Med reference til Broadway i New York, efter at gaden var blevet bilfri, påpeger Gehl således, at: “as soon as spaces are converted from traffic spaces to people spaces, people came along in thousands and settle down to enjoy the scenery, the city, and city attraction number one – the other people.” (Gehl 2010: 237). Anskuet med space syntaxen, forbigår en sådan forklaring, der betoner attraktionen ved andre mennesker, det fysiske byrums selvstændige bidrag til genereringen af byens attraktioner. At der overhoved er andre, socio-rituelt attraktive, mennesker tilstede i gaderummet, må forstås i lyset af, at den rumlige netværks- konfiguration i første omgang har fortættet byens bevægelsesstrømme i disse rum. Space syntax-læren af Dronning Louises Bro-eksemplet er, at

6 For Hillier og afhandlingen er det ikke afgørende, at rummets evne til at forme mikro- interaktioner begrebsliggøres med termen ’agency’/’handlen’. Den vigtige analytiske pointe er, at intentionelt handlende individer deler evnen til at forme interaktioner med tingene, herunder de rumlige netværksting. For at undgå en terminologisk balkanisering kan det handlingspotentiale, som humane og non-humane agenter har til fælles, med fordel formuleres med Turners (2010) begreb om ’forces’. Dette begreb referer til, at den sociale virkelighed medieres, uanset af hvem eller hvorfor dette sker: “A force is any property of the social universe – or any other domain of the biophysical universe – that drives the formation of reality. Whether gravity or the forces examined in physics, mutation and selection in biology, or roles and status in the sociology, these properties of various realms of the universe drive the formation and operation of various types and levels of reality (Turner 2010: 9-10).” I afhandlingen vil termerne ’forces’/’kræfter’ og ’agency’/’handlen’ således blive anvendt synonymt.

socio-rituelle attraktioner nok accelererer sig selv gennem tiltrækning af stadigt mere bevægelsesflow men, at dette er rumligt betinget af, at det rumlige netværk i første instans har fortættet det attraktive bevægelses- flow i de pågældende gaderum (Hillier 1996). At en sådan space syntax-inspireret forståelse af netværkshandlen er væsentlig for en rumsociologisk forståelse af broer og rummet i alminde- lighed kan ydermere illustreres ved sammenligning med Simmels (1994) og Heideggers (1971) indflydelsesrige rumteoretiske bidrag. De forklarer deres respektive sociologiske og fænomenologiske, men i mange henseender, konvergerende, syn på rummet med et broeksempel. For Heideg- ger er broen en “ting”, der samler “firefoldet”, denne mystiske enhed af “earth, sky, divinity and mortals”, hvilket gør, at en ellers upåagtet del af floden, eller i mit tilfælde søerne i København, bliver et “sted”:

The location is not already there before the bridge is. Before the bridge stands, there are of course many spots along the stream that can be occupied by something. One of them proves to be a location, and does so because of the bridge. Thus the bridge does not first come to a location to stand in for; rather, a location comes into existence only by virtue of the bridge. The bridge is a thing; it gathers the fourfold, but in such a way that it allows for the fourfold (Heidegger 1971: 151f, original kursivering).

Den metafysiske klangbund skal ikke skygge for den klare relationelle pointe, som Heidegger her lægger op til: Steder er ikke essentielt selvbe- roende, men gives relationelt i kraft af dets forbindelser, her illustreret ved broen, som én tings-relation mellem flodens to bredder. Denne principielle forståelse af det lokale sted, som givet af rumlige relationer, deler Heidegger for så vidt med Hilliers rum-relationelle perspektiv. Hei- degger – og i forlængelse af ham eksempelvis de Certeau (1998) – går dog ikke langt nok i sit relationelle perspektiv, der forbliver fænomeno- logisk forpligtet på det, der kan erfares lokalt. Havde Heidegger, som space syntaxen gør det, forladt den lokale skala og kvantificeret rummet fra oven, ville han måske have opdaget, at broen er en virksom lokal relation, som et resultat af dens placering i det globale netværk af vejrelationer (Hillier 2005). Hvis Heideggers bro havde en lav grad af netværksmæssig tilgængelighed, ville den sandsynligvis (og det som produkt af rummets netværkshandlen) være ringe benyttet af mennesker. Dette vil være definerende for brostedets karakter, dets sociale attraktion og stemning. Og endvidere, på det mest grundlæggende plan, vil det fåtallige antal mennesker, der passerede broen, erodere dette steds stedslighed, der i Heideg-

gers perspektiv, som Harman (2010) har påpeget, kun har en fænomeno- logisk eksistens, hvis der er mennesker til stede, der erfarer stedet som sted. I en næsten heideggeriansk tone har Simmel (1994) i essayet Bridge and Door ligeledes teoretiseret rummet med udgangspunkt i et broeksempel. “Broen” eksemplificerer her, hvordan det sociale liv forbinder, hvad der er adskilt i rummet. Dette står i et spændingsforhold til “døren”, der gives som grundeksempel på, hvordan det sociale liv fungerer gennem rumlige afgrænsninger. Centralt for Simmels argument er, at broen er en arketypisk menneskelig bedrift. Dette står i modsætning til dyret, der også kan overkomme rummet, men ikke etablere blivende forbindelser og således ikke er i stand til, at “(…) accomplish the miracle of the road: freezing movement into a solid structure that commences from it and on which it terminates.” (Simmel 1994: 6). Mens Simmel i lighed med Heidegger kan kritiseres for at glemme, hvordan den lokale bro-relations bevægelsespotentiale er givet i og med relationens indlejring i det globale netværksmønster, virker det ved første øjekast som om, at agency-aspektet ved rumnetværket er betonet tydeli- gere hos Simmel end det er hos Heidegger (jævnfør her eksempelvis Schillmeiers 2009 argument for at Simmel skal læses som en proto-ANT- position). En sådan læsning er imidlertid kun mulig, hvis der ses bort fra Simmels fundamentalt idealistiske forståelse af rum og broer, der afspejler en neo-kantiansk epistemologi, som han i øvrigt deler med Weber (Blegvad 1996). Det afgørende ved broen er således, hvordan den udtrykker en menneskelig “will to connection” (Simmel 1994). Det idealistisk definerende ved broen er, hvordan den er en rumlig-materiel for- længelse af menneskets meningsskabende virksomhed, og således ikke, hvordan broen, som en ting i sig selv har en fysisk effekt på det sociale liv. Space syntax-betragtningen af Dronning Louises Bro er i kontrast hertil en insisteren på en rumlige agency, der ikke beror på menneskelig mening og vilje. Simmels idealistiske, frem for materielle, forståelse af rummet kommer ydermere til udtryk i forbindelse med hans længste sammenhængen- de behandling af rumspørgsmålet (Simmel 2009). Her understreges det, hvordan “grænsen”, der er begrebslogisk analog til bro-forbindelsen “(…) is not a spatial fact with sociological effects, but a sociological reality that is formed spatially. The idealistic principle that space is our conception – more precisely, that it is realized through our synthesizing activity by which we shape sense material – is specified here in such a way that the spatial formation that we call a boundary is a sociological func-

tion.” (Simmel 2009: 551). Simmels idealistiske opfattelse af rummet opsummeres på rammende vis af Tonboe:

His main thesis, however, that space represents only a contingent or external form or appearance of the social world, with no social consequences of its own – a sociological version of Kant’s a priori perception space of the mind – lacked a persistency in the whole discussion, despite its purely theoretical character and eclectic choice of examples. Occasionally, material space broke through the formal-sociological surface, with differential social consequences of its own (Tonboe 1993: 527).

Tonboe (1993) fremdrager altså ligeledes Simmels neo-kantianisme og tilføjer endvidere, at de elementer af rumeffekter, der kan spores hos Simmel, viser sig på trods hans klare, idealistiske position. Det rumlig- materielle objekt trængte sig, så at sige, kun teoretisk ubelejligt på, men med en empirisk nødvendighed. Den rumsociologiske lære af Simmel er således først og fremmest, at en idealistisk tilgang til rummet ikke kan opretholdes i praksis.

Scener og ikke-scener Netop som afhandlingen nærmede sig sin afslutning, udkom den danske sociolog Ole B. Jensens (2013) Staging Mobilities, hvis rumsociologiske ærinde i mange henseender, er beslægtet med denne afhandlings. Jensens afsæt er ligeledes den mikro-durkheimianske tradition, specifikt Goff- man, som sammentænkes med en pragmatisk ANT-læsning og mobili- tetsteoriens fokus på de interaktioner og følelser som rumlig bevægelse er anledning til (Jensen 2006; 2010; Urry 2007; Cresswell 2006). Mobilitet forstås her som det mere end den rene fysiske bevægelse fra punktet A til B. Specifikt foreslår Jensen, at relationen mellem mobilitet, interaktion og den rumlige kontekst kan analyseres med to komplementære scene- metaforer, afledt af Goffmans dramaturgiske argument, at hverdagens interaktionsritualer sættes i scene rumligt: “Mobilities do not ‘just happen’ or simply ‘take place’. Mobilities are carefully and meticulously designed, planned and staged (from above). However, they are equally important acted out, performed and lived as people are ‘staging themselves’ (from below).” (Jensen 2013: 4). Dette begrebspar overlapper i betydeligt omfang med det begrebsapparat, som afhandlingen udvikler, men efterlader også en uforklaret rest som afhandlingens begrebsarbejde fokuserer på. Dronning Louises Bro- eksemplet, og afhandlingen som helhed kan læses som et argument for,

at Jensen kunne have optalt tre sceniske processer med reference til Goffman, fremfor de to, iscenesættelser hhv. “fra neden” og “fra oven”. Lad os starte med at tælle til to med Jensen – og så derefter lægge en til. Et. Det er eksempelvis evident, at der foregår en iscenesættelse “fra neden” på Dronning Louises Bro. Gaderummet sættes festligt i scene med medbragte artefakter (musikanlæg, øller, etc.), og der udspiller sig, især ud på aftenen og natten, en udpræget grad af sensuel selv-iscenesæt- telse. To. Det er endvidere oplagt, at selve broen er iscenesat “fra oven” af byens planlæggere. Den nuværende Dronning Louises Bro erstattede i 1887 den gamle Peblinge bro, der havde vist sig uegnet til at håndtere det stigende trafikale pres. I denne anledning foretog planlæggerne en lille men væsentlig netværksmæssig ændring, at den nye, bredere og mere dekorerede Dronning Louises Bro blev placeret lidt mere sydligt, og så- ledes på den netværksmæssigt velintegrerede akse, der forbinder bydelen Nørrebro med Københavns gamle bykerne (Wassard 1990: 175). Det må i høj grad forstås som resultat af denne rummorfologiske beslutning “fra oven”, at broen er et så netværksmæssigt centralt gadeforløb. Tre. At det er nødvendigt at lægge én til kan, så at sige, tilskrives det forhold, at den planlægningsmæssige iscenesættelse “fra oven”, der fandt sted for over et århundrede siden, ikke havde til hensigt at generere det urbane interaktionsritual, der i dag trives på Dronning Louises Bro. Det interaktionsritual, der finder sted her, synes netop blot “just [to] happen” eller simpelthen at “take place”, snarere end at være “carefully and meticulously designed, planned and staged”, som Jensen (2013: 4) formulerer det. Denne pointe er beslægtet med afhandlingens kritiske kommentarer til Simmels idealistiske rumforståelse, som Jensen (2013) i øvrigt trækker direkte på. Jensens begrebssæt synes at overbetone den sociale vilje og mening i rummet dér, hvor rummet, så at sige, blot gør noget af egen “vilje” og uden mening. At denne bro i dag er attraktivt interaktionsritual-marked skete bare, præcist ligesom Webers to cyklister, der kolliderede uden, at denne hændelse reflekterede et subjektivt og intentionelt meningsindhold. Det rumlige netværk kan gøre en ikke-intenderet forskel for det sociale liv, uden rummet behøver at reflektere en designet, planlagt eller iscenesat intention herom. Et scenedramaturgisk begreb om “stagings” indfanger glimrende den intentionelle organisering af rummet, men ikke rummets netværkshandlen, der emergerer, som et utilsigtet produkt. Det var, selvsagt planlæggere, der eksempelvis satte den nye streg på køben- havnerkortet, der blev til Dronning Louises Bro, men det ritual, der opstod et århundrede senere på dette sted, må også forklares som en utilsigtet konsekvens af en planlægningsmæssig handling: “[S]pace and society

seem to be for the most part (in spite of utopias and ideal cities) emergent systems arising from distributed processes, rather than design systems.” (Hillier og Raford 2010: 277). Når først gaderummet er blevet planlagt fra oven, tiltager det rumlige netværk sig en relativ autonomi til at handle videre på egen hånd – uden respekt for iscenesættelsen fra oven. Dette sker ofte som en tidslig proces, som i Dronning Louises Bros tilfælde, hvor det oprindelige formål forlades, glemmes eller erstat- tes af andre sociale formål, meninger og iscenesættelser. Og dette kan også, som vi skal se det i afhandlingens empiriske undersøgelser, ske ved en rumlig proces, hvor planlæggerne kun vanskeligt kan gennemskue de emergente eller netværkshandlingsmæssige konsekvenser for tilgængelig- hed, bevægelse og socialitet ved at ændre det rumlige netværks globale egenskaber (Portugali 2011; DeLanda 1997; Merton 1936). Spørgsmålet er nu hvilken tredje mikro-durkheimiansk scenemetafor der kan indfange dette virksomme, men utilsigtede, handlingspotentiale ved den rumlige morfologi. Her kan vi med fordel holde fast i Goffman og hans materielle grænsedragning for, hvor stor en del af den sociale virkelighed, der er symbolsk iscenesat. Som Goffman (1974: 1) under- stregede det ovenfor: “All the world is not a stage – certainly the theater isn’t entirely.” Særligt i Frame analysis diskuteres spændingsfeltet mellem den sociale virkeligheds iscenesatte aktiviteter og den reelle virkeligheds, rumligt-materielle substratum:

[A]ctivity interpreted by the application of particular rules and in- ducing fitting actions from the interpreter, activity, in short, that organizes matter for the interpreter, itself is located in a physical, biological, and social world. Fanciful words can speak about make- believe places, but these words can only be spoken in the real world. Even so with dreaming. When Coleridge dreamed his “Ku- bla Khan,” he dreamed it in an undreaming world: he had to begin and terminate his dreaming in the “natural” flow of time; he had to use up a bed, a good portion of the night, and apparently some supplies of a medicinal kind in order to be carried away into his dream; and a sufficient control of the environment was assumed, pertaining to air, temperature, and noise level so that he could go on dreaming (Goffman 1974: 247-248, original kursivering).

Hele virkeligheden er, hvad enten der er tale om den drømmende eller vågne virkelighed, ikke sat i scene, idet der, som Goffman opsummerer det i forlængelse af ovenstående citat, sker en socialmorfologisk “(…) intermeshing of framed activity in the everyday unstaged world (...).”

(Goffman 1974: 248). For mig at se er dette begreb, om den ikke- iscenesatte – unstaged – del af den hverdagslige verden, det tredje metafo- riske ben, som Jensens begrebspar om “staging from below” og “staging form above” med fordel kan suppleres med, for at få bedre greb om den rumlige morfologis reelle, netværksbaserede og ikke-intenderede handlen. Uden at det var iscenesat med henblik på dette formål, øger rummets netværks-distribuerede bevægelsesflow sommetider sandsynligheden for, at to cyklister støder sammen, eller at en bybro bliver centrum for et attraktivt interaktionsritual. Afhandlingens artikelbaserede kapitler uddyber og operationaliserer hver især denne sceniske trefoldighed ved det bebyggede miljø: Det morfologiske rum er sat i scene fra oven og fra neden, men udtrykker også en ikke-iscenesat handlen, der udgår fra rummets netværk af relationer.

Undersøgelsesformål Som det fremgår af ovenstående rids af afhandlingens tilblivelse og tilgang, rejser bestræbelsen efter at undersøge det bebyggede rum, som en fysisk mikro-kontekst for interaktionel følelsesintensitet, ikke ét, men et antal relaterede forskningsspørgsmål; spørgsmål der hver især adresserer forskningsmæssige mangler i den rumsociologiske og beslægtede litteratur om urban morfologi, analytisk arkitekturteori og human geografi. Afhandlingen har til formål:

 At fremvise det analytisk frugtbare ved at anlægge et durkhemi- ansk perspektiv på det fysiske mikro-rum som en kausal kontekst for den følelsesmæssige intensitet af interaktionsprocesser. Den- ne bestræbelse anfægter rumsociologiens, særligt den ANT- orienterede rumsociologis, afskrivning af den durkheimianske tradition som et relevant materialitets-sensitivt bidrag til rumsociologien.  At udvikle et kvantitativt metodedesign, der kan operationalisere mikro-rummet som en kausal kontekst for mikro-interaktionelle processer. Relevansen af en sådan mikro-kvantificering vil blive kritisk sammenholdt med et kvalitativt metode-perspektiv på mikro-rumlige processer.  At bygge tværvidenskabelig bro mellem space syntax-tilgangen og rumsociologien, hvilket litteraturen kun byder på få og kortfatte-

de forsøg på.7 Dette er overraskende og beklageligt, idet både space syntaxen og rumsociologien potentielt kan styrkes metodisk og teoretisk ved en sådan tværvidenskabelig udveksling.  At teste generaliserbarheden af afhandlingens centrale kausalhypotese som er, at mikro-rummet er med til at forme følelsesin- tensiteten af stedslige mikro-interaktionsprocesser. Undersøgel- sen af denne hypotese vil ske gennem tre empirisk-statistiske del- undersøgelser, der hver især adresserer specifikke mangler i den rumteoretiske litteratur. Disse undersøger: 1. Relationen mellem festivalgæsters medkreative (“co-creative”) til- egnelser af festivalrum og festivalrummets mikro-fysiske egenskaber, hvilket ikke tidligere er undersøgt kvantitativt. 2. Collins’ interaktionsritual-teoretiske forklaring af kollektiv opbrusen, der ikke tidligere er underkastet en direkte kvantitativ test. Det undersøges særskilt, hvorvidt mikro-rummet spiller en mere direkte rolle for følelsesintensitet end, hvad IR-teorien lægger op til. 3. Det oplevelsesmæssige aspekt af Hilliers teori om byer som be- vægelsesøkonomier, der ikke tidligere har været gjort til genstand for teoretisk diskussion og en direkte rumstatistisk test.

Det skal bemærkes, at afhandlingens tre empirisk-statistiske undersø- gelser tester den generelle hypotese om sammenhængen mellem det fysiske mikro-rum og mikro-interaktionsprocesser i to forskelle kontekster (festival og by) og på baggrund af to forskellige operationaliseringer af mikro-interaktionel følelsesintensitet (brug og oplevelser). Denne empirisk-operationelle variation øger, som jeg vender tilbage til, muligheden for at teste generaliserbarheden af afhandlingens fremsatte kausalhypotese.

Durkheimiansk rumsociologi: Bidrag til en rekonstruktion Den sociale morfologi er ikke en rumsociologisk metervare, men er en position, der kalder på rekonstruktion. Elementerne til en samtidig og konsistent rumsociologi er til stede, men disse er spredt over flere tekster, to århundreder og flere durkheimianske forfattere. Ydermere er det

7 Denne situation er eksempelvis påpeget af Netto (uden år: 15): “Space syntax tools have certainly been used in areas like archaeology, history and even in phenomenological studies. But its concepts have not been absorbed by other neighbouring areas, such as social theory or economics.”

nødvendige rekonstruktionsarbejde vanskeliggjort af det forhold, at ge- nerationer af sociologer har negligeret socialmorfologiens rumsociologiske bidrag. Vi kan ikke uden videre læne os op ad en eksisterende kon- sensusopfattelse af, hvordan en durkheimiansk rumsociologi ser ud, og hvordan den bør videreudvikles. Denne situation er, som vi har set det tidligere, beskrevet af Tonboe (1993), og hos Lévi-Strauss (1963) finder vi en rammende beklagelse over traditionens uforløste potentiale:

It has been Durkheim’s and Mauss’ great merit to call attention for the first time to the variable properties of space which should be considered in order to understand properly the structure of several primitive societies (…) There have been practically no attempts to correlate the spatial configuration with the formal properties of the other aspects of social life.” (Lévi-Strauss 1963: 290-291).

Denne karakteristik er stadig aktuel, og afhandlingens rekonstruktion af den durkheimianske rumsociologi, tager afsæt i en overbevisning om, at denne tradition rummer et stadig uforløst rumsociologisk potentiale. Først undersøges dette ved at placere og afklare socialmorfologien i forhold til Durkheims generelle sociologiske ærinde. Dernæst diskuteres nøglebegrebet om dynamisk densitet og det foreslås, at litteraturens reception af dette begreb er vanskeliggjort af, at den tidlige og sene Durk- heim refererede til to væsensforskellige densitetsdrevne mekanismer, nemlig hhv. social konflikt og følelsesmæssig smitte. I forlængelse heraf vises det, hvordan det sene begreb om følelsessmitte, i modsætning til hvad Durkheim-receptionen traditionelt har argumenteret for, kan forklares som en socialmorfologisk funktion af rumlig kausalitet. Herefter sættes den socialmorfologiske rumteori i relation til ANT-traditionen, med et argument om, at Durkheim og Latour ikke blot udgør diverge- rende bidrag, men på væsentlige punkter bidrager til sociologiens akkumulerede viden om rum-samfundsrelationen. Dette afvæbner yderligere Latours antidurkheimianisme. Endeligt vil jeg fremdrage det kvantitative metodespørgsmål, som socialmorfologiens akilleshæl: Den måske vigtigste hindring for socialmorfologiens analytiske succes – nu som da – er, at traditionen aldrig formåede at gøre perspektivet til operationel genstand for kvantitativ-statistisk undersøgelse. Dette argument fører over til ind- ledningens andet afsnit, hvor jeg specifikt vil diskuterer og forsøger at løse op for dette morfologiske metodeproblem.

Den sociale morfologi Den sociale morfologi udgør sammen med den såkaldte sociale fysiologi, de subdisciplinære ben, som Durkheims generelle sociologi – bestående af syntesen af disse to ben – hviler på. Forskellen mellem fysiologien og morfologien forklares med analogi til biologien: Fysiologien omhandler samfundets indre “funktioner”, mens morfologien angår samfundets ydre “form” (Durkheim 1960). Med morfologien vil Durkheim understrege, hvordan alt sociale liv, uanset dets institutionelle variationer, er betinget af gruppelivets rumligt-materielle form. Lad os her citere en længere central passus, hvor Durkheim præsenterer det socialmorfologiske forskningsfelt som en selvstændig sektion i sit tidsskrift L’Année sociologique:

Social life rests upon a substratum which is determinate both in its extent and in its form. It is composted of the mass of individuals who comprise the society, the manner in which they are disposed upon the earth, and the nature and configuration of objects of all sorts which affect collective relations. Depending on whether the population is more or less sizeable, more or less dense; depending on whether it is concentrated on cities or dispersed in the countryside; depending on the way in which the cities and the houses are constructed; depending on whether the space occupied by the society is more or less extensive; depending on the borders which define its limits, the avenues of communication which traverse it, and to forth, this social substratum will differ. (…) the constitution of this substratum directly or indirectly affects all social phenomena, just as all psychic phenomena are placed in mediate or immediate relationship to the brain. (…) It is this science which we propose to call social morphology (Durkheim 1994: 88, original kursivering).

Durkheim anser socialmorfologien som en nøgle til realiseringen af sociologien, som en forklarende videnskab. Det morfologiske substratum tilskrives en væsentlig kausal effekt på det sociale livs funktioner. “It can and must be explanatory (…)”, som Durkheim (1994: 88) endvidere slår fast. Ydermere fremgår det af citatets skitserede eksempler, at samfundets rumlige former, (dvs. individernes rumlige distribution, det sociales rumlige udstrækning, den rumlige konstruktion af byer og huse) ind- tager en central rolle i etableringen af sådanne socialmorfologiske kausalforklaringer. Når Durkheim taler om rumlige egenskaber, sker dette både med reference til rummets mest makro-strukturelle forhold og til det ariktektoniske rums mikro-skala organisering: “The disposition of streets

and squares, the architecture of the houses, and the structure of things made vary from village to town and form the large city to the small one, and so on.” (Durkheim 1960: 361). En sådan betoning af morfologiens rumligt-arkitektoniske aspekter kommer ligeledes til udtryk i nedenståen- de citat, der også er interessant, fordi det med al tydelighed viser, at Latours kritik af Durkheim, for at være blind for det rumligt-materielle objekts aktive rolle i forhold til det sociales konstitution, ikke er teorihistorisk holdbar. Når først de rumligt-arkitektoniske former er bygget af samfundet, handler det rumlige objekt tilbage på det sociale liv:

First, it is not true that society is made up only of individuals; it also includes material things, which play an essential role in the common life. The social fact is sometimes so far materialized as to become an element of the external world. For instance, a definite type of architecture is a social phenomenon; but it is partially embodied in houses and buildings of all sorts which, once constructed, become autonomous realities, independent of individuals. It is the same with the avenues of communication and transportation, with instruments and machines used in industry or private life which express the state of technology at any moment in history, of written language, etc. Social life, which is thus crystallized, as it were, and fixed on material supports, is by just so much external- ized, and acts upon us from without (Durkheim 2005: 278).

I mere formelle termer foreslår Durkheim, at socialmorfologien kan opdeles analytisk i to materielle komponenter: En hhv. rum- og en populations-komponent, sådan som dette fremgår i Durkheims (1978a: 83) optegnelse af sociologiens principielle opdeling (se også Durkheim 1984; 1982; Mauss 2003):

Figur 3. Durkheims principielle opdeling af sociologien

The study of the geographical base of various peoples in terms of its relationship with their social organization. Social Morphology The study of population: its volume, its density, and its disposition on earth.

Sociology of Religion Sociology of Morality Sociology of Law Social Physiology Economic Sociology Linguistic Sociology Aesthetic Sociology

General Sociology

Den sociale fysiologi angår spørgsmålet om samfundets (moralske, reli- giøse, etc.) funktioner eller institutioner, der er virksomme som følge af, at samfundsmedlemmerne har en kollektiv bevidsthed, sui generis. I afkla- ringen af relationen mellem morfologi og fysiologi, er det væsentligt for Durkheim at balancere det sociales ideale og materielle aspekter. Den sociale fysiologi er betinget af morfologien, mens det samtidigt underste- ges, at der er en “variation” og “kompleksitet” til forskel mellem fysiologien og morfologien (Durkheim 1960). Morfologien betinger emergen- sen af den fysiologiske kompleksitet, hvis “(…) whole very often has very different properties from those which its constituent parts possess.” (Durkheim, 1978a: 76). Dette er, hvad Durkheim mener, når han skriver, at “(…) collective consciousness is something other than a mere epiphe- nomenon of its morphological base, just as individuals conciseness is something other than mere product of the nervous system.” (Durkheim 1995: 426) Den sociale morfologi udgør, som anført, en nøgle for Durkheims sociologisk-positivistiske ambition om at forklare det sociale. Centralt for etableringen af morfologiske kausalmekanismer, er gruppers sociale densitet, eller dynamiske densitet, som Durkheim ofte benævner denne morfologiske egenskab (1982). Densitet (ρ) er, i sagens natur, en rumlig kate- gori, defineret som forholdet mellem den numeriske masse af mennesker

(m) og rummets arealmæssig volumen (V): ρ = m/V.8 Endvidere spiller rummet en afgørende rolle i reguleringen af densitetsgraden, idet eksempelvis fremkomsten af transportforbindelser vil være med til at forme, hvor mange mennesker det er muligt at transportere fra A til B. Urbani- seringen er med dens karakteristiske høje densitetsgrad tæt forbundet med den materielt-teknologiske udbygning af infrastrukturer, der mulig- gør og fremskynder til tilflytningen til byen, sådan som Durkheim (1984) eksempelvis diskuterer det i The Division of Labor in Society. Læsningen af Durkheims overvejelser om den dynamiske densitet vanskeliggøres af, at dens kausale effekt tilskrives eksistensen af to væ- sensforskellige mekanismer: Social konflikt og følelsesmæssig smitte. Mest berømt er utvivlsomt Durkheims (1984[1893]) argument i det tidligere hovedværk, Division, hvor han argumenterer for, at den øgede dynamiske densitet har en social effekt, eftersom sociale densitetsgrader er kausalt korreleret med samfundets niveau af social konflikt. Dette begrundes med reference til Darwin (2010[1872]) og hans evolutionsbiolo- giske argument om, at den øgede densitet af dyr og arter i en økologisk niche skærper “kampen for tilværelsen” (en vending som Darwin i øvrigt lånte fra sociologen Spencer, som Durkheim også refererer flittigt til i Division). Denne densitets-konfliktmekanisme udpeges slet og ret af Durkheim som “den sociale verdens tyngdelov” (Durkheim 1984). Og skiftet fra mekanisk solidaritet (dvs. solidaritet som resultat af ligheds- punkter, der forbinder individerne umiddelbart til samfundet) til organisk solidaritet (dvs. solidaritet, som et resultat af gensidig afhængighed mellem differentierede individer) sker således på kausal-mekanisk vis, idet “(…) a break in the equilibrium of the social mass gives rise to conflict that can only be resolved by a more developed form of the division of labor: this is the driving force for progress.” (Durkheim 1984: 212). I det sene hovedværk, Elementary Forms, fastholdes den morfologiske interesse for densiteten blot under andre betegnelser som “assemblies”, “congregations” eller gruppens “concentration” (Durkheim 1995[1912]). Hvad der derimod forlades er konfliktfortolkningen af densiteten, der nu sættes i forbindelse med en følelsessociologisk mekanisme: Når sociale

8 Bemærk, at vi her har at gøre med en georumlig definition af densitet, hvilket står i kontrast til en socialtopologisk/relationel forståelse af densitet, der er udbredt indenfor social-netværksanalyse. Sidstnævnte definerer typisk densitet som ratioen mellem antallet af faktiske sociale relationer og antallet af mulige sociale relationer (Giuffre 2013; Scott 2012). Forskellen lader sig illustrere ved en befærdet banegård, hvor der er en høj geo-rumlig densitet, men en lav socio-topologisk densitet, idet der ikke eksisterer sociale bånd mellem de tilstedeværende. Dette afhænger selvsagt af definitionen af sociale bånd, men sammenligningen holder stik, hvis denne er karakteriseret ved, at individerne ’kender’ hinanden (Urry 2004).

grupper er forsamlede, finder der en følelsesmæssig smitte sted mellem de kropsligt tilstedeværende individer. En sådan følelsesmæssigt selvfor- stærkende smitte betegner Durkheim som en “kollektiv opbrusen” og dette begreb befinder sig i hjertet af Durkheims argument om, at religiø- se institutioner er lige så sociale, som sociale institutioner er religiøst funderede. Det er således som resultat af en densitet-smittemekanisme, at gruppen bliver bevist om sig selv som gruppe, dvs. gruppen får intersubjektivitet eller en kollektiv bevidsthed, hvilket individet oplever, som at stå over for en eksternt tvingende social kraft; det lade sig være gud, samfundets moral eller situationens sanktioner:

The very fact of the concentration acts as an exceptionally power- ful stimulant. When they are once come together, a sort of electric- ity is formed by their collecting which quickly transports them to an extraordinary degree of exaltation. Every sentiment expressed finds a place without resistance in all the minds, which are very open to outside impressions; each re-echoes the others, and is re- echoed by the others (Durkheim 1995: 217-218).

Afhandlingens argument om, at en sådan durkheimiansk følelsessmit- te skal fortolkes som et socialmorfologisk begreb, der er begrundet i gruppens sociale densitet, er ikke ukontroversiel. En lang tradition af sociologiske Durkheim-fortolkere har således insisteret på, at den sene Durkheim i Elementary Forms brød med det tidlige forfatterskab og Divisi- ons strenge betoning af socialmorfologiske lovmæssigheder. Afhandlin- gen lægger imidlertid afstand til denne, af Tonboe (1993: 125ff) karakte- riserede, “brudlæsning” af Durkheims forfatterskab, som Alexander (2005), Lukes (1973) og Nemedi (1995) er indflydelsesrige eksponenter for. Alexander (2005) går endda så langt som til at hævde, at “(…) Durk- heim reached his maturity after a prolonged, if confused, flirtation with materialist forms of structural theory, and eventually a fierce struggle against them.” (Alexander 2005: 136). Det er dog sigende for denne betragtning, at den “fierce struggle”, som Alexander her foreslår eksistensen af, ikke kan anføres som et gyldigt teorihistorisk bevis for, at Durk- heim forkastede sin tidlige socialmorfologi. Durkheim havde rigtignok en “fierce struggle”, men nærlæses det argument hos Durkheim (1978b), som Alexander refererer til, fremgår det klart, at hans afvisning specifikt angik den marxistiske historiske materialisme og ikke sin egen sociale morfologi. Ved således at betone socialmorfologien som et kontinuerligt spor i Durkheims forfatterskab, er afhandlingen snarere på linje med Tonboe

(1993), Andrews (1993), Nielsen (1999), Strenski (2006) og Fourniers (1996; 2012) nyere teorihistoriske studier. Disse peger samstemmende på, at morfologien nok fik en anden, og i visse henseender mindre frem- trædende, rolle i Durkheims sensociologi, men ikke desto mindre var et konstant tema forfatterskabet igennem. Durkheim forkastede aldrig sin sociale morfologi. Den morfologiske egenskab af social densitet er et godt eksempel herpå. Durkheim fastholdt her, under anvendelse af forskellige termer for densitet, at det sociale livs tæthedsgrader har betydelige sociale effekter, men forklaringsmekanismen forskydes fra konflikt til smitte. Denne kontinuitets- frem for brudlæsning af Durkheims sociale morfologi bliver også tydelig, når Durkheim læses i konteksten af hans sociologiske skole, hvor flere centrale medlemmer fastholdt og opdyrke- de det socialmorfologiske spor. Som Fournier (2012) understreger det, er en sådan skolekontekst-læsning den biografisk set mest korrekte måde at læse den durkheimianske tradition:

There is a collective dimension to Durkheim’s approach. L’Année sociologique was obviously a collective undertaking. But what about the individual books? Who is the author of Elementary Forms of Reli- gious Life? Durkheim or Mauss? The relationship between uncle and nephew was very close – Marcel was Émile’s alter ego: It is impossible for me to extricate myself from the work of the school’, said Marcel Mauss (…), who set great store by the importance of collective work and collaboration – as opposed to isolation and the pretentious quest for originality. (Fournier 2012: 4).

At Durkheim i realiteten er “Durkheim, Mauss & Co”, som Fournier (2012) formulerer det, har specifikt betydning for, hvor rummorfologisk, der er belæg for at fortolke den sene durkheimianske skole og dennes forståelse af densitetsdrevet følelsessmitte. Afgørende i denne sammen- hæng er Mauss’ (1979[1906]) studie, Seasonal Variations of the Eskimo: a Study in Social Morphology, der sandsynligvis er det mest konsekvente rummorfologiske bidrag til traditionen – og det på trods af, at studiet er udgivet et årti efter at brudlæsningen antager, at den sociale morfologi var blevet forladt (Alexander 2005 argumenterer således for, at bruddet skete i 1896, ti år før udgivelsen af Mauss’ eskimostudie i 1906). Eskimo er imidlertid ikke blot af teorihistorisk interesse. I relation til afhandlingens teoriudviklende rekonstruktion af morfologiens rumsociologiske bidrag, inviterer og inspirerer Mauss’ ofte oversete studie mig til forstå intensiteten, altså den sociale følelsessmitte, som et morfologisk produkt af rumlig kausalitet. Det er netop denne rummorfologiske invitation, som af-

handlingen udfolder rumsociologisk og operationaliserer mikro-metodologisk, hvorfor jeg i det følgende vil skitsere Mauss’ argument.

Rum, densitet og følelsesmæssig smitte Mauss’ Eskimo er et casestudie, der tager empirisk-metodologisk bestik af, hvordan eskimokulturens sæsonmorfologi “(…) offer a favourable condition for a study of how the material form of human groups – the very nature and composition of their substratum – affects different modes of collective activity.” (Mauss 1979: 20). Eskimoerne lever med en ekstremt forskellig social densitet i sommer- og vinterhalvåret. Om sommeren lever de rumligt spredt ud over landskabet i små hytter og telte med en meget lav densitetsgrad, hvilket omsætter sig i eskimoernes lav-intense aktivitetsniveau i dette sommerhalvår. Dette ændrer sig med vinterens komme, hvor eskimoerne lever ekstremt rumligt fortættet i små hytter og sommetider i igloer. Denne forandring af eskimoernes rumligt- materielle livsbetingelser har en markant morfologisk effekt på karakteren af det sociale liv. Den ændrede morfologiske ratio mellem livsforhol- denes masse af individer og rumforholdenes knappe volumen accelererer og intensiverer eskimoernes sociale følelsesliv:

Winter is a season when Eskimo society is highly concentrated and in a state of continual excitement and hyperactivity. Because the individuals are brought into close contact with each other, their social interactions become more frequent, more continuous and more coherent; ideas are exchanged; feelings are mutually revived and reinforced. By its existence and constant activity, the group becomes more aware of itself and assumes a more prominent place in the consciousness of individuals (Mauss 1979: 76).

Hvad der understøtter, at vi har at gøre med Durkheim, Mauss & Co, er Pickerings (1984) teorihistoriske påpegning af, at Durkheims begreb om kollektiv opbrusen sandsynligvis er inspireret af Mauss’ Eskimo. Det er altså Mauss, der først fremdrog den densitets-smittemekanisme som Durkheim senere lagde til grund for Elementary Forms argument om, hvordan den følelsesmæssigt smittede og begejstrede gruppe, bliver be- vidst om sig selv og får kollektiv bevidsthed. I henhold til afhandlingens fokus på, hvordan det fysiske mikro-rum udgør en rumlig-kausal kontekst for mikro-interaktionelle følelsesintensi- teter, er det mest rumsociologisk interessante ved Mauss’ argument, hvordan det specifkt er rumlige egenskaber ved naturen og det bebyggede miljø, der udpeges, som den kausale kraft, der intensiverer følelsesli-

vet. Naturens sæsonmorfologiske rytme regulerer, hvilke arealmæssige forhold og, som rumlig konsekvens heraf, hvilke densitetsmæssige forhold som eskimo-populationen lever under: “The seasons are not the direct determining cause of the phenomena they occasion; they act, rather, upon the social density that they regulate.” (Mauss 1979: 79). Mens betydningen af rummet optræder i Durkheims Elementary Forms, som en del af definitionen af den forsamlede gruppes densitet (dvs. som nævne- ren i en matematisk bestemmelse af densitet, som: ρ = m/V), går Mauss altså rumsociologisk videre og forklarer denne sociale densitet som et kausalt produkt af rummet kausalitet. Mauss tilfører således socialmorfologien et non-humant twist: I Durkheims Elementary Forms forsamles gruppen, som følge af individernes formålsbevidste opsøgen, dvs. stammens medlemmer beslutter sig for at rejse til ritualet, mens Eskimo så at sige redistribuerer denne sociale handlen til rummets non-humane agency. Her skal det nævnes, at en sådan læsning af Eskimo ej heller er nor- men i den sociologiske litteratur; eksempelvis gør Douglas (1972) sig til talsperson for det modsatte synspunkt. Douglas forsøger hermed at presse brudlæsningen ned over Eskimo, der i hendes udlægning:

(…) is an explicit attack on geographical or technological deter- minism interpreting domestic organization. It demands an ecological approach in which the structure of ideas and of society, the mode of gaining livelihood and the domestic architecture are interpreted as a single interaction whole in which no element can be said to determine the others (Douglas 1972: 513-514)

Rigtignok tager Mauss i lighed med Durkheim afstand fra rumlig de- terminisme, idet det eksempelvis understreges, at rumfaktorer altid skal overvejes i relation til den sociale kontekst i al dens komplekse totalitet (Mauss 1979). Men Douglas forbigår dog, at Mauss opererer med en morfologisk kausalmekanisme, hvor rummet udgør den udløsende årsag til den effekt, at eskimoernes sociale liv følelsesintensiveres. Det er netop denne rummorfologiske mekanisme i Eskimo, som afhandlingen har fundet så rumsociologisk inspirerende. Hertil skal dog tilføjes, at Mauss’ rummorfologiske begrebsapparat lader flere af de rumsociologiske spørgsmål, der er specifik interesserer afhandlingen, stå ubesvaret hen: Hvordan og hvorvidt kan den kausale kraft, der, ifølge Mauss’ Eskimo, udgår fra naturens sæson-morfologi generaliseres til det bebyggede rums arkitektur-morfologi? Hvordan kan vi gøre sådanne mikro-rumlige egenskaber og effekter ved arkitektur-morfologien til genstand for kvantita-

tiv-metodologisk undersøgelse? Disse teoretiske og metodiske spørgsmål lader sig, som jeg vil argumentere for senere, først besvare fyldestgøren- de i lyset af space syntaxens neo-durkheimainske bidrag. Mens vi således midlertidigt sætter disse spørgsmål i parentes, vil jeg fremhæve særligt én central analytisk styrke ved Durkheim, Mauss og Co’s forståelse af social densitet som korreleret med følelsessmitte. Mens dette durkheimianske argument peger på, at den sociale densitet først og fremmest er kilde til intensive oplevelser i den positive ende af følelsesre- gisteret (opbrusen, medriven, etc.), har en lang tradition inden for rum- og bysociologien haft tendens til at overbetone densitetens problematiske sider og negative følelseskorrelater. Eksistensen af en sådan pessimi- stisk ensidighed er eksempelvis påvist i Boyko og Coopers (2010: 19ff) nyere tværvidenskabelige litteraturstudie, der konkluderer, at densitet i hovedsagen forstås som kilde til patologier, konflikt og negative følelser af stress og ubehag (se også Baldassare 1979; Liebst 2009). Særdeles indflydelsesrig i denne sammenhæng er Simmels (1976) argument om, at byens fortættede socialitet medfører en æstetisk over- stimulation eller “intensivering af nervelivet”, hvilket tvinger byboen til at anlægge en blasert attitude. Menneskets tendens til “imitation” (dvs. Simmels begreb om smitte), står i et psykologisk-problematisk spæn- dingsforhold til bylivets høje sociale densitet. I forlængelse af Simmels (1976) over-stimulations-hypotese er det Wirths’ (1938) mindst lige så indflydelsesrige hypotese, at byens høje densitet implicerer mentalt stress, social konflikt og anomi. Og i samme spor finder vi Milgram (1974), der, i et indflydelsesrigt studie af oplevelsen af at leve i byen, påpeger, hvordan byens densitetsmæssige forhold implicerer et sansemæssigt “overlo- ad”, der udkrystalliserer sig i dehumaniserede urbane omgangsformer. Litteraturen byder dog også på modsigelser af dette densitets-pessimistiske argument, hvilket socialpsykologen Freedmans (1975) densitets- intensitets-teori nok er det bedste eksempel på. I hans perspektiv er densitet ikke i sig selv kilde til negative følelser, sådan som det er litteraturens dominerende argument. I stedet påpeger Freedman “(…) that crowding by itself has neither good effects nor bad effects on people but rather serves to intensify the individuals reactions to the situation.” (Freedman 1975: 90, original kursivering).” Argumentet er altså, at densitet fungerer som en art intensiverings-mekanisme for den følelsesmæssige tilstand, som situationen har i forvejen. Dette forklares psykologisk som, at en højere situationel densitet, defineret med reference til individ-rums- ratioen, “(…) makes other people a more important stimilus and thereby intensifies the typical relation to them.” (Freedman 1975: 105). At den høje densitetsgrad, som Simmel, Wirth og Milgrams bysociologiske pes-

simisme baserer sig på, således i Freedmans (1975) udlægning ikke i sig selv anses som værende anledning til negative følelser betyder, at hans moralske stillingtagen til byen er langt mere optimistisk. Efter Freedman har fremlagt densitets-intensitet-teorien i sin Crowding and Behavior, følger således et kapitel, der slet og ret har titlen: “In praise of cities”! Freedmans densitets-intensitets-hypotese befinder sig tæt på det durkheimianske argument om den densitets-genererede følelsessmitte som kilde til intensivering af gruppers følelsesliv. Dette understreges også af, at det er en beslægtet intensiverings-mekanisme, som Collins (1993; 2004), i direkte forlængelse af Durkheims og Mauss’ socialmorfologiske argument, bygger ind i sin IR-teori: “The crucial dynamic of group experience is emotional. Which emotion is initially aroused is less important than the fact that group density, boundedness and focus produces an increase in this emotion.” (Collins 1993: 208). Dog skal det bemærkes, at den intensiverings-mekanisme, der for Freedman forløber som en mo- nokausal densitet-intensitet-proces med densiteten som den ene uaf- hængige variabel, i Collins neo-durkheimianske udlægning udfolder sig som en multikausal density/boundedness/focus-intensitets-proces. Nær- læses de tænkte eksempler, som Freedman (1975) benytter sig af til at illustrere sit teoretiske argument, lader konturen af en sådan multikausal forståelse af intensiveringsprocessen sig også ane; men Collins når altså videre end Freedman i den teoretiske formalisering og forklaring heraf. Jeg vil i denne sammenhæng ikke redegøre nærmere for Collins socialmorfologiske bidrag, da dette gøres detaljeret andetsteds. Her er det tiltrækkeligt at fastslå, at Collins må anses som den, der i den mikro- durkheimianske tradition klarest og sociologisk mest nuanceret, har ar- bejdet videre med Mauss og Durkheims’ forståelse af den sociale densitets positive korrelation med følelsesintensiverende smitte. En sådan densitet-intensitetskorrelation er således også impliceret i det mikro-sociologiske argument, som Collins afleder fra Goffman (1967; 1969) og som udgør et kerneargumentet for afhandlingen, nemlig: at det at være i en fortættet gruppe af mennesket, giver individet fornemmelsen af at være dér, hvor tingene sker, “where the action is” (Collins 2004: 82). Goffman og Collins må således henregnes som del af det nutidige ‘Co’, der konstituerer det durkheimianske subjekt: Durkheim, Mauss & Co.

Balkanisering eller akkumulering Negligeringen af socialmorfologiens rumsociologiske bidrag har et histo- risk og nutidigt gestalt. Ovenfor har vi håndteret dette arguments historiske gestalt, der havde rod i den teorihistoriske misforståelse, at socialmorfologien mest af alt er et umodent materialistisk fejlskud, som Durk-

heim forkastede, da hans tanker blev idealistisk modne. Nu må det overvejes, hvordan fremlæsningen af den morfologiske kontinuitet i den durkheimianske tradition, stiller os i forhold til den nutidige negligering af morfologien, der har rod i Latours kritik af Durkheim for at forbigå samfundets rumligt-materielle ting. Hvorvidt den Latour-inspirerede afvisning af Durkheim anses som legitim hænger tæt sammen med, om man forstår sociologien som en akkumulativ videnskab eller ej. Som påpeget af Collins (1999; 1989), baserer en antikumulativ tilgang til sociologien sig ofte på, at man lader de sociologiske teoriers uenighed skygge for den akkumulerede empiriske viden, som sociologien faktisk har præsteret. Sociologien lider, ifølge Collins, under en sådan socialt miskendt videnskumulation. Collins (1999) fremhæver særskilt Latour (1987) som en sociolog, der har bidraget til en sådan teoretisk balkanisering af sociologien. Og Collins beretti- gede argument er, i kontrast hertil, at sociologisk videnskabelighed ikke blot er “in-the-making”, men også akkumuleres empirisk til et relativt sikkert vidensfundament (for tilsvarende kritikker af Latours’ antiakku- mulations-epistemologi, se Osborne 1996; Turner og Kim 1999). Relate- res disse epistemologiske betragtninger dernæst til afhandlingens rummorfologiske argument, rejser følgende spørgsmål sig: Er Latours afvisning af Durkheim kun mulig fordi han miskender den akkumulation af viden om det rumligt-materielle objekts betydning for det sociale liv som de bidrager konvergerende til? Tag følgende citat i betragtning, hvor La- tour definerer ANT og dets non-humane perspektiv i kontrast til “det sociales” sociologer, der, som Durkheim, angiveligt overser, hvordan det sociale kun er stabilt og udstrakt når det er sammenføjet med det materielle objekt:

ANT is not the empty claim that objects do things ‘instead’ of human actors: it simply says that no science of the social can even begin if the question of who and what participates in the action is not first of all thoroughly explored, even though it might mean letting elements in which, for lack of a better term, we would call nonhumans. This expression, like all the others chosen by ANT is meaningless in itself. It does not designate a domain of reality. It does not designate little goblins with red hats acting at atomic levels, only that the analyst should be prepared to look in order to account for the durability and extension of any interaction. The project of ANT is simply to extend the list and modify the shapes and figures of those assembled as participants and to design a way to

make them act as a durable whole (Latour 2005: 72, original kursivering).

Læg først mærke til, at det vigtige ikke er, om etiketten “non-human” anvendes (hvilket durkheimianerne ikke gør) men, at sociologien har en analytisk åbenhed over for hvem og hvad, der deltager i sociale hand- lingsprocesser. Uden en sådan sociologisk åbenhed, som “det sociales” sociologer angiveligt ikke har, er det ikke muligt at redegøre for holdbarhe- den og udstrækningen af interaktioner. I citatet tales der ikke direkte om rummet som objekt, men dette berøres dog indirekte med termerne former og figurer (“shapes and figures”) og ANT-perspektivet er da også andetsteds, blevet sat i relation til arkitektoniske-rumlige ting og tematik- ker (Latour 1988; Latour og Yaneva 2008; Yaneva 2009). Således argumenterer Yaneva (2010) for, at et ANT-perspektiv på rumligt-arkitektoniske former implicerer en understregning af, hvordan de rumlige objekter fungerer som relationer (“connectors”), der netop øger interaktionens udstrækning og holdbarhed. Den teoretiske terminologi er anderledes, men dette ændrer ikke ved, at de rumligt-materielle konklusioner vi finder hos Durkheim, Mauss og Co i mange henseender er konvergerende med sådanne ANT-indsigter. Så frem for at fetichere de teoretisk-terminologiske forskelle positioner- ne imellem, foreslår afhandlingen, at vi læser morfologi- og ANT-traditionerne som konvergerende bidrag til sociologiens akkumulerede viden om, hvordan rumlige ting “(…) shapes, conditions, facilitates and makes possible everyday sociality.” (Yaneva 2009: 273). I dette lys er afhandlingen på linje med Lindemanns argument, der påpeger, at Latour – på trods af sin teoretiske intension om det modsatte – tilbyder et “(…) brilliant proof of Durkheim’s theory of the morphology of social facts.” (Lindemann 2011: 93; se også Lemieux 2012). Lindemann begrunder imidlertid ikke sin teorihistoriske betragtning med meget belæg. Det kommer her: Understregningen af rummets aktive formning af sociale interaktioner er, som anført, fremtrædende i Mauss’ Eskimo og Durkheims Division, hvor det eksempelvis noteres: ‘‘The communication network forcibly prescribes the direction of internal migrations or commercial exchanges, etc., and even their intensity.’’ (Durkheim 1984: 58). En sådan rummorfologisk betragtning er snævert sammenfaldende med Latours ovenfor citerede argument om, at det rumlige objekt er med til at sikre “udstræk- ningen” af samfundets interaktionsmønstre. Skiftet fra præmodernitet til modernitet er, i Durkhiems opfattelse, forbundet med en rumlig foran- dringsproces, hvor præmodernitetens “segmentære” organisering af det

sociale nedbrydes af fremkomsten af infrastrukturelle rum- og teknologi- relationer, der faciliterer mere udstrakte former for interaktion, kommu- nikation og mobilitet. Konvergerer socialmorfologien også med ANTs argument om det sociales objektbaserede holdbarhed? Her hjælper Latour, på trods af sin intention om det modsatte, mit kumulations-epistemologiske argument lidt på vej. Under overskriften “Durkheim having a Tardian moment” citerer Latour en central passus fra Elementary Forms, der får en ironisk ment kommentar med på vejen: “Does the totem express the group, facilitate its cohesion, or is it what allows the group to exist as a group?” Lad os citere fra det samme sted i Elementary Forms:

Without symbols, moreover, social feelings could have only an un- stable existence. Those feelings are very strong so long as men are assembled, mutually influencing one another. But when the gathering is over, they survive only in the form of memories that gradually dim and fade away if left to themselves. (…) But if movements by which these feelings have been expressed eventually become inscribed on things that are durable, then they too become durable. These things keep bringing the feelings to individual minds and keep perpetually aroused, just as would happen if the cause that first called them forth was still acting. Thus, while emblematizing is necessary if society is to become conscious of itself, so is it no less indispensable in perpetuating that consciousness (Durkheim 1995: 232-233).

Latour ironiserer over det forhold, at Durkheim ufrivilligt gør det, som Gabriel Tarde (udråbt af Latour 2004 som ANTs stamfader) kritiserede Durkheim for ikke at gøre: nemlig at indse, at “gruppen” ikke forklarer noget, men tværtimod er det, der skal forklares. Det er ANT- argumentet, at en sådan forklaring netop må inkludere tingenes bidrag, sådan som Durkheim på ironisk vis kommer til at indrømme i et øjebliks klarsyn. Gruppen ville ikke kunne eksistere stabilt uden mellemkomsten af gruppens totemobjekt. Problemet med Latours læsning af Durkheim er imidlertid, at der ikke blot er tale om et øjebliks klarhed fra Durkheims side i det citerede passus fra Elementary Forms. Durkheim er snarere vedholdende og konsistent i sin understregning af, at totemmet er et symbol, der er stofligt forankret i konkrete ting og objekter: “The totem is above all a symbol, a tangible expression of something else” (Durkheim 1995: 208), nemlig gruppens fejring af sig selv, hvilket følelsesoplader totemobjektet til at være grup-

pens “flag” (se også Durkheim 1995: 114). Schiermer (2011: 86) opsummerer dette på rammende manér: “In a word, Durkheim’s sociology of religion is not simply about social things but also about social things; about sacred objects.” (Schiermer 2011: 86, original kursivering; se endvidere Jerolmack og Tavory 2014). Den stoflige egenskab ved emblemet er af- gørende for, at gruppens sociale relationer kan stabiliseres, herunder som udstrakte relationer på tværs af tid og rum, når det fjernes fra den rituelle situation, hvor det er blevet følelsesmæssigt opladt. Symboler og følelser, der ikke er tingsligt forankrede, opleves som mere abstrakte og porøse og er derfor mindre virksomme totemobjekts-manifestationer af gruppens tingslige tvang. Disse overvejelser peger på, at det durkheimianske totemobjekt befinder sig tæt på Latours (1990) begreb om “immutable mobile”; forstået som et objekt, der kan transporteres uden at ændre form, og som i denne egenskab er afgørende for udstrækningen af viden- skabelig viden (i Latours tilfælde) og religiøse følelser (i Durkheims til- fælde). Durkhem, Mauss og Co giver ydermere utallige specifikke rummorfologiske eksempler på, hvordan det rumlige objekt er med til at stabilisere sociale interaktioner. Her vil jeg fremhæve Durkheim og Mauss’ (2009) Primitive Classification, hvor det påpeges, hvordan primitive stammers mentale klassifikationer af verden (“cosmic space”), reflekterer og korre- sponderer med stammens rumligt-sociale organisering (“tribal space”): “Cosmic space and tribal space are thus only very imperfectly distinguished, and the mind passes from one to the other without difficulty, almost without being aware of doing so.” (Durkheim og Mauss 2009: 39). Og det uddybes videre, at stammens klassifikationssystem skal for- stås som modelleret over stammens “morfologiske organisation” (se også Mauss 1979: 80; Thompson 2003). Ifølge Durkheim og Mauss (2009: 7) er en sådan klassifikations-morfologi-korrespondens med til at reproducere menneskets begrebsverden: Den rumlige morfologi spiller her en direkte rolle i vedligeholdelsen af sociale interaktioner. Termino- logien er igen anderledes fra ANTs, men grundindsigten er konvergerende. De sociale interaktioner stabiliseres af gruppens rumligt-materielle organisering.9

9 Der er flere konvergenser end de her anførte. En konvergens, der fortjener at blive nævnt, er mellem Latours (2012) We Have Never Been Modern, der fokuserer på den hy- bride natur-kultur-relation og Mauss’ eskimostudie. Eskimo kan netop forstås som et studie af natur-kultur-hybrider og i lighed med Latours argument er det Mauss’ argument, at sådanne hybrider ikke er unikke for de præmoderne eskimoer men, at vi, i det moderne også er præget af sammenføjningen mellem naturens sæsonmorfologi og samfundets sociokulturelle fysiologi (Mauss 1979).

Den metodiske tærskel Afhandlingens argument om, at socialmorfologien har et rumsociologisk potentiale, som traditionens kritikere har overset, må ikke forveksles med et argument om, at denne durkheimianske tradition er problemfri. Dette er ingenlunde tilfældet. Jeg vil dog argumentere for, at disse problemer ikke, som kritikerne oftest antager, er af teoretisk karakter, men først og fremmest skal tilskrives det forhold, at den sociale morfologi aldrig er blevet forløst metodisk-kvantitativt. I fremførelsen af dette argument vil jeg indledningsvis citere Tonboe (1993), der ligeledes sætter socialmorfologiens teorihistoriske problem i relation til metodespørgsmålet:

Det er det metodiske dogme – analogiseringen med naturviden- skaben – om at sociologiske fakta kun kan forklares med sociologiske fakta, der sammen med kravet om generelle love og relativt enkelte og klare kausale sammenhænge, stiller sig i vejen for en klar tematisering af samspillet mellem det rumlige og det sociale. (…) Dermed modsiges reelt Durkheims erklærede faglige interesse, men ikke det faginstitutionelle perspektiv, der altså kom til at do- minere. (…) Man kan måske sige det sådan, at Durkheims metodologisk-sociologiske (positivistiske) dogmer blokerer for udviklingen af en sekundær faglig interesse i rummets sociologi. Epistemologi- en blokerer for ontologien. Durkheim valgte, da det kom til stykket, at fastholde og endda profilere sin specielle sociologi, som han opfattede som kernen i al fremtidig sociologi, frem for at ekspan- dere eller fortsat gå på strandhugst med en ny underdisciplin ved selv at udvikle det faglige felt empirisk, når nu andre ikke ville (Tonboe 1993: 168).

Mens Tonboe på den ene side har ret i, at socialfysiologien gradvist (men på trods af Durkheims erklærede morfologiske interesse) får en stadig mere fremtrædende plads i Durkheims senforfatterskab, stiller jeg mig på den anden side skeptisk over for Tonboes epistemologiske fortolkning af denne tendens. Durkheims (1982) regel om, at sociale fakta skal forklares med forudgående sociale fakta, er ikke, som det synes at være Tonboes argument, epistemologisk uforenelig med Durkheims faginteresse i rummets ontologi. Durkheim (2005: 278) forstår tværtimod det morfologiske rum som et typisk socialt fakta, der på tingslig vis udø- ver en ekstern tvang på individet. Dette fremgår i The Rules of Sociological Method, hvor Durkheim slår fast, at “[s]ince the facts of social morphology are of the same nature as physiological phenomena, they must be explained by the principle just enunciated.” (Durkheim 1982: 112). Sociale

fakta inkluderer både samfundets immaterielle og materielle egenskaber (Durkheim, 1960: 363). Frem for Tonboes argument om, at “epistemologien blokerer for ontologien”, foreslår jeg, at rummorfologiens udfordring hænger sammen med, at metodologien blokerer for epistemologien. Fysiologien og morfologien kan og skal forklares positivistisk-epistemologisk som kausale sociale fakta, men den metodologiske blokering der konfronterede Durkheim var, at rum-morfologiske fakta er langt vanskeligere at måle som en ting end social-fysiologiske fakta. Durkheim, Mauss og Co havde ganske enkelt ikke de kvantitativt-metodiske værktøjer ved hånden, der ville have gjort det muligt at realisere den mest grundlæggende metode- regel i relation til rummet: “The first and most fundamental rule is: Consider social facts as things” (Durkheim 1982: 14, original kursivering). Casemetoden, der bliver anvendt i Eskimo, havde vist en mulig kvalitativ vej at måle det morfologiske rum som en ting. Men for Durkheim, Mauss og Co er det afgørende, at morfologien også gøres til genstand for statistisk-kvantitativ måling og forklaring. De anså statistikken som en uomgængelig måde at kontrollere for (“cancel out”) effekten af individu- elt-unikke forhold; kvantificering er således en metodologisk kongevej til at studere samfundets kollektive effekter og årsager (Durkheim 1982: 55).10 Som den ellers etnografisk orienterede Mauss (2003: 69) slår fast: “Basically, every social problem is a statistical problem.” (se også Mauss’ 2009 gentagende understregning af vigtigheden af statistiske morfologi- studier i hans Manual of Ethnography). At manglen på metodisk-kvantitative værktøjer specifikt udgør en akilleshæl for rummorfologiens empiriske udfoldelse, kommer måske klarest til udtryk i forbindelse med Mauss’ refleksioner over den durkheimianske tradition i årene efter Durkheims død. Her fremhæves det, hvordan måling af morfologiske egenskaber som “rum”, “antal” og “tid” “(…) demand the use of quantitative methods. To refer to material and social structures and the movements of these structures is to refer to things which can be measured.” Men samtidigt beklages det, at en sådan kvantificering ikke finder sted i tilstrækkeligt omfang: “Alas! Even in the Année sociologique, we are far from what is desirable.” (Mauss 2003: 69, original kursivering). Anvendelsen af de eksisterende kvantitative metoder er for snæver; sociologien tager ifølge Mauss ikke den fulde metodiske konsekvens af, at ethvert socialt problem er et statistisk problem. Mauss er dog metodologisk optimist på vegne af den fremtidige genera-

10 Jævnfør også de, efter min vurdering, korrekte indvendinger mod case-metoden, som er fremført af Lieberson (1991).

tion af socialteoretikere, idet disse ifølge ham vil være bedre teknisk “armed than we were” (Mauss 2003: 70). I relation til afhandlingens mikro-teoretiske tilgang er det interessant, at Mauss, i forbindelse med disse overvejelser, specifikt efterlyser kvantificeringer af det sociales morfologiske mikro-egenskaber frem for makro- skala kvantificeringer af samfundets egenskaber af “population”, “kriminalitet og civilstatus” og “økonomi” (Mauss 2003: 69). Dette skal ses i lyset af, at det netop var i relation til makro-kvantificeringen af sådanne forhold, at Durkheim og hans samtidige stod metodisk stærkest. Dette kommer eksempelvis til udtryk i Durkheims (2005) Suicide og Halbwachs (1930) senere opdatering af dette makro-orienterede selvmordsstudie samt Halbwachs (1960) udprægede makro-sociologiske opsummering af det socialmorfologiske program.11 Når Mauss eksemplificerer, hvilke morfologiske emneområder, der også burde kvantificeres, refererer han på denne baggrund til mere hverdagslige mikro-fænomener:

The frequency of the fact, the number of participating individuals, the repetition trough time, the absolute and the relative importance of the acts and their effect in relation to the rest of life, etc. – all this is measured and should be counted. The attendance at the theater or at sports events, or the number of editions of a book, tell us far more about the value attracted to a work or to a sport than do pages and pages by moralists or critics. The strength of a church is measured by the number and the wealth of its temples, the number of its believers and the magnitude of their sacrifices (…) (Mauss 2003: 69-70).

Mens Mauss her foreslår, hvordan en øget socialmorfologisk mikro- kvantificering kan finde sted, skal vi også hæfte os ved, hvad Mauss ikke kommer ind på i citatet. De anførte eksempler angår alene morfologiens

11 En sådan makro-sociologisk kvantificering af socialmorfologien har endvidere et teorihistorisk spor frem til den eksplicit makro-sociologiske del af den humanøkologi- ske skole, der, som Schnore (1958; 1961) og Duncan (1959), forstod sig som arvtagere af den sociale morfologi. Således er det også sigende, at Halbwachs (1960) makrosociologisk orienterede læsning af socialmorfologien blev oversat fra fransk til engelsk af Otis Duncan. Her skal det dog tilføjes, at den humanøkologiske skole fra starten af (f.eks. i Parks 1915 klassiske formulering), ikke byggede på socialmorfologien, sådan som Schnore (1958; 1961) forsøger at fremlæse teorihistorien. Som Tonboe (1993) har påvist, er det Simmel snarere end Durkheim, der ligger til grund for humanøkologien. Schnores forsøg på at knytte skolen an til morfologien, må ses som et efterrationalise- rende forsøg på et re-positionere skolen i en tid, hvor den var på vej ned, snarere end en teorihistorisk præcis læsning (se Snell 2010 for en nutidig variant af denne problematiske kobling af humanøkologien og Durkheim).

numeriske egenskaber, mens det er påfaldende, at Mauss ikke nævner morfologiens rumlige egenskaber. Dette kunne være et simpelt tilfælde, men jeg vil i stedet foreslå, at denne rummorfologisk fortielse fortolkes som symptom på morfologiens metodologiske mangel på kvantitative redskaber til at måle det fysiske mikro-rum. Mens de numeriske egenskaber ved det morfologiske substratum lader sig optælle og måle relativt problemfrit, var og er værktøjskassen anderledes tom, når det gælder kvantificeringen af mikro-rummets morfologi. Ikke uden grund har Mauss således svært ved at give et metodologisk klart eksempel på en mulig måde at mikro-kvantificere den rumlige morfologi. Hertil skal det tilføjes, at dette metodologiske måleproblem ikke er unikt for socialmorfologien, men mere generelt må forstås i lyset de principielle metodologiske udfordringer, der er forbundede ved at gøre de fysiske rum til rumsociologisk genstand for metodisk undersøgelse. Hvori disse metodologiske udfordringer består, og hvordan de lader sig løse, ser jeg nær- mere på i næste afsnit.

At måle mikro-rummet Mens man i hovedsagen søger forgæves efter redskaber til kvantificering af mikro-rummet blandt de nutidige, ofte ANT-inspirerede, rumsociologier, der påberåber sig den “rumlige vending” (Gieryn 2000; 2002; Gra- ham og Marvin 2001; Farias og Bender 2012; Thrift 2008; Amin og Thrift 2002; Warf og Arias 2008), findes der, uden for denne rumsociologiske hovedstrøm, ansatser til en sådan rumkvantitativ metodeudvik- ling. Det skal understreges, at der vitterligt er tale om ansatser, sådan som Logan (2010) konstaterer det, i forbindelse med sit litteraturstudie af det rumsociologiske felts metodiske status: “Unfortunately, we have little knowledge at this level of specificity.” (Logan 2010: 513). Vurderet på baggrund af Logans gennemgang af den kvantitativt orienterede rumsociologi skiller især to felter sig ud ved at bidrage til en metodisk kvantificering af mikro-rummet. For det første har et antal kriminologer udviklet metoder til at studere kriminalitet som et mikro-stedsfænomen (“micro-place”) (Sherman et al. 1989; Groff et al. 2010; Weisburd et al. 2012). Dette sker i erkendelsen af, at kriminalitet ofte har en stærk tendens til at være koncentreret i “hot spots”; over halvdelen af den kriminelle aktivitet i storbyer er således koncentreret i få procent af byers vejsegmenter som påvist af Weisburd

et al. (2012).12 Det er netop fordi kriminaliteten kvantificeres på dette niveau af rumlig specificitet, at det er muligt for disse kriminologer at fremsætte den banebrydende “law of concentrations of crime at place” (se også Weisburd og Amram 2014; Weisburd og Telep 2014). For det andet har et antal nyere bysociologiske studier af “naboskabseffekter” statistisk dokumenteret, hvordan nabolaget har betydelige sociale konsekvenser for en række sociale og individuelle forhold (f.eks. mønstre af kriminalitet, social mobilitet, sundhed og etnisk segregation). Der er tale om en mikro-stedslig effekt, der ikke blot skal tilskrives aggregerede individegenskaber, men som derimod må fortolkes og forklares som en kontekstuel effekt af nabolaget i sig selv (Sampson et al. 1999; Sampson 2011; 2012; Dietz 2002; Elliott et al. 1996; Grannis 1998; 2005; 2009). I lighed med de kriminologiske mikro-steds-analyser er studiet af naboskabseffekterne forudsat af, at det kvantificerende metodeblik zoomer ind på nabolagets mindste mikro-fysiske rumelementer. Sampson (2011) beskriver dette som en strategi, der “(…) begins with neighborhoods in physical space rather than elevating social interactions or identity to the definitional criteria.” (Sampson 2011: 229). Og i lighed hermed foreslår Grannis (1998; 2005; 2009), at det er de enkelte gadesegmenter og disses netværkskomposition, der udgør den centrale fysisk-rumlige byggeklods for nabolag. Den metodologiske fællesnævner for disse to sociologiske felter er altså et dedikeret fokus på sociale processer, sådan som disse udfolder sig på det fysiske rums mikro-skala. Her vil jeg dog indskyde, at et sådan mikro-metodologisk fokus på rummet ikke nødvendigvis er sammenfaldende med en særskilt interesse i det fysiske rum i sig selv. Sociale processer kan måles og modelleres på en rumlig mikro-skala uden, at der siges noget om mikro-rummets fysisk iboende egenskaber og effekter. Formuleret, med inspiration fra Jessop (2009), er der forskel på, om den mikro-rumlige skala blot betragtes som et metodologisk nyttigt sted at starte analysen fra, eller om det fysiske mikro-rum også i sig selv anses for et ontologisk nøgleelement for struktureringen af den sociale verden (se også Herod 2011: 37ff). I lyset af denne skelnen er det mit argument, at naboskabseffekt- og kriminalitet-studiernes fokus på den rumlige mikro-skala først og fremmest tages som et metodologisk afsæt for at foretage mere præcise analyser af sociale processer. Fokus er ikke særskilt på, hvad det fysiske mikro-rum gør ved det sociale liv. Det ontologiske omdrejnings-

12 “Vejsegment” eller blot “segment” defineres her, som det stykke vej, det segment, der ligger mellem to vejkryds. Som jeg skal vende tilbage til nedenfor, er det den samme rumlige mikro-skalaenhed, som afhandlingen, i forlængelse af space syntax-metoden, måler og modellerer det fysiske rum ud fra.

punkt er sociale processer, sådan som disse udspiller sig på den mikro- rumlige skala. Formuleret mere alment kan vi sige, at den metodologiske tilgang til den mikro-rumlige skala kan være anledning til at studere tre ontologiske effekter, hvorfra kun én har rummet i sig selv som det ontologiske hovedfokus. Dette argument vil jeg fremføre med reference til Hedström (2005: 46ff), der argumenterer for, at sociale gruppefænomener kan forklares principielt, som et produkt af tre hovedtyper af ontologiske effekter: Effekter af interaktion (social smitte), selektion (social komposition) og det rumlige miljø (social morfologi).13 Relaterer vi denne skelnen til naboskabseffekt- og kriminalitet-studierne viser det sig, at når det kommer ontologisk til stykket, fokuseres der på de to sociale effekter af interaktion og selektion, mens den non-sociale effekt af rummet i sig selv tildeles en sekundær betydning. Dette socio-ontologiske omdrejnings- punkt kommer eksempelvis til udtryk hos Grannis (2009) på følgende måde:

Neighborhood effects, however, are not produced by neighborhood geography. Nor are they – at least most of them – merely spatial effects, a byproduct or spurious confound of the geographic location of residents with particular demographic characteristics or psychological profiles. I argue that cataloguing neighborhood effects, by definition, hypothesizes that there exists a thing, a social entity, a neighborhood community, that has effects. Neighborhood effects are the product of these neighborhood communities. I argue that neighborhood communities and their effects emerge from neighboring interactions among their constituent residents (Gran- nis 2009: 4).

13 Denne skelnen mellem interaktions/selektions/rum-effekter er beslægtet med Fi- schers (1976: 25ff) indflydelsesrige klassifikation af de centrale teorier indenfor det urbansociologiske felt som hhv. ’deterministiske’, ’kompositionelle’ og ’subkulturelle’. Den deterministiske position dækker over byteorier (f.eks. Wirth 1938), der antager, at byen og bylivet i sig selv former en bestemt urban livsform. Formuleret med Heds- tröms (2005) vil en sådan effekt af byen kunne være af interaktionel og/eller rumlig karakter. Kompositions-position betoner i kontrast hertil den sociale sammensætning af byens grupper (f.eks. Gans 1962), hvilket i Hedströms terminologi svarer til en selektionseffekt. Endeligt dækker den subkulturelle position over byteorier, der fremhæver betydningen af social komposition (Fischer 1975) kombineret med indsigter fra den deterministiske position: Byens densitet (rumeffekt) og kontakten mellem subkulturer (interaktionseffekt) intensiverer i sig selv storbyens subkulturelle aktiviteter (selektionseffekt).

Grannis fokuserer metodologisk på rummets mikro-skala, men dette er altså ikke ensbetydende med, at det fysiske rum anses som et ontologisk nøgleelement for struktureringen af nabolagets sociale verden. For ham er den rumlige mikro-skala først og fremmest en metodologisk anledning til at kortlægge nabolaget som social enhed og fællesskab konsti- tueret ved de sociale effekter af interaktion og selektion. Effekten af mikro-rummet er, i hans socio-ontologiske perspektiv, ikke konstitutiv for naboskabseffekter. Om end Sampson (2011: 230) lægger mindre vægt på selektionseffekter end Grannis, er hans ontologiske vurdering af rummet identisk med Grannis (2009). Sampsons (2011: 229) ovenfor citerede fysisk-rumlige definition af nabolag viser sig således primært at være en metodologiske mikro-skalaanledning til at studere nabolagets socio-interaktionelle processer; ikke nabolaget, som rumfysisk fakta og effekt (Sampson 2012: 47). Den centrale socio-interaktionelle proces, der driver nabolagseffekterne, uddyber Sampson (2012: 152f), som en “collective efficacy”, forstået som den sociale sammenhængskraft mellem naboer, kombineret med deres villighed til at intervenere i naboernes liv med henvisning til et fælles værdisæt. Den rumlige mikro-skala er ikke et ontologisk mål i sig selv, men et metodologisk middel til at analysere de ansigt-til-ansigt situationer, der konstituerer nabolagene som virksomt fællesskab:

“Neighborhood-ness” (e.g. as reflected in spatially bounded social interactions, place identity, or social controls) is a contingent or variable event; neighborhoodness or community– defined by the social features associated with location– is not the same thing as a physical neighborhood. It is the intersection of practices and social meanings with spatial context that is at the root of neighborhood effects (Sampson 2011: 230).

Vender vi os dernæst til de kriminologiske studier af mikro-stedslig kriminalitet, tegner der sig et lignende billede af en metodologisk begrundet, ikke-ontologisk, tilgang til mikro-rummet. Dette lader sig identificere i en række studier (Eck 1997; Yang 2010; Weisburd et al. 2009), hvor undersøgelsen af de kriminelle hot spots ganske enkelt er begrundet i et rent metodologisk argument om, at en kvantificering på den mikro- stedslige skala producerer mere præcise statistiske resultater. De ontologiske effekter af interaktion, selektion eller rummorfologi, der kunne have været overvejet og anført som mulige forklaringer på disse metodisk kortlagte hot spots, er her sat i parentes. Groff et al. (2010) konkluderer således deres empirisk-metodiske studie med bemærkningen, at

“(…) something is going on at the micro level that requires explanation.” (Groff et al. 2010: 26). I Weisburd, Groff og Yangs (2012) nyere arbejder knytter de ontologisk an til Sampsons begreb om “collective efficacy” og udskifter således deres ikke-forklaring med en social-interaktionel forklaring. Mikro-stedet er nu metodologisk anledning til at studere, hvordan struktureringen af kriminalitets-koncentrationen sker som effekter af de interaktions-kontekster som de fysiske vejsegmenter er indlejrede i. Det er altså vejseg- menternes sociale indhold, der udgår en effekt fra; ikke fra deres fysisk- iboende egenskaber. Karakteristisk for denne ikke-ontologiske tilgang til det fysiske rums egenskaber og effekter er ydermere det forhold, at disse studier taler om “micro-places” snarere end micro-spaces. Dette signalerer, at rummet her forstås som en stedslig udkrystallisering af sociale processer snarere end en fysisk lokalitet, der former disse processer (jævnfør her Logan 2012 for en problematisering af sociologiens ontologiske fokus på “steder” frem for på egenskaber ved og effekter af rummet som fysisk lokalitet).

Space syntax At afhandlingen finder inspiration i space syntax metoden hænger sammen med, at denne tilgangs kvantificering af mikro-rummet netop sker med henblik på at forklare rummet som et ontologisk substratum for det sociale. I kontrast til de to ovenfor skitserede positioner er mikro-skala- rummet for space syntaxen ikke alene en metodologisk anledning til at foretage højopløselige studier af sociale processer, men er også knyttet til en ontologisk antagelse om, at mikro-rummets fysisk-iboende egenskaber er afgørende for strukturingen af det sociale. På linje med det ovenfor fremførte argument er det space syntax-grundlæggeren Hilliers (2008) argument, at enhver metodologisk undersøgelse af rum-samfundsrelationen er koblet til og konfronteret med et fundamentalt ontologisk valg:

It can either be approached from society to space, or from space to society, that is by working from social theory towards the spatial environment, or from the spatial environment toward social theory. To most social scientists it has always seemed self-evident that one must take the former route, since it is surely society that determines space and not space that determines society. The approach to the city that this generates is one of trying to see the spatial environment as the spatial output, and so as the by-product, of social, economic and perhaps cognitive processes. The ‘society-first’ assumption might reasonably be called the spatiality paradigm, since

it does not question the idea the link from society to space should be sought through an examination of the spatiality of social processes. (Hillier 2008: 221, original kursivering).

Sampson (2011) og Grannis’ (2009) naboskabsstudier og Weisburd et al.’s (2012) mikro-stedslige kriminalitetsstudier kan betragtes som en del af det samme “spatiality paradigme”; dvs. en rumsociologisk tilgang, der i lyset af en “society first”-antagelse, fokuserer på den rumlige mikro- skalaudfoldelse af selektions- og interaktionsprocesser. Det ontologiske fikspunkt for analysen er, hvad der gøres socialt i og ved rummet ikke, hvad rummet eventuelt måtte gøre ved det sociale. Konsekvensen af et sådan society-first-fokus er, som Hillier påpeger det, at man ikke forholder sig til rummet som fysisk form og fænomen i egen ret: “Is space completely amorphous, and so nothing, until given shape by social agency?” (Hillier 2008; 223, original kursivering). Paradigmatiske eksempler på society-first-rumteoretikere er, ifølge Hillier: Lefebvre (1991: 90), Soja (1989: 79), Giddens (1984: 368) og Harvey (1973: 203). Til denne række af navne tilføjer afhandlingen altså Sampson (2011), Grannis (2009) og Weisburd et al. (2012). Andre centrale teoretikere kunne også tilføjes denne liste, såsom en rumsociolog som Gans (2002) eller humangeogra- ferne Massey og Allen (1984), hvis indflydelsesrige – og arketypisk society-first-baserede – argument er, at det ikke er rummets form, men den rumlige form af sociale processer, der har sociale effekter: “It is not spatial form in itself (nor distance, nor movement) that has effect, but the spatial form of particular and specified social processes and relationships.” (Massey og Allen 1984: 5). Hilliers “space-first”-modtræk til “society-first”-tilgangen indebærer, at rummet beskrives og forklares som havende sin egen indre fysisk- ontologiske logik, som de sociale logikker virker igennem og præges af: “(…) for there to be a social logic of space, there must in the first instance be a some kind of logic of space, on which this can built.” (Hillier 2008: 224, original kursivering; se også Hillier og Vaughan 2007; Hillier og Hanson 1984). Karakteren af denne rumontologiske forskydning kan måske bedst forklares med en analogi til studiet af sprogets underliggende grammatik, hvilket også er forklaringen på det lingvistisk klingende navn: space syntax. Mennesket gør socialt brug af sproget, men det betyder omvendt ikke, at sproget er et komplet amorft produkt af social handlen. Sprog har også en underliggende eller indre grammatisk logik; et sæt af semantiske og syntaktiske regler, der strukturerer enhver social brug af sproget, og som det socialvidenskabelige studie af sproget også må studere i egen ret. At analysere det sociales relation til rummet uden

metoder specialiseret i at analysere rummet i egen ret, svarer til at studere en diskursiv konstruktion af den sociale virkelighed uden også at forholde sig metodisk til sprogets grammatiske logik. I space syntaxens perspektiv er rummet og det sociale i denne forstand bundet symmetrisk op på hverandre, og det metodologiske studie af denne ontologiske relation påbegyndes i rummets form. Space syntaxen er, som Hillier og Netto (2002) formulerer det, en teori om relationen mellem samfund og rum “set gennem rummets prisme”. Pointen er altså “(…) to look at the society space relationship ‘space first’, by examining the patterns of real space found in built environment and asking in what sense these could be seen to be outcome of social and economic processes.” (Hillier 2008: 224, original kursivering). Denne betragtning peger ydermere på, at space syntaxen space-first-tilgang ikke udgør et alternativ til en sociologisk analyse af rummet, men bedst kan forstås som et genuint rumsociologisk bidrag, hvor de sociologiske fakta studeres gennem deres rumligt-morfologiske former:

The analysis of real space is not then the slide-rule alternative to the social analysis of space, but the means to it. To understand the space of different fields of human activity, one must first investigate the ‘space of space’. By clarifying how space is manipulable, it can be seen how it is manipulated, and why it works for a particular social purpose. This is to say no more than that to understand what space is saying one must learn its language. Space has an active and structured engagement with social life, and without understanding this one cannot fully realize the theoretical promise of the social study of space (Hillier 2008: 228, original kursivering).

Som det fremgår, er det et væsentligt rum-ontologisk argument for space syntaxen, at det fysiske rum er “aktivt” og “struktureret” involveret i det sociale liv, der således ikke har monopol på handlen. Rummets “logik” og “love”, der eksisterer “uafhængigt af social handlen”, er ikke blot en fysisk-materiel træghed, sådan som eksempelvis Sartre (2004) og Østerberg (1975) argumenterer for, men spiller snarere direkte og pro- duktivt ind på det sociale liv. Rummet har, som Hillier (2005) formulerer det, en “real agency”. I lyset af nutidens rumsociologiske jargon vil en sådan ontologisk udmelding typisk blive læst som en latouriansk inspiration, hvilket imidlertid ingenlunde er tilfældet med space syntaxen, der er

udviklet uden berøring og inspiration fra ANT (Kärrholm 2010).14 På linje med afhandlingens rumkausale hypotese er Hilliers inspiration til argumentet om rummets evne til agency socialteoriens “foundings fathers” og det specifikt Durkheim og den sociale morfologi: “The theoretical consequence [af society-first tilgangen, LSL] is that the fundamental preoccupation of the founding fathers with the possible agency of spatial transformation in social morphology, and so an independent role in creating the society-space nexus, is more or less excluded from thought.” (Hillier 2008: 224, original kursivering).15 Så vidt i første omgang space syntaxens ontologiske forpligtigelse på rummet. Men ved hvilke rumdedikerede mikro-metoder kan en sådan space-first-tilgang realiseres? Det er, som anført, det rummetodiske spørgsmål, som rumsociologien, for sjældent, stiller sig selv og/eller som besvares uden at være forpligtet på rummet som ontologisk fakta. Space syntaxens metodologiske ambition angår mest grundlæggende den tekniske evne til at beskrive forskellen mellem ét rumligt design og ét andet, hvilket hverken rummorfologien, arkitekturteorien eller rumsociologien giver noget entydigt bud på: “(…) in order to compare dwellings with one another and to interpret their sociological significance, we have to solve a prior problem, that of identifying the elements and relations which make up the space pattern.” (Hanson 1998: 22; se også Hillier et al. 1976; Hillier og Hanson 1984). Som Hillier og Raford (2010) påpeger, og som det også er denne afhandlings argument, er det i høj grad manglen på metoder til at beskrive rummets mikro-egenskaber, der optegner nutidens gab mellem socialteori og rumligt design:

[T]here is a gap between social theory and spatial design, and that the inability of contemporary theory to join them is not a function of their lack of relationship, but instead a function of our lack of ability to measure them properly and thus construct evidence- based theories and testable hypothesis. The gap between space and

14 I forbindelse med mit ph.d.-ophold ved UCL Bartlett havde jeg flere samtaler med professor Hillier, der bl.a. blev anledning til at afklare dette spørgsmål. Mange har igennem årene gjort Hillier opmærksom på lighederne mellem space syntaxens argument om rummets (netværks-)agency og ANTs begreb om non-human agency. Men Hillier har aldrig selv fulgt denne kobling op og har udviklet sit arbejde uden berøring hermed. 15 En begrebslig kilde til forvirring skal nævnes her: Indenfor space syntax-traditionen anvendes begrebet “morfologi” både med reference til den durkheimianske skole og den urban-morfologiske skole indenfor planlægningsstudier, der forstår morfologi som studiet af formen, formationen og transformationen af menneskelig beboelse (Conzen 1960). Når Hillier og space syntaxen referer til social/spatial/urban morfologi, sker det almindeligvis med reference til denne planlægningsskole, men altså ikke i det citerede passus, hvor ’sociale morfologi’ refererer specifik til Durkheims sociologi.

society, in other words, is a measurement challenge first and foremost (Hillier og Raford 2010: 266).

Mens en forståelse af rummet som et ontologisk nøgleelement for struktureringen af sociale processer udgør en nødvendig betingelse for realiseringen af en space-first-orienteret rumsociologi, vil en sådan rumsociologi kun kunne blive tilstrækkeligt realiseret gennem anvendelse af metoder dedikeret til at måle rummet. En sådan rum-måling er imidlertid vanskelig. Den måske mest grundlæggende udfordring udgår fra det forhold, at det hverdagslige, såvel som videnskabelige, sprog kun i meget begrænset omfang kan beskrive forskellen mellem et rumligt layout og et andet. Præpositioner som ‘mellem’, ‘indenfor’ eller ‘igennem’ udtrykker en vis præcision til at beskriver relationen mellem to til tre rumlige ting. En term som ‘blandt’ øger antallet af rumlige relationer, der kan beskrives, men gør det med mindre præcision: “In general, language lacks terms to describe complex patterns of spatial relations, and in fact com- plexes of relations of any kind.” (Hillier og Raford 2010: 272). Dette argument taler, som jeg vil vende tilbage til, klart imod en pri- mær metodisk anvendelse af sproglige repræsentationer i rumsociologiske analyser. Et sådan argument er da ej heller fremmed for rumsociologien, der ofte argumenterer for anvendelsen af visuelle, frem for diskursive, metoder til at beskrive det sociales rumlige former (Halford 2008; Whyte 1988; Harpers 2012). Denne metodologiske erkendelse skinner også igennem i flere af sociologiens klassiske rumanalyser, der ofte betje- ner sig af visuelle illustrationer af de rumlige layouts til at understøtte deres ellers diskursivt baserede beskrivelser af rummet. Tre klassiske eksempler herpå er Lévi-Strauss’ (1963: 128) bororo-landsby, Foucaults (2002: 212) panoptikon-fængsel og Bourdieus (1990: 272) kabyler-hus:

Figur 4. Visuelle illustrationer af rumlige layout indenfor sociologien: Lévi-Strauss’ bororo-landsby (venstre); Foucaults panoptikon-fængsel (midt); og Bourdieus kabyler- hus (højre).

Anvendelsen af sådanne simple visuelle repræsentationer af rummet øger utvivlsomt den metodologiske evne til at beskrive komplekse rumlige relationer ud over sprogets begrænsede formåen. Men samtidig må det dog noteres, at denne formåen, også for den visuelle metodes vedkommende, hurtigt møder dens øvre kompleksitetsgrænse. Det er såle- des ikke uden grund, at panoptikon-fængslet, bororo-landsbyen og kabyler-huset er yderst simple rumlige layouts. Alle tre layouts optegnes reelt af én binær relation mellem: hhv. de visuelt isolerede fangeceller og den visuelle overvågning fra layoutets center (panoptikon-fængsel); en dual organisering af stammens socio-rumlighed (bororo-landsby); og en rumlig organisering, der afspejler en binær modsætning mellem det maskuline og feminine (kabyler-hus). I disse analyser fungerer den visuelle repræ- sentation således kun, fordi de rumlige layouts er så simple; dvs. i realiteten så simple, at analyserne problemfrit kunne være gennemført metodisk med sproget som eneste repræsentations-medium. Øges antallet af rumlige relationer, vil den visuelle repræsentation hurtigt komme metodologisk til kort. Dette er tydeligst i byen, hvis net- værks-layout bogstaveligt talt består af et tusindtal af relationer mellem gaderum. Her ville et visuelt gadekort over gadenetværket højst givet et kursorisk overblik, mens det ville være mere end vanskeligt at beskrive forskellen på, hvordan to gader er forskelligt indlejrede i gadenetværket. Tænkt her på Dronning Louises Bro, som vi kun vanskeligt ville have kunne identificere som en af de gader med den højeste netværksmæssige tilgængelighed i København, hvis vi kun havde haft et simpelt visuelt bykort over byen til metodisk rådighed.

Rumlig netværksanalyse – eller at gætte den goffmanske restaurant I bestræbelsen på at overkomme de metodologiske begrænsninger der er forbundede med at beskrive komplekse rumlige layouts ved eksempelvis sproglige og/eller visuelle repræsentationer, foreslår space syntaxen anvendelsen af en matematisk repræsentation af rummet (Marcus, Westin og Liebst 2013; Batty 2007; 2013; Hillier et al. 1976). I henhold til metodespørgsmålet en “(…) key insight in space syntax was to articulate the connectedness of space as a network of relations, which could be objectively measured using mathematical tools of graph theory, network analysis and topology.” (Hillier og Raford 2010: 273; original kursivering). Space syntaxen kan hermed forstås som en rumsociologisk pendant til social netværksanalyse, der, i de seneste år, har opnået en stigende popu- laritet inden for det sociologiske felt (Scott 2012). Hvad der taler for denne analogi og karakteristik af space syntaxen som en rumlige net- værksanalyse, er det forhold, at begge netværkstilgange tager matematisk

afsæt i den såkaldte grafteori til at beskrive sociale og rumlige netværk af relationer (Freeman 1977; Crucitti et al. 2006). Grafteorien er det matematiske studie af grafer, forstået som stukturer bestående af parvise relationer (“vertices”) mellem objekter (“noder”). Perspektivet er strengt relationelt eller topologisk i den forstand, at grafteorien, som udgangspunkt, ignorerer forhold som objekternes areal og relationernes metriske længde. Fokus er på de rene relationer. Figur 5 illustrerer to simple, rumlige layouts med tilhørende grafer. Per konvention arrangeres graferne sådan, at rummet udenfor er grafens rod. Når grafen er arrangeret sådan, kaldes den en “justified graph” (eller blot en “j- graf”):

Figur 5. J-grafter over to simple layouts. Egen tilvirkning.

På baggrund af disse to simple layouts kan jeg tydeliggøre grafteoretisk, hvad space syntaxen forstår ved rumlige netværk eller netværks- konfigurationer, der defineres som “(…) relations which take account of other relations on a complex.” (Hillier 1996: i). Formalt betragtet har vi at gøre med konfigurationer, når relationen mellem to rum er betinget af, hvordan disse rum relaterer sig til et tredje rum (eller ethvert andet rum). Forskellen på om a og b er symmetrisk relateret til roden (som i layout x) eller a og b er asymmetrisk relateret til roden (som i layout y), er netop en konfigurationel forskel de to rumlige layouts imellem. Det væsentlige er altså, at der er tale om en forskel, der ikke kan uddrages lokalt fra de(t) enkelte rum, men kun ud fra hvordan rummene er relaterede i et globalt netværksmønster (Hanson 1998: 22f; Hillier 1996: 23ff). Den matematiske grafteori gør det endvidere muligt at beskrive mere eller mindre komplekse netværks-konfigurationer med et antal kvantitative mål. Her vil jeg i første omgang fokusere på det meget simple mål “dybde”, der kan defineres som det mindste antal relationelle “skridt”, der adskiller to rumlige punkter i netværket (Klarqvist 1993). Dybde kan også forstås

som et mål for rumlig distance, hvor distance ikke er målt med euklidisk metrik, men som antallet af topologiske skridt mellem rummene. Hvordan muliggør en sådan grafteoretisk kvantificering af rummet en space-first-analyse af samfunds-rumrelationen; dvs. en rumsociologisk analyse, der opsporer eksistensen af en social logik indlejret i det bebyggede rums netværksmæssige former? Lad mig illustrere dette med et lille rumsociologisk tankeeksperiment. Antag at vi får stillet den simple rumsociologiske opgave at identificere, hvilket af de to rumlayouts i figur 5 – x eller y – der er socialt indrettede, som en restaurant. Fra Goffman (1956) ved vi, at restauranter sociologisk set er indrettet med et klart mikro-rumligt skel mellem en front- og bagscene. Der er en frontregion, hvor serveringen foretages og performes af tjenerne og en bagregion, hvor maden tilberedes og tjenere kan få sig et tiltrængt afbræk fra at være på: “In general, of course, the back region will be the place where the performer can reliably expect that no member of the audience will intru- de.” (Goffman 1956: 70).16 Spørgsmålet er nu, hvorvidt space syntax-metoden gør det muligt at fremhæve konfigurationelle egenskaber ved et af de to rumlige layouts, der relaterer direkte til den måde, hvormed restauranter typisk fungerer goffmansk. Dybdemålet efterlader ikke megen tvivl om, hvilket af de to rumlige layouts, der har bygget en social restaurant-logik ind i sin net- værksform. Layout y må være restauranten, idet dette konfigurationelt asymmetriske layout udkrystalliserer og understøtter det sociale behov for at have en bagscene (b) afskærmet fra en frontscene (a), der åbner ud mod gaden, så gæsterne let kan bevæge sig til og fra restauranten. Layout x’s symmetriske konfiguration modvirker modsat denne afskærmning, ved i stedet at lade gæsterne trænge direkte ind på bagscenen. Rumligt betragtet kan den sociale forskel mellem front- og bagscenerne således kvantificeres, identificeres og måske ligefrem defineres matematisk, som en konfigurationel forskel mellem, at bagscener er dybere lokaliserede i netværket end frontscener. Dette goffmanske eksemplet viser således, hvordan samfundet kan identificeres fra selve formen på det rumlige artefakt og, at dette kan udtrykkes på en simpel grafteoretisk måde.17

16 Det metodiske potentiale ved space syntax-teknikken kunne også være vist i forhold til de tre simple layouteksempler, som jeg diskuterede foroven. Og relevansen af space syntax-metoden er faktisk blevet diskuteret i forhold til både Foucaults panoptikon- fængsel (Peatross 1994), Lévi-strauss bororo-landsby (Hillier 1996) og Bourdieus kabyler-hus (Osman og Suliman 1994). 17 En påfaldende illustration af space syntaxens metodologiske relevans for nutidige socialmorfologiske studier finder man hos Dawson (2002), hvis studie af eskimo-igloer er et af litteraturens første metodisk-systematiske opfølgninger på Mauss’ (1979) studie heraf. Dawson benytter sig netop, og med god grund, af space syntax-tilgangen som en beskrivende metode til at afkode igloernes rumligt-iboende sociale logik.

Endvidere skal det bemærkes, at dette goffmanske eksempel også kaster lys over, hvilken sociologisk forståelse af det sociale, som space syntaxen trækker på, når der refereres til den “sociale logik”, der indlejres i den rumlige form. Som anført er Hillier og Hansons (1984) sociologiske inspiration i væsentlig grad durkheimiansk. Men mere specifikt synes de at knytte an til den funktionalistiske makrofløj indenfor den durkheimianske tradition (Collins 1994), og space syntaxens sociologiske position kan bedst beskrives som en art rumligt orienteret struktur-funktionalisme. Rummets indlejring af sociale logikker sker således, på karakteristisk vis, med henblik på, at den rumlige form opfylder et socialt behov: “(…) different types of society required different kinds of control on encounters in order to be that type of society; because if this were so, we could reasonably expect it to be the deepest level at which society generated spatial form.” (Hillier og Hanson 1984: 18). I dette perspektiv organise- res rum ikke socialt kontingent, men med henblik på opfyldelse af et funktionelt behov, der eksempelvis kan reproduceres gennem en bestemt socio-rumlig opdeling i front- og bagscener, sådan som det var tilfældet med den goffmanske restaurant. De sociologiske problemer ved en så- dan rumligt orienteret struktur-funktionalisme diskuteres systematisk nedenfor i det artikelbaserede kapitel 1.18

Distancer og tilgængeligheder Lad os vende tilbage til metodespørgsmålet. Mens j-grafens strukturelle elementer, noder og kanter, tilbyder en frugtbar måde at repræsentere og analysere bygningers indre layouts (som i det goffmanske restauranteksempel), er j-grafen ikke altid det metodologisk mest oplagte valg, når vi bevæger os op i rumlig skala. Dette bliver tydeligt, når det eksempelvis gælder festival- eller bymæssige uderum, hvor gadernes rumlige udstræk- ning skaber tvivl om, hvor langs vejforløbet at grafens node skal placeres og således om, hvordan gadenettets relationer overhovedet kan repræ- senteres grafteoretisk. For at imødegå dette repræsentationsproblem har Hillier og Hanson (1984: 93) foreslået en space syntax-variant af den traditionelle grafteori. Dette indebærer, at forholdet mellem noder og

18 At space syntaxens rumligt orienterede struktur-funktionalisme her går i spænd med Goffman er i sig selv en teoretisk påmindelse om, at Goffman er en art durkheimiansk funktionalist; specifikt en mikro-funktionalist, der forstår hverdagens mikro-rituelle ceremonier og herunder disses rumligt-sceniske strukturering som funktionelt nødven- dige for opretholdelsen af den sociale samhandlingsorden (Strong 2006; Collins og Makowsky 1972). Som Goffman selv formulerede det: “I guess I’m as much what you call a symbolic interactionist as anyone else. But I’m also a structural functionalist in the traditional sense (…).” (Goffman 1993: 318).

kanter omvendes således, at gadeforløbene i stedet repræsenteres med en kant (altså en linje) og relationen mellem gaderne (altså vejkrydsene) re- præsenteres med kanternes skæringspunkter. Netværkets grafteoretiske mål (dybde, etc.) udregnes da med udgangspunkt i kanterne/linjerne. I space syntax terminologien benævnes kanterne som “axiallinjer”, og den netværksmæssige repræsentation af gadenetværk som helheden kaldes et “axialkort”. Følgende figur 6 viser de axialkort, der svarer til de to j-grafer fra det goffmanske restaurant-eksempel:

Figur 6. Topologisk identiske j-grafter og axialkort

De mest tekniske aspekter ved at analysere axialkortet uddybes i de empiriske analyser. I denne sammenhæng vil jeg nøjes med at fokusere på intuitionen og ræsonnementerne bag teknikken. Igennem de seneste tre årtier er det blevet udviklet en række mere eller mindre sofistikerede grafteoretiske mål, der beskriver axiallinjers egenskaber og netværksmæs- sige tilgængelighed i konfigurationer. I space syntaxens tidligere år var det mest almindelige mål “integration”, som beskriver den gennemsnitlige dybe fra axiallinjerne til alle andre axiallinjer i systemet (Peponis og Wineman 2002). Integrations-målet beskriver, så at sige, hvor mange gange man skal dreje om hjørnet for at nå til alle andre rum i systemet. Integration udregnes typisk inden for en given topologisk radius, forstået på den måde, at man udregner, hvor mange axiallinjer der er tilgængelige fra hver af systemets linjer, hvis man kun må dreje eksempelvis 3 gange om hjørnet (dette noteres typisk som radius-3 integration).19 Space syntaxens rene topologiske tilgang er enkelt, men som flere har argumenteret for, er den måske også for enkel. Ratti (2004) og Batty (2004; 2013) har således anført, at den topologiske afskrivning af rummenes metriske egenskaber er uforenelig med, hvordan mennesker faktisk bevæger sig i og perciperer rummet. Det giver et urealistisk billede af rummenes relationelle sammenhæng, når det ignoreres, at det ikke har

19 I tillæg til axiallinje-repræsentationen af de rumlige layouts kan også anvendes en såkaldt synlighedsgraf-analyse (“visibility graph analysis”), hvor den netværksmæssige tilgængelighed udregnes på baggrund af rummenes visuelle relationer. Synlighedsgrafen og axialkortet vil almindeligvis ende ud i konvergerende rumlige repræsentationer og grafteoretiske mål, hvilket ikke er overraskende i lyset af, at axiallinjerne ligeledes defineres visuelt som rumlige ’lines of sight’ (Hillier 1996; Turner 2001). Synlighedsgraf- teknikken anvendes og introduceres mere detaljeret nedenfor i det artikelbaserede kapitel 5. Se endvidere appendiks 3.

samme “omkostning” (målt i tid, energi, etc.) at bevæge sig gennem to rum af forskellig metrisk længde. Denne situation kan vi illustrere i figur 7, hvor x er den rene toplogiske repræsentation af sammenhængen mellem tre rum. Layout y er topologisk identisk hermed, men axiallinjen b er her betydeligt metrisk længere, hvorfor det oplagt vil være mere omkost- ningsfyldt at bevæge sig gennem design y sammenhold med design x. Dette peger på, at det rene topologiske distancemål, hvor rumlig distance måles i skridt må kombineres med et metrisk afstandsmål.

Figur 7. Rum målt topologisk, metrisk og geometrisk

x y z c c c b b a b a a

Endvidere er det blevet anført (Turner 2001), at den rene topologiske repræsentation giver et upræcist indtryk af rummenes relationelle sam- menhæng, når det ignoreres, hvilken gradsvinkel som linjerne skærer hinanden med. Denne situation illustreres i z, hvor linjerne a og c skærer b med en betydelig mere skarp vinkel end i design x og y. At bevæge sig igennem layout z opleves således som en mere kompliceret og omkost- ningsfuld vej at tilbagelægge. Denne betragtning, der i øvrigt lader sig underbygge eksperimentelt i rumkognitive studier (Dalton 2003), foreslår således, at der i tillæg til de topologiske (skridt) og metriske (meter) di- stancemål, også må anvendes et geometrisk eller gradsmæssigt distance- mål.20 En sådan gradsdistance kan også forstås som en gradsmæssig væg- tet topologisk distance, hvor omkostningen ved at dreje ét skridt om hjørnet vægtes i forhold til, hvor mange grader, der skal drejes. Disse kritikker der angår hvilket matematisk mål for rumlig distance, der giver den mest socialt og kognitivt virkelighedstro repræsentation af de rumlige layouts, er igennem de sidste år blevet indarbejdede i space syntax-metoden. Dette har utvivlsomt øget den empiriske præcision af visse former for rumanalyse (Hillier og Iida 2005), men har også gjort metoden mere teknisk og utilgængelig for eksempelvis rumsociologiske ikke-specialister og har således bidraget til den tendens til “self- enclosure” af space syntax-paradigmet, som Netto (uden år) identificerer (se også Marcus, Westin og Liebst 2013).

20 På det eksperimentelle plan har enkelte forskere endvidere argumenteret for (Schro- der et al. 2007; Ratti 2004; 2005), at space syntaxens 2D-repræsentation af rummet kunne suppleres med en 3D-repræsentation. Det analytiske potentiale ved dette er imidlertid stadig forskningsmæssigt uafklaret.

På godt og ondt er space syntaxens tekniske benchmark i dag, at der foretages en såkaldt “segment angular analysis” af axialkortet (Turner 2001; Hillier og Iida 2005; Hillier et al. 2012). Dette involverer, at axiallinjerne dekomponeres i mellem-vejkrydssegmenter (dette er øvrigt den samme mikro-rumlige analyseenhed, der anvendes af Weisburd et al. 2012). Dernæst udregnes segmenternes netværksmæssige tilgængelig; dvs. den gradsmæssige distance fra alle segmenter til alle andre inden for en given metrisk radius. Mere specifikt skelnes der almindeligvis mellem to grafteoretiske mål: Segment angular “integration” og segment angular “choice”, der hhv. kan forstås som rumteoretiske pendanter til “close- ness”- og “betweenness”-målene, sådan som de kendes indenfor social netværksanalyse (Freeman 1977; Scott 2012). Den centrale og ikke- tekniske forskel mellem disse to mål kan formuleres som, at choice-målet typisk vil identificere de dele af netværket, som fungerer som de “broer”, der binder systemet sammen på tværs. I en space syntax-analyse af Kø- benhavn kommer dette bogstaveligt talt til udtryk ved, at Dronning Lou- ises Bro netop træder markant frem som et gadeforløb med en ekstraor- dinær høj choice-værdi. Typisk vil sådanne passager, som i Dronning Louises Bros tilfælde, være gaderum med en høj grad af gennemfarts- bevægelse (“through-movement”). I sammenligning hermed identificerer integrations-målet typisk netværkets mere lokale tilgængeligheder, der ofte hænger sammen med former for bevægelse, hvor man bevæger sig inden for et nabolag; eksempelvis når man bevæger sig hen til kiosken fra hvor man bor (“to-movement”) (Hillier og Iida 2005). Denne teoretiske og empiriske sammenhæng mellem gadenetværkets egenskaber og faktiske bevægelsesmønstre vender jeg tilbage til senere. Ovenfor har vi set, hvordan space syntaxen har inkorporeret kritikken af det rene topologiske distancemål, og således i dag kombinerer et topologisk (skridt), metrisk (meter) og geografisk (grader) distancemål, når rummenes netværks-tilgængelighed udregnes. Her vil jeg tage et sidste måleteknisk kritikpunkt i betragtning, der adresserer det forhold, at rummenes brug (“land use”) ikke inkluderes i space syntaxens analyser (Ratti 2004; Batty et al. 1998; Jiang 1999; Ståhle et al. 2005). Dette kan formuleres som, at space syntaxen overfokuserer på rummets netværksmæssige tilgængelighed, mens spørgsmålet om geografisk tilgængelighed eksklude- res; dvs. hvor høj grad af tilgængelighed – gennem netværket – der er til lokationer med et givent socialt eller økonomisk brugsindhold (Ståhle et al. 2005). Rumsociologisk vurderet er dette uheldigt, og der er på det seneste gjort forsøg på at integrere space syntax-teknikkerne med GIS- teknologier med henblik på at koble spørgsmålet om netværks-tilgænge- lighed med spørgsmålet om den geografisk-tilgængelige brug af rummet

(Jiang og Claramunt 2002). Specifikt benytter afhandlingen sig teknisk af Ståhle et al.’s (2005) “place syntax”-redskab, der tillader en række avance- rede udregninger af geografisk tilgængelighed gennem netværket; det lader sig eksempelvis være tilgængelighed til butikker gennem netværket fra beboelsesadresser.

Netværkshandlen I space syntaxens first-space-perspektiv er det ontologisk afgørende, at det rumlige netværk ikke reduceres til en funktion af social handlen, men også selv tilskrives et reelt handlingspotentiale. Rummet netværkshand- ler. Men hvordan, og med hvilke rumontologiske midler, kan rummet påvirke den sociale brug i og af rummet? Space syntaxen formulerer sig bemærkelsesværdigt klart om dette rumkausale spørgsmål. Menneskelige bevægelsesmønstre udpeges som den mellemliggende variabel igennem hvilken rummets netværkshandlen kan påvirke sociale interaktioner. Net- værkshandlen handler for space syntaxen altid om, hvordan rummet på- virker bevægelsesmønstre, og hvordan de rumligt distribuerede bevægel- sesstrømme dernæst er med til at forme sociale interaktioner. I det omfang, at rummet påvirker sociale udvekslinger, sker dette således hverken symbolsk eller som et direkte produkt af rummet men, så at sige, som et biprodukt af rummets effekter på bevægelse:

However much we may prefer to discuss architecture in terms of visual styles, its most far-reaching practical effects are not at the level of appearances at all, but at the level of space. By giving shape and form to our material world, architecture structures the system of space in which we live and move. In that it does so, it has a direct relation – rather than a merely symbolic one – to social life, since it provides the material preconditions for the patterns of movement, encounter and avoidance which are the material reali- zation – as well as sometimes the generator – of social relations (Hillier og Hanson 1984: ix).

Durkheims sociale morfologi er her en væsentlig inspiration; særligt nøglebegrebet om dynamisk densitet som Hillier og Hansons (1984) begreb om “bevægelse” (og synonymt hermed begreberne om “encounters”, “avoidance”, “co-presence” og “co-awareness”) er teoretisk mod- ellerede over: “Durkheim actually located the cause of the different solidarities in spatial variables, namely the size and density of populations. In the work of Durkheim, we found the missing component of a theory of space, in the form of the elements for a spatial analysis of social for-

mations.” (Hillier og Hanson 1984: 18). Argumentet om, at rummet ikke påvirker sociale interaktioner direkte, men indirekte gennem rummets distributionseffekter på bevægelsesdensiteten, kan således forstås som afledt af Durkheim, Mauss og Co’s sociale morfologi. Det er netop en sådan morfologisk mekanisme, der er virksom i Division, Elementary Forms og tydeligst i Eskimo med Mauss’ (1979: 79) kausale understregning af, at den rumlige sæsonvariation ikke er den direkte årsag til den rituelle intensitet; rummet påvirker snarere den rituelle intensitet indirekte gennem sæson-morfologiens effekt på den sociale densitet. Hillier og Mauss’ argument er altså, at den morfologiske mekanisme, hvormed rum påvirker socialitet, ikke sker direkte (rum → socialitet), men indirekte gennem dens effekt på bevægelsesdensiteten (rum → bevægelsesdensitet → socialitet). Hvordan en sådan rum-bevægelsesmekanisme folder sig ud, kan vi illustrere i relation til det goffmanske restauranteksempel. Front- og bags- cenernes konfigurationelle forskel i dybde er ikke blot virksom, fordi layoutet kommunikerer et symbolsk meningsindhold eller en symbolsk norm for korrekt adfærd. Rummorfologisk anskuet er pointen snarere, at rummets fysiske layout bidrager til reproduktionen af restaurantens duale samhandlingsorden ved rum-fysisk at segregere gæsternes bevægelses- strømme fra bagscenen. I space syntaxens rumsociologiske terminologi kan rummet fungere “konservativt” eller “generativt”, og den goffmanske restaurant illustrerer rummet som en funktions-konserverende kraft: (…) space can be used both in a conservative mode to structure and reproduce existing social relations and statuses, usually by using space to seg- regate, and in a generative mode to create the potential for new relations by using space to create co-presence through integration.” (Hillier og Vaughan 2007: 212f, original kursivering). Space syntaxens strenge betoning af rum-bevægelsesprocessen implicerer en eksplicit kritisk stillingtagen til arkitekturteoriens (og dele af rumsociologiens) ofte upræcise forklaringer af, hvordan rummet overhoved kan have sociale effekter. I denne litteratur antages blot eksistensen af en effekt, der forløber

(…) from physical forms directly to individuals. The building itself is seen as the machine, and the physical form of the building the determinant of behaviour. Such relations do not exist, or at least not in any interesting sense. Belief in their existence really does vi- olate common sense. How can a material object like a building impinge directly on human behaviour? Even so, it is exactly this that we are expected to believe if we abandon spatial configuration as

the intervening variable. The paradigm of the machine in effect asks us to believe that the relation of form to function in architecture passes not, credibly, from a pattern of space created by the building to a pattern of co-awareness and copresence, but, incredi- bly, directly from building to individual (Hillier 1996: 293).

Et sådan argument findes i mange varianter, sådan som Hillier (1996) videre diskuterer det. Nogen rumteoretikere antager, at en direkte relation mellem det byggede rum og social adfærd sker gennem anvendelse af rumligt passende aktiviteter, sådan som det eksempelvis er tilfældet med Newmans (1972) Defensible Space. Andre betoner betydningen af rumlig kognition og foreslår, at det byggede miljø fungerer gennem en række tegn og ansporinger (“clue and clues”) til bestemte former for social ad- færd, sådan som det eksempelvis er tilfældet i Wilson og Kellings (1982) indflydelsesrige teori om smadrede vinduer. Sådanne ansporinger kan også have en socio-symbolsk karakter, hvor rummet udgør en art scenisk baggrundstæppe (“theatrical back cloth”), som er passende for de sociale aktiviteter, der udspiller sig på denne rum-symbolske baggrund. I forhold til sidstnævnte synes Hillier at tænke på Goffmans scenemetafor, som han imidlertid mistolker som en society-first-sociolog, hvor rummets alene udgør en kommunikativ container for et symbolsk meningsindhold.21 Problemet med disse positioner er samlet set, at de overser konfi- gurationens netværkshandlen:

All have in common that they presuppose a relation between built form and behaviour unmediated by spatial configuration. That such relations do not really exist in any systematic sense seems amply confirmed both by the lack of research results which show such relations, and by the fact that the only relations we can find are those that pass through spatial configuration (Hillier 1996: 293).

Grundet space syntaxens ophav indenfor arkitekturteorien er denne kritik rettet skarpest imod en rumlig overfokusering på arkitektonisk “stil”, forstået som “dekorationer”, “udsmykninger” og former for “mo- difikation” af det byggede rums “udseende” (Hillier og Hanson 1984: ix,

21 At Hillier misforstår Goffman som værende en society-first-teoretiker er illustrativt for det argument, som fremføres senere i afhandlingen: nemlig, at space syntaxen pri- mært, og i problematisk grad, er inspireret af den durkheimianske traditions funktionalistiske makro-fløj, mens dialogen med mikro-fløjen, og herunder Goffmans mikro- sociologi, er stort set fraværende.

2). Sådanne æstetiske “overflade-egenskaber” ved det arkitektoniske objekt må dog anses for socialmorfologisk sekundære sammenholdt med “(…) the empty volumes of space resulting from that object into a pattern. It is this ordering of space that is the purpose of buildings, not the physical object itself.” (Hillier og Hanson 1984: 1). Kritikken retter sig her mod arkitekturteorien, men det skal tilføjes, at en sådan overflade- æstetisk tilgang til rummet også findes udbredt indenfor dele af rumsociologien. Dette kommer klarest til udtryk blandt fænomenologisk orienterede rumsociologier, der antager, at rummets overflade har en iboende æstetisk karakter eller fysiognomi, der “stråler” ud i rummet og således fylder mellem-rummet med en atmosfærisk fornemmelse. Som resultat af sådanne æstetiske overflade-ekstaser vil det blå rum således, eksempelvis, fyldes med en distinkt atmosfærisk fornemmelse af blåhed (Albertsen 1999; Degen 2008; Löw 2008; Böhme 2006; Heinskou 2010: 91). Er Hilliers rummorfologiske skepsis overfor sådanne overflade-æste- tiske egenskaber og effekter berettiget? Hans kritik balancerer unægtelig på kanten af at være kategorisk, men samtidigt må hans empiriske argument om, at et sådan overfladefokus ofte mangel empirisk evidens også tages alvorligt. Wilson og Kellings (1982) ovenfor omtalte teori om smadrede vinduer er et illustrativt og indflydelsesrigt eksempel herpå. Der er tale om en kriminologisk tilgang til rummets overflade-æstetik, hvor rummets visuelle tegn antages at invitere til kriminel adfærd. Det antages, at “fysisk uorden” forårsager “social uorden” i den forstand, at overflade-æstetiske tegn på uorden (smadrede vinduer, graffiti, etc.) signalerer, at naboerne er uvillige til at konfrontere fremmede, blande sig i kriminalitet og kontakte politiet. Denne hypotese har imidlertid været mere end vanskelig at underbygge empirisk (Holst 2013). Sampson og Raudenbuch (1999) har gennemført den hidtil mest opfattende test af teorien baseret på et omfattende surveystudie, der inkluderer geokodning af faktiske smadrede vinduer og andre æstetisk-visuelle tegn på fysisk uorden (se også Sampson 2012). Dette studie peger på, at sociale former for uorden ikke kan forklares som et kausalt produkt af smadrede vinduer eller lignende uordens-æstetikker, men først og fremmest må forklares som en social interaktionseffekt af nabolagets collektive efficiacy.22

22 Dette naboskabsstudie er det tætteste Sampson kommer på et rum-ontologisk fokus på rummet. Ovenfor viste jeg, hvordan hans interesse for mikro-rummet primært tages som et metodologisk udgangspunkt for at analysere de social-ontologiske interaktionsprocesser på mikro-skalaen. Man kan måske sige, at Sampsons ontologiske interesse i mikro-rummet aftager, fordi han fanges i socialteoriens (og herunder teorien om smadrede vinduers) empiriske-problematiske ontologi om overflade-æstetiske egenskaber og effekter. Sampson forholder sig omvendt ikke særskilt til rummet som en netværksef- fekt, sådan som afhandlingen og space syntaxen gør det.

Baggrunden for Hilliers stærke afvisning af, at der udgår relevante sociale effekter fra rummets overflade-egenskaber er endvidere, at der modsat eksisterer en betydelig empirisk evidens for, at rummets net- værksegenskaber har række sociale effekter. Space syntaxens empiriske rygrad udgøres af teorien om naturlig bevægelse, forstået som den andel af fodgængerbevægelse, der er produkt af gadenettet i sig selv; dvs. den andel af bevægelsesstrømmen, der skal tilskrives rummets netværkshand- len, snarere end tilstedeværelsen af sociale attraktioner eller “magneter” sådan som en traditionel (rum)sociologisk intuition ville foreslå vigtigheden af (Gans 2002; Bourdieu 1996). En lang række empirisk-statistiske studier har påvist eksistensen af en ekstraordinær stærk kausalkorrelation mellem rummets netværksmæssige tilgængelighed og faktiske bevægel- sesmønstre (Hillier et al. 1993; Hillier og Iida 2005; Penn et al. 1998; Peponis et al. 1997). “The discovery was that that the spatial configuration of a street network was in and of itself a major factor – probably the major factor – in shaping movement flows.” (Hillier og Raford 2010: 275). For Hillier (1996: 303) er det centralt at understrege, at denne rum- bevægelsessammenhæng ikke skal forstås i deterministiske termer; dvs. som en lovmæssig påvirkning af det enkelte individs bevægelser. Der er snarere tale om en aggegeret “system effekt”, der forløber fra den rumlige struktur til en probabilistisk distribution af bevægelsesstrømme.23 Når jeg i afhandlingen refererer til eksistensen af en rumlig netværks- handlen sker dette, som i space syntaxens tilfælde, på empirisk baggrund af denne empirisk robuste rum-bevægelsessammenhæng. Dette betyder ydermere, at styrkeeffekten af denne netværkshandlen først og fremmest må vurderes, som et empirisk spørgsmål. Således noterer afhandlingen sig, at de stærkeste og mest robuste sammenhænge kan identificeres i bymæssige kontekster. Til sammenligning er de estimerede sammenhæn- ge betydeligt svagere (om end ikke fraværende), når det gælder bevægelse indenfor i bygninger, hvor aktivitet og bevægelse snarere må “(…) be understood primarily in terms of specific purposefulness rather than spatial regularity.” (Peponis og Wineman 2002: 280). Spørgsmål om hvor, hvornår og hvem, der må bevæge sig rundt i huse, er tæt regulerede og sanktionerede af sociale normer og vaner (Hanson 1998). Interessant nok kan der identificeres en lignende empirisk tendens til, at en social

23 Af denne probabilistiske understregning følger det endvidere, at space syntaxen ikke i udgangspunktet opererer med en hypotese om, at rummet entrerer individernes mentale verden og ansporer individet til at bevæge sig i denne og hint retning. Indenfor de seneste år har space syntax-forskere dog udvist en stigende interesse i rumlig kogniti- ons-psykologi, herunder med henblik på at udvikle en individ-psykologisk forklaring på rummets effekter på bevægelsesdistributioner (Hillier 2009; Dalton et al. 2012).

formålsrettethed suspenderer rummets mulige netværkshandlen, når det gælder bevægelse på den inter-urbane og regionale skala:

(…) as dispersion increases, the movement system becomes more a pure origin-destination system. Instead of one journey accom- plishing a number of purposes, more journeys, each one accom- plishing fewer purposes, must be made to attain the same goals. These are the basic reasons why people travel farther in the country, and why most of this extra travel is in private cars (Hillier 1996: 134).

De empiriske rum-bevægelsesstudier peger således samlet set på, at når vi har at gøre med bevægelsesstrømme på en rumlige skala mindre og større end gaderummets skala, er der en empirisk tendens til, at net- værkshandlen er sekundær i forhold til formålsrettede sociale handlinger. Dette kan vi opsummere weberiansk: Bevægelsen indenfor i huse afspejler aggregerede mønstre af værdi-rationel handlen i overensstemmelse med sociale normer for adfærd, mens bevægelsen på den inter-urbane og regionale skala afspejler en instrumentel-rationel handlen fokuseret på at tilbagelægge afstanden A til B så effektivt som muligt (Weber 1978: 24f). Dette fører os til at de mulige sociale biprodukter af rum-bevægelses- korrelationen. Fællesnævneren for de former for sociale processer, der potentielt kan påvirkes af de rumdistribuerede bevægelsesflow er, at vi har gøre med “bevægelsesafhængige” former for socialitet (Hillier 1996). Dette ekskluderer, på den ene side, en lang række sociale fænomener fra rum-påvirkning (f.eks. social adfærd på internettet), mens space syntax- studier, på den anden side, har påvist, hvordan rummet er med til at forme bevægelsesafhængige fænomener indenfor en lang række sociale sfærer. Dette gælder eksempelvis sociale fænomener som fedme og sundhed (Baran et al. 2008), fodgængersikkerhed (Raford og Ragland 2004) og kriminalitet (Hillier 2004; Hillier og Sahbaz 2012).24

24 De kriminologisk orienterede space syntax-studier fortjener, i lyset af ovenstående diskussion af mikro-stedslig kriminalitet, et ekstra ord med på vejen. Mens f.eks. Weis- burd et al. (2012) anlægger et society-first-perspektiv, studerer space syntaxen kriminalitet ud fra en space-first-tilgang, hvor fokus er på den rumligt formede fortætning af bevægelsesmønstre i bestemte gader. Denne fortætning skaber en ‘safety in numbers’ og en ‘virtual community’ af ’co-present pedestrians’, hvilket angiveligt mindsker sandsynligheden for bl.a. indbrud og røverier (Hillier og Sahbaz 2012). Det skal dog tilføjes, at dette argument ikke er ukontroversielt indenfor det kriminologiske felt og forsk- ningsfeltet Crime Prevention Through Environmental Design (Holst 2013; Johnson og Bowers 2010).

I tillæg til sådanne studier af mere specifik karakter, foreslår space syntaxen også mere generelt, at rum-bevægelsessammenhængen har betydning for, hvad vi overhoved skal forstå ved byer: “Movement emerges as the “strong force” that holds the whole urban system together, with the fundamental pattern of movement generated by the urban grid itself.” (Hillier 2002: 154). Dette argument udfoldes i teorien om byer som bevægelsesøkonomier (Hillier et al. 1993; Hillier 1996: 111-136; 1999; 2002), hvilket udgør en central inspiration for afhandlingens hypotese om, at det fysiske mikro-rum kan agere som en ikke-iscenesat kraft, der faciliterer skabelsen af følelsesmæssigt attraktive intensiteter. Mens en bysocio- logisk society-first-tilgang typisk ser byen som er resultat af rummenes sociale brug (Gans 2002), har Hilliers space-first-perspektiv på byen øje for, hvordan rumlige brug, og i sidste ende byen som sådan, også er funktion af netværkshandlen. Det er alene ikke sociale processer, der fører til, at et byrum er særligt attraktivt. Brugs-attraktive rum genereres også af den netværks-fortættede bevægelsesdensitet. “Configuration generates attraktion”, sådan som Hillier (2002: 154) spidsformulerer sit be- vægelsesøkonomiske argument, der opsummeres på følgende måde:

[N]atural movement – and so ultimately the urban grid itself – im- pacted on land-use patterns by attracting movement-seeking uses such as retail to locations with high natural movement, and send- ing non-movement-seeking uses such as residence to low natural movement locations. The attracted uses then attracted more movement to the high movement locations, and this in turn attracted further uses, creating a spiral of multiplier effects and resulting in an urban pattern of dense mixed use areas set against a background of more homogeneous, mainly residential development (Hillier 2002: 153).

Et paradigmatisk og empirisk velunderbyggede eksempel på en sådan bevægelsesøkonomisk proces er butikker i storbyer. Som jeg berørt tidligere, eksisterer der en betydelig evidens for, at butikker klynger sig sammen i konfigurationelt integrerede vejsegmenter, hvis rumligt generede bevægelsesdensitet tilbyder butikkerne en solid kundebase (Hillier 1999; Chiaradia et al. 2009; Ortiz-Chao og Hillier 2007). Dette bevægelsesøko- nomiske princip er endvidere succesfuldt blevet anvendt på en række andre byøkonomiske fænomener som huslejepriser (Matthews og Turn- bull 2007) og kontor- og butikslejepriser (Enström og Netzell 2008). Væsentligt for afhandlingens mikro-sociologiske argument er dog, at Hillier (1996) også antyder, at bevægelsesøkonomiteorien også tilbyder

en forklaring på byens “urban buzz”; dvs. byens oplevelsesmæssige at- traktionsværdi. Dette oplevelsesmæssige aspekt af teorien er imidlertid ikke tidligere blevet teoretisk diskuteret og direkte testet empirisk. Af- handlingens undersøgelse af mikro-rummet som netværks-motor for den følelsesmæssige intensitet af mikro-interaktioner, adresserer netop dette underbelyste aspekt af bevægelsesøkonomiteorien.25

Ikke-diskursiv og ikke-lokal I egenskab af at være en space-first-metode, er det oplagt, at space syntaxen ikke kan stå metodologisk alene i et rumsociologiske studie, men må suppleres med traditionelle, rumsociologiske society-first-metoder: “(…) the space syntax methodology alone is incapable of eliciting the cultural norms of societies governing the use of building. It should be supplemented by social science methods; they are much more equipped for such a purpose.” (Osman og Suliman 1994: 201). Tilsvarende har Lawson (2001: 246) påpeget, at space syntax-metodens fokus på stedets “geometriske fysikalitet” giver os indsigt i det sociales rumlige organisering, men kun i begrænset omfang kan sige noget om menneskelig perception af rummet og social organisation. Hvilken metodologisk konklusion kan der uddrages fra disse, i mange henseende korrekte, indvendinger til space syntaxen? En metodisk position vil hævde, at disse problemer bør føre til en skinlæggelse af ethvert sådan forsøg på at kvantificere rummet, idet dette uafværgeligt indebærer en uholdbar reduktionisme. I dette lys er vi ganske enkelt bedre tjent med at studere rummet ved hjælp af kvalitative metoder. En aggressiv version af denne metodologiske position indtages, eksempelvis, af hu- mangeografen Edward Soja, der specifik afviser space syntax-metoden og enhver beslægtet form for kvantificering og forklaring af rummet, idet “(…) these physicalist methodologies are fixed too exclusively on the

25 Space syntax-traditionens forståelse af “movement” er godt eksempel på Cresswells (2006) kritiske karakteristik af “movement”, som “abstracted mobility”; dvs. abstraheret fra “mobility” som en fænomenologisk “way of being in the word” og fra det praktise- rede, oplevede og kropsliggjorte meningsindhold, der også finder sted, når mennesker bevæger sig fra A til B. Hvordan, synes det rimeligt at spørge, er afhandlingens følelses- sociologiske interesse forenelig med en sådan “abstracted mobility”? Her skal det understreges, at afhandlingen ikke angår de fænomenologiske mobility-oplevelser af at bevæge sig, sådan som et mobility-perspektiv ville fokusere på. Fokus er derimod på de følelsesmæssige oplevelser, som ikke-bevægelse individer har, når disse eksponeres for andre individers bevægelsesflow. Endvidere er det spørgsmål om temperament, om en sådan forståelse af movement nødvendigvis er antitesen til et mobility-perspektiv; Cresswells (2006: 3) argument om, at mobility også involverer et “potentially observable” og “brute fact” synes snarere at foreslå, at vi har at gøre med et kontinuum. Space syntaxens begreb om “movement” kan således lige så vel karakteriseres som det fysisk kvantificerbare aspekt af “mobility” (Shaw og Hesse 2010).

formal properties of materialized spatial configurations, giving too little attention to the complex social forces that exist behind their appearance.” (Soja 2001: 1.4). Sojas argument er tæt beslægtet med fænome- nologiens principielle kritik af tendenser til at reducere og hypostasere det levede liv og sted til meningsløs materie og geometriske rum (De Certeau 1998; Heidegger 1971; Schmitz 2011; Simonsen 1996; Holloway et al. 2010). Jeg er grundlæggende uenig i en sådan anti-kvantificerings-position og foreslår alternativt, at for så vidt angår det fysiske rums egenskaber, er en kvantificering af rummet ikke til at komme ude om. Afhandlingen anerkender, at en sådan kvantificering af rummet, som anført af Osman og Suliman (1994), ikke bør stå alene, men må indarbejdes som metode- komponent i et design, der også inkluderer andre sociologiske metoder; hvad enten disse er kvantitative eller kvalitative. Før vi ser nærmere på, hvorfor jeg mener, at kvantificering af rummet ikke er til at komme metodologisk uden om, lad mig først opridse de punkter, hvor den anti- kvantitative position og afhandlingens position faktisk er metodologisk enige. Denne enighed angår, at diskursivt baserede kvalitative metoder har en begrænset gyldighed i rumsociologiske studier. Afhandlingen følger her Hilliers argument, som jeg berørte tidligere. Sproget kan kun vanskeligt repræsentere komplekse rumlige relationer; dvs. rummets underliggende “grammatiske” struktur: “We deal with complex spatial patterns competently but intuitively, and again as with language, we don’t really understand how we do this.” (Hillier og Raford 2010: 272, original kursivering). Netværks-konfigurationer har i denne forstand karakter af en “hidden structure” eller en “unconscious configu- rationality” og skal med Hilliers (1996: 29) rammende ord, forstås som idéer vi tænker med (som sprogets grammatik), snarere end idéer, som vi tænker på (som sprogets mening). Med dette argument befinder vis os tæt på Polanyis’ (1966) klassiske definition af tavs viden (“tacit knowledge”), der fremdrager det forhold, at vi ofte ved mere end vi kan verbali- sere. Konfigurationer har en regularitet, der befinder sig hinsides det diskursive domæne. De er, slet og ret, “non-diskursive”: “Configuration is in general ‘non-discursive’, meaning that we do not know how to talk about it and do not in general talk about it even when we are most actively using it.” (Hillier 1996: 3). Dette taler klart imod anvendelse af diskursive metoder og foreslår i stedet en metodologisk kvantificerings- strategi: “non-discursive regularities” fordrer, som Hillier formulerer det, anvendelsen af en “non-discursive technique”, “(…) that is, a technique for handling those matters of pattern and configuration of form and space that we find it hard to talk about.” (Hillier 1996: 59).

Anti-kvantificerings-positionen fremfører et argument imod diskursivt baserede kvalitative metoder, der ikke er ulig Hilliers non-diskursivi- tets-argument: “These [spatial, LSL] elements are difficult to catch in words and indeed an increasing number of sociologists interested in spatiality are beginning to draw on the contribution that visual methods can make in discovering and representing spatial knowledge.” (Halford 2008: 939). Et sådan argument for visuelle metoder er på linje med Whyte (1988) og Harpers (2012) argument om, at visuelle metoder er velegnede til at studere den “sete men upåagtede” og den “tavse viden”, der karak- teriserer rum og rumlige praksisser. Gieryns (2000) refleksion over rumsociologiens metodologiske mangler indfanger rammende forståelsen af visuelle metoder som et alternativ til diskursivt (og kvantitativt) baserede studier af rummet:

What I lacked were tools to analyze place in its given two and three dimensions. I am a victim, perhaps, of trained incompetence in a discipline that cultivates statistics and words as means to grasp the social. Sociologists could become more adept with maps, floor, plans, photographic images, bricks and mortar, landscapes and cit- yscapes, so that interpreting a street or forest becomes as routine and as informative as computing a chi-square. That visualizing (I think) is the next step (Gieryn 2000: 483-484).

Det skal her tilføjes, at anvendelse af visuelle metoder ofte integreres med etnografisk deltagerobservation, der af mange rumsociologer også udpeges som en privilegeret måde at få adgang til de rumlige elementer ved menneskelig praksis (Degen og Rose 2012; Hall 2009; Holloway et al. 2010; Urry 2007: 39ff; Larsen og Meged 2012). Vurderet i et space syntax-perspektiv lider den visuelle og/eller etnografiske tilgang imidlertid under samme metodologiske begrænsning. Mens sådanne metoder nok er velegnede til at analysere, hvordan rummet opleves og bruges lokalt (dvs. i de lokale rum, hvor der observeres og/eller tages billeder), kan metoderne kun vanskeligt indfange, hvordan disse lokale rum er relationelt indlejrede i et globalt netværksmønster, der strukturer, hvilke former for brug og oplevelse, der finder sted i de lokale rum. Globale relationer lader sig ikke observere i eller visualisere fra

rummets lokale egenskaber.26 Konfigurationer er en emergent eller ikke- lokal egenskab ved det byggede rummet (Hillier 2005; Nedeau og Kafa- tos 1999). En rumsociologisk anvendelse af visuelle og/eller etnografiske metoder risikerer følgeligt at hypostasere det lokale rum, hvis iboende sociale logik og potentiale for bevægelse, brug og oplevelse, også må forklares som produkt af de rumlige konfigurationers globale netværksegen- skaber.27 Dette argument om, at de kvalitative metoders akilleshæl specifik er de relationelle egenskaber ved rummet, er ikke sociologisk ukendt. In- denfor social netværksanalyse argumenterer Granovetter (1973) tilsvarende for, at mikro-data utvivlsomt kaster lys over små-gruppeprocesser: “But how interaction in small groups aggregates to form large-scale patterns eludes us in most cases.” (Granovetter 1973: 1360). I forlængelse heraf har Hedström (2005: 75) påpeget, at “[s]ocial outcomes, like other emergent phenomena, are difficult to anticipate because the outcome depends to such a high degree on how the individual parts are interrelated. (…) social outcomes cannot simply be ‘read off’ from the properties of the individuals that generate them.” Netværk lader sig kun vanskeligt kortlægge i et lokalt og/eller kvalitativt metodeperspektiv. I stedet har man, som Hedström formulerer det videre, brug for “formal analytical tools” med et kvantitativt tilsnit, hvis man ønsker at kortlægge det sociales netværksstrukturer (se endvidere Schelling 1978; Sampson 2011: 231; Grannis 2009; Goldthorpe 2000). Space syntaxen kan netop forstås som et sådan formalt analytisk redskab, der, ved hjælp af grafteoretisk kvanti-

26 Som illustration af denne pointe, kan jeg også nævne Jacobs’ (1961) klassiske argument om, at korte boligblokke er en rumlig betingelse for dynamisk byliv. Jacobs analyse af sammenhængen mellem bevægelsesmønstre og blokke foregriber i mange henseender space syntaxens netværksteoretiske blik på byens rum (Seamon 2012). Men at Jacobs ikke benytter sig af rumkvantificering, sætter imidlertid en natulig øvre grænse for, hvor komplekse rumlige netværk, som det er mulige for Jacobs at analysere under anvendelse af hendes håndterende visualiseringer af gadenettet. Space syntax-tilgangen muliggør således en metodologisk-kvantitativ generalisering af Jacobs’ visuelle beskrivelser og forklaringer af byrummets egenskaber og effekter. 27 Det skal understreges, at dette problem ikke alene kan løses, ved at rummet visualise- res fra oven med anvendelse af flyfotoer; sådan som Harper (2012) overvejer det. En sådan visualisering betyder blot, at den her kritiserede hypostasering af det lokale rum forskubber sig til en hypostasering af det lokale blik på rummet: “The aerial perspective may be most interesting because it reminds us that what we see is a matter of how we look at it. Our knowledge, our values and our data are a product of seeing.” (Harper 2012: 70). Space syntax-metoden er i kontrast hertil ikke en visualisering af rummet, men en visualisering af grafteoretiske/matematiske værdimønstre.

ficering, muliggør en aflæsning af rummets netværksmæssige egenskaber og effekter.28

At måle sig med rumstatistikken Space syntaxens beskrivende mål af rummet lægger op til en statistisk modellering af sammenhængen mellem rumlige og sociale faktorer, og den store overvægt af empiriske space syntax-studier anvender da også statistiske metoder. Her står vi imidlertid overfor det metodologiske pa- radoks, at space syntaxens insisteren på, at rum-samfundsrelationen først og fremmest udgør en måleudfordring, ikke smitter af på valget af statistiske metoder. Stort set uden undtagelse anvender space syntax litteraturen ikke de dedikerede rumstatistiske teknikker, der bl.a. er blevet udviklet indenfor socialvidenskaben igennem sidste tre årtier (Anselin 1988; 2002; Logan 2012; Raudenbush og Sampson 1999; Bernasco og Elffers 2010). Udfordringen ved at foretage statistisk modellering af socio-rumlige data er, at den traditionelle OLS-regressionsanalyses centrale antagelse om observationernes uafhængighed er uforenelig med “geografiens før- ste lov”, sådan som denne er fremsat af Tobler (1970: 236): “Everything is related to everything else, but near things are more related than distant things.”29 Rumlige sociale data er oftest “rumligt klyngede” og det er velbeskrevet i den rumstatistiske litteratur, at en negligering af en sådan rumlig autokorrelation resulterer i for lave standartfejl og hermed øger sandsynligheden for at begå fejl af type-I (Ward og Gleditsch 2008). Den space syntax-baserede litteratur forholder sig ikke tilstrækkeligt til dette rumstatistiske problem. Når Hillier og Vaughan (2007) eksempelvis skriver, at “(…) just as a statistical data table takes social phenomena out of real space and into logical space, so the addition of space syntax values puts the world of real space and materiality back into the statistics” (Hil- lier og Vaughan 2007: note 7), forholder de sig ikke til, at en sådan lokali-

28 Den her fremførte kritik af kvalitative metoder angår vanskeligheden ved at foretage en fyldestgørende beskrivelse af rummets relationelle egenskaber uden kvantificering. Hvad der yderligere taler for en kvantificerende beskrivelse af mikro-rummet er, at disse beskrivelsers matematiske udtryk, dernæst kan omsættes i en statistisk forklaring af socio-rumlige fænomener. Den rene visualisering forbliver i kontrast hertil på det beskrivende (ikke-forklaringsmæssige) plan, som påpeget af Porter and Howell (2012: 82): “Spatial statistical methods allow spatial analysis of social data to move beyond visual inspection and exploratory analysis to fit probability-based models of the role space plays in social relationships.” (se endvidere Goodchild and Janelle 2004). 29 Dette klassiske argument indenfor geografien har en næsten lige så klassisk formulering indenfor sociologien: “Perhaps the most basic source of homophily is space: We are more likely to have contact with those who are closer to us in geographic location than those who are distant.” (McPherson et al. 2001: 429, min kursivering).

sering af sociale fakta i rummet som grundregel implicerer, at data har en rumligt klynget struktur, der ikke bør modelleres med traditionelle regressions-modeller. Multilevel-modeller og rumlige regressions-modeller (eventuelt anvendt i kombination) er mulige bud på statistiske teknikker, der tager højde for rumlig autokorrelation og som muliggør en korrekt estimering af rumlige konteksteffekter (Logan 2012; Morenoff 2003). Alvoren af space syntaxens manglende håndtering af de særlige problemer, der angår statistisk modellering af socio-rumlige data, er fornylig blevet opsummeret af Johnson og Bowers (2010):

[T]he methods of statistical analysis employed elsewhere [indenfor space syntax-forskningsmiljøet, LSL] have not taken account of the nesting and hence dependency in the data analysed. That is, homes are located on street segments, street segments are located in neighbourhoods and so on. Failing to take account of the nesting in the data – and consequently assuming that there is independence in the data when in fact there is not – violates the assumption of most statistical tests inviting errors of inference. More problematic still, a number of studies have not reported inferential statistics at all, relying instead on descriptive statistics alone (John- son og Bowers 2010: 94).

I lyset af denne påpegning er det eksempelvis sigende, at centrale space syntax-studier som Hillier et al. (1993) og Hillier & Iida (2005) benytter traditionelle OLS-regressionsmodeller til at estimere sammenhæn- gen mellem bevægelsesmønstre og byrummets netværkstilgængeligheder. Problemet er imidlertid, at der oplagt finder et “spatial spillover” (Anse- lin 2002; Anselin og Griffith 1988) af fodgængerstrømme sted mellem rumligt tilstødende tælleporte, hvor disse fodgængerdata er målt. Disse bevægelsesdata er således netop rumligt klyngede og ville følgelig modelleres mere korrekt med eksempelvis rumregressionsanalyse.30

30 Denne rumlige klyngedannelse af bevægelsesflow behøver dog ikke at blive betragtet som et metodologisk pålidelighedsproblem, men kunne også fortolkes som et space syntax-teoretisk resultat i sig selv: Antallet af fodgængere, der tælles ved rumligt tilstø- dende porte vil være rumligt klyngede som følge af, at en gadernes niveau af rumligt genererede bevægelses-attraktioner (f.eks. butikker) igangsætter et selvforstærkende bevægelses-attraktions ’spillover’ mellem tilstødende gader. Butikkerne i en given gade vil tiltrække bevægelsesflow, der smitter af på de tilstødende gader, hvilket også vil gøre det mere økonomisk attraktivt at have en butik her. Et sådan teoretisk argument, der er afledt af Hilliers (1996) teori om byer som bevægelsesøkonomier, fører dog til samme metodologiske konklusion: Tælledatene kan ikke antages at være rumligt uafhængige, men bør modelleres som rumligt klyngede.

I dette lys er det en væsentlig metodologisk ambition for afhandlingen at kombinere space syntax-metoden med dedikerede rumstatistiske teknikker. Til at modellere punkt-mønster-data (“point-pattern data”) (dvs. fordelingen af geokodede sociale begivenheder i rummet) anvender afhandlingen således rumregressionsanalyse, der indarbejder datas rumlige klyngestruktur ved at inkludere en såkaldt “spatially lagged” version af den afhængige variabel som en forklarende variabel. Den spatially-lagged værdi for området i udregnes som gennemsnittet (eller antallet af begivenheder) af y i de områder, der støder op til området i. En sådan spatially-lagged version af y er særlig aktuel i situationer, hvor det må formodes, at værdien af y i et givent i-område, er direkte påvirket af værdierne af y i de områder, der støder op til området i. Et godt eksempel på et sådant direkte rumligt spillover kunne være mønstre af kriminalitet, hvor det er oplagt, at kriminaliteten i en bydel smitter af på tilstødende bydele (Ward og Gleditsch 2008; Anselin og Griffith 1988; Burt et al. 2009). Til at modellere områdedata (“lattice data”) (dvs. sociale data, der er afgrænsede af et rumligt areal) anvender afhandlingen en mulitilevel- regressionsmodel, der tager højde for rumlige klynger i data ved at be- handle observationerne som hierarkisk indlejrede i rumlige kontekster. Intuitionen bag denne teknik kan lettest forklares med et uddannelsesso- ciologisk eksempel: Mens traditionel OLS-regressionsanalyse ville modellere skoleelevers karakterpræstation under antagelse om disses uafhæng- ighed, vil et multilevel-design tage afsæt i, at eleverne er (rumligt) indlejrede klasseværelser, skoler, etc. De deler eksempelvis en dygtig klasselæ- rer, hvilket har en positiv konteksteffekt på klassens gennemsnitlige præ- station. Multilevel-modellen tager således højde for rumlig autokorrelation og muliggør estimering af kontekstens selvstændige effekt på individuelle udfald; det lade sig være karakterpræstationer, oplevelsen af følel- sesmæssig opstemthed, etc. (Hox 2010; Logan 2012).31

Kontekst og kvantificering Det er usædvanligt at arbejde metodisk-kvantitativt indenfor den mikro- durkheimianske tradition. At dette er tilfældet, må ses i lyset af, at Goff-

31 Som antydet er det også muligt at kombinere multilevel- og rumregressionsmodeller, hvilket sker ved at inkludere en spatially-lagged version af y som kovariat i multilevel- modellen. Dette er specifikt aktuelt, hvis det formodes, at værdierne af y i en givent rumlig kontekst er direkte påvirket af værdierne af y i de tilgrænsende rumkontekster (Morenoff 2003; Chaix et al. 2005; Logan 2012). Dette er, som vi skal se, tilfældet i den statistiske analyse, der præsenteres i kapital 4, mens dette ikke synes tilfældet i kapital 3, der således hhv. benytter sig af en multilevel-model med og uden en spatially-lagged version af y som kovariat.

mans (1989) skoledannende mikro-durkheimianisme i hovedsagen er ud- foldet ved hjælp af etnografiske metoder og som et eksplicit alternativ til en ensidig makrosociologisk kvantificering. Når det specifik gælder morfologiske spørgsmål, holder Goffman dog, interessant nok, den metodisk-kvantitative dør på klem:

I myself believe that there is nothing wrong at all in counting bodies or houses, but that any study involving attitudes is likely to be worthless. So I myself have never been against quantification but only against the use of quantification linked to instruments that purported to get at subjective experiences. Well that stance was largely Blumerian, in that he was the one who provided early on in the late fifties and sixties the clearest statements of the shortcomings of quantitative sociology (Goffman 1993: 332).

Goffman anerkender altså relevansen af kvantificering af det morfologiske substratum (“bodies or houses”), men opponerer specifik og etnografisk imod, at (inter-)subjektive oplevelser gøres til genstand for statistisk undersøgelse. Sidstnævnte afstandstagen er, som det også frem- går, på linje med Blumers (1956) kritik af “variabel”-sociologien. Denne kritik går på, at måling og modellering af sociale interaktioner, som variabler, implicerer en bortabstraktion af interaktionernes kontekstuelle “her-og-nu”. “The variable relation is a single relation, necessarily strip- ped bare of the complex of things that sustain it in a ‘here and now’ context.” (Blumer 1956: 131). Variabelsociologien er, som konsekvens heraf, blind overfor subjektive og sociale fortolkningsprocesser. Den blumerianske kritik af kvantificering har en vis berettigelse, men kun når det gælder den snævrest definerede, positivistiske variant af variabelsociologien (Collins 1986; 1989). Problemet med den generaliserede version af Blumers kritik er, at det overses, hvordan ikke alle former for statistisk kvantificering er blinde overfor det situationelle her-og-nu. Fremfor at rette skytset mod kvantificering som sådan, burde kritikken i stedet rettes mod den måde, der kvantificeres; dvs. om vi har at gøre med en variabelsociologi med eller uden metodologisk opmærksomhed på det situationelle her-og-nu. Dette argument knytter an til Goffmans undtagelse af morfologien fra Blumers kritik: Det er muligt at bedrive en sociologisk lødig variabelkvantificering af situationens morfologiske egenskaber af numeriske kroppe og byggede former. Afhandlingens argument om relevansen af rumlig kvantificering kan forstås som en radi- kalisering af denne goffmanske erkendelse: Uden kvantificeringsværktø-

jer falder de relationelle egenskaber ved det lokale rums her-og-nu bort i den rumsociologiske analyse. Dette metodologiske argument kan jeg endvidere tydeliggøre med reference til Abbott (1988; 1997), der fremhæver, at variabelsociologiens problem specifik angår tendensen til at dekontekstualisere de sociale fakta fra deres placering i tid og rum. Men i stedet for at fremture med en generaliseret kritik af kvantificering, peger Abbott på, at problemet afgrænser sig til bestemte dekontekstualiserende statistiske værktøjer (f.eks. generaliserede linære modeller). Løsningen på dette problem er ikke en forkastelse af kvantificering som sådan, men udviklingen af en mere kontekstuel sociologi; uanset om denne realiseres vha. kvalitative og/eller kvantitative metoder. Det afmålte og konstruktive ved Abbotts argument viser sig endvidere ved, at han selv aktivt har bidraget til meto- deudviklingen af kontekst-sensitive kvantificeringsværktøjer (Abbott 1995 har bl.a. gjort et pionerarbejde ved at introducere den tids-kontekst- senstive sekvensanalyseteknik til sociologien). Efter min vurdering er space syntaxen, multilevel-modeller og rumregressionsanalyse alle gode eksempler på kvantificeringsteknikker, der klart inkorporerer det rumligt-kontekstuelle aspekt i analysen. Disse metoder indfrier netop Abbotts (1997: 1166) generelle forespørgsel efter teknikker, der tilbyder “ways of investigating complex spatial interde- pendence”, med henblik på, at “arrive at the description and measurement of interactional fields”. At den nyere udvikling af kvantitative metoder således har overhalet Blumers kategoriske kritik indenom, kommer endvidere tydeligt til udtryk i forbindelse med Raudenbush og Sampsons (1999) økometrisk (“ecometric”) tilgang til statistik, som afhandlingen i flere henseender er inspireret af. Denne tilgang udvikles i kontrast til de dekontekstualiserede økonometriske og psykometriske traditioner, idet der fokuseres på “(…) the social and physical properties of ecological settings, especially neighborhoods.” (Raudenbush og Sampson 1999: 2). Raudenbush og Sampson foreslår specifikt anvendelse af multilevel- og rumregressionsanalyse og kontekstuelt baserede samplingsstrategier. Vi har allerede set, hvordan afhandlingen benytter sig at sådanne rumstatistiske redskaber i kombination med space syntax-metoden; en teknik der må vurderes som en tiltrængt rummetodisk tilføjelse til den økometriske værktøjskasse. Ydermere har jeg fundet inspiration i Raudenbush og Sampsons (1999) overvejelser om økometriske samplingsstrategier. En sådan samplingstrategi involverer udførsel nabolagssurveys og teknikker til systematisk social observation af de rumlige kontekster (Reiss 1971). Bemærk at begge teknikker blev anvendt i forbindelse med deres økometrisk baserede test af teorien om smadrede vinduer (Samp-

son og Raudenbuch 1999). Anvendelsen af sådanne samplingsteknikker sker med henblik på at “(…) collect data that more directly reflects the sights, sounds, and feels of the street.” (Sampson et al. 2002: 471). Den metodologiske ambition er altså, at øge kvantificeringens kontekstuelle sensitivitet, hvilket ligefrem formuleres som, at “(…) it is perhaps not unreasonable to think about neighborhood effects in conceptual terms as socio-ecological interactionism; Erving Goffman meets social ecology, as it were.” (Sampson 2012: 283). At det økometriske perspektiv rettelig kan betragtes som goffmansk, lader sig understrege af, at neo-goffmanianeren Collins slår til lyd for en identisk sampling-strategi. Collins (1983) foreslår således anvendelse af en “mikro-sampling”-strategi, som er fokuseret på indsamling af mikro- interaktioner, sådan som de udfolder sig deres “naturlige kontekst”: “What I am advocating (…) is that we engage in systematic sampling of certain kinds of microsituations. Instead of sampling individuals as in conventional survey research, we sample encounters.” (Collins 1983: 195). Ifølge Collins er en sådan systematisk mikro-samplingstrategi ofte at foretrækker frem for etnografisk deltagerobservation, hvis “løse” tilgang til data, stiller sig i vejen for, at der kan foretages systematiske teori- tests. Mikrosampling-strategien er således en metodologisk nøgle for, at Collins (1975; 1986; 1987; 1989; 1999; 2004) kan indfri sit vedholdene epistemologiske argument om, at sociologiens videnskabelige berettigelse står og falder med evnen til formulering af “(…) testable, generalized explanatory principles.” (Collins 1989: 134). Raudenbush og Sampson (1999) og Sampson (2011; 2012) deler Collins’ opfattelse af sociologiens epistemologiske rolle. Og afhandlingens metodologiske design er ligeledes et forsøg på efterleve en sådan epistemologisk forpligtigelse på at teste generaliserbarheden af sociologiske forklaringsprincipper. Alt afhandlingens datamateriale er således mikro-samplet med henblik på foretagelse af et antal statistiske tests af afhandlingens centrale mikro- durkheimianske kausal-hypotese: At mikro-rummet er med til at forme følelsesintensiteten af stedslige mikro-interaktioner. For at øge muligheden for at teste generaliserbarheden af dette forklaringsprincip (Collins 1989; Fuchs 2001: 47ff), gennemføres afhandlingens statistiske tests med data indsamlet i forskellige typer socio-rumlige kontekster og på baggrund af forskellige måder at operationalisere de statistiske variabler:

Figur 8. Oversigt over afhandlingens data Rapporteret (survey) Observeret (tælledata, flyfoto) By Oplevelse (y) Rum/bevægelse (x) (lavintens) Selektion (x)

Festival Oplevelse (y) Brug (y) (højintens) Brug (x) Rum/bevægelse (x) Selektion (x)

Med hensyn til den empiriske variation over socio-rumlige kontekster er data indsamlet i en festival- og bykontekst; specifik Roskilde festivalen og kvarteret indre Nørrebro i København. Kriteriet for at udvælge disse kontekster er, at de har meget varierende grader af interaktions- og følel- ses-intensitet. Storbylivet er mere rutineret, formaliseret og følelsesmæs- sigt lav-intenst sammenholdt med festivallivet, der er mere midlertidigt, informaliseret og følelsesmæssigt høj-intenst. Ved at anvende data fra disse forskelligartede kontekster øges muligheden for at afgøre, hvorvidt det kun er i bestemte typer af sociale og følelsesmæssige situationer, at det fysiske mikro-rum spiller en rolle i formingen af mikro-interaktionsprocesser. Endvidere har afhandlingen både operationaliseret mikro-interaktionernes følelsesintensitet (dvs. artiklernes afhængige y-variabel) som et spørgsmål om, hvordan individerne oplever den rumlige mikro-kontekst og på baggrund af forskellige måder at mikro-rummet bruges. At følelses- intensiteten både operationaliseres som et spørgsmål om oplevelser og brug kan teoretisk begrundes i Mauss’ påpegning af, at den kollektive opbrusen er “(…) a state of continual excitement and hyperactivity.” (Mauss 1979: 76, min kursivering). Ved at variere operationaliseringerne af y over disse forhold øges muligheden for at afgøre, hvorvidt det kun er under bestemte oplevelsesmæssige og/eller brugsmæssige betingelser, at mikro- rummet bidrager til den mikro-interaktionelle intensitet. Der er forskel på, hvordan de forskellige datatyper er mikro-sample- de. De rumlige oplevelser er indsamlede med et surveyredskab i både festival- og bykonteksten (se appendiks 1 og 2). En styrke ved surveymeto- den er, at den ligeledes muliggør en måling af, hvem individerne er (selektion) og hvad de gør i rummet (brug). Afhandlingens surveys i byen og på festivalen er udført på samme måde; inspireret af Raudenbush og Sampsons (1999) økometriske naboskabssurveys. Disse er gennemført som ansigt-til-ansigt-interviews i de mikro-rumlige kontekster. Denne samplingstrategi er statistisk skræddersyet til anvendelse af multilevel- modellering, hvor informanternes behandles som indlejrede i den samme rumlige mikro-kontekst.

Målingen af, hvordan individerne tilegner sig mikro-rummet brugs- mæssigt, er sket ved direkte observation og ved hjælp af det netop be- skrevne surveyredskab. Den direkte observation er sket gennem fjern- observation (“remote sensing”) under anvendelse af de høj-opløselige flyfotoer, som Roskilde Festival ligger inde med, og som jeg venligst har fået adgang til. Konkret er der foretaget en geokodning af festlige forsamlinger i festivalrummet ud fra en mikro-durkheimianske antagelse om, at disse repræsenterer mikro-rituelle forsamlinger eller intensiverin- ger af festivalens sociale liv. Fordelen ved at mikro-sample disse forsamlinger gennem direkte observationer er, at der opnås en mere objektiv (i kontrast til subjektivt afrapporteret) viden om, hvor præcist i mikro- rummet at hele den distributionen af forsamlinger befinder sig på et givent tidspunkt. Denne mikro-sampling-strategi er i øvrigt beslægtet med den strategi som Weisburd et al. (2012) anvender i studiet af den mikro- stedslige distribution af kriminalitet. Endvidere er der, som det også fremgår fra dataoversigten, blevet mi- krosamplet forskellige typer af rum- og bevægelsesdata. Dette gælder dels forskellige typer space (og herunder place) syntax-analyser af festivalens og byens rumlige layouts, der er konstruerede på baggrund af flyfotoer (se appendiks 3 og 4); og dels indsamling af data over faktiske bevægelses- mønstre. Sidstnævnte observerede bevægelsesdata er indsamlet i overensstemmelse med Grajewskis (2001) metodiske anvisninger. Disse be- vægelsesdata korreleres med space syntax-analyserne af rummets net- værksegenskaber, og hvis korrelationen er stærk og højsignifikant vil space syntax-målene blive behandlet som en gyldig proxy for bevægelse i de statistiske analyser. Endeligt skal der knyttes en etisk overvejelse til afhandlingens mikro- samplingstrategi. Som Rindfuss og Stern (1998) påpeger, rejser det skær- pede krav til anonymiseringen af sociale data, hvis disse inkluderer en georeference, sådan som det er tilfældet med afhandlingens empiri. Pro- blemet opstår, hvis der georefereres til en fast (beboelses-, arbejds-, etc.) adresse, hvilket muliggør identifikation af informanter. Afhandlingens data muliggør imidlertid ikke en sådan identifikation, idet data ikke georefereres til et permanent og/eller administrativt fast punkt i rummet, men til et geografisk punkt, hvor individerne kun opholder sig midlertidigt og uformelt; (informanternes bevæger sig igennem byrummet; de har midlertidig festivallejr, forsamling, etc.). Afhandlingens georefererede sociale data vurderes således ikke at udgøre et etisk problem for informanternes anonymitet.

Skitse af artikelbaserede kapitler Afhandlingens fem kapitler er udarbejdede som videnskabelige artikler, der således, på en og samme tid, kan og skal læses som selvstændige bidrag og delundersøgelser. To af artiklerne er teoretiske og tre af artiklerne er empiriske-statistiske. Det skal bemærkes, at kapitlernes rækkefølge ikke er begrundet i, hvornår i udforskningsprocessen de er udarbejdede, men i argumenternes indholdsmæssige progression. I det første kapitel, Durkheim’s Social Morphology and Space Syntax: A Re- assessment, sættes Durkheims sociale morfologi i relation til space syntaxens rummorfologi, hvilket karakteriseres som et neo-durkheimiansk bidrag som rumsociologien og den sociologiske Durkheim-tradition har været påfaldende tavse omkring. Der argumentres for, at denne tavshed er teoretisk ufrugtbar for den sociologiske Durkheim-tradition og space syntax-teorien. Durkheim-traditionen mister space syntaxen som en tværvidenskabelig samtalepartner, der kunne afklare, hvorvidt og hvordan rummet skal behandles som en agent, og hvordan rummet kan måles som et socialt fakta. Endvidere er tavsheden skadelig for space syntax- traditionen, der synes fastlåst i en forældet rum-funktionalistisk durkheimianisme, der implicerer en problematisk teleologisk ontologi og hvis makro-strukturelle perspektiv stiller sig i vej for, at space syntaxen kan få det fulde analytiske udbytte af metodens mikro-kvantificering af rummet. Som modsvar til disse udfordringer foreslås det, at space syntaxens sam- mentænkes med den post-funktionalistiske mikro-durkheimianske tradition; specifikt Collins’ interaktionsritual-teori. Denne artikel har været præsenteret på Dansk Sociolog kongres 2012, Aarhus Universitet, og er inviteret til at blive resubmitted (major changes) med henblik på udgivelse i Journal for the Theory of Social Behaviour. Det andet kapitel, Ambiance of the Machine: Towards a Post-Phenomenology of Urban Spatial Atmospheres, undersøger teoretisk, hvilke egenskaber ved byens rum, der er konstituerende for byens rumlige atmosfærer. Argumen- tet taget kritisk afsæt i rumsociologiens fænomenologisk inspirerede argument om, at rummets bidrag til de urbane atmosfærer udgår fra rummets iboende karakter eller fysiognomi. Atmosfærer forstås i dette fæno- menologiske perspektiv, som æstetiske ekstaser, der “stråler” fra rummets overflader. Som et kritisk alternativ til denne fysiognomiske fortolkning, foreslås en morfologisk forklaring på rumligt-urbane atmosfærer, der, med inspiration hentet fra Hilliers space syntax-teori, forskyder blikket fra rummets overflader til det rumlige netværk, som overfladerne danner. Således foreslås det, at det atmosfærisk givne er givet af varierende densiteter af fodfængerbevægelser, der fortættes af byrummets net- værkshandlen. Atmosfærer opstår som en kropslig affekt af en rumlig ef-

fekt, når fodgængernes fortættede kroppe synkroniseres mere eller mindre rytmisk. Dette affekt-teoretiske argument trækker på en deleuziansk forståelse af intensive kropsaffekter, som “smittende” kropsrytmer, hvilket sættes i relation til Collins’ social-fysiologiske argument om, at kroppe er biologisk disponerede for til at falde følelsesmæssigt ind i og blive synkroniserede med hinandens krops-rytmer. Artiklen er hermed et for- søg på at fremlæse det “rent” kropsligt-fysiologiske aspekt af densitet- smitterelationen som allerede Durkheim og Mauss, men særligt Collins anfører som en nøgleforklaring på situationelle følelsesintensiteter.32 Denne artikel har været præsenteret ved 8th International Space Syntax Symposium, Santiago Chile, og er inviteret til at blive resubmitted med henblik på udgivelse i tidskiftet Space and Culture. I afhandlingens tredje kapitel, Phenomenology of the Movement Economy: A Multilevel Analysis, anvendes surveydatasættet, indsamlet på Nørrebro, til at undersøge de oplevelsesmæssige aspekter af Hilliers indflydelsesrige teori om byer som bevægelsesøkonomier. Mens der eksisterer en betydelig evidens for denne teoris relevans i forhold til forklaring af økonomi- ske attraktioner, f.eks. den rumlige distribution af butikker i storbyer, er det overset i litteraturen, at teorien ligeledes tilbyder sig som en mulig forklaring af byens oplevelsesmæssige eller fænomenologiske attraktions- værdi. Efter kontrol for selektionsvariabler peger artiklens statistiske analyse på, at byens bevægelsesøkonomi har en signifikant konteksteffekt på individernes fænomenologiske oplevelse af mikro-rummets attraktions- værdi. Set i afhandlingens bredere sammenhæng taler dette empiriske resultat for hovedpåstanden om, at det morfologiske mikro-rum er med til at forme den mikro-stedslige intensitetsoplevelse. Denne artikel er præsenteret ved Spatial Cultures Workshop, The Bartlett University Col- lege London, og skal indgå som kapitel i den bog, der udgives på baggrund af denne workshop.

32 Det er overset i litteraturen, at Collins’ (2004: 65-78: 2009: 26) IR-teori og den nyere deleuziansk inspirerede affektteori (Thrift 2008; Massumi 2002) tilbyder konvergerende forståelser af, at følelseslivet har et kropsligt-affektivt moment, der udfolder sig præ- kognitivt og prækulturelt. At disse følelsesteoretiske traditioner sjældent sættes i direkte relation kan tilskrives to forhold: Det underbetones, at Collins udtrykkeligt udpeger det kropsligt-affektive moment som én af de virksomme “variable” i den multivariate IR- teori om intersubjektive følelsesbånd. Og der overfokuseres (bl.a. med inspiration fra Latour 1996) på affektteorien som værende et postsubjektivt modbegreb til alskens former for intersubjektivitet (Massumi 2002; Wetherell 2012). Beslægtet med afhandlingens ovenfor skitserede ærinde om at læse ANT- og Durkheim-traditionerne som konvergerende rum-materielle bidrag, insisterer afhandlingen også i denne følelsessociolo- giske sammenhæng på, at affekt- og IR-teorierne bidrager til en akkumuleret viden om, at følelser ikke alene er socialt regulerede, men også udfolder sig og skal studeres som et affektivt-autonomt moment (se endvidere Sahni 2012; Turner 1996).

Det fjerde kapitel, Dissecting Collective Effervescence: A Quantitative Inquiry into Randall Collins’ Interaction Ritual Theory, anvender festivalsurveyen til at foretage en statistisk test af Collins’ kausalforklaring af det durkheimianske nøglebegreb: Kollektiv opbrusen. Mens dette fænomen er kvalitativt- etnografisk velskrevet, byder litteraturen ikke på tidligere forsøg på at teste Collins’ forklaring direkte og kvantitativt. Specifik undersøges det, hvorvidt Collins’ skelnen mellem effekten af interaktion som den primære forklaring af kollektiv opbrusen og effekterne af selektion og rumlig morfologi som vigtige, men sekundære forklaringer, holder til en empirisk test. Empirien bekræfter Collins’ forventning om, at interaktionsprocesser er af primær betydning og selektionsprocesser er af sekundær betydning, men peger også på, at rummet har en mere direkte effekt på den kollektiv opbrusen end Collins umiddelbart lægger op til. Dette resultat er sammenfaldende med resultatet i den forrige artikel og bekræfter yderligere afhandlingens hovedpåstand om mikro-rummet som en agent for situationel følelsesintensitet. Artiklen er præsenteret på Dansk Sociolog- kongres 2014, Københavns Universitet og er målrettet tidsskiftet Social Forces. I det femte kapitlet, A Spatial Statistical Analysis Hot Spots of Festival Co- Creativity: A Spatial Statistical Analysis, præsenteres resultaterne af en rumstatistisk analyse af festivalgæsters medkreative (“co-creative”) brug af festivalrummet; dvs. hvordan gæsternes tilegner sig festivalrummet gennem kreative iscenesættelser “fra neden”. Analysen baserer sig empirisk på festlige forsamlinger i festivalrummet, hvis distribution er geokodet ved hjælp af flyfotodata. Den socialteoretiske litteratur om festivalbegivenhe- der indeholder ikke tidligere forsøg på at forklare festivalgæsters faktiske brug af festivalrummet ved hjælp af kvantitative og rumstatistiske værk- tøjer. Med inspiration fra den kvantitative kriminologis hot spot-analyser, undersøger artiklen således, hvorvidt festivalens medkreative hot spots skal tilskrives effekter af selektion, interaktion og/eller festivalens rummorfologi. I relation til selektions- og interaktionseffekter er resultaterne i samklang med de to forrige artikler: Den rumlige koncentration af med- kreativitet kan dels tilskrives en selektionsproces, hvor feststemte gæster opsøger særligt festlige dele af festivalrummet, og kan dels forklares med henvisning til spillover-effekter af interaktion, hvor gæsterne imiterer hinandens medkreative festivaladfærd. Studiets rummorfologiske resultat adskiller sig imidlertid, ved første øjekast, fra afhandlingens to andre empiriske artikler: Frem for at bevæ- gelsesdensiteten bidrager positivt til den stedslige intensitet (som i de to øvrige studier), er sammenhængen i denne artikel negativ. Dette negative resultat peger på, at bevægelsesdensiteten også kan være for hektisk eller

stressende til at understøtte dannelsen af et attraktivt mikro-rum til at tage ophold i. I afhandlingens konkluderende bemærkninger diskuteres det, hvordan dette hhv. positive og negative resultat ikke nødvendigvis modsiger hinanden, men snarere kaster lys over, hvordan måden, hvorpå individer opholder sig kropsligt i mikro-rummet, regulerer om bevægel- sesdensiteten er negativt eller positivt korreleret med attraktive følelses- intensiteter. Artiklen er præsenteret på Nordisk Sociologkongres 2012, Reykjavik Island og Dansk Sociolog kongres 2013, Roskilde Universitet, og er i peer-review i tidsskiftet Environment and Planning B: Planning and Design. Bibliografi Abbott, A. (1988). Transcending general linear reality. Sociological Theory, 6(2), 169-186. Abbott, A. (1995). Sequence analysis: new methods for old ideas. Annual Review of Sociology, 21(1), 93-113. Abbott, A. (1997). Of time and space: The contemporary relevance of the Chi- cago School. Social Forces, 75(4), 1149-1182. Alexander, J. C. (2005). The inner development of Durkheim’s sociological theory: from early writings to maturity. i: Alexander, J. C., & Smith, P. (red.). The Cambridge Companion to Durkheim. Cambridge: Cambridge University Press. Albertsen, N. (1999). Urbane atmosfærer. Sociologi i dag, 4: 5-29. Amin, A., & Thrift, N. (2002). Cities: Reimagining the Urban. Cambridge: Polity Press. Andrews, H. (1993). Durkheim and Social Morphology. i: Turner, S. (Red.). Emile Durkheim: Sociologist and Moralist. London & New York: Routledge Anselin, L. (1988). Spatial Econometrics: Methods and Models. New York: Springer. Anselin, L. (2002). Under the hood issues in the specification and interpretation of spatial regression models. Agricultural economics, 27(3), 247-267. Anselin, L., & Griffith, D. A. (1988). Do spatial effects really matter in regression analysis? Papers in Regional Science, 65(1), 11-34. Appelyard, D. (1981). Livable Streets, Berkeley: University of California Press. Ball, P. (2004). Critical Mass. How one Thing Leads to Another. London: Arrow Books. Baldassare, M. (1979). Residential Crowding in Urban America. Berkeley: University of California Press. Baran, P. K., Rodríguez, D. A., & Khattak, A. J. (2008). Space syntax and walking in a new urbanist and suburban neighbourhoods. Journal of Urban Design, 13(1), 5-28. Bernasco, W., & Elffers, H. (2010). Statistical analysis of spatial crime data. i: A. R. Piquero & D. Weisburd (red.). Handbook of quantitative criminology. New York: Springer. Batty, M. (2004). A New Theory of Space Syntax. Centre for Advanced Spatial Analysis. Working paper series, nr. 75. Batty, M. (2007). Cities and Complexity. Cambridge: MIT Press. Batty, M. (2013). The New Science of Cities. Cambridge: MIT Press.

Batty, M, B. Jiang & M. Thurstain-Goodwin (1998). Local movement: agent- based models of pedestrian flows. Centre for Advanced Spatial Analysis. Working paper series, nr. 4. Batty, M., Desyllas, J., & Duxbury, E. (2003). Safety in numbers? Modelling crowds and designing control for the Notting Hill Carnival. Urban Studies, 40(8), 1573-1590. Bech, Henning (2005). Teksten eller livet: fire begreber om kultur. i: Povlsen, K. K. (red.). Kunstkritik og kulturkamp. Århus: Klim. Bjerre, H. J., & Fabian, L. (2010). Rummet finder sted: redaktionelt forord. Slagmark, 57, 9-18. Blegvad, M. (1996). Max Webers sociologi i teorihistorisk lys. Dansk Sociologi, 4(7), 19-30. Blok, A., & Jensen, T. E. (2011). Bruno Latour: Hybrid Thoughts in a Hybrid World. London & New York: Routledge. Blumer, H. (1956). Sociological analysis and the ‘variable’. American sociological review, 21(6), 683-690. Blumer, H. (1986). Symbolic Interactionism: Perspective and Method. University of California Press. Boden, D. and Molotch, H. (1994). The Compulsion to Proximity. i: Friedland, R., & Boden, D. (Red.). Nowhere: Space, Time and Modernity. University of Cali- fornia Press. Bourdieu, P. (1990). The Logic of Practice. Palo Alto: Stanford University Press. Bourdieu, P. (1996). Physical space, social space and habitus. Vilhelm Aubert Memorial Lecture, Institutt for Sosiologi og samfunnsgeografi, Rapport 10. Boyko, C. T., & Cooper, R. (2011). Clarifying and re-conceptualising density. Progress in Planning, 76(1), 1-61. Burt, J. E., Barber, G. M., & Rigby, D. L. (2009). Elementary Statistics for Geogra- phers. Guilford Press. Böhme, G. (1993). Atmosphere as the fundamental concept of a new aesthetics. Thesis Eleven, 36(1), 113-126. Castells, Manuel (1996). The Rise of the Network Society: The Information Age: Econ- omy, Society, and Culture. Volume I. John Wiley & Sons. Chaix, B., Merlo, J., Subramanian, S. V., Lynch, J., & Chauvin, P. (2005). Com- parison of a spatial perspective with the multilevel analytical ap-proach in neighborhood studies: the case of mental and behavioral dis-orders due to psychoactive substance use in Malmö, Sweden, 2001. American Journal of Epi- demiology, 162(2), 171-182. Chiaradia, A., Hillier, B., Schwander, C., & Wedderburn, M. (2009). Spatial centrality, economic vitality/viability: compositional and spatial effects in Greater London. Proceedings of the 7th International Space Syntax Sympo- sium, Stockholm: KTH Collins, R., & Makowsky, M. (1972). The Discovery of Society. New York: Random House. Collins, R. (1975). Conflict sociology: Toward an Explanatory Science. New York: Academic Press. Collins, R. (1983). Micromethods as a Basis for Macrosociology. Journal of Con- temporary Ethnography, 12(2), 184-202. Collins, R. (1986). Is 1980s sociology in the doldrums?. American Journal of Sociol- ogy, 91(6), 1336-1355.

Collins, R. (1987). Looking Forward or Looking Back? Reply to Denzin. Ameri- can Sociological Review, 93(1), 180-184. Collins, Randall (1988). Theoretical Sociology. San Diego: Harcourt Brace Jo- vanovich Collins, R. (1989). Sociology: Proscience or antiscience? American Sociological Review, 54(1), 124-139. Collins, R. (1993). Emotional energy as the common denominator of rational action. Rationality and Society, 5(2), 203-230. Collins, R. (2004). Interaction Ritual Chains. Princeton: Princeton university press. Collins, R. (1994). Four Sociological Traditions. New York: Oxford University Press. Collins, R. (1999). Socially unrecognized cumulation. The American Sociologist, 30(2), 41-61. Conzen, M. R. G. (1960). Alnwick, Northumberland: a study in town-plan analysis. Transactions and Papers, 27, iii-122. Cresswell, T. (2006). On the Move: Mobility in the Modern Western World. CRC Press. Crucitti, P., Latora, V., & Porta, S. (2006). Centrality in networks of urban streets. Chaos: an Interdisciplinary Journal of Nonlinear Science, 16(1). Dalton, R. C. (2003). The Secret Is To Follow Your Nose Route Path Selection and Angularity. Environment and Behavior, 35(1), 107-131. Dalton, R. C., Hölscher, C., & Turner, A. (2012). Understanding space: the nascent synthesis of cognition and the syntax of spatial morphologies. Envi- ronment and Planning B: Planning and Design, 39(1). Darwin, C. (2010). The Origin of Species. Cricket House Books. Dawson, P. C. (2002). Space syntax analysis of Central Inuit snow houses. Jour- nal of Anthropological Archaeology, 21(4), 464-480. De Certeau, M. (1998). The Practice of Everyday Life: Living and Cooking. University of Minnesota Press. Degen, M. (2008). Sensing Cities. London & New York: Routledge. Degen, M., & Rose, G. (2012). The sensory experiencing of urban design: the role of walking and perceptual memory. Urban Studies, 49(15), 3271-3287. DeLanda, M. (1997). A Thousand Years of Nonlinear History. New York: Zone Books. Deleuze, G. (1994). Difference and Repetition. New York: Columbia University Press. Dietz, R. D. (2002). The estimation of neighborhood effects in the social sciences: An interdisciplinary approach. Social Science Research, 31(4), 539-575. Douglas, M. (1972). Symbolic Orders in the Use of Domistic Space. i: Ucko, P. J., Tringham, R., & Dimbleby, G. W. (red.). Man, Settlement and Urbanism. London: Duckworth. Duncan, O. D. (1959). Human Ecology and Population Studies. i: Hauser, P. M., & Duncan, O. D. (red.). The Study of Population. Chicago: University of Chicago Press. Durkheim, É. (1960). Sociology and Its Scientific Field. i: Wolff, K. H. (red.). Emile Durkheim, 1858-1917: A Collection of Essays. Columbus: The Ohio State University Press. Durkheim, É. (1978a). Sociology and the Social Sciences. i: Traugott, M (red.). Emile Durkheim on Institutional Analysis. Chicago: University of Chicago Press.

Durkheim, É (1978b). Antonio Labriola, Essays on the Materialist Conception of History (review). i: Traugott, R. (red.). Emile Durkheim on Institutional Anal- ysis. Chicago: University of Chicago Press. Durkheim, E. (1982). Rules of Sociological Method. New York: The Free Press. Durkheim, É (1984). The Division of Labor in Society (translated W. D. Hall). New York: The Free Press. Durkheim É. (1994). Note on Social Morphology. i: Traugott, M. (red.). On Institutional Analysis. Chicago: University of Chicago Press. Durkheim, É. (1995). The Elementary Forms of Religious Life. New York: Free Press. Durkheim, É. (2005). Suicide: A study in sociology. London & New York: Routledge. Durkheim, É., & Mauss, M (2009). Primitive Classification. Taylor & Francis e- Library Eck, J. E. (1997). What do those dots mean? Mapping theories with data. Crime mapping and crime prevention, 8, 379-406. Elliott, D. S., Wilson, W. J., Huizinga, D., Sampson, R. J., Elliott, A., & Rankin, B. (1996). The effects of neighborhood disadvantage on adolescent development. Journal of Research in Crime and Delinquency, 33(4), 389-426. Enström, R., & Netzell, O. (2008). Can space syntax help us in understanding the intraurban office rent pattern? Accessibility and rents in downtown Stockholm. The Journal of Real Estate Finance and Economics, 36(3), 289-305. Emirbayer, M. (1996). Useful Durkheim. Sociological Theory, 14, 109-130. Evans, D. (1997). Michel Maffesoli’s sociology of modernity and postmoderni- ty: an introduction and critical assessment. The Sociological Review, 45(2), 220- 243. Farias, I., & Bender, T. (2012). Urban assemblages: How Actor-Network Theory Changes Urban Studies. London & New York: Routledge. Fischer, C. S. (1975). Toward a subcultural theory of urbanism, American journal of Sociology, 80(6): 1319-1341. Fischer, C. S. (1976). The Urban Experience. New York: Harcourt Brace Jo- vanovich. Fogle, N. (2011). The Spatial Logic of Social Struggle: A Bourdieuian Topology. Lexing- ton Books. Foucault, M. (2002). Overvågning og staf. Frederiksberg: Det lille Forlag. Fournier, M. (2006). Marcel Mauss: A Biography. Princeton: Princeton University Press. Fournier, M. (2012). Emile Durkheim: A Biography. Cambridge: Polity Press. Franzén, M. (2009). Matters of urban segregation. Proceedings of the 7th In- ternational Space Syntax Symposium. Stockholm, KTH. Freedman, J. L. (1975). Crowding and Behavior. WH Freedman. Freeman, L. C. (1977). A set of measures of centrality based on betweenness. Sociometry. 40(1), 35-41. Fuchs, S. (2005). Against Essentialism: A theory of Culture and Society. Cambridge: Harvard University Press. Gans, H. J. (1962). The Urban Villagers: Group and Class in the Life of Italians- Americans. New York: The Free Press. Gans, H. J. (2002). The sociology of space: a use-centered view. City & Commu- nity, 1(4), 329-339. Garfnkel, H. (1967). Studies in Ethnomethodology. Englewood Cliffs: Prentice-Hall.

Gehl, J. (2010). For You Jane. i: Goldsmith, S. A., & Elizabeth, L. (red.). What We See. Advancing the Observations of Jane Jacobs. Oakland: New Village Press. Gehl, J. (2003). Livet mellem husene. København: Købehavn: Arkitekternes forlag. Geyer, J., Raford, N., Ragland, D., & Pham, T. (2005). The Continuing Debate about Safety in Numbers. Data From Oakland. Safe Transportation Re- search and Education Center, UC Berkeley. Giddens, A. (1984). The Constitution of Society. Outline of the Theory of Structuration. Cambridge: Polity Press. Gieryn, T. F. (2000). A space for place in sociology, Annual Review of Sociology, 26, 463-496. Gieryn, T. F. (2002). What buildings do. Theory and Society, 31(1), 35-74. Giuffre, K. (2013). Communities and Networks: Using Social Network Analysis to Rethink Urban and Community Studies. Cambridge: Polity Press. Goffman, E. (1956). The Presentation of Self in Everyday Life. Edinburgh: Universi- ty of Edinburgh Press. Goffman, E. (1963). Behavior in Public Places: Notes on the Social Organization of Gatherings. New York: The Free Press. Goffman, E. (1967). Interaction Ritual: Essays on Face-to-Face Interaction. Oxford: Aldine. Goffman, E. (1969). Where the Action Is: Three Essays. London: Allen Lane Goffman, E. (1971). Relations in Public: Microstudies of the Public Order. New York: Basic Books. Goffman, E. (1974). Frame Analysis: An Essay on the Organization of Experience. Cambridge: Harvard University Press. Goffman, Erving (1989). On Fieldwork. Journal of Contemporary Ethnography, 18(2), 123-132. Goffman, E. (1993). An interview with Erving Goffman, 1980 (ved Verhoeven, J. C.). Research on Language and Social Interaction, 26(3), 317-348. Goldthorpe, J. H. (2000). On sociology: Numbers, Narratives, and the Integration of Research and Theory. Oxford: Oxford University Press. Goodchild, M. F., & Janelle, D. G. (2004). Thinking spatially in the social sciences. i: Goodchild, M. F., & Janelle, D. G. (red.). Spatially Integrated Social Sciences. Oxford: Oxford University Press. Graham, S., & Marvin, S. (2001). Splintering Urbanism. Networked infrastructures, technological mobilities and the urban condition. London: Taylor & Francis. Grajewski, T. (2001). Space Syntax Observation Manual (upubliceret manus, revideret ved Vaughan, L.). London: UCL Bartlett and Space Syntax Lim- ited. Grannis, R. (1998). The Importance of Trivial Streets: Residential Streets and Residential Segregation. American Journal of Sociology, 103(6), 1530-1564. Grannis, R. (2005). T-Communities: Pedestrian Street Networks and Residen- tial Segregation in Chicago, Los Angeles, and New York. City & Community, 4(3), 295-321. Grannis, R. (2009). From the Ground Up: Translating Geography into Community Through Neighbor Networks. Princeton: Princeton University Press. Granovetter, M. S. (1973). The strength of weak ties. American Journal of Sociology, 78(6), 1360-1380. Groff, E. R., Weisburd, D., & Yang, S. M. (2010). Is it important to examine crime trends at a local “micro” level?: A longitudinal analysis of street to

street variability in crime trajectories. Journal of Quantitative Criminology, 26(1), 7-32. Haggett, P., Chorley, R. J. (1969). Network Analysis in Geography. London: Ed- ward Arnold Halbwachs, M. (1930). Les causes du suicide. Paris: F. Alcan. Halbwachs, M. (1960). Population and Society: Introduction to Social Morphology. New York: Free Press. Halford, S. (2008). Sociologies of space, work and organisation: From frag- ments to spatial theory. Sociology Compass, 2(3), 925-943. Hall, T. (2009). Footwork: Moving and knowing in local space(s). Qualitative Research, 9(5), 571-585. Hanson, J. (1998). Decoding Homes and Houses. Cambridge: Cambridge University Press. Harman, G. (2009). Dwelling with the fourfold. Space and Culture, 12(3), 292- 302. Harper, D. (2012). Visual sociology. London & New York: Routledge. Harvey, D. (1973). Social Justice and the City, Baltimore: Baltimore: Johns Hop- kins University Press. Hastrup, K. (2010). Introduktion. Den antropologiske videnskab. i: Hastrup, K (red.). Ind i Verden. En grundbog i antropologisk metode. København: Hans Reit- zels forlag. Hedström, P. (2005). Dissecting the Social: On the Principles of Analytical Sociology. Cambridge: Cambridge University Press. Heidegger, M. (1971). Building Dwelling Thinking. i: Hofstadter, A. (red.). Poet- ry, Language, Thought, Perennial Classics. Heinskou, M. (2010). En kompleks affære. Anmeldte voldtægter i Danmark. Ph.d.-afhandling. Sociologisk Institut. Københavns Universitet. Herod, A. (2011). Scale. London & New York: Routledge. Hillier, B. (1996). Space is the Machine. Cambridge: Cambridge University Press. Hillier, B. (1999). Centrality as a process: accounting for attraction inequalities in deformed grids. Urban Design International, 4(3-4), 107-127. Hillier, B. (2002). A theory of the city as object: or, how spatial laws mediate the social construction of urban space. Urban Design International, 7(3), 153- 179. Hillier, B. (2004). Can streets be made safe?. Urban Design International, 9(1), 31- 45. Hillier, B. (2005). Between social physics and phenomenology: explorations towards an urban synthesis?. Proceedings of the 5th International Space Syntax Symposium, TU Delft. Hillier, B. (2008). Space and spatiality: what the built environment needs from social theory. Building Research & Information, 36(3), 216-230. Hillier, B. (2009). Studying cities to learn about minds: some possible implications of space syntax for spatial cognition. Environment and Planning B: Plan- ning and Design, 39(1) 12-32. Hillier, B., Leaman, A., Stansall, P., & Bedford, M. (1976). Space syntax. Envi- ronment and Planning B: Planning and Design, 3(2), 147-185. Hillier, B., & Hanson, J. (1984). The Social Logic of Space. Cambridge: Cambridge University Press.

Hillier, B., Penn, A., Hanson, J., Grajewski, T., & Xu, J. (1993). Natural movement-or, configuration and attraction in urban pedestrian movement. Envi- ronment and Planning B: Planning and Design, 20(1), 29-66. Hillier, B., & Netto, V. (2002). Society seen through the prism of space: outline of a theory of society and space. Urban Design International, 7(3), 181-203. Hillier, B., & Iida, S. (2005). Network and psychological effects in urban movement. Spatial Information Theory, 3693, 475-490. Hillier, B., & Vaughan, L. (2007). The city as one thing. Progress in Planning, 67(3), 205-230. Hillier, B., & Raford, N. (2010). Description and discovery in socio-spatial analysis: the case of space syntax. i: G. Walford, E. Tucker, & M. Viswanathan (red.). The SAGE Handbook of Measurement, London: Sage. Hillier, B., & Sahbaz, O. (2012). Safety in Numbers: High-Resolution Analysis of Crime in Street Networks. i: The Urban Fabric of Crime and Fear. New York: Springer. Hillier, W. R. G., Yang, T., & Turner, A. (2012). Normalising least angle choice in Depthmap-and how it opens up new perspectives on the global and local analysis of city space. Journal of Space Syntax, 3(2), 155-193. Holloway, I., Brown, L., & Shipway, R. (2010). Meaning not measurement: Using ethnography to bring a deeper understanding to the participant experience of festivals and events. International Journal of Event and Festival Manage- ment, 1(1), 74-85. Holst, R. S. (2013). Kriminalpræventiv og utryghedsforebyggende miljø- og byplanlægning. SBI forlag, nr. 27. Hox, J. J. (2002). Multilevel Analysis: Techniques and Applications. London: Taylor & Francis . Jacobs, J. (1961). The Death and Life of Great American Cities. New York: Random House Digital. Jensen, O. B. (2006). ‘Facework’, flow and the city: Simmel, Goffman, and mobility in the contemporary city. Mobilities, 1(2), 143-165. Jensen, O. B. (2010). Erving Goffman and everyday life mobility. i: Jacobsen, M. H. (2009). The Contemporary Goffman. London & New York: Routledge. Jensen, O. B. (2013). Staging Mobilities. London & New York: Routledge. Jerolmack, C., & Tavory, I. (2014). Molds and Totems Nonhumans and the Constitution of the Social Self. Sociological Theory, 32(1), 64-77. Jessop, B. (2009). Avoiding traps, rescaling states, governing Europe. i: R. Keil and R. Mahon (red.). Leviathan Undone? Towards a Political Economy of Scale, Vancouver: University of British Columbia Press. Johnson, S. D., & Bowers, K. J. (2010). Permeability and burglary risk: are cul- de-sacs safer?. Journal of Quantitative Criminology, 26(1), 89-111. Jiang, B., Claramunt, C., & Batty, M. (1999). Geometric accessibility and geographic information: extending desktop GIS to space syntax. Computers, En- vironment and Urban Systems, 23(2), 127-146. Jiang, B., & Claramunt, C. (2002). Integration of space syntax into GIS: new perspectives for urban morphology, Transactions in GIS, 6(3), 295-309. Klarqvist, B. (1993). A space syntax glossary, Nordisk Arkitekturforskning, 2, 11- 12. Krarup, T. M., & Blok, A. (2011). Unfolding the social: quasi actants, virtual theory, and the new empiricism of Bruno Latour. The Sociological Review, 59(1), 42-63. ‐

Kärrholm, M. (2010). Space Syntax and Meta Theory. Journal of Space Syntax, 1(1), 251-253. Larsen, J., & Meged, W. (2012). At fotografere og filme byen - visuelle etnogra- fier. i: Andersen, J. Freudendal-Pedersen, M., Koefoed, L., & Larsen, J. (red.). Byen i bevægelse. Roskilde Universitetsforlag. Latour, B. (1987). Science in Action: How to Follow Scientists and Engineers through Society. Cambridge: Harvard University Press. Latour, B. (udgivet under pseudonymet: Johnson, J.) (1988). Mixing humans and nonhumans together: The sociology of a door-closer. Social problems, 35(3), 298-310. Latour, B. (1990). Drawing Things Together. Lynch, M., & Woolgar, S. (red.). Representation in Scientific Practice. Cambridge: MIT Press. Latour, B. (1996). On interobjectivity. Mind, Culture, and Activity, 3(4), 228-245. Latour, B. (2004). Gabriel Tarde og det sociales endeligt. Distinktion: Scandinavi- an Journal of social theory, 5(2), 33-47. Latour, B. (2005). Reassembling the Social. An Introduction to Actor-Network-Theory. Oxford: Oxford University Press. Latour, B. (2012). We Have Never Been Modern. Cambridge: Harvard University Press. Latour, B., & Yaneva, A. (2008). Give me a gun and I will make all buildings move: An ANT’s view of architecture. i: Geiser. R (red.). Explorations in architecture: Teaching, design, research, Basel: Birkhäuser. Law, J., & Mol, A. (2001). Situating technoscience: an inquiry into spatialities, Environment and Planning D: Society and Space, 19: 609-621. Lawson, B. (2005). The Language of Space. Oxford: Architectural Press. Lefebvre, H. (1991). The Social Production of Space. Oxford: Blackwell. Lemieux, C. (2012). What Durkheimian thought shares with pragmatism: How the two can work together for the greater relevance of sociological practice. Journal of Classical Sociology, 12(3-4), 384-397. Lévi-Strauss, C. (1963). Structural Anthropology. New York: Basic Books. Lieberson, S. (1991). Small N’s and big conclusions: an examination of the reasoning in comparative studies based on a small number of cases. Social Forces, 70(2), 307-320. Liebst, L. S. (2009). Etikken i den maskerede by. Om Baumans etiske afvisning af den postmoderne bys æstetisering. Dansk Sociologi, 20(1), 7-23. Lindemann, G. (2011). On Latour’s social theory and theory of society, and his contribution to saving the world. Human Studies, 34(1), 93-110. Logan, J. R., Zhang, W., & Xu, H. (2010). Applying spatial thinking in social science research. GeoJournal, 75(1), 15-27. Logan, J. R. (2012). Making a place for space: Spatial thinking in social science. Annual Review of Sociology, 38, 507-524. Lukes, S. (1985). Emile Durkheim: His Life and Work. A Historical and Critical Study. Palo Alto: Stanford University Press. Löw, M. (2007). Raumsoziologie. Frankfurt am Main: Suhrkamp. Löw, M. (2008). The Constitution of Space: The Structuration of Spaces Through the Simultaneity of Effects and Perception. European Journal of So- cial Theory, 11(1): 25-49. Löw, M. (2009). The Intrinsic Logic of Cities. Proceedings of the 7th Interna- tional Space Syntax Symposium, Stockholm: KTH

Maffesoli, M. (1995). The Time of the Tribes: The Decline of Individualism in Mass Society. London: Sage. Massey, D., & Allen, J. (1984). Introduction: Geography Matters. Massey, D., & Allen, J. (red.). Geography Matters! A Reader. Cambridge: Cambridge Universi- ty Press. Massumi, B. (2002). Parables for the Virtual: Movement, Affect, Sensation. Durham: Duke University Press. Matthews, J. W., & Turnbull, G. K. (2007). Neighborhood street layout and property value: the interaction of accessibility and land use mix. The journal of real estate finance and economics, 35(2), 111-141. Marcus, L., Westin, S., & Liebst, L. S. (2013). Network buzz: conception and geometry of networks in geography, architecture and sociology. Proceedings of the 9th International Space Syntax Symposium, Seoul: Sejong University Press. Mauss, M. (1979). Seasonal Variations of the Eskimo: a Study in Social Morphology. London: Routledge & Kegan Poul. Mauss, M. (2003). Concrete Divisions of Sociology. Jeffrey, W., Gane, M. (red.).The Nature of Sociology. New York & Oxford: Durkheim Press/Berghahn Books. Mauss, M. (2009). Manual of Ethnography. New York & Oxford: Durkheim Press/Berghahn Books. McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Ho- mophily in social networks. Annual Review of Sociology, 27, 415-444. Merton, R. K. (1936). The unanticipated consequences of purposive social action. American Sociological Review, 1(6), 894-904. Milgram, S. (1974). The experience of living in cities. i: Loo, C. M (red.) Crowd- ing and Behavior, Ardent Media. Morenoff, J. D. (2003). Neighborhood mechanisms and the spatial dynamics of birth weight. American Journal of Sociology, 108(5), 976–1017. Morgan, M. (2007). Festival Spaces and the Visitor Experience. Casado-Diaz, M., Everett, S., & Wilson, J. (red.). Social and Cultural Change: Making Space(s) for Leisure and Tourism. Eastbourne: Lesiure Studies Association. Morrill, R. (1970). The Spatial Organisation of Society. Belmont: Wadsworth Nadeau, R., & Kafatos, M. (1999). The Non-Local Universe: The New Physics and Matters of the Mind. Oxford: Oxford University Press. Narvaez, L., Penn, A., & Griffiths, S. (2012). Configurational Economies: The Value of Accessibility in Urban Development. International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies, 3(3), 293-309. Nemedi, D. (1995). Collective consciousness, morphology, and collective representations: Durkheim’s sociology of knowledge, 1894-1900. Sociological Per- spectives, 38(1), 41-56. Netto, V. (uden år). What space syntax is not. Upubliceret arbejdspapir. down- loaded fra academia.edu. Newman, O. (1972). Defensible Space. New York: Macmillan. Nielsen, D. A. (1999). Three Faces of God: Society, Religion, and the Categories of Total- ity in the Philosophy of Emile Durkheim. New York: SUNY Press. Oppenheim, R. (2008). Kyngju Things: Assembling Place. Michigan: University of Michigan Press.

Ortiz-Chao, C., & Hillier, B. (2007) In search of patterns of land-use in Mexico City using logistic regression at the plot level. Proceedings of the 6th Inter- national Space Syntax Symposium, Istanbul. Osborne, J. F. (1996). Beyond constructivism. Science Education, 80(1), 53-82. Osman, K. M., & Suliman, M. (1994). The Space Syntax Methodology: Fits and Misfits. Architecture & Comportement/Architecture & Behaviour, 10(2), 189-204. Park, R. E. (1915). The city: Suggestions for the investigation of human behavior in the city environment. The American Journal of Sociology, 20(5), 577-612. Peatross, F. D. (1994). The Spatial dimensions of control in restricted settings. Ph.d.-afhandling, Georgia Institute of Technology. Peponis, J., Ross, C., & Rashid, M. (1997). The structure of urban space, movement and co-presence: the case of Atlanta. Geoforum, 28(3), 341-358. Peponis, J., & Wineman, J. (2002). Spatial structure of environment and behavior. i: Bechtel, R. B., & Churchman, A. (red.). Handbook of environmental psychology, New Jersey: Wiley. Penn, A., Hillier, B., Banister, D., & Xu, J. (1998). Configurational modelling of urban movement networks. Environment and Planning B: Planning & Design, 25(1), 59-84. Pickering, W. S. F. (1984). Durkheim's Sociology of Religion: Themes and Theories. London: Routledge & Kegan Paul. Polanyi, Michael (1966). The Tacit Dimension. Chicago: University of Chicago Press: Chicago. Porter, J. R., & Howell, F. M. (2012). Geographical Sociology: Theoretical Foundations and Methodological Applications in the Sociology of Location. New York: Springer. Portugali, J. (2011). Complexity, cognition and the city. New York: Springer. Raford, N., & Ragland, D. (2004). Space syntax: innovative pedestrian volume modeling tool for pedestrian safety. Transportation Research Record: Journal of the Transportation Research Board, 1878(1), 66-74. Ratti, C. (2004). Urban texture and space syntax: some inconsistencies. Environ- ment and Planning B: Planning and Design, 31(4), 487-499. Ratti, C. (2005). The lineage of the line: space syntax parameters from the analysis of urban DEMs. Environment and Planning B: Planning and Design, 32, 547- 566. Raudenbush, S. W., & Sampson, R. J. (1999). Ecometrics: toward a science of assessing ecological settings, with application to the systematic social observation of neighborhoods. Sociological methodology, 29(1), 1-41. Reiss, A. J. Jr. (1971). Systematic Observations of Natual Social Phenomena. Costner, H. (red.). Sociological Methodology. San Francisco: Jossey-Bass. Rindfuss, R. R. and Stern, P. C. (1998). Linking Remote Sening and Social Sci- ences. The Need and the Challenges. i: Liverman, D., Moran, E. F., Rind- fuss, R. R., & Stern, P. C. (red.). People and Pixels: Linking Remote Sensing and Social Science. Washington: National Academy Press. Sahni, I. P. (2012). More than a Façade: Somatic and Structural Determinants in Erving Goffman's Theory of the Perduring Self. Symbolic Interaction, 35(2), 162-185. Sampson, R. (2011). Neighborhood effects, causal mechanisms and the social structure of the city. i: Demeulenaere, P. (red.). Analytical Sociology and Social Mechanisms. Cambridge: Cambridge University Press. Sampson, R. J. (2012). Great American city: Chicago and the Enduring Neighborhood Effect. Chicago: University of Chicago Press.

Sampson, R. J., Raudenbush, S. W., & Earls, F. (1997). Neighborhoods and violent crime: A multilevel study of collective efficacy. Science, 277(5328), 918-924. Sampson, Robert J & Raudenbush, Stephen W. (1999). Systematic Social Ob- servation of Public Spaces: A New Look at Disorder in Urban Neighbor- hoods. American Journal of Sociology. 105(3), 603–651. Sampson, R. J., Morenoff, J. D., & Gannon-Rowley, T. (2002). Assessing ‘neighborhood effects’: Social processes and new directions in research. An- nual Review of Sociology, 28, 443-478. Sartre, J. P. (2004). Critique of Dialectical Reason: Theory of practical ensembles (Vol. 1). London & New York: Verso. Scheff, T. J. (1990): Microsociology. Discourse, Emotion, and Social Structure. Chicago: The University of Chicago Press. Schelling, T. C. (1978). Micromotives and Macrobehavior. New York: W. W. Norton & Company. Schiermer, B. (2011). Quasi-objects, Cult Objects and Fashion Objects On Two Kinds of Fetishism on Display in Modern Culture. Theory, Culture & Society, 28(1), 81-102. Schillmeier, M. (2009). Assembling money and senses: revisiting Georg Simmel and the city. i: Farias, I., & Bender, T. (red.). (2012). Urban Assemblages: How Actor-Network Theory Changes Urban Studies. London & New York: Routledge. Schmitz, H. (2011). Emotions outside the box - the new phenomenology of feeling and corporeality. Phenomenology and the Cognitive Sciences, 10(2), 241-259. Schnore, L. F. (1958). Social morphology and human ecology. American Journal of Sociology, 63(6), 620-634. Schnore, L. F. (1961). The Myth of Human Ecology. Sociological Inquiry, 31(2), 128-139. Schroder, C., Mackaness, W., & Reitsma, F. (2007). Quantifying urban visibility using 3D space syntax. Proceedings of GISRUK, Maynooth, Ireland. Scott, J. (2012). What is Social Network Analysis? London: A&C Black. Scollon, R., & Scollon, S. W. (2003). Discourses in Place: Language in the Material World. London & New York: Routledge. Seamon, D. (2012). ‘A jumping, joyous urban jumble’: Jane Jacobs’s Death and Life of Great American Cities as a phenomenology of urban place. The Jour- nal of Space Syntax, 3(1), 139-149. Shaw, J., & Hesse, M. (2010). Transport, geography and the ‘new’ mobilities. Transactions of the Institute of British Geographers, 35(3), 305-312. Sherman, L. W., Gartin, P. R., & Buerger, M. E. (1989). Hot Spots of Predatory Crime: Routine Activities and the Criminology of Place. Criminology, 27(1), 27-56. Simmel, Georg (1976). The Metropolis and Mental Life. i: Wolff, K (red.) The Sociology of Georg Simmel. New York: Free Press. Simmel, G. (1994). Bridge and door. Theory Culture and Society, 11(1), 5-10. Simmel, G. (2009). Sociology: Inquiries into the Construction of Social Forms. Vol. 2. Leiden & Boston: Brill. Simonsen, K. (1996). What kind of space in what kind of social theory? Progress in Human Geography, 20(4), 494-512. Smith, R. W., & Bugni, V. (2006). Symbolic interaction theory and architecture. Symbolic Interaction, 29(2), 123-155.

Smith, P. (1999). The elementary forms of place and their transformations: a Durkheimian model. Qualitative Sociology, 22(1), 13-36. Snell, P. (2010). From Durkheim to the Chicago school: Against the ‘variables sociology’ paradigm. Journal of Classical Sociology, 10(1), 51-67. Soja, E. (1989). Postmodern Geographies: The Reassertion of Space in Critical Social Theory. London & New York: Verso. Soja, E. (2001). In Different Spaces, Interpreting the spatial organization of societies. Proceedings of the 3rd International Space Syntax Symposium, At- lanta, Georgia Institute of Technology. Stevens, Q. & Shin, H. (2012): Urban Festivals and Local Social Spaces. Planning Practice and Research. 29(1), 1-20. Strenski, I. (2006). The New Durkheim. New Brunswick: Rutgers University Press. Strong, P. M. (2006). Two Types of Ceremonial Order. Sociology and Medicine: Selected Essays by PM Strong. Aldershot: Ashgate. Stouffer, S. A. (1955). Communism, Conformity and Civil Liberties. Garden City: Doubleday. Ståhle, A., Marcus, L., & Karlström, A. (2005). Place Syntax: Geographic accessibility with axial lines in GIS. Proceedings of the 5th International Space Syntax Symposium, The Faculty of Architecture, TU Delft. Thrift, N. (2008). Non-representational Theory: Space, Politics, Affect. London & New York: Routledge. Thompson, K. (2003). Emile Durkheim. London & New York: Routledge. Tobler W., (1970) A computer movie simulating urban growth in the Detroit region. Economic Geography, 46(2): 234-240. Tonboe, J. (1993). Rummets sociologi. København: Akademisk Forlag. Turner, A. (2001). Angular analysis. Proceedings of the 3rd International Space Syntax Symposium, Atlanta, Georgia Institute of Technology. Turner, J. H., & J. E. Stets (2005). The Sociology of Emotions. Cambridge: Cam- bridge University Press. Turner, J. H., & Kim, K. M. (1999). The disintegration of tribal solidarity among American sociologists: Implications for knowledge accumulation. The American Sociologist, 30(2), 5-20. Turner, J. H. (1996). The evolution of emotions in humans: A Darwinian– Durkheimian analysis. Journal for the Theory of Social Behaviour, 26(1), 1-33. Turner, J. (2010). Theoretical Principles of Sociology. Vol. 2. Microdynamics. New York: Springer. Urry, J. (2000). Sociology Beyond Societies. Mobilities for the Twenty-First Century. Lon- don & New York: Routledge. Urry, J. (2004). Small worlds and the new ‘social physics’. Global Networks, 4(2), 109-130. Urry, J. (2005). The complexity turn. Theory Culture and Society, 22(5), 1-14. Urry, J. (2007). Mobilities. Cambridge: Cambridge: Polity Press. Vandenberghe, F. (2002). Reconstructing humants: a humanist critique of act- ant-network theory. Theory, Culture & Society, 19(5-6), 51-67. Ward, M. D., & Gleditsch, K. S. (2008). Spatial Regression Models (Vol. 155). London: Sage. Warf, B., & Arias, S. (2008). The Spatial Turn: Interdisciplinary Perspectives. London: Taylor & Francis

Wassard, Erik (1990): Nørrevold, Østervold og Kastellet. i: Bramsen, B. (red.). København før og nu - og aldrig. Bind 10. København: Palle Fogtdals Forlag. Weber, Max (1978). Economy and Society. An Outline of Interpretive Sociology. Berke- ley: University of California Press. Weisburd, D. L., Bruinsma, G. J. N., & Bernasco, W. (2009). Units of Analysis in Geographic Criminology: Historical Developments, Critical Issues and Open Questions. i: Weisburd, D., Bernasco, W., & Bruinsma, G. (red.). Put- ting crime in its place. New York: Springer. Weisburd, D. L., Groff, E. R., & Yang, S. M. (2012). The Criminology of Place: Street Segments and our Understanding of the Crime Problem. Oxford: Oxford Uni- versity Press. Weisburd, D., & Telep, C. W. (2014). Law of Crime Concentrations at Places. i: Bruinsma, G., & Weisburd, D. (red.). Encyclopedia of Criminology and Criminal Justice. New York: Springer. Weisburd, D., & Amram, S. (2014). The law of concentrations of crime at place: the case of Tel Aviv-Jaffa. Police Practice and Research, (ahead-of-print), 1-14. Wetherell, M. (2012). Affect and Emotion: A New Social Science Understanding. Lon- don: Sage. Whyte, W. H. (1988). City. Rediscovering the Center. New York: Doubleday. Wilson, T. C. (1985). Urbanism and Tolerance: A Test of some Hypothesis Drawn from Wirth and Stouffer. American Sociological Review, 50(1), 117-123. Wilson, J. Q., & Kelling, G. (1982). The police and neighborhood safety: Bro- ken windows. Atlantic Monthly, 249(3), 29-38. Wirth, L. (1938): Urbanism as a way of life. American Journal of Sociology, 44(1), 1- 24. Yaneva, A. (2009). Making the social hold: Towards an Actor-Network theory of design. Design and Culture, 1(3), 273-288. Yaneva, A. (2010). The Architectural as a Type of Connector. Perspecta. The Yale Architectural Journal, 42, 138-143 Yang, S. M. (2010). Assessing the spatial-temporal relationship between disorder and violence. Journal of Quantitative Criminology, 26(1), 139-163. Østerberg, D. (1975). Essays i samfunnsteori. Oslo: Pax forlag.

Kapitel 1 Space Syntax Theory and Durkheim’s Social Morpholo- gy: A Reassessment

Abstract. When Bill Hillier and Julienne Hanson outlined their space syntax approach in The Social Logic of Space, they acknowledged the affinity of their space morphological oeuvre to Durk- heim’s classical sociology. Taking Durkheim’s social morphology as their point of departure, their ambition was to develop a theory of the hitherto under-theorized relationship between society and space. In the light of this explicitly Durkheimian inspiration, it is surprising that mainstream sociology, has been so silent on the subject of space syntax theory. This silence ends up being counterproductive for space syntax and sociology alike: while sociology neglects the original contribution of space syntax to a neo-Durkheimian theory of spatial morphology, space syntax theory loses a valuable sociological interlocutor. Hence, after revisiting Durkheim’s social morphology and reviewing the analytical strengths and deficits of Hillier and Hanson’s Durkheimian reception, the objective of the article is to propose that the micro-sociological face of the Durkheimian tradition as a much- needed corrective to space syntax.

Keywords. Émile Durkheim; social morphology; space syntax; interaction ritual theory

Introduction S PACE IS ON THE SOCIOLOGICAL AGENDA. Increasingly, sociology describes itself as undergoing a ‘spatial turn’ (Warf and Arias, 2009), suggesting that space should be regarded as a new dimension of sociology. However, following Tonboe’s (1993) systematic examination of discussions of spatiality found within classic and modern sociology, we might well question the actual novelty of the spatial dimension. It would be more accurate to consider space as a forgotten dimension, given that most of the classical writers in the sociological tradition, from Marx and Weber to Simmel and Durkheim, conceptualized modern society in the context of space. Similar to the situation two decades ago when Tonboe (1993) suggested that the notion of space had been ‘lost and found’, today’s hypothesis of ‘newness’ is based on a historical under-emphasis of these space theoretical contributions. Such neglect may well be most conspicuous with regard to the founding father of French sociology,

Émile Durkheim, whose notion of social morphology introduced the question of space into the very heart of sociological inquiry. Now as then, the purported ‘newness’ of the spatial turn is buttressed by an absence of references to this Durkheimian notion of social morphology. Remarkably enough it is outside mainstream sociology that we find a contemporary contribution that fully acknowledges Durkheim’s theoriz- ing on space. The analytical architecture theory of space syntax, para- digmatically outlined in Bill Hillier and Julienne Hanson’s (1984) The Social Logic of Space, was explicitly built on the spatial shoulders of Durk- heim, making this line of reception one of the most fully elaborated theoretical continuations of Durkheim’s social morphology in the literature, including the sociological literature (cf. also the preliminary statement in Hillier et al. 1976). Elaborating on Tonboe’s sociological thesis on the ‘lost and found’ nature of space, we might say that the fate and future of Durkheim’s social morphology is contained in the history of how it was lost by sociology and then found by the space syntax paradigm. Unfor- tunately, Hillier and Hanson’s reappraisal has not prompted sociologists in general, or the proponents of the spatial turn in particular, to recon- sider their lost interest in Durkheim’s social morphology, or its contemporary reception. Despite Hillier and Hanson’s original and systematic contribution to Durkheimian scholarship, the sociological literature offers only a few scattered remarks on space syntax, often with a disdainful under-tone, and without acknowledgment of its Durkheimian nature or relevance (Saunders, 1986; Soja, 2001; Lawson, 2001). The aim of this article is to take some initial steps to break the sociological silence, which is counterproductive for both space syntax and sociology: the space syntax paradigm loses a valuable discussion partner, one that could clarify the nature of spatially embedded social life, while sociology overlooks the original contribution of space syntax to a neo- Durkheimian theory of spatial morphology. First, the article will revisit Durkheim’s social morphological argument that social life rests upon, and thus is marked by, a material substratum made up of the two morphological components of space and population. Second, we will examine how Hillier and Hanson elaborate and ultimately go beyond Durk- heim’s original framework by resolving Durkheim’s ambivalent account of how these two morphological components are causally related. Third, in the light of this original reception, the article outlines the theoretical and methodological nature of the Durkheimian research agenda as outlined by Hillier and Hanson. Fourth, we identify a number of deficits of this reading of social morphology, specifically originating from Hillier and Hanson’s theoretical proximity to the structural-functionalist ‘mac-

ro-wing’ of Durkheimian scholarship. Finally, we will suggest that these macro-wing deficits can be dealt with by establishing a closer theoretical link between space syntax and the competing ‘micro-wing’ reading of Durkheim, specifically Randall Collins’ (2004) micro-sociology of the interaction ritual.

Revisiting Durkheim’s Social Morphology Durkheim was an extraordinarily self-conscious scholar, as reflected by the fact that he was well aware of his role as one of the principal found- ers of sociology. Accordingly, he devoted considerable effort to explain the foundation of this new science of the social; these formal inquiries testify to the great significance Durkheim ascribed to the ‘social morphology’ that, along with what is called ‘social physiology’, constitute the sub-disciplinary pillars on which his ‘general sociology’ rests (Durkheim 1978; Mauss, 2005). Essentially, social morphology is the sociological inquiry into the ‘material substratum’ of society, defined by two socio-morphological components. The first component is an examination of the numerical properties of the social population, understood as the volume, distribution, and social density of the societal population, although Durkheim, as we shall see, especially emphasizes the explanatory importance of the property of density. The second component of social morphology concerns the geographical base of social life, that is, the spatial surface on which the population is distributed. The spatial surface includes the properties of natural geography and the organization of the built environment, that is, how “(…) the disposition of streets and squares, the architecture of the houses, and the structure of things made vary from village to town and from the large city to the small one, and so on” (Durkheim, 1960: 361). In summary, Durkheim describes these two social morphologies as an analytical inquiry into the ‘anatomy’ of society, analogous to the biological study of the morphological ‘form’ – the structure and composition of ‘tissues’ and ‘organs’ – of living creatures. Keeping to this biological analogy, Durkheim defines the other sub- disciplinary pillar of his general sociology, social physiology, as the study of the societal ‘functions’ inherent to this tissue and these societal organs: “Besides the social ways of being, there are the social modes of doing; besides the morphological phenomena, there are the functional or physiological phenomena” (Durkheim, 1960: 363). These ‘functional doings’ suggest a rich variety of sociological subject fields, ranked by their importance to society as a whole. For example, aesthetic physiog-

nomies are considered of secondary importance in comparison to moral and religious functions, which, as we shall later describe, were the respective physiological focal points for the early and the late Durkheim. Careful to strike a balance between the real and ideal ontologies of society, Durkheim emphasizes that although the functional doings of a society are contingent upon its form-specific manner of material being, there is a ‘variety’ and ‘complexity’ in the differences between physiology and morphology (Durkheim, 1960). The material composition of ‘people and things’, which by necessity are ‘connected in space’, conditions the emergence of a societal complexity whose “(…) whole very often has very different properties from those which its constituent parts possess” (Durkheim, 1978: 76). Thus, although morphologically derived, society has a (religious, moral, etc.) existence sui generis, whose social physiological manifestation acts autonomously as the ‘collective consciousness’ of society:

While the different forms of collective activity also have their own substratum and while they derive from it in the last instance, once they exist, they become, in turn, creative sources of action, they have an effectiveness all their own, and they react on the very causes on which they depend (Durkheim, 2005a: 17).

Durkheim (1984) most fully elaborated the empirical implications of this morphology-physiology nexus in his early masterpiece, The Division of Labor in Society, which is also the Durkheimian reference that Hillier and Hanson (1984) take as their point of departure. The moral-theoretical and, in this sense, social physiological, question that Durkheim sets out to explore in Division is as relevant today as ever: “Why does the individual while becoming more autonomous, depend more upon society? How can he become at once more individual and more solidary?” (Durkheim, 1984: 37). Famously, Durkheim provides the socio-physiological answer that pre-modern and modern societies are regulated by two distinct manners of performing socio-moral functions. The ‘mechanic solidarity’ of the pre-modern society is succeeded by the ‘organic solidarity’ of modernity, which does not arise due to infinite points of resemblance immediately connecting the individual to the society, as it is the case in the weakly differentiated and segmented pre-modern society. Rather, modern and organic solidarity is a product of the growth of the reciprocal interdependences among differentiated individuals, especially in the form of contractual relations on the urban labor market.

In keeping with sociologists’ general neglect of social morphology, it may be less familiar that Durkheim suggests a clear-cut morphological explanation of this historical evolution. Hence, the “(…) causes explaining the progress of division of labor (…)” (Durkheim, 1984: 256), which, in turn, transforms the societal nature of solidarity, are found in the ‘material substratum’. The division of labor develops as the undifferentiated population segments of pre-modernity gradually expand their social relations through space: “The increase in the division of labor is therefore due to the fact that the social segments lose their individuality, that the partitions dividing them become more permeable” (Durkheim, 1984: 200). This emphasis on the gradual disappearance of the ‘boundaries’, which ensured the segmenting of the social population during pre-modern times, highlights how Durkheim perceives this morphological cause of history as having an unmistakably spatial cause. It is as the spatial partitions of social life become less marked that the social segmentation of pre-modernity is succeeded by a dense population structure, so characteristic of urbanized modernity; moreover, Durkheim emphasizes how this spatial progress is technologically accelerated by the growth of larger numbers of faster communication and transportation connections across space and across former spatial boundaries, especially in the form of infrastructure within and between metropolitan cities. In relating the two morphological properties of spatial organization and population characteristics, the crux of matter is how this spatially initiated process increases the volume and especially the ‘dynamic density’ of social life. The level of labor division is essentially a function of the numerical growth of social relations defined as “(…) the number of individuals who are actually having not only commercial but also social relations” (Durkheim, 1966: 114). Advancing this argument, Durkheim (1984) draws on a controversial social-Darwinist schema: the increasing level of dynamic density intensifies ‘the struggle for existence’, and this conflictual situation is, in turn, socio-physiologically resolved by the emergence of organic solidarity, generated by the same dynamic density. Thus, explicating how ‘everything occurs mechanically’ Durkheim (1984: 212) concludes that “(…) a break in the equilibrium of the social mass gives rise to conflict that can only be resolved by a more developed form of the division of labor: this is the driving force for progress”. Although the hypotheses outlined in Division are indubitably among the most influential within sociology, they also remain controversial (Til- ly, 1981). As I will suggest, the originality of Hillier and Hanson’s Durk- heimian contribution can initially be ascribed to their formulating a corrective to one of the remaining controversies that limits Durkheim’s the-

oretical inquiry into the morphological mechanics of social progress; that is, by which ontological means does spatial morphology affect the property of population density?

Explaining Durkheim’s Non-Explanation As the title of Hillier and Hanson’s (1984) The Social Logic of Space suggests, the space syntax paradigm is of at least as much interest to sociology as to architecture theory. Their principal focus is on the ‘society- space relation’, reflected in the purpose of establishing a systematic theory on “(…) how spatial pattern can, and does, in itself carry social information and content” (Hillier and Hanson 1984: xi). This explicitly addresses a paradigmatic deficit in contemporary theories of the society- space relation from the humanities and the social sciences: “The [pre- dominant] paradigm in effect conceptualizes space as being without social content and society without spatial content. Yet neither can be the case, if there is a lawful relation between them” (Hillier and Hanson, 1984: x). Facing this theoretical impasse, Durkheim’s social morphology offers, according to Hillier and Hanson, a theoretical exception so extraordinary that they adopt it as the cornerstone for their space syntax paradigm:

Durkheim actually located the cause of the different solidarities in the spatial variables, namely the size and density of populations. In the work of Durkheim, we found the missing component of a theory of space, in the form of the elements for a spatial analysis of social formations. But to develop these initial ideas into a social theory of space, we had to go back once again into the foundations, and consider the sociology of the simplest spatial structure we had found useful to consider: the elementary cell (Hillier and Hanson 1984: 18, italics added).

At first, Hillier and Hanson seem utterly loyal to Durkheim’s original morphological scheme. Yet the small but paramount ‘but’ signifies that Hillier and Hanson’s contribution is far from a simple reiteration. Durk- heim’s material substratum will have its elementary foundation spatially reconsidered. As mentioned, Hillier and Hanson are addressing one of the controversies that have continued to constrain Durkheim’s scheme. This Achilles heel in the original version of social morphology appears when we scrutinize the theoretical (in)consistency of the explanatory chain provided by Durkheim in Division. Specifically, this concerns the

ontological nature of the causal relation between the two socio-morphological components of space and population density; Durkheim appears to reveal an ambivalent or ‘shifting attitude’ (Schlanger 2006) about stick- ing to the spatial explanation of the dynamic density that he, in fact, suggests in Division. This space morphological causal explanation is cut off just as Durkheim’s analysis approaches its conclusion. The transforming spatial boundaries that had been the first link in Durkheim’s socio- morphological chain of explanation of organic solidarity is replaced by something that has the appearance of a self-imposed non-explanation:

We do not here have to look to see if the fact which determines the progress of the division of labor and civilization, growth in social mass and density, explains itself automatically (…). We content ourselves with stating this law of gravitation in the social world without going back any farther (Durkheim, 1964: 339).

After this surprising argument in Division (why would a thinker with a positivist mind such as Durkheim’s let something as graceful as the law of gravitation of the social world remain unexplained?), Durkheim continues in the following paragraph by making the self-contradictory statement that we should, after all, hold onto the spatial explanation of the dynamic density as provided:

The walls which separate different parts of society are torn down by the force of things, through a sort of natural wear (…). The movements of population thus become more numerous and rapid and the passage-lines through which these movements are effected – the means of communication – deepen. They are more particularly active at points where several of these lines cross; these are cities. Thus, social density grows (Durkheim, 1964: 339).

Why this ambivalence? I would like to suggest that the reason for Durkheim’s ambivalence might be ascribed to the fact that in providing the spatial explanation in Division, Durkheim is balancing on the ontological edge of his own morphology-physiology distinction, assigning a mo- nopoly of societal agency to the physiological face of society. As we recall, the social physiology relates to the functional ways of ‘doing’, whereas social morphology relates to the material modes of ‘being’. Moreover, this is reflected in Durkheim’s explicit ‘sociocentrism’ (Durkheim and Mauss, 2009), which – despite being a collective alternative to purely individual ‘anthropocentrism’ – persists in the anthropocentric view that

rejects ascribing agency akin to that of humans to any material or ‘non- human’ actors (Tonboe, 1993; Latour, 2005). As regards ‘things’, which would include the spatial artifacts of the social morphology, Durkheim thus argues in his The Rules of Sociological Method that “(…) they contain nothing of what is required to put things in motion. They are matter upon which social forces of society acts, but by themselves they release no social energy (Durkheim, 1966: 113). Yet, this inability to release social energy is not so evident in Division, where the moving, and in this sense, agential, forces seem to be ascribed to morphological space as well. As we saw above, the increase in dynamic density is spatially initiated ‘by the force of things, through a sort of natural wear’, stimulating the emergence of the ‘passage-lines through which these movements are effected’. Or, in an even more ambivalent manner, Durkheim (2005b) argues in On Suicide that once the ‘material objects’ of society are ‘built’, these materializations of an architectural, infrastructural, or technological kind become an autonomous reality, independent of individuals and socio-physiological doings: “Social life, having as it were crystallized itself in this way and fixed itself on material props, is by that very fact exteriorized and acts upon us from outside” (Durkheim, 2005b: 348). In sum, the self-imposed spatial non-explanation simply reflects Durkheim’s ambivalent stance towards the potential agency of spatio-material things. Returning to space syntax, I would suggest that Hillier and Hanson’s ‘but’ is prompted by Durkheim’s ambivalence about space morphology: the effort to ‘go back once again into the foundations’ can be read as an attempt to explicate how the socio-physiological solidarities of society are contingent upon and, thus, explainable by the ‘real agency’ (Hillier, 2005) of the spatial morphology. Durkheim had the right intuition regarding morphological space, but his ambivalence diverted him from giving a full-fledged spatial explanation for the dynamic density and, hence, for the ‘law of gravitation in the social world’. The remarkable Durkheimian contribution of Hillier and Hanson’s space syntax is to re- assemble these space morphological elements into a social theory of space.

A Sociology of the Elementary Cell In outlining the contribution of the space syntax paradigm to Durkheim- ian thought, we should take our point of departure where Hillier and Hanson began, by reconsidering the simplest spatial structure in Durk- heim’s material substratum: the elementary cell. Hillier and Hanson’s

(1984) ‘sociology of the elementary cell’ meets the dual purpose of providing a theoretical and an operational account of the space-society relation. Let us outline these two purposes separately, starting with their theoretical contribution. The elementary cell is essentially defined as a spatial difference that makes a social difference as it establishes a sociological distinction between an ‘inside’ and ‘outside’ (Hillier and Hanson, 1984). Hence, space carries a sociological reference: while interior spaces have well-defined social roles and norms, exterior spaces are characterized by having fewer and less explicit categorical differences. Insiders are recognized as the ‘inhabitants’, spatially demarcated from the outside ‘world of (urban) strangers’ (Hillier 1989). At the most general level, social relationships are spatially distinguished as being either ‘trans-spatial’ or ‘spatial’, reflecting whether social identities are merely based on spatial proximity (e.g. between urban strangers on the labor market), or instead, on a social bond that persists across space (e.g. family members who maintain their identity as ‘daughter’ or ‘father’ despite their spatial non-proximity). Elaborating these socio-spatial considerations into a general sociology of the elementary cell, Hillier and Hanson suggest two pathways of growth from the elementary cell: “it can be by subdividing a cell, or ac- cumulating cells, so that internal permeability is maintained; or by aggre- gating them independently, so that the continuous permeability is maintained externally” (Hillier and Hanson, 1984: 19). This result is two socially as well as spatially distinct phenomena, ‘buildings’ and ‘settlements’, which follow the respective sociological logic of the inside and outside. Keeping Durkheim’s physiology-morphology nexus in mind, the syntactical differences between these spaces should not be confused as having a merely symbolic nature, as in the influential case of Lévi-Strauss’ (1963) physiologically based Durkheimianism. Instead, the elementary cell has to do with the material substratum and thus with how the symbolic difference between inside and outside is realized and reproduced through space morphology. This view has a closer affinity to Goffman’s (1956) morphologically grounded difference between the ‘front’ and ‘back’ regions of space. As Hillier and Hanson address Durkheim’s spatial non-explanation of the dynamic density, the heart of their explanatory argument is that the spatial difference between inside and outside makes a social difference due to the fact that the spatial morphology – that is, the network produced by the two pathways of growth from the elementary cell – holds an agential capacity to generate and regulate patterns of ‘movement’ and ‘co-presence’. As emphasized by Hillier and Hanson, the spatial morphology

100

(…) has a direct relation – rather than a merely symbolic one – to social life, since it provides the material preconditions for the pattern of movement, encounters and avoidance which are the material realiza- tion – as well as sometimes the generator – of social relations (1984: ix)

In a remarkable manner, Hillier and Hanson thus transcend Durk- heim’s ambivalence: space strongly affects society because of the network agency of space on the dynamic density (Hillier, 1996). While Hilli- er and Hanson spatially reappraise the explanans of Durkheim’s morphological chain of (non-)explanation, they hold on to the functionalist ex- planandum outlined in Division, emphasizing that this spatial agency specifically serves to reproduce the physiological function of social solidarity. The ‘principal axiom’ of the space syntax is, accordingly, that the difference between the ‘trans-spatial’ relationships of the inside and the ‘spatial’ relationships of the outside reflects the existence of disparate morphological organizations and spatial means by which social solidarities are reproduced. Thus, continuing from Durkheim’s distinction between mechanical and organic solidarity, there is a duality in which society generates space and, as Hillier and Hanson put it (1984: 20, original italics), this duality is a

(…) function of the different forms of solidarity. (…) One requires a strong control on boundaries and strong internal organization in order to maintain an essentially transspatial form of solidarity. The other requires weak boundaries, and the generation rather than the control of events.

Although Hillier and Hanson never explicate the nature of their Durkheimian reception, I would suggest that their sociological inquiry into the elementary cell strongly suggest a theoretical proximity to the so-called ‘macro wing’ within the Durkheimian tradition (Collins 1994). Affiliation with the ‘macro wing’, as exemplified by Parsons (1951), implies an emphasis on the structural-functionalist aspect of Durkheim’s sociology, which argues that society is an ordered totality holding a number of ‘functional requisites’ or ‘social needs’, which the structural parts of society exist in order to reproduce (Turner and Maryanski 1979). Regarding Hillier and Hanson’s sociology of the elementary cell, the notion of social solidarity seems to be elaborated precisely in accordance with such a notion of functional requisites: morphological space is per-

101

ceived as being organized in order to reproduce a transspatial and spatial need for mechanical and organic solidarity, respectively:

Could it be that different types of society required different kinds of control on encounters in order to be that type of society; because if this were so, we could reasonably expect it to be the deepest level at which society generated spatial form (Hillier and Han- son, 1984: 18).

Moreover, in the light of these consideration, it should not come as a surprise that Hillier and Hanson (1984), in keeping with Durkheim’s (2005a) socio-biological theory of anomie, describe the space morphological (un)fulfillment of the solidarity requisite in term of ‘normal’, ‘healthy’, and ‘pathological’. This reflects the most structural-functionalist side of Durkheim, and should, along with space syntax, be regarded as a token of a macro wing reading of Durkheim’s thought. This being so, neither Netto’s (2007) categorization of Hillier and Hanson as simply ‘Durkheimian’, without specifying any particular line of reception, nor Seibert’s (2006) suggestion that space syntax displays certain ‘functionalist undertones’ goes far enough. Hillier and Hanson’s space syntax should be described as an example of genuine structural-functionalism, and as belonging to the macro wing of the Durkheimian tradition. Leaving these theoretical considerations, we are next going to consider space syntax as a methodological corrective to Durkheim’s social morphology that makes it possible to correlate the objective network properties of morphological space with actual patterns of movement and co- presence as found in different types of building and settlement morphologies. This allows for an empirical test of Hillier and Hanson’s non-ambivalent account of spatial agency. Durkheim gives no methodological guidelines for such an operationalization of morphological space. While Durkheim, however ambivalently, did provide some theoretical considerations on the spatial ‘passage-lines’ through which movement and densities are effected, it should be considered the novel achievement of the space syntax methodology to operationalize these passage-lines into graph-theoretical networks of ‘axial-lines’ (Hillier and Hanson, 1984). Most importantly, it is network ‘integration’, measuring how ‘deep’ each axial-line is located in the spatial network, that proves to be the strongest predictor of actual movement, especially in urban morphologies (Hillier et al., 1993). Thus, contrary to predominantly non-morphological views of urban and space sociology (Saunders, 1986), which explain such variations in human movement as a function of some kind of

102

social attraction or aversion, space syntax proposes that the spatial network or ‘configuration’ confounds the relation between ‘movement’ and ‘attraction’; moreover, due to its spatio-material inertness, this configuration is not similarly shaped by the patterns of movement and attraction:

Figure 1. Configuration as Confounder (Hillier et al., 1993: 31).

Here, we should note that space syntax is not the first Durkheimian attempt to make a methodological contribution to the space morphological study of the material substratum, as evidenced by the endeavors of Durkheim’s contemporary students, especially Mauss (1979) and Halb- wachs (1960). However innovative, these contributions were, as acknowledged by Mauss (2005), constrained by the undeveloped status of contemporary quantitative methods, thus making it difficult to meet Durk- heim’s strict positivist dictum: “The first and most fundamental rule is: Consider social facts as things” (Durkheim, 1966: 14, original italics). Accord- ing to Mauss, it thus falls upon the future generations of Durkheimians to meet this methodological challenge to measure morphological space as a thing, since “(…) we already know how far the historian and sociologist of the coming generations will be better armed than we were” (Mauss, 2005: 70). Simply put, space syntax belongs to this emerging quantitative future of Durkheimian scholarship. Methodologically, space syntax allows us measure the ‘non-discursive regularity’ (Hillier, 1996) of societal space as a ‘social fact’; that is, as prescribed by Durkheim’s methodological dictum, considered as an exterior and restraining ‘thing’, which holds a real agential capacity to affect actual patterns of movement and co-presence. Consequently, Durkheim’s spatial theory that Hillier and Hanson found it so ‘useful’ to reappraise is more closely affiliated to Emirbayer’s (1996) Useful Durkheim than to Tilly’s (1981) Useless Durkheim, but with an original methodological slant. By measuring the social fact of space as a thing, Hillier and Hanson offer a methodological approach for using the social morphology, which Durkheim and his contemporary proponents should have, but did not succeed, in fully opera- tionalizing.

103

Space Syntax and the Shortcomings of the Macro Wing All paradigmatic positions involve certain limitations. In the case of space syntax, this is manifested by its paradigmatic affinity to the Durk- heimian macro wing, whose analytical shortcomings Hillier and Hanson do not merely reassess, but also inherit. As I will argue in this section, the deficiencies of the macro wing create a serious problem for the validity of Hillier and Hanson’s theoretical and methodological contributions, as I have just outlined them. One of the defining characteristics of the Durkheimian macro wing is a strong, and sometimes almost exclusive, emphasis on Durkheim’s early sociology and masterpiece, Division, thus downplaying the importance of Durkheim’s (1995) later sociology of religion, outlined in the late masterpiece, The Elementary forms of Religious Life. Moreover, this structural-functionalist reading often includes a du- bious use of teleological explanation, a tendency that Durkheim explicitly rejected (Turner and Maryanski, 1979). It is in relation to this kind of teleological schema that problems initially arise in Hillier and Hanson’s theoretical approach to Durkheim. Space syntax inherits, faces, but ultimately fails to adequately deal with the questionable teleological framework in the structural-functionalist macro wing tradition. This contro- versy can be seen in the following citation in which Hillier and Hanson explain their spatially based structure-functionalism with an empirical analogy to Durkheim’s concept of division of labor:

The pubs are analogous to the kivas of Hobi society, in the sense that they operate not only in a localized way, but as a means of generating a higher order of system (…). The same type of morphological principles, though with a very different social mechanism, is illustrated by the relation between division of labour and the wider system in the medieval town (Hillier and Hanson, 1984: 254).

In other words, Hillier and Hanson suggest that the spatial organization of social entities as diverse as a modern pub culture and pre-modern Hobi culture serves as a morphological means to fulfill a higher-order functional requisite, analogous to Durkheim’s argument that the division of labor fulfills a functional need for social solidarity, as in the case of medieval town guilds. The problem with this explanation is that Durk- heim would be more than skeptical about the teleological scheme that underlies Hillier and Hanson’s argument and, furthermore, read into

104

Durkheim so as to justify the analogy. Insisting on the contingency as opposed to the teleology of historical evolution, Durkheim argues that “(…) social phenomena do not generally exist for the useful results they produce” (Durkheim, 1966: 95). The progressive division of labor did not develop in order to fulfill a functional need for conciliatory solidarity; one must draw a sharp line between teleological (and thus illegitimate) ‘functional explanations’ and scientifically valid ‘causal explanations’: “We cannot use ‘aim’ or ‘purpose’, and speak of the goal of the division of labor, because that would suppose that the division of labor exists for the sake of results that we shall determine” (Durkheim, 1984: 11). Yet, this is precisely what Hillier and Hanson assume in their teleological schema and their analogy: the socio-functional logic of space orches- trates the spatial morphology so as to bring about the result of reproducing the functional need for social solidarity; accordingly, this spatial organization of society is to be explained in functional terms as being a “(…) means of generating a higher order of system” (Hillier and Han- son, 1984: 254). Durkheim’s anti-teleological stance can be explained with reference to his effort to legitimize sociology as a positive science, rather than regarding it as being based on philosophical speculation. Thus, it is no coincidence that Division quotes Aristotle’s Politics on its title page: “A city is not made up of people who are same; it is different from an alliance” (Durkheim, 1964: viii). That is to say, while Durkheim praises Aristotle for being the originator of the philosophical distinction between ‘same- ness’-based mechanical solidarity and organic solidarity based upon urban heterogeneity, Durkheim indicates at the same time that his sociological treatment of the subject is evidence-based. Contrary to the metaphysical nature of Aristotle’s teleological explanation offered in Politics, he now proposes to explain social solidarity on the basis of socio-morphological facts and mechanics (Durkheim 2004). Returning to Hillier and Hanson, it is important to emphasize that they also seem to understand teleological explanations as being an Aris- totelian problem (Hillier and Hanson, 1984; Hillier 1996). This is evident in their critique of Aristotle’s position about the ways that nature manag- es to design successful form-function relationships. Drawing an analogy with architecture, Aristotle argues that built ‘forms’ should not be explained mechanically with reference to material causes, but instead, teleo- logically with reference to their ‘functional’ purposefulness. According to Hillier and Hanson, this explanation is illegitimate because Aristotle can only establish a causal sequence connecting such functional purposes with built form by applying a metaphysical notion of ‘final causes’: “On

105

this architectural foundation the whole fallacious structure of Aristoteli- an ‘science’ was erected” (Hillier, 1985: 163). Although Durkheim and space syntax appear to converge in concluding that Aristotelian teleology implies an illegitimate metaphysics, they diverge in the solutions they suggest to resolve this problem. While Durkheim deliberately chose to adopt an anti-teleological position, Hilli- er and Hanson’s approach is to legitimize structural-functionalist teleology by means of space morphology, that is, by proposing a socio-physical explanation of the causal sequence that translates societal function into built form. They explain the logic of their space syntax solution in the following way. Having rejected the existence of any Aristotelian ‘meta- structure’ capable of coordinating the evolution of reality by tautological mean, the crux of the matter is to find an equally efficient, but strictly socio-physical, explanation of how morphological space is coordinated into a built form that can reproduce the functional requisites for social solidarity. Essentially, Hillier and Hanson’s solution is to localize a ‘description centre’ located within reality itself rather than in a metaphysical space beyond reality, and yet still capable of ‘programming’ the behavior of individuals in a manner that supports the reproduction of what is functionally required:

The structured information on which the system runs is not carried in the description mechanism but in reality itself in the spatio-temporal world. The programme does not generate reality. Reality generates a programme, one whose description is retrievable, leading to self- reproduction of the system under reasonable stable conditions (Hillier and Hanson, 1984: 44, original italics).

In other words, metaphysical teleology is to be explained in socio- physical terms as a ‘retrievability of descriptions’, which emerges as reality ‘embodies’ its own output. While Aristotelian teleology assumes that reality is ‘dragged’ metaphysically towards a teleological description center, that is, reality2 → description1, Hillier and Hanson (1984; cf. also Hillier and Netto, 2001) replace this logic with a ‘reality sandwich’ schema: reality internalizes and, thus, builds the coordinating information needed in order to reproduce the solidarity requisite into its own time- spatial organization: ‘reality1 → description → reality2’. The question, then, is whether a reality sandwich schema of this kind provides a fruitful solution to the Aristotelian problem of teleology. Ex- amining this, we should initially note that the problem of teleology is but one of two logical illegitimacies of structural-functionalism. As Turner

106

and Maryanski (1979) point out, teleological illegitimacies often appears alongside a similarly illegitimate tautology, which is applied in an attempt to resolve the problem of teleology, and “(…) in which variables are defined in terms of each other, thus making the causes and effects ob- scure and difficult to assess” (Turner and Maryanski, 1979: 124). Hillier and Hanson’s reality sandwich scheme perfectly reassembles this kind of circular reasoning or a statement that contains its own proof. The morphological organization of society meets the integrative need of the society as a whole, while the existence of the same societal whole allows this organization to persist. Such circular logic is often an indication of an illegitimate ex post facto explanation of social life. Tautological explanations of why society functions say nothing more than what is already implied in the initial description of society having a certain social need. Social structures are defined in such a way as to make these needs seem like conceptually necessary parts, rather than explanatory variables, of the societal whole. Instead of providing a logical solution to the problem of teleology, tautological reasoning is only a circular doubling and, thus, a hypostatization of its illegitimate necessity (Merton, 1968; Giddens, 1979). Consequently, the substantive deficiency of this logically illegitimate line of argument is that Hillier and Hanson’s explanation about the way that the functional requisite for social solidarity is tautologically reproduced does not adequately explain why social structures function so as to produce such social need in the first place. The reality sandwich scheme hypostatizes the structural-functionalist assumption that society is an organic whole that is capable of having teleological social needs. Howev- er, in the strict sense, society is not comparable to a reproducing organ- ism given that only humans are capable of having teleological purposes, reasons, and social needs: “Any explanation of social reproduction which imputes teleology to social systems must be declared invalid” (Giddens, 1979: 7 original italics). Moreover, refusing to accept the non-teleological nature of society implies an inadequate understanding of the ways that social life evolves along conflictual and/or contingent trajectories, rather than developing along a teleological stream towards a higher level of functional order and solidarity (Collins, 1994; Merton, 1968; cf. also Durkheim’s reservations as cited earlier). Thus, before asking where the description center has to be located to allow social reality to be tautologically retrievable in a non-metaphysical manner, Hillier and Hanson should have considered the possibility that society, following such non-teleological trajectories, simply evolves without being coordinated by a pervasive and undisputed description center.

107

One confirmation of the fact that we are speaking about a salient problem of space syntax is that, at least indirectly, it is Hiller and Han- son’s tautologically hypostatized teleology that constitutes the object of one of the most frequent criticisms of the space syntax paradigm. That is, neither buildings nor settlements are far from always evolving along a socio-teleological trajectory. As a first example, in relation to the evolution of building layouts, Bafner (2003: §1) argues that that the fact that these are ‘not entirely explainable on the grounds of generic social function’, but rather have to be accounted for as ‘unintended consequences’, “(…) presents a considerable challenge to space syntactical accounts for why buildings take the forms that they do, since the basic premise in such accounts is that sociological factors are the primary forces in determining the internal forms of buildings”. We should add to Bafner’s criticism that such a blind spot regarding the contingent evolution of built space is hardly characteristic of all sociological perspectives (cf., e.g., Williams 2009). In fact, this blind spot should be specifically ascribed to Hillier and Hanson’s structural-functionalist sociology, which assumes that the organization of built space evolves in accordance with a societal intent to reproduce a generic social function. Without this kind of teleological schema, the ‘sociological problem’ identified by Bafner would simply vanish. Second, a similar teleological problem comes up in relation to urban settlements, as argued by Griffiths (2011) and O’Sullivan (2000). While space syntax assumes that the socio-spatial structures of the city are ‘nearly ordered’, not ‘nearly chaotic’ (Hillier, 1997), reflecting that these structural properties are defined in such a way as to make them conceptually necessary parts of the societal whole, this argument can be criti- cized for underestimating that cities are also ‘complex systems’, characterized by their contingent existence on the ‘edge-of-chaos’. In this sense, cities also evolve without being coordinated by any pervasive and undisputed description center that ensures an approximation toward order and functional equilibrium; that is to say, complex(c)ities are also far-from-functional-equilibrium systems, and they develop in a non- coordinated, contingent and possibly conflictual manner. Next, let us examine the related methodological deficiency that is also the result of Hillier and Hanson’s affinity with the macro wing. This deficiency consists of an analytical gap between the micro-methodological resolution of space syntax techniques and the macro-theoretical nature of the Durkheimian concepts applied by Hillier and Hanson. Earlier on, we acknowledged space syntax as a micro-methodological operationalization of Durkheim’s social morphology. In comparison with existing quan-

108

titative techniques measuring space, space syntax works on a ‘finer level of resolution’, allowing it to measure “(…) the ways in which space ‘works’ at the level of patterns of movement” (Hillier, 1996: 140). How- ever, an analytical gap arises at this point, because this micro-methodological approach is not accompanied by a micro-sociological theory that would clarify the reason that variations in micro-scale patterns of movement and co-presence makes such a significant difference for the reproduction of society and solidarity. This situation leaves space syntax wearing a ‘theoretical lens’ that is too macroscopic to take full sociological advantage of the design-level resolution provided by its micro-methodology. This should be ascribed to the fact that Hillier and Hanson’s key morphological notions of movement and co-presence are derived from Durkheim’s original notion of dynamic density, that is, from a notion that is one of the most clear- cut macro-theoretical concepts in Durkheim’s sociology. This macro- sociological perspective is especially evident in Division, in which the central morphological categories of ‘population’, ‘volume’ and ‘density’ are macro-sociological, and thereby detached from the face-to-face sociality that might be characteristic of such morphological phenomena. Durk- heim’s notion of dynamic density provides no micro-sociological explanation of how these phenomena, in themselves, are a morphological means for reproducing social solidarities. Organic solidarity, for example, is specifically a macro-social product of an increasing division of labor in society. Yet, with reference to Durkheim, Hillier and Hanson suggest that the emergence of such solidarities can be adequately explained by micro-variations in the spatially driven patterns of movement and co- presence. In sum, while space syntax offers a refined micro-methodology to explain and predict the where of movement and physical co-presence in relation to the spatial configuration, it fails to provide a micro- sociological explanation of why such morphological variations are capable of producing solidarity.

The Micro-Ritual Logic of Space Summarizing the critical remarks above, I would conclude that Hillier and Hanson’s engagement with Durkheim, was stuck in a dual impasse from the outset. From its affiliation with the macro wing of the Durk- heimian tradition, space syntax inherits the logical illegitimacies of the structural-functionalist notion of functional requisites, and in addition, it becomes removed from a micro-sociological approach that could have allowed space syntax to take full analytical advantage of its micro-

109

methodological findings. I would suggest that this dual problem could best be resolved by means of a closer alliance with the Durkheimian micro wing, given that this tradition is defined by its rejection of the structural- functionalist focus on macro-level requisites and integration in favor of a micro-sociological explanation on why social solidarities are produced in face-to-face situations (Collins 1994; 2004; Goffman, 1956). This shift of perspective is also reflected in the way that proponents of the micro wing downplay the importance of Division, unlike Hillier and Hanson, in favor of Durkheim’s The Elementary Forms:

“(…) the strength of the Durkheimian tradition has been its contribution to micro-sociology, rather than as a theory of the macro- level societal integration or social evolution. Especially in The Ele- mentary Forms, Durkheim provides a model of how solidarity and shared symbols are produced by interaction in small groups” (Col- lins, 2004: 14f, original italics).

Thus, contrary to Hillier and Hanson (1984: 202), who emphasize that “[w]hatever it is, society is not a dance or a ritual,” the central proposition of the Durkheimian micro wing, especially as outlined by Collins’ (2004) contemporary interaction ritual theory, is that society is to be explained as chains of ritual situations, which, among other things, are defined by their dance-like nature! Consequently, Hillier and Hanson forfeit two much-needed micro-sociological metaphors that could have helped them explicate why morphological variation in movement and co-presence might generate social solidarity. The first of these is precisely the dance-like, or as Collins (2004) put it, the ‘rhythmically entrained’ nature of ritual activity, which emphasizes that the hardwired capacity of human bodies to enter into a state of synchronization when co-present is a vital ‘micro-morphological’ condition to explain why interaction rituals produce social solidarity: “Rhythmic synchronization is correlated with solidarity” (Collins, 2004: 76). Seen from a micro-sociological perspective, this is the essential solidarity-generating potential inherent in the sheer fact of individuals moving into each other’s bodily proximity, something that Hillier and Hanson overlook because their notion of co-presence is derived from Durkheim’s macro-morphology. Second, this consideration of micro-ritual also suggests the explanatory limit of physical co-presence. While we may consider the bodily nature of the ritual situation as a necessary micro-morphological condition for ritual activity – because it is human “(…) bodies moving into the same place [that] starts off the ritual process” – Collins (2004: 53) also empha-

110

sizes that this variable is not a sufficient condition to explain sociality and solidarity. This specifically contradicts Hillier and Hanson’s argument that proposed the spatially regulated patterns of movement and co-presence as a sufficient morphological means to reproduce the functional requisite for social solidarities. When we consider Collins’ micro-sociological scheme, we can see that Hillier and Hanson over-emphasize the explanatory importance of the sheer fact of human movement and co- presence, which is only one of the four ritual ingredients necessary for establishing a causal explanation of why such micro-morphological phenomena contribute to the production of social solidarity. Let us sketch out Collins’ multi-variable theory of the interaction ritual, and subsequently consider how it might be linked to the space syntax approach, and thus, how Collins’s paradigm might lend micro-sociological clarity to space syntax theory:

Figure 2. Interaction Rituals. Slightly simplified from Collins (2004: 48).

Let us first look at the ritual ingredients. (1) Two or more people are physically gathered in the same space. (2) The participants in the ritual share a focus of attention on an object or an activity. (3) They share a common mood or feeling. (4) The ritual is demarcated to ensure a sense of who is taking part in the ritual event. Moreover, the ritual ingredients reinforce one another through feedback effects. The reciprocal effect between the joint focus and the mood is especially important, and it proceeds in the form of micro-sequences of rhythmic entrainment among the bodily co-present participants, as described earlier. Depending on the composition of the ritual ingredients, the interaction ritual generates varying degrees of ‘collective effervescence’, which acts as a ‘sort of electric- ity’ (Durkheim, 1995) to intensify the ritual situation. This produces a number of ritual outcomes. (a) An emotional energy (EE) in the individual creates motivation for involvement in the ritual activity. (b) Group solidarity and a feeling of membership. (c) Symbols represent the group and are emotionally arousing. (d) Finally, feelings crystalize into a group- specific moral frame.

111

In attempting to connect space syntax and Collins, a profitable point of departure would be the least macro-theoretically deficient and most micro-methodologically well-founded part of the space syntax paradigm. Hence, we return to Figure 1, summarizing the robust result that spatial configuration confounds the relation between movement and attraction. Considering the full consequence of this model for urban theory, Hillier (1996) elaborates this finding into his theory on cities as ‘movement economies’, proposing that the configuration, at first, generates movement patterns “(…) which then influence land use choices, and these in turn generate multiplier effects on movement with further feed-back on land use choices and the local grid as it adapts itself to more intensive development” (Hillier, 1999: 06.2). Hillier continues to insist on the re- versal of sociological intuition: urban land use attractions are not the explanatory variable for such movement economies, but on the contrary, attractions can be spatially explained with reference to configurational effects on movement. Hence the axiom (Hillier, 1999): ‘configuration generates attraction’. When considered in reference to the analytical gap in the space syntax approach, the analytical strength of the movement economic theory is its micro-methodological explanation of the correlation between configuration and movement, that is, the ability to predict where the movement economic activity intensifies, while the micro-sociological why appears under-examined: Hillier (1996) does not explain the theory of human motivation that must inevitably be assumed to operate in a movement economic notion such as ‘attraction’ or ‘choice’. However, it is far from self-explanatory why individuals would find something motivationally attractive and choice-worthy in the first place. Are we dealing with the rational behavior of a homo oeconomicus, localizing shops where the number of pedestrians optimizes earnings; or, alternatively, with a more irra- tional and non-economic motivation to partake in and simply enjoy the ‘urban buzz’ of densely used spaces? When it leaves these questions un- answered, space syntax is not, as Penn (2001: 11.1) suggests, characterized by the fact that it “(…) assumes anything about individuals or their cognitive capacities,” but rather, that it assumes more about the cognitive nature of human motivation than it provides micro-sociological arguments to support. This deficiency leads us back to Collins (2004), whose interaction ritual theory includes a micro-sociological ‘model of micro-situation cognition’, which suggests that the most fundamental motivational force of individual behavior is search for emotional energy (EE) produced by successful interaction ritual (cf. also Turner 1996). This is particularly

112

relevant for the theory on movement economies, because it is the cognitive component of the ritual situations that gives them a ‘market-like’ character. Individuals are drawn to partake in interaction rituals due to an ‘EE-seeking’ motivation to maximize emotional energy on ‘markets for interaction rituals’: “(…) one seeks EE according to what is immediately attractive, and what is emblematic of past EE payoffs” (Collins, 2004: 174). This micro-sociological model of micro-situational cognition offers a possible (albeit not the only possible) way to explicate the absent why in Hillier’s movement economical theory. The common denominator of human attractions is their capability to arouse emotional energy in the individual. From this perspective, the movement economic axiom suggested by Hillier, ‘configuration generates attraction’, can be specified in micro-sociological terms as describing a process by which the configurational effects on movement and co-presence contribute to the production of more or less EE-attractive and, in this specific sense, choice- worthy market economies for interaction rituals. This reassessment of movement economies as interaction ritual markets can be summarized into a synthesis of Hillier’s Figure 1 and Collins’ Figure 2:

Figure 3. Movement Economies as Interaction Ritual Markets. Author’s creation.

Two simple conceptual linkages make this synthesis between Hillier and Collins’ models possible. As we have just argued, the outcome variable of ‘attraction’ is first explicated in micro-sociological terms as a cognitive component of ‘individual emotional energy’ (EE) as it unfolds in the market-like context of interaction rituals. Second, the ritual variable of ‘physical co-presence’ corresponds to ‘movement’, since co-presence is simply a gathering of human bodies moving into the same space. Again, it should be emphasized that this notion of physical co-presence implies a dance-like rhythmicity occurring among dense bodies, an element that is neglected by Hillier and Hanson’s macro-theoretically diverted and non-ritual approach to this micro-morphological fact. Finally, I should point out that this new theoretical framework does not merely constitute a micro-sociological corrective to the space syntax

113

paradigm; at the same time, it nudges Collins’ notion about the markets for interaction rituals in a more space-morphological direction. This has the important implication that the EE-seeking individuals must share the agential capacity to start off the ritual process with the spatial configuration as it moves human bodies into physical co-presence. Thus densified as humans by human or non-human agency, a more or less attractive market situation arises as a feeling of being ‘where the action is’; “the lure of the ‘bright lights of the city’ is not so much the visual illumination but the minimal excitement of being within a mass of human bodies” (Collins, 2004: 82). The movement economic capacity of space does not determine the market event, but it does contribute to shaping the situational possibility of co-present action and hence the EE-attractiveness of the ritual market economies.

Conclusion Readers familiar with the literature of space morphology might find the sociological reasoning in this article old-fashioned, because the space syntax paradigm has undergone a paradigmatic change of perspective, especially in the last 10 years. Today, its analytical focus is progressively less on the relationship between space and society, as was Hillier and Hanson’s original ambition, but instead, on the nexus between space and the individual. Social structures have been supplanted by cognitive processes. Spatiality is approached as a question of psychology rather than sociology: “(…) the centre of space syntax research starts to shift its emphasis from society as a whole (ie, the origins of space syntax) to a line of enquiry that is firmly focused on individuals and their cognitive processes” (Dalton et al., 2012: 10). Recalling the Durkheimian genealogy of space syntax, is it striking that no one (at least to my knowledge) within the space morphological literature has addressed the ways that this ‘cognitive turn’ might risk dividing space syntax into two incommensurable paradigms. Here, we should recall that the key proposition of Durkheim’s sociology, that society is more than its individual parts, implies an anti-psychological ontology, insisting that social facts have to be explained by preceding social facts, not by individual cognition. “Consequently, every time that a social phenomenon is directly explained by a psychological phenomenon, we may be sure that the explanation is false” (Durkheim, 1966: 104). This is not to say that we simply should accept Durkheim’s dogmatism about sociological explanation, which is most evident in his macro-sociological writings. Rather, the fruitful and forward-looking question is how this

114

growing incommensurability between research programs based in sociology and psychology might be reversed in a manner that maintains space syntax as a coherent body of theory. This possibility provides us with one final reason to advocate the proposed alliance between space syntax and the Durkheimian micro wing tradition. Transcending Durkheim’s sociological dogmatism, Collins’ micro-sociology deliberately includes cognitive processes in its description of human social life: “Interaction ritual is a full-scale social psychology, not only of emotion and situations, but of cognition. Ritual generates symbols; experience in rituals inculcates those symbols in individual mind and memory” (Collins, 2004: 44). The micro-ritual situation is the interface between the structural properties of society and the cognitive processes in the individual, which, as we have already seen, includes the cognitive component of emotional energy. In addition, Collins’ micro wing approach is a social-physiological invitation to include such considerations as how ‘spatial cognition’ might affect ‘wayfinding’ (Dalton et al., 2012) and hence the individual’s EE-seeking effort and ability to (re)discover attractive ritual markets. The Durkheimian micro wing not only offers a possible way to reassess the macro wing deficiencies of the space syntax paradigm, but also opens a still unexplored opportunity to investigate the micro-ritual situation as common Durkheimian ground for Hillier and Hanson’s original society-space scheme and the novel emphasis on individual cognitions within space syntax theory.

References Bafner, S. (2003). The role of corporeal form in architectural thinking. The 4th International Symposium on Space Syntax. London, 17-19 June, www.spacesyntax.net. Collins, R. (2004). Interaction Ritual Chains. Princeton: Princeton University Press. Collins, R. (1994). Four Sociological Traditions. Oxford: Oxford University Press. Dalton, R. C., Hölscher, C., & Turner, A. (2012). Understanding Space: the nascent synthesis of cognition and the syntax of spatial morphologies. Envi- ronment and Planning B: Planning and Design, 39, 7-11. Durkheim, É. (1960). Sociology and Its Scientific Field. In K H Wolff (Eds.), Emile Durkheim, 1858-1917: A Collection of Essays. Columbus: The Ohio State University Press. Durkheim, É. (1964). The Division of Labor in Society (translated G. Simpson). London: Collier-Macmillan. Durkheim, É. (1966). The Rules of Sociological Method. New York: Free Press. Durkheim, É (1984). The Division of Labor in Society (translated W. D. Hall). New York: Free Press.

115

Durkheim, É (1978) Sociology and the Social Sciences. In M. Traugott (Eds.), Emile Durkheim. On Institutional Analysis. Chicago & London: The University of Chicago Press. Durkheim, É (1995). The Elementary Form of Religious Life. New York: Free Press. Durkheim, É (2004). Durkheim’s Philosophy Lectures. Cambridge: Cambridge Uni- versity Press. Durkheim, É (2005a) Review of Antonio Labriola, Essays on the Materialist Conception of History. Thompson, K (red.). Readings from Emile Durkheim. London & New York: Routledge. Durkheim, É (2005b). On Suicide. London: Penguin Press. Durkheim, É, & Mauss, M. (2009). Primitive Classification. Taylor & Francis e- Library. Emirbayer, M. (1996). Useful Durkheim. Sociological Theory, 24, 109-130. Giddens, A. (1979). Central Problems in Social Theory. Hampshire & London: Macmillan Goffman, E. (1956). The Presentation of Self in Everyday Life. Edinburgh: Universi- ty of Edinburgh. Griffiths, S. (2011). Temporality in Hillier and Hanson’s Theory of Spatial De- scription. Journal of Space Syntax, 2(1), 73-96. Halbwachs, M. (1960). Population and Society. Introduction to Social Morphology. New York: Free Press. Hillier, B (1985). The nature of the artificial: the contingent and the necessary in spatial form in architecture. Geoforum, 16(2), 163-178. Hillier, B. (1989). The Architecture of the Urban Object. Ekistics, 334(5), 5-21. Hillier, B. (1996). Space is the Machine. Cambridge: Cambridge University Press. Hillier, B. (1997). The Hidden Geometry of Deformed Grids. Proceedings. The 1st International Space Syntax Symposium. London, April, www.spacesyntax.net. Hillier, B. (1999). Centrality as a Process. Proceedings. The 2nd International Space Syntax Symposium. Brasilia, March, www.spacesyntax.net. Hillier, B. (2005). Between social physics and phenomenology. Proceedings. The 5th International Space Syntax Symposium. Delft, 13-17 June, www.spacesyntax.net. Hillier, B., Leaman, A., Stansall, P., & Bedford, M. (1976). Space Syntax. Envi- ronment and Planning B: Planning and Design, 3(2), 147-185. Hillier, B., & Hanson, J. (1984). The Social Logic of Space Cambridge: University Press. Hillier, B., Penn, A., Hanson J., Grajewski T., & Xu J. (1993). Natural Move- ment: Or, configuration and attraction in urban pedestrian movement. Envi- ronment and Planning B: Planning and Design, 20(1), 29-66. Hillier, B., & Netto, V. (2001). Society seen through the prism of space. Proceed- ings. The 3rd International Space Syntax Symposium. Atlanta, 7-11 May, www.spacesyntax.net. Latour, B. (2005). Reassembling the Social. Oxford: Oxford University Press. Lawson, B. (2001). Language of Space. Oxford: Architectural Press. Lévi-Strauss, C. (1963). Structural Anthropology. New York: Basic Books. Merton, R. K. (1968). Social Theory and Social Structure. New York: FreePress. Mauss, M. (2005). The Nature of Sociology. New York & Oxford: Durkheim Press. Mauss, M. (1979). Seasonal Variations of the Eskimo. A Study in Social Morphology. London: Routledge & Kegan Paul.

116

Netto, V. (2007). Practice, Communication and Space. PhD thesis, The Bartlett, University College London. O’Sullivan, D. (2000). Graph-based cellular automaton models of urban spatial processes. PhD thesis, The Bartlett, University College London. Parsons, T. (1951). The Social System. New York: The Free Press. Penn, A. (2001). Space Syntax and Spatial Cognition. Or, why the axial line?. Proceedings. The 3rd International Space Syntax Symposium. Atlanta, 7-11 May, www.spacesyntax.net. Saunders, P. (1986). Social Theory and the Urban Question. London & New York: Routledge. Schlanger, N. (2006). Introduction. In M. Mauss, Techniques, Technology and Civili- zation. New York & Oxford: Durkheim Press. Seibert, J. (2006). Introduction. In E. C. Robertson, J. D. Seibert, D. C. Fer- nandez, & M. U. Zender (Eds.), Space and Spatial Analysis on Archaeology. Cal- gar: University of Calgary Press. Soja, E. (2001). In Different Spaces. Interpreting the spatial organization of societies. Proceedings. The 3rd International Space Syntax Symposium. Atlanta, 7-11 May, www.spacesyntax.net. Tilly, C. (1981). As Sociology Meets History. New York: Academic books. Tonboe, J. (1993). The Sociology of Space. In J. Tonboe, Rummets Sociologi. Kø- benhavn: Akademisk forlag. Turner, J. H. (1996). The Evolution of Emotions in Humans: A Darwinian– Durkheimian Analysis. Journal for the Theory of Social Behaviour, 26(1), 1-33. Turner, J., & Maryanski, A. (1979). Functionalism. San Francisco: Cummings Publishing Company. Warf, B., Arias, S. (2009). The Spatial Turn: Interdisciplinary Perspectives. London & New York: Routledge. Williams, M. (2009). Social Objects, Causality and Contingent Realism. Journal for the Theory of Social Behaviour, 39(1), 1-18.

117

Kapitel 2 Ambiance of the Machine: Towards a Post-Phenomeno- logy of Urban Spatial Atmospheres

Abstract. The phenomenological notion of atmospheres is increasingly being applied as an aesthetic description of urban space. The article welcomes this notion, but proposes a post-phenomenological reassessment: while phenomenology subscribes to a physiognomic ontology, perceiving atmospheres as aesthetic ecstasies emana- ting from the facades of space, the article suggests a morphological ontology. The atmospheric ‘given’ is given by varying densities of pedestrians, compressed by the urban spatial configuration; atmospheres emerge as a bodily affect of this spatial effect, producing a rhythmic entrainment within the dense body of pedestrians. This morphological corrective to phenomenology combines Bill Hilli- er’s urban morphological paradigm of ‘space as the machine’ and Gilles Deleuze’s philosophy of intensity, which explicates the affective effects of the spatial machine. It is argued that the difficult but pivotal argument for the aesthetic quasi-universalism of the atmospheric phenomenon is better explained within a post-phenomenological framework of this kind.

Keywords: atmospheres; phenomenology; urban morphology; Gilles Deleuze

Colors do not move a people (Deleuze and Guattari, 1987, p. 348).

Introduction ARCHITECTURAL SPACE IS CLOSELY intertwined with the atmosphere of the city. A city as a whole, as well as its smallest spatial parts, is often characterized by a distinct atmosphere. Arriving in a metropolis, entering a neighborhood, emerging on the street level from the depths of the un- derground, or just turning around a street corner is often experienced as entering another urban atmospheric world. This atmospheric character of urban space has been especially highlighted by a phenomenological sensitivity towards the city as it is experienced; indeed, a description of the contemporary city would be half-hearted if it failed to acknowledge this sensitivity (Degen, 2008; Löw, 2008; Böhme, 2006; Albertsen, 1993).

118

Notwithstanding this atmospheric fact, the phenomenological approach to the urban spatial atmospheres comes along with a problematic ontology. This is especially true regarding the assumption of aesthetic quasi-universalism; that is, the provocative hypothesis that everyone feels the atmospheric phenomenon in a more or less similar manner, despite differences in the socio-cultural backgrounds of the affected individuals. This kind of phenomenological ontology fails, however, to stand up to a sociological critique, unveiling aesthetic universalism as social particular- ism (Bourdieu, 2008). While generally agreeing with the premises of this sociological critique, the current article will argue that this critique contradicts the phenomenological argument for an aesthetic quasi-universalism, but does not altogether rule out the possible existence of such universalism. As we shall see, the particular problem of the phenomenological notion of urban spatial atmospheres is its failure to adequately clarify the urban spatial conditions that underlie the atmospheric phenomenon with its quasi-universal gestalt. At first sight, this critical proposition might seem to be a misconcep- tion of the phenomenological oeuvre to the extent that this descriptive approach to lived life might be defined negatively as a refraining from explaining the conditions of the phenomenon. Going ‘back to the things themselves,’ as the dictum of phenomenology prescribes, implies a ‘pure’ description of the phenomenon as it ‘shows itself,’ and, accordingly, a looking away from the conditions that give the ‘given’ phenomenon (Heidegger, 1986, §7). However, the problem is not that phenomenology fails to include spatial conditions in its notion of atmospheres but that it fails to explicate, and thus validate them (Bourdieu, 2008; DeLanda, 2011). That is to say, descriptions are never ‘pure’ and phenomenology always operates with a set of implicit assumptions about where (not) to go searching for the atmospheric in urban space. Does the ambiance emerge from the colors, the architectural scale, or maybe the ornamenta- tion of space? As unavoidable as such conditional questions may be, phenomenology is often reluctant to penetrate behind the ‘pureness’ of the phenomenon; consequently, it too often fails to explain why it specifically is this, rather than that, spatial property from which the atmospheric phenomenon emanates. In the face of these problems, and with the aim of sustaining the assumption of aesthetic quasi-universalism, the post-phenomenological objective of the article is to reassess the spatio-ontological conditions of the atmospheric phenomenon. The article is divided into four sections. First, the article examines the problematic nature of the physiognomic ontology used by phenomenology. Second, Hillier’s urban morphological

119

paradigm of space as the machine is suggested as a post-phenomenological path to the urban spatial atmospheres. Third, accounting for the fact that Hillier’s urban morphology appears more functionalist than vitalist, the article suggests that Deleuze’s philosophy of intensity allows us to explicate the atmospheric phenomenon as an affective effect of Hillier’s spatial machine. Fourth, the quasi-universal nature of the atmospheres is attributed to the bodily-rhythmic shape of these affective effects, unfolding partially autonomous from social mediation and, thus, beyond the reach of sociological particularization.

Path and Method Urban space makes itself felt as atmospheres. This thesis recapitulates the phenomenological argument as it is summarized, for example, by the phenomenologically inspired sociologist, Martina Löw (2008, p. 44): “Spaces develop their own potentiality which can influence feelings. This potentiality of spaces I call ‘atmosphere’.” The most important sources of inspiration for such phenomenological reasoning are Heidegger’s (1986, §29) proposition that being in space always-already implies a ‘Stimmung,’ understood as a certain mood or emotional attunement, and Böhme’s (1993; 2006) closely associated existential phenomenology of ‘aesthetic atmospheres’. Without making a sharp distinction between these affiliated phenomenologies (cf. also Norbert-Schulz, 1991), we can illustrate the phenomenological approach to atmospheres with a simply example taken from Böhme: a plain blue cup (Böhme, 1993, p. 121). Located in time and space, the blue cup makes its presence percepti- ble. The ‘blueness’ of the cup is not restricted to the cup; rather, this blueness should be understood as something that ‘radiates’ from the character of the cup, giving the spatial environment a certain atmospheric haze of blueness. A thing, as Böhme (1993, p. 121) puts it, ‘is tuned as ecstasies,’ and aesthetic atmospheres are, hence, to be understood as ‘tuned spaces’ (Böhme, 2006, p. 16). In more general terms, we are dealing in this case with a physiognomic ontology by which the blue cup (or for that matter, any object, architectural facade, or constellation of spaces and humans) has an inherent character and, thus, potential for expression, which is experienced atmospherically due to its aesthetic ecstasies. That is to say, phenomenology is the art of judging the atmospheric character of the urban space as derived from its aesthetic ‘face-value’ features and ecstasies. In attempting to track down the problem inherent in such a physiognomy, we should initially pay attention to its tautological character: argu-

120

ing that the blue ‘blues’ is a double enunciation of one and the same. However, this is no coincidence, as Böhme is following in Heidegger’s phenomenological path, when he highlighted that “(…) tautological thinking is the primordial sense of phenomenology. (…) To understand this, we need to learn to distinguish between path and method.” (Heideg- ger, 2003, p. 80, original italics). Phenomenology must follow and locate the phenomenon along a ‘path,’ which should remain ‘methodically’ non- validated by analytic procedures. That is to say, one should follow the path and tautologically insist on the phenomenon in its face-value givenness, beyond mediation or validation (Heidegger, 1986, p. 25). Posi- tioned anti-methodologically in this way, the tautological formula of the blueness of the blue corresponds exactly to Heidegger’s (2005) enigmatic endeavor to let the phenomenon speak for itself: space spaces, the thing things, language speaks. However, the fact that tautologies are logically true per definition should not be confused as the real truth of the description (Wittgenstein, 1961, §4.461, §4.462). Due to this tautological reasoning, phenomenology is, as Kwan (2005, p. 330) puts it, “(...) extremely poor in its ability to build up concrete references to worldly affairs and phenomena.” De- scription loops into iteration and thus moves away from the thing; the argument that the atmosphere of ‘blueness’ emanates from the blue character of the cup becomes a pure claim. Is it, really, the blue cup that ‘blues,’ or is it, rather, the wooden table on which the cup is standing, the dim light in the room, the facial expression of the other person co- present, or simply something entirely different? An additional reason to address these critical questions are the high theoretical stakes in phenomenology’s physiognomic reasoning. The absence of concrete references to worldly affairs and phenomena to validate which spatial properties that produce the given phenomenon makes it difficult, in turn, to give an comprehensive account for why this spatial (every)thing has the capacity to evoke a quasi-universal aesthetic sensation. And yet, this is precisely the claim made by phenomenology: it assumes that the atmospheric aesthetic has a quasi-universal character, evoking ‘spatially discharged, quasi-objective feelings’ (Böhme, 2006, p. 16) with a capacity to grab hold of an individual, partially independent of socio-cultural schemes or personal states of feeling. Although phenomenologists do acknowledge that societal and interpretative processes can reframe the aesthetic experience, the quasi-universal nature of the atmosphere should be regarded as its ontological hallmark (Löw, 2008, p. 45; Borch, 2009, p. 234).

121

The sociological critique of physiognomic aesthetics and its quasi-universal ambitions has been ruthless – and in large part, justified. Here, Bourdieu’s sociology (2008, p. 73) may be hit home most tellingly due to its wealth of concrete references to socio-worldly affairs and phenomena that cast doubt on such aesthetic universalism. Examined from this sociological perspective, the inherent and universal character of the physiognomy is simply false, reflecting a ‘scholastic fallacy’ that universalizes the particular, thereby neglecting the relational conditions of the societal world that, taken as a whole, make the phenomenon possible. From the sociological perspective, the ‘universal’ (or, what amounts to the same thing, ‘pure,’ ‘authentic, ‘immediate,’ ‘face-value’) character of the atmospheres is simply a distinctive judgment of taste: a canonization of the aesthetic preferences of a particular social class into a ‘universal’ norm. For the sake of clarity, we can decompose Bourdieu’s critique into two dimensions of his praxeological key concept of ‘habitus’: the aesthetic particularization occurs at the level of movement and at the level of perception. First, the Bourdieuian perspective would question whether people from different social classes move into, co-exist, and thus share a potentially universal atmosphere in and of the same urban space. Such ‘physical spaces’ of the city are appropriated as ‘social spaces,’ and particular classes and class-specific preferences and aversions dominate them. Social groups tend to move into, through and gather in space homologous with their habitual dispositions, reflecting an aversion to those socio-urban spaces appropriated by different social classes: “(…) socially distanced people find nothing more intolerable than physical proximity (experienced as promiscuity).” (Bourdieu, 2003, p. 128). This is how the social structures are inscribed into the bodies, ‘via moves and movement of the body’ (Bourdieu, 1996, p. 16), reproducing a habitual (dis)taste for certain urban spatial ambiances. In sum, before the aesthetic experience takes place in and of a certain urban space, the particularization of the ‘universal’ would thus have occurred at the level of the habitual moves and movement of the body. Second, even assuming that different social groups, despite this pattern of habitual behavior, happened to be co-present in same urban space, Bourdieu would insist on the particularity of their experiences at the level of perception. All aesthetic perceptions, including the atmospheric sensation of urban space, are subject to a habitual logic of social valuation: the idea of casting a ‘fresh eye’ on the ‘face-value’ character of urban space is a physiognomic fallacy. What is assumed inherent is always-already embedded in a relational social space of positions. Aesthetic

122

judgments are, in essence, a way of positioning one’s class-specific habitus: aesthetic perception is social valuation (Bourdieu, 2008). If phenomenology is to defend the quasi-universalist ontology, it has to overcome these two counterarguments. Hence, the post-phenomenological (rather than anti-phenomenological) hope in this article is to address these two counterarguments (Lea 2009, p. 374). That is to say, the problem is not phenomenology as such, but rather its problematic ontology, and our hope is to use a post-phenomenologic approach as a way to get beyond this difficulty. The first question that comes up as we reassess the spatio-ontological possibility of quasi-universalism is why phenomenology chose the problematic Heideggerian ‘path,’ instead of the ‘method,’ as a way to approach urban spatial atmospheres.

Physiognomy and Morphology Although Heidegger (2005) preferred his forest paths, die Holzwege, phenomenology urges us to walk along the urban paths, describing how the street, sidewalk, or town square is passed, dwelled and felt by human beings. In this horizontal and lived perspective, the urban dweller moves through and along the urban paths, encircled by the towering facades of built space, presenting their inherent character as aesthetic ecstasies: the ‘despondently’ modernistic, the ‘monumentally’ sacred, the ‘romantically’ ornamented. “The surface-level expressions (…) provide unmediated access to the fundamental substance of the state of things.” (Kracauer, 1995, p. 75). Thus, perceived from this phenomenological path-perspective, the physiognomy – the manner by which the architectural facades tower up along the urban path – seems to be the very manifestation of the atmospheric (Bech 1998). Following Seamon’s (2007) and Hillier’s (2005) call to combine phenomenology and morphology leads us to the possibility that a post- phenomenological ‘method’ might call attention to other sources of the atmospheric than these physiognomic surface-level ecstasies. To explore this possibility, we must temporarily abandon the horizontal world of the path-perspective in favor of a vertical view from above. Yet, this is what phenomenology finds most objectionable; it would diagnose this shift in perspective as a modernist perversion where an observer soars high above the city to objectify, enjoy, and exploit the quality of lived life. Recall de Certeau’s famous critique of the voyeuristic gaze from the tow- ers above New York: “His elevation transfigures him into a voyeur. It puts him at a distance. It transforms the bewitching world by which one was ‘possessed’ into a text that lies before one’s eyes. It allows one to

123

read it, to be a solar Eye, looking down like a god.” (de Certeau, 1988, p. 92). However, we should not let de Certeau’s phenomenologically inspired aversion for heights annihilate the post-phenomenological insight that de Certeau makes, contrary to his explicit intentions: when the city is viewed from above, it can be read as an ‘urban text’ that the pedestrians “(...) write without being able to read it. These practitioners make use of spaces that cannot be seen.” (de Certeau, 1988, p. 92). Voyeurism or not, this makes us aware of a crucial property of space that extends beyond its physiognomic properties: the pedestrians are phenomenologically una- ware of the syntax of space, even though it directly shapes their urban practices. Paradoxically, de Certeau’s critique highlights the limits of the path-perspective. Urban space is also an emergent syntax structure, which is more than the sum of urban pedestrian experiences. This urban syntax text cannot be experienced in its network-totality and, thus, cannot be analyzed with a pure phenomenological description. The syntax of space has to be mapped with a method – from above. This insight implies a post-phenomenological change of perspective. The spatio-ontological matter that matters is not the physiognomic forms of space – the aesthetic facade, face-value, and surface-level ecstasies of the built environment – but rather the morphological forms of space, understood as ‘the empty volumes of space’ produced by the as- semblage of built objects into a network pattern (Hillier and Hanson, 1984, p. 1). Above and beyond the physiognomy of the space, urban life is morphologically embedded in a spatial configuration whose ‘emergent structures’ entail a ‘real agency’ on urban practices (Hillier, 2005, p. 13). Urban life is subject to a spatial grammar of the city; and the space syntax approach developed by Hillier and Hanson (1984) is a specifically rigor- ous method-perspective that maps this syntactical regularity. With de Certeau’s understanding of the ‘textual’ character of urban space in mind, it is no coincidence that Hillier and Hanson (1984) denote their approach as space syntax. Their focus is on the underlying grammat- ical order of space, analogous to a linguistic mapping of the underlying grammar of language. However, space syntax is a dedicated non-discursive approach: its groundbreaking empirical insight consists of proving how the movement economy of the city is primarily a probabilistic function of the agential forces of the spatial configuration (Hillier, 1996, p. 111). This configurational mediation of movement is the fundamental way that space affects social life. Accordingly, the vague but widespread notion in the social sciences and humanities that built form has an inherent (and in our context physiognomic) potential to affect social life in

124

an unmediated manner “(…) simply does not exist.” (Hillier, 1996, p. 294). “How can a material object, like a building impinge directly on human behavior? Even so, it is exactly this that we are expected to believe if we abandon spatial configuration as the intervening variable.” (Hillier, 1996, p. 293). Essentially, this is the blind spot of the physiognomic approach to urban spatial atmospheres. In the face of a contentious configurational proposition of this kind, phenomenology might object that physiognomy, as it is employed by phenomenologists, includes not merely the character of the built facades, but also the ‘constellations’ between built environment and persons present in space in a manner not entirely different from the concept of configurations. As Böhme argues, aesthetic atmospheres “(…) are spaces insofar as they are ‘tinctured’ through the presence of things, of persons or environmental constellations, that is, through their ecstasies.” (Böh- me, 1993, p. 121). However, rather than serving to clarify the manner by which space contributes to the emergence of atmospheres, this notion of constellations is actually a telling example of the kind of thinking that Hillier is questioning. Recapitulating the flaw in phenomenology, Böhme never explicates the specific properties of the built environment that contribute to these constellations. As a consequence, the issue of how these constellations orchestrate an unmediated relationship between space and people remains unexplained. The phenomenological notion of constellation turns out to be yet another face of the physiognomic ontology, and Hillier offers us the appropriate response: we should be thinking in terms of configurations rather than constellations. Hence, the fact that ‘space is the machine’ (Hillier, 1996) provides us with a strong counterargument against Bourdieu’s first particularization of the quasi-universal, that is, the sociological argument that the homology between physical spaces and social spaces is embodied as a result of habitually orchestrated moves and movements. What is lacking in Bour- dieu and in urban sociology in general (Gans, 1968), is an understanding that the spatial machine is not merely subsumed to a habitual logic of social appropriation, where the behavior of the social classes is explained in reference to socially stratified attraction-preferences. As it is exposed to the spatial machine, bodily movement is conditioned by the network agency of space itself, not only by the existence of social and class- specific attractions. In this way, the spatial machine can challenge the boundaries of social space, reassembling what the social machinery divides into homologous territories: “Space may not be structured to correspond to social groups, and by implication to separate them, but on the contrary to create en-

125

counters among those whom the structures of social categories divide from each other.” (Hanson and Hillier, 1987, p. 265). Contrary to Bour- dieu’s insistence of the ‘homology’ (or in Hanson and Hillier’s words, ‘territorial correspondence’) between social and physical spaces, urban space also works due to a ‘structured non-correspondence’ (Hanson and Hillier, 1987, p. 251). That is, there is a spatial relationship between co- present urban dwellers produced by space-mechanical means, going beyond the logic of habitual identities and homologous territories – as hypothesized by Bourdieu. The spatial machine is a ‘mixing mechanism’ that physically compresses and transforms ‘trans-spatial’ into ‘spatial’ groupings, thereby allowing them to share a co-present experience in and of the same urban physical space. In considering Bourdieu’s second criticism, we should ask how these considerations about urban spatial morphology are related to aesthetic sensation. Would it be possible to argue in favor of a morphological aesthetics by which the urban atmospheric phenomenon is ascribed to the mediating agency of spatial configuration, rather than to the ecstasies of the physiognomic facades? Although Hillier (1996, p. 119) suggests that “(…) many of the properties of urban space that we value aesthetically are a product of this functional shaping of space,” the space syntax paradigm has not fleshed out the idea of a morphological aesthetics. Howev- er, in the light of the critique cited above, it seems clearly appropriate that an inquiry of this kind should elaborate the potential aesthetic affect emerging from the spatio-mechanical compression of pedestrian bodies. In other words, if urban space conditions the emergence of quasi-universal atmospheric sensations, a post-phenomenological approach would have to ascribe them to differences in co-presence and bodily density, compressed by the agential forces of the spatial configuration. In the following section, I will suggest that Deleuze’s vitalist philosophy of intensity may be used to provide a theoretical link to explicate this affective dimension of bodily density, compressed by the spatial machine.

Quality and Quantity What initially justifies a linkage between Hillier and Deleuze is the fact that they each address problems of Aristotle’s classical functionalism by means of a density ontology. Hillier’s (1996, p. 295) emphasis on co- present density mediated by configuration is an explicit attempt to resolve the architectural ‘form-function’ question, originating from Aristo- tle’s suggestion that successful form-functions relationships found in nature should essentially be perceived as a ‘design problem,’ answered by

126

making an analogy with architecture (Hillier, 1996, p. 296). However, Aristotle had great difficulties accounting for the trajectory of this functional design-process: Aristotle’s ‘teleology’ and notion of an ‘unmoved mover’ should both be abandoned as illegitimate metaphysics. To resolve this Aristotelian problem, Hillier suggests a socio-physical solution “(…) that the form-function relation in architecture, and the relation of space to society, is mediated by the spatial configuration.” That is to say, the configurational ‘form,’ which is designed purposefully by society and is thus anything but unmoved, is related to ‘functions’ by virtue of its capacity as a ‘mover’ of co-present densities, a process that serves to fulfill certain socio-functional needs of society (Hillier and Hanson, 1984, p. 18). While Hillier re-establishes Aristotle’s functionalism in reference to the ways that density is subject to the network agency of the spatial configuration, Deleuze addresses the ontology of density with a view of rejecting Aristotelian functionalism (DeLanda, 2011, p. 26). Instead of applying density as an ontological element that permits the seamless integration of form and function into an organic whole, Deleuze invokes density as part of a vitalist alternative to all functionalism of this kind: “the essential thing is no longer forms and matters (…) but forces, densities, intensities.” (Deleuze and Guatteri, 1987, p. 342). In this light, I would suggest that Deleuze’s understanding of density as an intensive force offers an important corrective to Hillier’s functionalist undertones and non-affective notion of density. Density concerns the affectively felt, rather than functionally integrated, fact of becoming human. In other words, we should plug Hillier’s spatial machine into Deleuze’s ‘becoming-machine,’ whose force fields of affective flows are the ontological key for understanding the emergence of the urban spatial atmosphere (Deleuze and Guattari, 1987). What remains to be clarified is how the notion of density – a quantitative property – might help explain the emergence of phenomenological atmospheres, which are presented as qualitative experiences. To examine this interconnection of quantity and quality we have to penetrate deeper into Deleuze’s understanding on how quantitative differences of forces, densities, and intensities constitute the hidden ground for the emergence of the qualitatively given. Deleuze formulates the ontology of the inner relationship between quantity and the emergence of qualitative phenomena as follows:

Difference is not diversity. Diversity is given, but difference is that by which the given is given, that by which the given is given as di-

127

verse. Difference is not phenomenon but the noumenon closest to the phenomenon (…). Everything which happens and everything which appears is correlated with orders of differences: differences of level, temperature, pressure, tension, potential, difference of intensity (Deleuze, 1994, p. 222, original italics).

The qualitative diversity, which in this case denotes the phenomenological experience as it is ‘given’ (‘phenomenon’), has to be distinguished from the quantitative difference of intensity (‘noumenon’), which constitutes the hidden ground that gives the given. The post-phenomenological nature of Deleuze’s oeuvre is evident: reality is not dependent on the intentionality of consciousness or the occurrences of being. Rather, reality is morphogenetically produced by the real and virtual intensities of reality “(…) revealing [the] properly qualitative content of quantity” (Deleuze, 1994, p. 222), independent of any phenomenological idealism (DeLanda, 2011). At first hand, to perceive such quantities as productive might seem to be an odd kind of metaphysics. Here, one should not neglect the physical undercurrent of Deleuzian thought: elaborating the difference of intensity, Deleuze’s argument parallels the notion of emergent properties of complex physical systems, particularly as they are described in far-from- equilibrium thermodynamics. As DeLanda (2006, p. 82) has pointed out, Deleuze was explicitly inspired by thermodynamics (Deleuze, 1994, pp. 223-235). Hence, what Deleuze is addressing is how a quantitative intensification or densification in a given system might reach a certain threshold that forces it to undergo a ‘phase transition,’ producing a new set of qualitative properties of the system – on the chaotic edge of further phase transitions. Simply put, more is different (Anderson, 1972). As argued by, among others, Urry (2005), the analysis of physical and social systems converge in such complexity processes. Moreover, it is widely acknowledged that cities are complex systems with emergent properties and regular phase transitions, and thus, similar to thermodynamic systems (Portugali, 2011). Here, I will argue that Deleuze’s anticipation of this contemporary ‘complexity turn’ (Urry, 2005) allows us to address the complexity of the urban aesthetic experience, which, is not merely a given quality, but instead, is given by a quantitative phase transition. That is to say, physical and urban atmospheres share a similar thermodynamic complexity. Let us illustrate this suggestion with an example borrowed from literature, an art form that sometimes has a unique ability to ‘open up’ the complexity of everyday experience, in this case the complex interconnection between quantity and quality (Harrington,

128

2002). The opening passage from Robert Musil’s novel, The Man Without Qualities, offers a brilliant post-phenomenological revelation about the thermodynamics of the urban atmosphere:

A barometric low hung over the Atlantic. It moved eastward toward a high-pressure area over Russia without as yet showing any inclination to bypass this high in a northerly direction. (…) The water vapor in the air was at its maximal state of tension, while the humidity was minimal. In a word that characterizes the facts fairly accurately, even if it is a bit old-fashioned: It was a fine day in Au- gust 1913. Cars shot out of narrow, deep streets into the shallowness of bright squares. Pedestrian darkness formed cloudy strings. Where thick lines of speed transgressed their loose haste, they thickened, then trickled faster and after just a few vibrations assumed their former even pulse. (…) This noise, the peculiarity of which cannot be described, is all one needs, even after years of absence and with closed eyes, to recognize that one is in the royal capital city of Vi- enna. Cities, like people, can be recognized by their walk. Opening his eyes, he would know the place by the rhythm of movement in the streets long before he caught any characteristic detail (Musil 1996, p. 1).

The first paragraph sets the thermodynamic scene. The quantitative differences of intensity – the differences of ‘level, temperature, pressure, tension, potential’ as Deleuze described it above – are the quantitative noumenon that produces weather change and, in this sense, generate the qualitative phenomenon: ‘It was a fine day in August 1913.’ The phenomenological precedence accorded to the qualitative is replaced by an asymmetrical relationship between quality and quantity: the quantitative difference that conditions the atmospheric ‘vanishes’ in the atmospheric quality. Deleuze (1994, p. 223) explicates this crucial asymmetry using the thermodynamic distinction between ‘quantitative intensity’ and ‘qualitative extensity’ (DeLanda, 2006). The quantitative-intensive change of pressure produces the atmospheric cloud formation with a particular extension (drawing in the sky, etc.) and filled with a certain quality (dark, rainy, etc.), thereby constituting the phenomenon experienced as a ‘fine day in August’. Although the atmospheric cloud is given by differences of pressure intensity, the only phenomenological manifestation is the qualitative extension of a cloud. The catalyst of the atmosphere is, to express it with a notion borrowed from Jameson’s (1973) dialectics, a

129

‘vanishing mediator’; the quantitative-intensive producer – neglected by phenomenology – is ‘cancelled’ or ‘hidden’ in its qualitative-extensive product:

It is cancelled in so far as it is drawn outside itself, in extensity and in the quality which fills that extensity. However, difference creates both this extensity and this quality (…). Difference of intensity is cancelled or tends to be cancelled in this system, but creates this system by explicating itself (Deleuze, 1994, p. 228, original italics).

In the second paragraph, the thermodynamics explicate the emergence the urban atmosphere. Metropolitan life is a pulsating flow of densities, compressed by the configurational properties of ‘deep streets’ and the ‘shallowness of bright squares’. In effect, Musil is approaching the city from above, viewing it as a configurational network that gives rise to the unique atmospheric character of the city; that is, “(…) like a boiling bub- ble inside a pot made up of the durable stuff of buildings (…),” as Musil (1996, p. 2) puts it in a consecutive passage. It is important to note that the urban atmospheric character is not reflected in any ‘characteristic details’ of the physiognomic facades of the city; rather, the atmospheric phenomenon is created morphologically by the spatially densified and in- tensified movements of the city: “Cities, like people, can be recognized by their walk.” Whereas physiognomic aesthetics presents itself in terms of visual ecstasies, quantitative intensity is an embodied affect through which the ‘the rhythm of movement in the streets’ makes the body pul- sate, whether your eyes are open or closed. In sum, cities are characterized by their rhythm of movements. Configurationally compressed and thermodynamically potent, the rhythm of movement is the quantitative intensity that gives the urban atmosphere its phenomenological quality and extensity.

Emotion and Affect Our next concern will be the implications of linking Hillier’s urban morphology and Deleuze’s philosophy of intensity for Bourdieu’s second particularization of universal aesthetics. Here, we should note that the urban spatial literature include other, similarly post-phenomenological, inquiries about the quasi-universal nature of the atmospheric. Brennan’s (2004) socio-biological affect theory has been extraordinarily influential in stimulating and substantiating such inquiries (Thrift 2008, p. 16, 221; Borch, 2009, 233; Anderson, 2009). Those who adopt Brennan’s ap-

130

proach attribute the ‘near-universalism’ (Rice, 2008, p. 202) of the atmospheric to the presence of airborne hormones, which “(…) is critical in how we ‘feel the atmosphere’ (…) I also suggest that hormonal interactions account for how the hormonal processes situate people in different as well as similar emotional places (…).” (Brennan, 2004, p. 9). However, it seems reasonable to question the generalizability of this atmosphere-hormone thesis across different spatial scales. Maybe indoor spaces might allow such hormone substances to concentrate sufficiently to have an affect (Kotler, 1973), but this idea seems less compelling for outdoor urban spaces, which literally – and due to their natural flow of air and bodies – are spaces with another level of ‘ventilation’ (Sloterdijk, 2006). To continue our consideration of the intensive quantities of the urban movement rhythm, we might instead concur with another explanation of the atmospheres that Brennan mentions only in passing: “Rhythm also has a unifying, regulating role in affective exchange between two or more people.” (Brennan, 2004, p. 70). Moreover, this notion of rhythms is compatible with a Deleuzian framework, which emphasizes how the bodily affects of the quantitative intensities have a rhythmic shape (McCor- mack, 2008; Deleuze and Guattari, 1987, p. 311). The existence of such an affective ‘rhythm of movement in the streets’ (Musil, 1996, p. 1), pulsating through the spatially densified bodies, should not be understood metaphorically. Rather, it refers to the socio-physiological fact that co- present bodies have a hard-wired capacity to enter into a state of fine- grained rhythmic entrainment, partially beyond conscious awareness and social regulation (Marsh et al., 2009): “This is rhythmic synchronization at a level much more fine-grained than the 0.2-second segments of which humans can be consciously aware.” (Collins, 2004, p. 77). Literally speaking, co-present bodies have the capacity to communicate non-discursively, that is, affectively, by means of rhythmic synchronization (Collins, 2004, p. 65-78; 2009, p. 26). In this sense, rhythmic synchronization among densified bodies can be understood as a quantitative condition for the qualitative arousal of a shared, and in this sense, atmospheric affect. In addition, the emergence of such rhythmic entrainment follows the Deleuzian thermodynamics of the urban spatial atmosphere: as the movement-intensity between co- present bodies increases, rhythmic synchronization is involuntary produced through a sudden and non-linear phase transition (Schmidt et al., 1990; Portugali, 2011, p. 130). This transition is spatially conditioned: transposed to Hillier’s spatial machine, the emergence of these affective rhythms is contingent on the network agency of the configuration, which

131

compresses density, thereby creating the bodily proximity that starts the rhythmic entrainment. Rhythmically, it is the spatial effect of the machine that redeems the bodily “(...) capacity for affecting and being affected” (Deleuze, 1988, p. 123); and the urban spatial atmosphere is simply the phenomenological experience of this rhythmic relation between bodies and spaces. Here, it is important to distinguish between the social nature of emotions and the ‘asocial’ affects of the bodily rhythm (Massumi, 2002, p. 30). Whereas emotions, as is well documented by the sociology of emotions, are identical with their social mediation, affects are characterized by some degree of bodily autonomy from social structures, significations and codings. Compared to emotions, about which Bourdieu is right to emphasize their habitual appropriation, “(…) affect is unqualified. As such, it is not ownable or recognizable and is thus resistant to critique.” (Massumi, 2002, p. 28). Embodied affects might be articulated as a socially appropriated emotion, but this possibility does not erase the affective ground of the atmosphere, as McCormack (2008, p. 1827) elegantly summarizes: “Affect is a kind of vague yet intense atmosphere; feeling is that atmosphere felt in a body; and emotion is that felt intensity articulated as an emotion.” Let us suggest two further examples of the affective rhythmicity of the atmospheric phenomenon. First, the post-phenomenological perspective of this article resembles Jacobs’ (1993) iconic notion of the ‘sidewalk ballet,’ which, should be understood as Jacobs’ way to describe the emergence of a ‘good’ urban atmosphere. According to Jacobs (1993, p. 65), this ‘complex order’ “(…) is all composed by movement and change, and although it is life, not art, we may fancifully call it the art form of the city and liken it to the dance (…) an intricate ballet in which the individual dancers and ensembles all have distinctive parts which mi- raculously reinforce each other and compose an orderly whole.” Jacobs perceives the city and the sidewalk ballet as systems of ‘organized complexity’ (Jacobs, 1993, p. 558), which correspond to the patterns of thermodynamic emergence that we presented earlier. Consequently, Jacobs’ emphasis on how large urban changes can emerge as a result of a very small ‘quantitative unaverage’ (Jacobs, 1993, p. 574), is analogous to the Deleuzian argument we have presented that suggests that the difference of quantitative intensities might pass a critical, and in this sense ‘unaverage,’ threshold that produces a new qualitative order of the system. In the case of the sidewalk ballet, this quantitative ‘unaverage’ is made up of the movement and changing densities of the pedestrians along the side-

132

walk, thus constituting the quantitative difference that gives the atmospherically given quality. Furthermore, Jacobs goes on to explain this urban complexity as being spatially conditioned by the existence of ‘short blocks’ (Jacobs, 1993, p. 233) that stimulate pedestrian movement, diversity, and density. In essence, the configurational approach of space syntax is a radicalization of this proto-configurational argument (Seamon, 2007). What qualifies such considerations to become a morphological aesthetics, an ‘art form of the city,’ is the way that the urban flow of pedestrians resembles a ballet; that is to say, the dense body of pedestrian dancers entails an affective rhythmicity. As McCormack emphasizes, in a Deleuzian sense, ‘dance’ is the example par excellence of “(…) the rhythmic relations between bodies and spaces.” (McCormack 2008, p. 1829). Second, while Jacobs’ image of the sidewalk ballet exemplifies the emergences of a ‘good’ atmospheric order, this is evidently not always the case: space also holds the morphological capacity to ‘tune’ the bodily rhythm of the sidewalk ‘out of step,’ producing an uncomfortable, turbu- lent, and atonal intensity. The complex order of ‘polyrhythmia’ balances at the chaotic edge of ‘arrhythmia’ (Lefebvre, 2004). Here, we should recall Simmel’s (1997) classic description of how urbanism, when compared to rural life, has a ‘rhythm of life’ that quantitatively intensifies urban life to the point of over-stimulation. As Simmel (1997, p. 175, original italics), again in a proto-Deleuzian manner, formulates it, the “human being is a ‘creature of difference’ with a vulnerability towards the (...) intensification of nervous stimulation, which results from the swift and uninterrupted change of outer and inner stimuli.” Thus, the rhythm of the city is capable of over-affecting the ‘nervous system of the individual,’ thereby becoming the quantitative ground for a metropolitan atmosphere at the unpleasant edge of the physically overwrought. To summarize, in mapping the affective rhythm of the urban spatial atmosphere the emphasis should be on the ontological absence of social ‘structure’ and presence of embodied ‘events,’ which hold the capacity to affect the involved bodies in an immediate manner (Massumi, 2002, p. 27). The bodily rhythm affects the co-present individuals partially autonomous from awareness or social regulation, and thus, free from Bour- dieu’s sociological particularization. Unmediated, and in this sense, quasi- universal, the rhythm affects the feeling body as a sort of affective contagion, which proceeds by ‘‘(…) vibrations, rotations, whirlings, gravita- tions, dances or leaps which directly touch the mind.’’ (Deleuze, 1994, p. 8; cf. furthermore Thrift, 2008, p. 177). Being in the spatial machine is becoming affected by a ‘metropolitan rhythm of events’ (Simmel, 1997,

133

p. 179), which is the post-phenomenological path to the quasi-universal: “The correlation between universal and event is fundamental.” (Badiou, 2009, p. 31). The ambiance of the machine is, in essence, an expression of a ‘singular universality’ – a felt absolute – by which any-body becomes rhythmically affected (Massumi, 1997, p. 748; DeLanda, 2011, p. 29).

Conclusion Phenomenology is unavoidable, especially when it comes to the urban space, which is, indeed, a felt absolute – an atmosphere. However, the article has identified a number of problems stemming from a self- contained phenomenology. Initially, we critically addressed the physiognomic ontology of phenomenology and placed it in question by invoking Bourdieu’s particularization of a quasi-universal aesthetics on the level of movement and perception. As a way of dealing with this problem, the article went on to suggest a post-phenomenological reassessment that would examine morphology rather than physiognomy as the spatial condition of urban atmospheres. The article first suggested that Hillier’s spatial machine, due to its logic of structured non-correspondence, holds the agential capacity to move and mix different social classes across their homological social spaces. Thus, mediated by the spatial configuration, the habitually (in)different individuals become co-present, and this lets them share an experience of the same urban space. Second, it was argued that these differences of density – compressed by the configuration – could be elaborated into a morphological aesthetics of the urban spatial atmospheres. Here, a connection between Hillier and Deleuze was suggested in order to explain the affective intensities of spatially compressed density differences. De- scribed in terms of an urban thermodynamic complexity, the quantitative intensity constitutes the hidden ground of the qualitative atmosphere. The quasi-universal potential of this reassessment specifically consists in the bodily-rhythmic nature of the affective intensity: rhythmic entrainment among affective bodies unfolds autonomously from social mediation and, thus, beyond the reach of Bourdieu’s sociological particularization.

References Albertsen, N. (1999). Urbane atmosfærer. Sociologi i dag, 4, 5-29. Anderson, P. W. (1972). More Is Different. Science, 177(4047), 393-396. Anderson, B. (2009). Affective Atmospheres. Emotion, Space and Society, 2(2), 77- 81.

134

Badiou, A. (2009). Thinking the Event. In A. Badiou, & S. Zizek, Philosophy in the Present (pp. 1-48). Cambridge & Malden: Polity. Bech, H. (1998). Citysex. Representing Lust in Public. Theory, Culture & Society, 15(3), 215-214. Borch, C. (2009). Organizational Atmospheres: Foam, Affect and Architecture. Organization, 17(2), 223-241. Bourdieu, P. (1996). Physical Space, Social Space and Habitus. Rapport 10. Insti- tutt for sosiologi og samfunnsgeografi. Oslo: Universitetet i Oslo. Bourdieu, P. (2003). The Weight of the World. Palo Alto: Stanford University Press. Bourdieu, P. (2008). Pascalian Meditations. Cambridge: Polity. Brennan, T. (2004). The Transmission of Affect. New York: Cornell University Press. Böhme, G. (1993). Atmosphere as the Fundamental Concepts of a new Aes- thetics. Thesis Eleven, 36, 113-126. Böhme, G. (2006). Architektur und Atmosphäre. Munich: Wilhelm Fink Verlag. Collins, R. (2004). Interaction Ritual Chains. Princeton: Princeton University Press. Collins, R. (2009). Violence: A Micro-Sociological Theory. Princeton: Princeton Uni- versity Press. de Certeau, M.. (1988). The Practice of Everyday Life. Berkeley: University of Cali- fornia Press. Degen, M. (2008). Sensing Cities. London & New York: Routledge. DeLanda, M. (2006). Space: Extensive and Intensive, Actual and Virtual. In: Ian Buchanan, & Gregg Lambert (Eds.), Deleuze and Space (pp. 80-88). Edin- burgh: Edinburgh University Press. DeLanda, M. (2011). New Philosophy of Society. New York: Continuum. Deleuze, G., & F. Guattari. (1987). A thousand plateaus. Minnesota: University of Minnesota Press. Deleuze, G. (1988). Spinoza. San Francisco: City Lights Books. Deleuze, G. (1994). Difference and Repetition. New York: Columbia University Press. Gans, H. J. (1968). People and Plans: Essays on Urban Problems and Solutions. New York: Basic Books. Harrington, A. (2002). Knowing the social world through literature: Sociologi- cal reflections on Robert Musil’s The Man Without Qualities. International Jour- nal of Social Research Methodology, 5(1), 51-59. Heidegger, M. (1986). Sein und Zeit. Tübingen: Max Niemeyer Verlag. Heidegger, M. (2003). Four Seminars. Bloomington: Indiana University Press. Heidegger, M. (2005). Off the Beaten Track. Cambridge: Cambridge University Press. Hillier, B., & J. Hanson. (1984). The Social Logic of Space. Cambridge: Cambridge University Press. Hanson, J, & B. Hillier. 1987. The Architecture of Community. Architecture and Behaviour, 3(3), 251-273. Hillier, B. (1996). Space is the Machine. London: Space Syntax. Hillier, B. (2005). Between social physics and phenomenology. Proceedings of the 5th International Space Syntax Symposium. Delft. Jacobs, J. (1993). The Death and Life of Great American Cities. New York: The Modern Library.

135

Jameson, F. (1973). The Vanishing Mediator: Narrative Structure in Max We- ber. New German Critique, 1, 52-89. Kotler, P. (1973). Atmospherics as a Marketing Tool. Journal of Retailing, 49, 48- 64. Kracauer, S. (1995). The Mass Ornament. Cambridge: Harvard University Press. Kwan, T.-W. (2005). Hegelian and Heideggerian Tautologies. Analecta Husserli- ana, 88(III), 317-336. Lea, J. (2009). Post-Phenomenology/Post-Phenomenological Geogra-phies. In R. Kitchin & N. Thrift (Eds.), International Encyclopedia of Human Geography. Oxford: Elsevier. Lefebvre, H. (2004). Rhythmanalysis. New York: Continuum. Löw, M. (2008). The Constitution of Space Through the Simultaneity of Effect and Perception. European Journal of Social Theory, 11(25), 25-49. McCormack, D. P. (2008). Geographies for Moving Bodies: Thinking, Dancing, Spaces. Geography Compass, 2/6, 1822-1836. Marsh, K. L., M. J. Richardson, & R. C. Schmidt (2009). Social Connection Through Joint Action and Interpersonal Coordination. Topics in Cognitive Sci- ences, 1, 320-339. Massumi, B. (1997). Deleuze, Guattari, and the Philosophy of Expression. Ca- nadian Review of Comparative Literature, 24(3), 745-782. Massumi, B. (2002). Parables for the Virtual. Durham: Duke University Press. Musil, R. (1996). The Man Without Qualities Vol. 1. New York: Vintage. Norbert-Schulz, C. (1991). Genius Loci. Towards a Phenomenology of Architecture. New York: Rizzoli. Portugali, J. (2011). Complexity, Cognition, and the City. New York: Springer. Rice, J. E. (2008). The New “New”: Making a Case for Critical Affect Studies. Quarterly Journal of Speech, 94(2), 200-212. Schmidt, R. C., C. Carello, & M. T. Turvey (1990). Phase Transitions and Criti- cal Fluctuations in the Visual Coordination of Rhythmic Movements Be- tween People. Journal of Experimental Psychology, 16(2), 227-247. Seamon, D. (2007). A Lived Hermetic of People and Place. Phenomenology and Space Syntax. Proceedings for the 6th International Space Syntax Symposium. Is- tanbul. Simmel, G. (1997). The Metropolis and Mental Life. In D. Frisby, & M. Feath- erstone (Eds.), Simmel of Culture: Selected Writings. Thousand Oaks: Sage Pub- lications. Sloterdijk, P. (2006). War on latency: on some relations between surrealism and terror. Radical Philosophy, 137, 14-19. Thrift, N. (2008). Non-Representational Theory. Space, Politics, Affect. London & New York: Routledge. Urry, J. (2005). The Complexity Turn. Theory, Culture, and Society, 22, 1-14. Wittgenstein, L. (1961). Tractatus Logico-Philosophicus. New York: Humanities Press.

136

Kapitel 3 Phenomenology of the Movement Economy: A Multi- level Analysis

Abstract. The theory of cities as movement economies constitutes a theoretical cornerstone of the space syntax paradigm, and the urban morphology literature offers considerable empirical evidence to support the key proposition of the theory about the urban economy: spatial accessibility correlates with economic land use values. However, the theory does not concern only the emergence of such spatio-economic attractions, but also provides an intriguing explanation of the city as lived or experienced attractions, that is, of the kinds of well-being and excitement often associated with urban life at its best. This ‘phenomenological’ rather than ‘economic’ aspect of the theory has received only limited attention in the literature. This article addresses the gap in the literature, and reports the results of a multilevel regression model, which combines individual survey data on urban attraction experiences and morphological data on the movement economy context of these experiences, measured with space syntax tools. After controlling for the effect of individual-level variables, the model suggests that the movement economy has a significant contextual effect on the experienced level of urban attraction.

Keywords: cities as movement economies; space syntax; phenomenology; urban sociology; spatial regression modeling

Introduction Bill Hillier’s movement economic theory (1996a; 1996b) should be considered the most important contribution toward making the space syntax approach a self-reliant ‘paradigm’. This theory defines the nuts and bolts that make up the framework for this kind of space morphological ‘puzzle solving’ (Kuhn 1962). Hence, the original publication of the theory was awarded the prestigious AESOP Prize for the best paper published in European planning journals in 1996, and the recently published Urban Design Reader (Tiesdell and Carmona 2007) includes this paper as the paradigmatic example of what the space syntax approach has to offer. In addition, within the space syntax community, a considerable number of studies have refined and accumulated a growing stock of puzzle solutions to the theory. This would also include this paper, which proposes an empirically based puzzle solution to the ‘phenomenological aspect’ of

137

the movement economic theory, an element that has evaded analytical attention in the literature when compared to the theory’s specifically economic propositions. The article is divided into four sections. First, the movement economy theory is outlined and reassessed as providing an explanation for the phenomenologically experienced attraction value of urban space. Second, the article presents its space statistical design and data base, which consists of survey and movement count data collected in the Nørrebro neighborhood of Copenhagen in 2011. Third, the article presents a movement economic analysis and statistical estimation of experienced attraction as a function of movement economic micro- contexts. Finally, the analysis is concluded, and the paper considers the implications of the results for space syntax as a normative architectural theory of the good city.

The Movement Economy Theory In essence, the movement economic theory consists of three urban morphological components, ‘configuration’, ‘movement’ and ‘attraction’, which are inter-related through a spiral of multiplier effects (Hillier 1996a; Hillier et al. 1993; Hillier 2008). The key proposition of the theory that is provocative for the field of sociology is that these reciprocal effects are first initiated by the spatial configuration itself, thereby assigning a secondary role to the ‘social production’ of the city (Lefebvre 1991). This morphological argument is advanced against the backdrop of the empirically well-founded ‘principle of natural movement’, suggesting that spatial configuration has a strong independent effect on movement patterns (Hiller et al. 1993; Hillier 1996a: 120; Peponis, Ross and Rashid 1997). Hillier recapitulates these urban morphological considerations in the following manner:

The ‘movement economy’ theory built on this [principle of natural movement], and proposed that evolving space organisation in settlements first generates movement patterns, which then influence land use choices, and these in turn generate multiplier effects on movement with further feed-back on land use choices and the local grid as it adapts itself to more intensive development (Hillier 1999: 06.2).

In other words, configurationally accessible street segments have a systematic tendency to be movement intensive, and such spatially driven movement patterns are key elements in making local street segments attractive for land use. Subsequently, a positive spiral is set off, by which

138

the land use attractions attract more movement, thereby intensifying the value and ability to attract even more movement, and so forth. Hence, the movement economic axiom: ‘configuration generates attraction’ (Hillier 2002: 154). As emphasized by Hillier (1996: 125), this process is most clearly illustrated by the spatial distribution of retail land use. Shops tend to cluster in street segments where the spatial configuration has already allocated enough movement to support a profitable customer base. Given this paradigmatic example, it is not surprising that the theory has been mainly applied to explain how spatial configuration, in itself, generates economic land use values. Accordingly, the literature includes extensive evidence for this spatio-economic aspect and predictive element of the theory. Studies have documented that spatial accessibility is correlated with spatial distribution of retail commerce (Hillier 1999; Chiaradia et al. 2009; Ortiz-Chao and Hillier 2007), population and employment density (Wang et al. 2011), property and housing values (Matthews and Turnbull 2007; Topçu 2009; Narvaez, Penn and Griffiths 2012; Rui-lan and Xingi 2004), office and retail rental patterns (Enström and Netzell 2008) and land use density of office buildings (Kim and Sohn 2002).

Economic and Phenomenological Attractions However, this impressive body of evidence does, not exhaust the explanatory relevance of the theory. Besides suggesting a spatial nature of economic attraction values, the theory also offers an intriguing explanation of the spatial mechanisms that create urban spaces rich in eventful or lived attractions. Despite its movement economic name, the theory also pertains to spatial configuration as a driving force behind such phenomena as the ‘urban buzz’, ‘pleasure’ and ‘experience’ of the city (Hillier 1996). That is, the theory also applies to other phenomena that are related to, but not identical with, economic attraction values of the urban barter market. In a nutshell, the fact that rental values of neighboring houses are more expensive does not, in itself, contribute to the lived attraction of an urban space! So we might say that the theory involves two different notions of ‘attraction’, one economic and one phenomenological, which share the spatially driven movement as their driving force. Spatially facilitated movement flows simultaneously enhance retail economic profits and the urban experience of being on a street rich with impressions, crowd buzz, and excitement. The movement economically intensive space often gives one a sense of being ‘where the action is’, to borrow an expression from the micro-sociologist Randall Collins (2004: 82). Even though Hillier

139

does not spell out this ‘economy-phenomenology’ distinction, it seems nonetheless to be implicitly operative within the framework of the theory (Hillier 2005). Take, for example, Hillier’s argument that the space- movement correlation both constitutes the basis of economic “(…) patterns of land use and building densities (…) and give rise to the sense that everything is working together to create the special kinds of well- being and excitement that we associate with cities at their best.” (Hillier 1996a: 113-114, italics added). That is to say, configuration generates economic and phenomenological attraction. In placing emphasis on the phenomenological aspect of the theory of cities as movement economies, this article is inspired by the phenome- nologist David Seamon (2007; 1994; 2012) who has discussed and developed the link between space syntax theory and phenomenology on several occasions. With reference to movement economy theory, Seamon points out that “(…) space syntax offers a striking picture of how the physical world – specifically, its configurational qualities – contributes to place experiences and place making, particular the discovery that relative place vitality of an urban district of city appears to be grounded in particular path structures.” (Seamon 2007: iii-10). However, in connection to this observation, Seamon (2007: iii-11) also emphasizes that this intriguing aspect of the theory lacks empirical examination and he explicitly proposes the need for phenomenologically inspired space syntax studies that explore the ‘lived aspects’ of physical design. In sum, compared with the extensive evidence regarding the economic aspects of the movement economy, the urban morphological literature contains only a few articles that note the phenomenological aspects of the theory, and never in an empirical manner (Turner 2005; Lu and Zhou 2013). To my knowledge, this article is the first attempt to directly test the phenomenological expectations of the theory by means of space syntax and space statistical methods. It should be added that the literature includes a number of studies that provide indirect evidence for the theory, not least Jacobs’ (1961) classical analysis of how blossoming urban life is spatially conditioned by the existence of short blocks that facilitate dynamic movement flows (Seamon 2012). To the extent that this article’s statistical analysis speaks in favor of the phenomenological propositions of movement economic theory, this result should also be regarded as empirical proof of Jacobs’ contemporary relevance.

Statistical Hypothesis Hillier’s (1996: 125) argument that the movement economic city is constituted by a ’multiplicity of inter-relationships’ between the morphologi-

140

cal components of configuration, movement and attraction, presents a methodological challenge for testing the theory statistically. Instead of measuring and modeling the independent effects of each these components, that is, abstracted from their morphological inter-relationships, we should instead examine specifically how the urban experience is conditioned by the inter-relationship between these morphological components. Thus, the movement economic hypothesis of the article is that the urban morphological inter-relationships between configurational accessibility, movement intensity and clustered shops contribute positively to the phenomenological experience of being in an attractive urban place.

Data and Methodology The quantitative design of the article is as follows. The inter-relatedness of the morphological components is operationalized by combining space syntax measures of configurational accessibility, pedestrian movement data and systematic on-site observations of land use. The correlation between these morphological items is assessed with regression analysis; moreover, polychoric principal component analysis is applied to extract a movement economic index variable from the latent interrelationship between the items. The phenomenological aspect of the movement economy, i.e. the outcome variable, is then operationalized with survey data. Data is analyzed applying a multilevel regression model that includes the movement economic index as the key context-level predictor and a number of individual-level control covariates.

Spatial Data A key methodological achievement of space syntax is the articulation of space as a network of relations that can be objectively measured by mathematical tools of graph theory (Hillier and Raford 2010). When applying the space syntax method, the first step is to construct an axial map, which represents the fewest and longest lines that are needed to cover all the ways of moving around the layout and reach all spaces. For this article, the axial map was constructed against the backdrop of high- resolution aerial photographs of central Copenhagen. Next, the axial map is used to calculate different measures of configurational accessibility. The article focuses on the configurational measure of segment angular choice. This choice measure may be understood as a spatial equivalent to Freeman’s (1977) graph theoretical measure, ‘betweenness centrality’, commonly applied in social network analysis. A growing body of space syntax studies suggests that the choice measure outperforms the

141

other and older graph theoretical measures (e.g. axial integration) as a predictor of pedestrian movement patterns and economic land use values (Hillier and Iida 2005; Hillier and Raford 2010). Taking the ‘segment’ rather than the ‘axial line’ as the smallest topo- logical unit of analysis, calculating choice involves decomposing axial lines into the segments between junctions, and choice is then calculated as the shortest paths between all segments within a given metric radius; that is, the journey with the lowest angular ‘cost’ for each possible origin and destination pair of segments. The fact that actual movement patterns and choice are strongly correlated in urban morphologies suggests that movement through the configurational network tends to densify in segments that lie on ‘shortest’ paths between pairs of segments, i.e. the bridges that link different areas of the street network together (Turner 2001; Hillier, Yang and Turner 2012). Technically, the choice measure is calculated in Mapinfo, applying the place syntax tool (Ståhle, Marcus and Karlström 2005). Following Grajeski’s (2001) methodological suggestions, movement data was collected at 65 gates to validate whether choice is correlated with actual movement patterns in the Copenhagen case. Each gate was simultaneously observed twice for 5 minutes each time. The gates were located in the same segments in which the survey data was collected. Inspired by Raudenbush and Sampson’s (1999) suggestion to apply on- site systematic observations as a tool to quantify the ecological setting of urban neighborhoods, the 65 gate locations were then geocoded as mainly retail ‘foreground’ spaces, or as residential ‘background’ spaces (Hillier 2012). Table 1 summarizes the measures of three morphological variables (choice, movement, and land use), and the context- and individual- level predictors and the dependent variable to be included in the multilevel model:

Table 1. Descriptive statistics Meassures Min. Max. Mean Lived attraction (1 = agree) 0 1 0.70 Movement economy index -3.24 2.71 -0.01 Age 11 90 37 Gender (1 = male) 0 1 0.46 Ethnicity (1 = Non-Danish) 0 1 0.19 Choicea 10.5 134162.5 22785.7 Movement per houra 8 848 243 Land use (1 = retail) 0 1 0.35 a The analysis uses the logged transformed versions of these variables due to skewness.

142

Measuring the Movement Economy As noted, the movement economy should be understood as an interrelated ‘whole’, which is more than the additive sum of its morphological component ‘parts’. Interrelationships among socio-spatial phenomena are often difficult to measure directly. This suggests that the movement economy can be measured more accurately indirectly, as a ‘latent’ structure existing between the directly observable morphology components (Dunteman 1989). For this purpose, the three morphological items (choice, movement and land use) are analyzed by polychoric principal component analysis, allowing us to extract the latent movement economic structure shared by the items (calculations are conducted using Stata’s polychoric module; cf. Kolenikov and Angeles 2004). The extracted latent variable is considered a movement economic index, and this index variable is included as the key context-level predictor in the multilevel regression model. Moreover, the application of principal component analysis has an advantage in terms of statistical estimation: a simultaneous inclusion of the three morphological items as separate model predictors would, given their interrelatedness, cause problems of multicollinearity by which the overlapping effects of the items would be erroneously estimated and/or controlled away.

Survey and Sampling The survey data (N=412) was collected in central Copenhagen, specifically the neighborhood of Nørrebro. This is an ethnically diverse neighborhood with a varied social morphology where only a few streets separate some of Copenhagen’s most popular areas from the most socio- spatially deprived. Data was collected on two weekdays in the daytime in the fall of 2011. Interviews were conducted on-site as face-to-face interviews with randomly selected pedestrians in 26 of the 65 aforementioned streets in which the movement gate counts also were conducted. Thus, the data sample should be considered a geographical cluster sample as the respondents were selected in the capacity of moving through one of the 26 street segments. The response rate is 61%. Due to this sampling strategy, the limited sample size, and the fact that the neighborhood of Nørrebro has a socio-morphological ‘variation’ greater that the Copen- hagen ‘average’ (e.g. in terms of ethnic composition), the sample cannot be regarded as strictly representative for Copenhagen and/or cities as such; more modestly considered, the estimated results should be read as indicating trends regarding the nature of movement economies.

143

Analytic Strategy Given the hierarchical data structure (i.e. individuals nested into street contexts), data is analyzed with a multilevel regression model (Hox 2010; Raudenbush and Sampson 1999). This type of generalized linear regression model is developed to model how social (or, in this case, socio- spatial) contexts affect individual outcomes. Contrary to traditional regression models that assume independency of observations (uncorrelated errors), multilevel models relax this assumption by modeling observations as being dependent due to their contextual embeddedness. In the specific case of spatially clustered social phenomena, multilevel models can be understood as one way to statistically address Tobler’s (1970: 236) first law of geography: “Everything is related to everything else, but near things are more related than distant things.” In multilevel regression terms, the proposition of the article – that the movement economy shapes the lived attraction value – is modeled as a context effect on the individual attraction outcome. The article applies a random intercept model, allowing the intercept to vary across contexts; that is to say, the random intercept model examines the extent to which the experienced attraction value varies across street contexts and how the movement economic properties of these street contexts affect the individual attraction outcome. Technically, the model is estimated in MLwiN with a Markov chain Monte Carlo (MCMC) procedure (Browne 2012), which is preferable over an ordinary least squares and/or maximum likelihood type of estimation: a MCMC estimation is more robust with smaller sample sizes, as in the case of this article. Moreover, the MCMC procedure relaxes the assumption of the normality distribution of the outcome variable, allowing us to estimate a conceptually simpler linear model while applying the binary attraction outcome (Hox 2010: 273; Hellevik 2009).33

Measures and Definitions The dependent variable is measured according to agreement with the statement: ‘This street is an exciting place’. This variable measures the phenomenological attraction value of urban micro-spaces, described above by Hillier as a sense of ‘well-being and excitement’. The variable has a binary outcome, ‘agree’ (coded 1) and ‘disagree’ (coded 0). The movement economy index is included as the model’s key context-level predictor variable. The technical aspects of calculating this

33 A parallel set of analyses has been performed applying a logistic multilevel model. This does not alter the main estimated results or conclusions.

144

index variable are outlined in the results section below. In terms of theory, a higher index score indicates that the street segment is more movement economically potent. In addition, the model includes four individual-level control variables in order to separate the movement economic context effect from the effect of social composition (Duncan, Jones and Moon 1998). For example, it could be the case that some streets are perceived more attractive not because of the context properties of these streets but because these segments are crowded with ‘bright minded’ young individuals (Gans 1968; Hedström 2005: 45ff). The control variables included are: age, gender (reference category: male) and a dummy variable that measures whether the respondent has a Danish or non-Danish ethnic background (reference category: Danish ethnicity). The ethnicity dummy is included in the expectation that respondents who are ‘ethnically homologous’ with Nørrebro’s ethnically diverse population express a more favorable opinion about the emplaced attraction value of the neighborhood (Collins 2004: 151; Bourdieu 1985).

Results

Movement Economic Analysis Figure 1 presents the segment angular choice analysis. The grey tone spectrum from dark to bright corresponds to a low and high level of spatial accessibility. The neighborhood of Nørrebro is seen to the West; and to the East is the old city center, separated from Nørrebro by the lakes running across the morphological heart of Copenhagen. Given the fact that the choice measure often picks up the ‘bridges’ that link the street network together, we should not be surprised that the analysis suggests that the actual bridges connecting the city center and Nørrebro are among the most spatially accessible segments. From these bridge segments the angular ‘cost’, i.e. the minimum number of turns around corners that are needed in order to reach all other segments, is very low.

145

Figure 1. Nørrebro section of the segment angular choice map

The next question is whether this choice representation of Nørrebro may be read as a sort of movement economic ‘hot spot’ map of spatially distributed movements and attractions. As a first examination of this question, linear regression models are applied to test choice as a predictor of actual pedestrian movement and retail land use patterns. The regression models include a spatial lag of the dependent variables in order to control for the effect of spatial autocorrelation (Ward and Gleditsch 2008). The relevance of applying this ‘spatial regression model’ can be explained by the fact that a Moran’s I statistics of movement, I = 0.20, p < .001, and land use patterns, I = 0.22, p < .001, suggests moderate levels of spatial autocorrelation. Technically, the Moran’s I statistics and the spatial lagged covariates are calculated with the software package GeoDa (Anselin 2003). Results are reported as semi partial coefficients in order to obtain a comparable effect size metric. In accordance with the well-established findings in the space syntax literature, the first regression model suggests that a global choice measure (radius 2,500 meters) is a strong predictor of pedestrian movement, sr = 0.67, p < .001. The spatial lag of pedestrian movement is also highly significant, sr = 0.38, p <.001, suggesting the existence of spatial spillover of movement flows between proximate street segments. Such spillover effect may arise because the same movement flows (i.e. the same individual pedestrians) are measured several times at neighboring movement count gates. Overall, this regression model suggests that a considerable proportion of the variance in the dependent movement variable, R2 =

146

0.58, can be explained by the two independent variables, choice and the spatial lag of movement. The next regression model suggests that choice is also significantly correlated with retail land use, sr = 0.37, p < .001. This corresponds with the expectations of space syntax regarding movement economy. The spatial lag of retail patterns is also highly significant, sr = 0.391, p <.001, suggesting that the presence of shops on one street segment lends economic value to the shops in neighboring street segments. When compared to the first regression model, the second model explains a significantly smaller share of variance in the dependent land use variable, R2=0.29. This r-squared difference may be attributed to the fact that in this instance, we are estimating a derived or byproduct effect of the space-movement correlation. Following these regression results, we use a polychoric principal component analysis to extract the joint variation of the three morphological items into a single movement economic index component. This analysis suggests that only one component has an eigenvalue above Kai- ser’s criterion of 1, explaining 74% of the variation of the three items. The component loadings for the choice (0.60), movement (0.57) and land use (0.65) items indicate strong coherence with the extracted component. These results strongly suggest that the morphological items share a latent or inter-related structure, thereby justifying the inclusion of the extracted component in the multilevel model as a movement economic index variable.

Multilevel Estimation We calculated an intraclass correlation (ICC) in order to assess the proportion of the variance of the individual outcome which may be attributable to the context, i.e. how similar are the attraction experiences within the street context. In space statistical terms, an intraclass correlation is simply a measure of spatial autocorrelation (Kreft and Leeuw 1998: 7). Calculated on the basis of the empty Model 0, the ICC suggests that approximately 10.4% of the variation in the dependent experience variable is due to variation between the street contexts. This supports the use of a multilevel design to examine whether this between-context variance may be explained as a context effect of the movement economy. Note that the Bayesian Information Criterion (BIC) suggests that Model 2 has the best overall data fit.

147

Table 2. Multilevel regression analysis of experienced attraction M0 M1 M2 Intercept 0.699** 0.661** 0.677** Fixed effects Non-Danish ethnicity (ref: Danish ethnicity) 0.154** 0.144** Age 0.001 0.001 Gender (ref: male) -0.026 -0.034 Movement economy index 0.119** Random effects Intercept variance component 0.023** 0.022** 0.006 BIC 505.96 496.00 486.89 N 410 401 401 *p<0.10; **p<0.05. All model estimates are obtained with 30,000 MCMC iterations.

In Model 1 it can be initially established that the individual-level variables age and gender do not contribute significantly to the experienced attraction level. Second, we note that respondents with a non-Danish ethnicity tend to regard the urban space as significantly more attractive than ethnic Danish respondents. This finding should be understood in the context of the fact that Nørrebro is often portrayed in Danish media as a neighborhood marked by social problems related to ethnicity. The fact that respondents who are ethnically homologous with the neighborhood have a more positive opinion may be explained as a result of their effort to portray a good image of the neighborhood and its ethnic diversity. In other words, they may be exercising what we, with Goffman (1955), might call a sort a neighborhood ‘face-work’ (cf. also Jensen and Christensen 2012). Regarding the article’s key hypothesis, we can establish that the individual-level control variables included in Model 1 do not account for the majority of the between-context variation. Calculating the proportional reduction of between-context variance suggests that only approximately 5% is accounted for by individual-level variables. In Model 2, the movement economic index is included in as an attempt to account for this unexplained between-context variance. The movement economic index predictor contributes significantly and positively to the experience of urban space as being more attractive. To further highlight this finding, the intercept variance component becomes insignificant once this context-level predictor is included, indicating that no significant context- level variation has been left unconsidered. This result clearly speaks in favor of the key hypothesis of the article: holding all other variables con-

148

stant, the movement economy is positively associated with the experience of urban space as being an attractive place.

Discussion and Concluding Remarks The article took as its point of departure the argument that the movement economic axiom that ‘configuration generates attraction’ relates to both economic land use values and the lived experience of urban space attractions. This phenomenological, as opposed to a strictly economic, aspect of the theory is implied by Hillier’s original presentation of the theory; yet, this aspect of theory has evaded empirical examination. The results presented in the article generally confirm that phenomenologically experienced urban attraction is positively shaped by the movement economy. This effect is more than the sum of individual characteristics. The attraction value of urban space should also be considered as a context effect of the movement economic city. More broadly, the article may be read as a space syntax contribution to the ongoing discussion about the frequent discrepancy between architects’ and laypersons’ aesthetic valuation of built space. Although the literature contains some studies, e.g. Llinares et al. (2011), where these aesthetical valuations converge, Gifford et al. (2000: 167) emphasize that in general, architects and non-architects differ in their aesthetic assessment of the urban built environment. Attraction excitement is understood in this context as part of what defines aesthetic appraisals (Mitias 1988; 146; Gifford et al. 2000: 169). Does the current study illuminate whether the aesthetic valuation of urban space implied by the architecture theory of space syntax is mismatched with that of the layperson? Here, we should note that Hillier’s movement economy theory is one of the few instances when he (and, more broadly, the space syntax paradigm) explicitly addresses urban space as an aesthetic phenomenon: “(…) many of the properties of urban space that we value aesthetically are the product of this functional shaping of space.” (Hillier 1996: 199). The movement economy theory implies what we might call an ‘aesthetic of urban morphology’, reflecting a descriptive as well as normative approach to urban space: “Good space is used space.” (Hillier 1996: 127). Although this article does not test whether the aesthetic assessment of urban space by the interviewed non-architects converges with that of architects trained in space syntax, (as a repetition of Giffords et al.’s 2000 study would require), it is noteworthy that what the respondents’ value as being aesthetically attractive not only coincides with the explanatory ex-

149

pectation of movement economy theory, but also with the theory’s aesthetic-normative idea of the good city.

References Anselin, L. (2003). GeoDA 0.9 User’s Guide. Urbana-Champaign, Spatial Anal- ysis Laboratory, University of Illinos. Abbott, A. (2001). Time Matters: On Theory and Method. Chicago: University of Chicago Press. Bourdieu, P. (1985). The social space and the genesis of groups. Theory and Socie- ty, 14(6), 723-744. Browne, W. J. (2012). MCMC estimation in MLwiN. Version 2.25. Centre for Multilevel Modelling. University of Bristol. Collins, R. (2004). Interaction Ritual Chains. Princeton: Princeton University Press. Chiaradia, A., Hillier, B., Schwander, C., & Wedderburn, M. (2009). Spatial centrality, economic vitality/viability: compositional and spatial effects in Greater London. Proceedings of the 7th International Space Syntax Sympo- sium, Stockholm: KTH. Duncan, C., Jones, K., & Moon, G. (1998). Context, composition and heterogeneity: using multilevel models in health research. Social Science & Medicine, 46(1), 97-117. Dunteman, G. H. (1989). Principal Components Analysis. London: Sage. Enström, R., & Netzell, O. (2008). Can space syntax help us in understanding the intraurban office rent pattern? Accessibility and rents in downtown Stockholm. The Journal of Real Estate Finance and Economics, 36(3), 289-305. Freeman, L. (1977). A set of measures of centrality based on betweenness. Soci- ometry 40(1), 35–41. Gans, H. J. (1968). People and Plans: Essays on Urban Problems and Solutions. New York: Basic Books. Goffman, E. (1955). On face-work. An analysis of ritual elements in social interaction. Journal for the study of Interpersonal Processes, 18, 213-231. Gifford, R., Hine, D. W., Muller-Clemm, W., D’Arcy, J. R., & Shaw, K. T. (2000). Decoding Modern Architecture. A Lens Model Approach for Un- derstanding the Aesthetic Differences of Architects and Laypersons. Envi- ronment and Behavior, 32(2), 163-187. Grajewski T. (2001). Space Syntax Observation Manual. Bartlett School of Graduate Studies, University College London. Hedström, P. (2005). Dissecting the Social: On the Principles of Analytical Sociology. Cambridge: Cambridge University Press. Hellevik, O. (2009). Linear versus logistic regression when the dependent variable is a dichotomy. Quality & Quantity, 43(1), 59-74. Hillier, B., Penn, A., Hanson, J., Grajewski, T., & Xu, J. (1993). Natural Move- ment-or, configuration and attraction in urban pedestrian movement. Envi- ronment and Planning B: Planning and Design, 20(1), 29-66. Hillier, B. (1996a). Space is the Machine. Cambridge: Cambridge University Press. Hillier, B. (1996b). Cities as movement economies. Urban Design International, 1(1), 41-60.

150

Hillier, B. (1999). Centrality as a process: accounting for attraction inequalities in deformed grids. Urban Design International, 4(3-4), 107-127. Hillier, B. (2002). A theory of the city as object: or, how spatial laws mediate the social construction of urban space. Urban Design International, 7(3), 153- 179. Hillier B (2005). Between social physics and phenomenology: notes towards and urban synthesis. Proceedings of the 5th International Space Syntax Symposium, TU Delft. Hillier, B. (2012). The Genetic Code for Cities: Is It Simpler than We Think? In J. Portugali, H. Meyer, S. Egbert, & E. Tan. (Eds.), Complexity Theories of Cit- ies Have Come of Age. Berlin Heidelberg: Springer. Hillier, B., & Iida, S. (2005). Network and psychological effects in urban movement. Spatial Information Theory, 3693, 475-490. Hillier, B. (2008). Space and spatiality: what the built environment needs from social theory. Building Research & Information, 36(3), 216-230. Hillier, B., Yang, T., & Turner, A. (2012). Normalising least angle choice in Depthmap-and how it opens up new perspectives on the global and local analysis of city space. The Journal of Space Syntax, 3(2), 155-193. Hillier, B., & Raford, N. (2010). Description and discovery in socio-spatial analysis: the case of space syntax. In G. Walford, E. Tucker, & M. Viswanathan (Eds.), The SAGE Handbook of Measurement. London: Sage. Hox, J. (2010). Multilevel Analysis: Techniques and Applications, Routledge. Jacobs, J. (1961).The Death and Life of Great American Cities. Random House Digi- tal. Jensen, S. Q., & Christensen, A. D. (2012). Territorial stigmatization and local belonging: A study of the Danish neighbourhood Aalborg East. City, 16(1- 2), 74-92. Johnson, S. D., & Bowers, K. J. (2010). Permeability and burglary risk: are cul- de-sacs safer?. Journal of Quantitative Criminology, 26(1), 89-111. Kim, H. K., & Sohn, D. W. (2002). An analysis of the relationship between land use density of office buildings and urban street configuration: case studies of two areas in Seoul by space syntax analysis. Cities, 19(6), 409-418. Kolenikov, S. & Angeles, G. (2004). The use of discrete data in PCA: Theory, simulations, and applications to socioeconomic indices, Working paper WP- 04-85, Carolina Population Center, University of North Carolina. Kuhn, T. (1962). The Structure of Scientific Revolutions, Chicago: University of Chi- cago Press Kreft, G. G. & Leeuw, J. d (1998). Introducing Multilevel Modeling. London: Sage. Lefebvre, H. (1991). The Production of Space. Oxford: Blackwell. Llinares, C., Montañana, A., & Navarro, E. (2011). Differences in Architects and Nonarchitects’ Perception of Urban Design: An Application of Kansei Engineering Techniques. Urban Studies Research, Article ID 736307. Lu, Z., & Zhou, X. (2013). The Interpretation of Place Phenomenology Based on Space Syntax Theory. Advanced Materials Research, 664, 422-428. Matthews, J. W., & Turnbull, G. K. (2007). Neighborhood street layout and property value: the interaction of accessibility and land use mix. The Journal of Real Estate Finance and Economics, 35(2), 111-141. Mitias, M. H. (1988). What Makes an Experience Aesthetic?. Amsterdam: Rodopi.

151

Narvaez, L., Penn, A., & Griffiths, S. (2012). Configurational Economies: The Value of Accessibility in Urban Development. International Transaction Journal of Engineering, Management, & Applied Sciences & Technologies, 3(3), 293-309. Netzell, O. (2013). The effect of accessibility on retail rents: testing integration value as a measure of geographic location. Journal of Property Research, 30(1), 1-23. Ortiz-Chao, C. and Hillier, B. (2007) In search of patterns of land-use in Mexi- co City using logistic regression at the plot level. Proceedings of the 6th In- ternational Space Syntax Symposium, Istanbul. Peponis, J., Ross, C., & Rashid, M. (1997). The structure of urban space, movement and co-presence: the case of Atlanta. Geoforum, 28(3), 341-358. Raudenbush, S. W., & Sampson, R. J. (1999). Ecometrics: toward a science of assessing ecological settings, with application to the systematic social observation of neighborhoods. Sociological Methodology, 29(1), 1-41. Rui-lan, D. U. A. N., & Xinqi, Z. (2004). The relation of the city road structure and the land price based on the space syntax. Science of Surveying and Mapping, 29(5), 76–79. Sampson, R. J. (2011). Neighborhood effects, causal mechanisms and the social structure of the city. Analytical sociology and social mechanisms. In: P. Demeulenaere (Ed.). Analytical Sociology and Social Mechanisms. Cambridge: Cambridge University Press. Seamon, D. (1994). The Life of the Place: A Phenomenological Commentary on Bill Hillier's Theory of Space Syntax. Nordic Journal of Architectural Research, 1: 35-48 Seamon, D (2007). A Lived Hermetic of People and Place: Phenomenology and Space Syntax. Proceedings of the 6th International Space Syntax Symposi- um, Istanbul. Seamon, D. (2012). ‘A jumping, joyous urban jumble’: Jane Jacobs’s Death and Life of Great American Cities as a phenomenology of urban place. The Jour- nal of Space Syntax, 3(1), 139-149. Ståhle, A., Marcus, L., & Karlström, A. (2005). Place Syntax: Geographic accessibility with axial lines in GIS. Proceedings of the 5th International Space Syntax Symposium, The Faculty of Architecture, TU Delft Tobler W., (1970). A computer movie simulating urban growth in the Detroit region. Economic Geography, 46(2): 234-240. Topçu, M. (2009). Accessibility effect on urban land values. Scientific Research and Essays, 4(11), 1286-1291. Tiesdell, S. & Carmona, M (2007). Urban Design Reader. London & New York: Routledge. Turner, A. (2001). Angular analysis. Proceedings of the 3rd International Space Syntax Symposium, Atlanta, Georgia Institute of Technology. Turner, A. (2005). Being in space and space in being. Proceedings of the 5th International Space Syntax Symposium, The Faculty of Architecture, TU Delft Wang, F., Antipova, A., & Porta, S. (2011). Street centrality and land use intensity in Baton Rouge, Louisiana. Journal of Transport Geography, 19(2), 285-293. Ward, M. D & Gleditsch, K. S. (2008). Spatial Regression Models. London: Sage Westheizen, D. L. (2012). Concepts of Space and Place. Neigborhood Access, Pedestrian Movement, and Physical Achtivity in Detroit: Implications for Urban Design and Research. PhD dissertation. The University of Michigan.

152

Kapitel 4 Dissecting Collective Effervescence: A Quantitative In- quiry into Randall Collins’ Interaction Ritual Theory

Abstract. In recent years, Randall Collins’ neo-Durkheimian interaction ritual (IR) theory has gained influence across a range of sociological subject fields. This mainly ethnographically based literature lacks studies specifically directed at testing the key principles of the theory. Yet Collins has asked for such tests. IR theory reflects the aim of sociology to formulate testable, generalized explanatory principles. Accordingly, the objective of this article is to put to the test IR theory’s explanation of Durkheimian collective effervescence. This test involves comparing the ‘interactional’ effect emphasized by Collins with the effects of ‘selection’ and ‘spatial environment’, i.e. mechanisms that offer competing explanations of why a group of individuals might share an effervescent experience of this kind. Methodologically, the article is based on survey and geospatial data collected at Northern Europe’s largest music festival. Data is analyzed applying multilevel regression modeling. Em- pirical findings confirm IR theory regarding the effects of interaction and selection, but suggest that the theory underestimates how the spatial environment may directly shape the experience of collective effervescence.

Keywords. interaction ritual theory; ritual intensity; multilevel models; analytical sociology; festival events

Introduction RANDALL COLLINS’ INTERACTION RITUAL (IR) theory is among the most influential contemporary micro-sociological theories. Across a wide range of sociological subject fields, a growing number of studies have applied the IR theory (Wollschleger 2012; Reeves and Bylund 1989; Zajac 1999; Summers-Effler 2004; Heider and Warner 2010; Coleman 2010; Hallett 2003; Börjesson 2013; Brown 2011). However, as Turner and Stets (2005) have pointed out, these are mainly ethnographically based studies without the specific aim of testing the generalizability of the theory’s explanatory principles, and consequently “[t]here have been no systematic tests of Collins’s theory.” (Turner and Stets 2005: 85). While the social interactionist wing of the micro-sociological tradition rejects such explanatory testability as the criterion for validating social theory (Blumer 1956; Denzin 1987; Garfinkel 1967), Collins has consist-

153

ently argued over four decades that the scientific justification of sociology consists in the ability to formulate “(…) testable, generalized explanatory principles.” (Collins 1989: 134; cf. also Collins 1975: 2ff; 1986; 1987; 1989; 2004). This stance should, however, not be confused with naïve positivism, as Collins also suggests that “(…) a complete and ridge for- malization, operationalization and measurement of anything in a scientific theory is a chimera.” (Collins 1989: 128). Scientific sociology works on the ‘fuzzy edge’ of uncertainty, tactic and informal assumptions, and this is why “[o]ne needs to work nonpositivistically, so to speak, to be a successful positivist.” (Collins 1989: 129). This article seeks to take Collins at his positivist word and put one of the key explanatory principles of IR theory to the test. This IR principle may be summarized as follows. When we observe a group of individuals who share feelings of excitement, that is, an experience of Durkheimian ‘collective effervescence’, this emotional condition should, first and foremost, be explained as a result of micro-interactional processes among co- present individuals (Collins 2004: 47-87). At its most general level, this explanatory principle reflects Collins’ (1981; 2004: 5) methodological situationism, suggesting that all social phenomena could and should be ‘translated’ to interaction ritual chains. Buehler’s (2012) recent literature review on collective effervescence concludes that the phenomenon is surprisingly understudied, both in terms of the causes, effects and the phenomenology of the phenomenon. This is also the case for IR theory’s explanation of collective effervescence, which, given the general lack of systematic tests of the theory, has not been specifically tested until now. Thus, the empirical question remains whether Collins does not place sufficient emphasis on other explanations, or may even be conflating the effect of interactional processes with other mechanisms that lead a group of individuals to share collective feelings of excitement. In addressing these question, this article follows Hedström’s (2005: 45) argument that a group of individuals might act, think, and feel in a similar manner without this being the product of interactional processes. With respect to specific groups at a single point in time, he suggests distinguishing between three types of processes that can result in individuals in a group acting in a similar manner, and only the last of these three effects is attributable to social interaction: “an environmental effect is operative if we do what we do because we are where we are. A selection effect is operative if we do what we do because we are who we are. And finally, a social-interaction effect is operative if we do what we do because others do what they do.” (Åberg and Hedström 2011: 203).

154

Collins does not, however, totally reject the effects of these two non- interactional processes of selection and environment. Regarding the question of selection, IR theory suggests that the level of collective effervescence in a situation is conditioned by the degree of ‘match up’ between the symbolic nature of the ritual and the cultural capital of the ritual participants (Collins 2004: 151ff). Concerning effect of the spatial environment, Collins argues that IR theory heralds “(…) a return to an older Durkheimian formulation in which social morphology [or, ecological environment, LSL] shapes social symbols” (Collins 2004: 32). Note that the word ‘return’ signifies that the social morphological aspects of the older Durkheimian tradition (Durkheim 1994) have often been neglected, if not explicitly rejected as immature materialism (Alexander 2005). Thus, although it does not ignore the other elements, IR theory as- signs explanatory precedence to the effects of interactional processes: the selection matchup and the spatio-morphological environment are treated as variables of secondary importance, mainly functioning as potential constraints to the interactional process that is the ultimate driving process for ritual exaltation. That is to say, effects of selection and environment may alter, weaken or intensify collective effervescence, but from the theoretical perspective of IR, it is interactional processes that primarily create the phenomenon. Given this imbalance, the present article seeks to test whether this ranking of causal effects can be empirically verified. Beginning from the a priori assumption that the effects of interaction, selection, and environment are equally important, the article draws on two network theoretical perspectives. First, the article applies the space morphological theory of ‘space syntax’ (Hillier and Hanson 1984; Hillier 1996). This theory suggests that the spatial network properties of built environment shape social interactions more directly than Collins or most sociologists of space suggest (Gans 2002). Second, the article finds inspiration from the sociological ‘homophily’ hypothesis, which suggests that most behavioral uniformities are the result of selection processes by which individuals enter network connections with similar others (McPherson et al. 2001; Lazarsfeld and Merton 1954). Giv- en these two network theoretical arguments, the article proposes the empirical question as to which, if any, of the three effects, are of primary and secondary importance in explaining the uniform group experience of collective effervescence. The article is divided into three sections. First, the article outlines the interactional explanation of collective effervescence in IR theory, and discusses the effects of selection and of spatial environment as compet-

155

ing explanations for this phenomenon. Based on these theoretical considerations, the article proposes five testable hypotheses. Second, the nature of the data and methodology of the article are presented. A geocoded survey data set has been collected at Northern Europa’s largest music festival. Data is analyzed using multilevel regression modeling. Finally, the article presents the statistical results and discusses their implications for IR theory.

Theoretical Frameworks and Expectations The best-known description of collective effervescence in the literature is undoubtedly found in Durkheim’s (1995[1912]) late major work, The Elementary Forms of Religious Life. However, as pointed out by Pickering (1984), it was probably Durkheim’s nephew and close associate, Mauss (1979[1906]), who first outlined the inner logic of the phenomenon in his less known case study, Seasonal Variations of the Eskimo: A Study in Social Morphology. Durkheim’s and, in turn, Collins’ ritual theoretical contributions, may thus be understood as a gradual refinement of Mauss’ argument advanced against the backdrop of his study on the season morphology of Eskimo culture. During the summer, the Eskimos live in tents and huts quite geographically dispersed across the countryside. This low social density facilitates a sociality that is of low emotional intensity. This changes as winter arrives. During the winter, the Eskimos live with in an extraordinary high social density in small huts and igloos. This intensifies social activity, especially those religious activities characterized by emotional energy – i.e. collective effervescence – that spreads among the densely gathered individuals:

Because individuals are brought into close contact with one another, their social interactions become more frequent, more continuous and more coherent; ideas are exchanged; feelings are mutually received and reinforced. By its existence and constant activity, the group becomes more aware of itself and assumes a more prominent place in the consciousness of the individuals (Mauss 1979: 76).

Here, we find the key elements of Collins’ IR theory and situational explanation of collective effervescence as the micro-ritual experience by which any group of individuals becomes aware of itself as a group, i.e. develops a collective consciousness. Compared with the classic Durk- heim-Maussian formulation, perhaps Collins most important contribu-

156

tion is his formulation of the notion of collective effervescence as a variable within a testable causal model. Otherwise stated, this is an IR model formulated to meet Collins’ epistemological claim that the scientific status of sociology is conditional on the formulation of testable, generalized explanatory principles.

Interactional Effect Collins’ (2004: 47-87) interaction ritual model is composed of four ritual ingredients and four ritual outcomes, causally connected by the experience of collective effervescence, as summarized in Figure 1:

Figure 1. Simplified outline of the IR model (Collins 2004: 48).

Starting from the left-hand side of the figure: (1) two or more people are physically gathered in the same place, so that they affect each other by their bodily presence. (2) The participants in the ritual share a focus of attention on an object or an activity. (3) The participants share a common mood or feeling. (4) There are boundaries to outsiders ensuring a sense of who is (and who is not) taking part in the ritual event. Moreo- ver, the ritual ingredients reinforce one another through feedback effects; a reciprocal effect between the shared focus and mood is especially important, proceeding as mostly unconscious micro-sequences of rhythmic entrainment by which the bodily movements of the participants become synchronized. Depending on the nature and intensity of these ingredients there takes “(…) place a process of intensification of shared experience, what Durkheim called collective effervescence, and the formation of collective conscience or collective consciousness.” (Collins 2004: 35). The result of the collective effervescence is, in other words, a condition of ‘heighted inter-subjectivity’, crystallizing into four ritual outcomes: (a) an emotional energy (EE) in the individual creating a motivation for involvement in the ritual activity. (b) Group solidarity and a feeling of membership. (c) Symbols that represent the group and are emotionally arousing. (d) Moral feelings that crystalize into a group- specific moral frame.

157

Applying Hedström’s topology to the three main sources of behavioral uniformity, IR theory’s explanation of collective effervescence is a clear-cut example of a social-interactional effect, defined above as operative if we do what we do because others do what they do. In accordance with this definition, Collins specifies collective effervescence as an ‘emotional contagion’ among co-present persons: “[B]ecause they are focus- ing attention on the same thing and are aware of each other’s focus, they become caught up in each other’s emotions. As a result, the emotional mood becomes stronger and more dominant; competing feelings are driven out by the main group feeling.” (Collins 2004: 108). Contagion- like processes are social-interactional effects par excellence (Hedström 1998; Manski 1993). Based on Collins’ interaction-centered explanation of collective effervescence, we can derive the following testable hypothesis: H1: A group of individuals with an intense level of interaction is likelier to experience collective feelings of effervescence.

Selection Effect The effect of selection is introduced into the IR theory in connection with Collins’ discussion of the ‘market-like’ nature of interaction rituals (Collins 2004: 151-158; 1981: 1004f; 1993). Individuals are assumed to move towards situations that feel like the highest-intensity interaction ritual currently available. That is to say, individuals maximize emotional energy (EE), understood as the subjective side of the collective effervescence. To some extent, such EE-seeking behavior unfolds, as a selection process, since individuals tend to seek interaction rituals that match their cultural capital (i.e. membership symbols) and prior level of EE: “(…) some forms of cultural capital do not match up well in some interactions: the interaction ritual does not reach a high level of intensity, and the EE payoff is low. Individuals are motivated to move away from such interactions.” (Collins 2004: 151). Thus, collective effervescence and potential EE-return is, mediated by who is participating in the ritual; according to Hedström, this is the defining characteristic of a selection effect, which is operative if we do what we do because we are who we are. When Collins directly compares the effects of selection and interaction, he relegates the importance of social composition to a parenthesis: “A situation is not merely the result of the individuals who comes into it, not even of a combination of individuals (although it is that, too). Situa- tions have laws or processes of their own; and that is what IR theory is about.” (Collins 2004: 5). The level of collective effervescence may be

158

negatively influenced by who is taking part in the ritual, but the emotional condition of excitement is positively determined by the interaction process as individuals are caught up in each other’s emotional rhythm. However, it is questionable whether it is justified to assign such a secondary role to the effect of selection in explaining collective effervescence. To examine this question the article draws upon the empirically well-established homophily hypothesis. This hypothesis suggests that (the vast majority of) behavioral similarities are the result of individuals associating with similar others (McPherson et al. 2001). Moreover, McPherson and Smith-Lovin (2002) suggest that such homophily driven sorting plays a crucial part in creating behavioral uniformities in small- groups. This argument directly counters the micro-sociological emphasis on influence-based contagion, as is found in IR theory. A homophily perspective does not reject a possible effect of emotional contagion, but anticipates that such collectively effervescent group feeling is, first and foremost, a result of preferences leading individuals to connect and in- teract with other individuals who are emotionally similar to them. The level of collective effervescence simply coincides with the degree of emotional homophily in the social group. The weak (matchup) and strong (homophily) interpretation of selection processes explaining collective effervescence may be summarized in the following hypothesis: H2a: A group of individuals with a low matchup with ritual symbols are less likely to experience contagion-based feelings of collective effervescence. H2b: A group of individuals will share collective feelings of effervescence in parallel to the group’s level of emotional homophily.

Environmental Effect These considerations lead us to Hedström’s environmental effects, which are defined as operative if we do what we do because we are where we are. A paradigmatic example of this is Mauss’ morphological study on Eskimo culture, where the group’s emotional intensity is conditioned by the seasonal-cyclical ‘where’ of the natural environment. Ac- cording to IR theory, the morphological substratum of social life has a special ontological status as the only ‘pure’ macro-variable permissible within the theory’s otherwise uniformly radical micro-situational perspective. Specifically, this applies to the number of persons/situations, the way they are dispersed in physical space, and the way they are organized in time (Rössel and Collins 2001: 510; Collins 1983: 187). This suggestion resembles Durkheim (1960: 260-262;1982: 119-146) and Mauss’ (2007:

159

19-23) principal argument that “(…) social morphology envisages society not only in space and numbers, but also in time.” (Mauss 2003: 69). All interaction rituals, including IR markets are morphologically conditioned, to some extent, by these macro properties: “The market for social interactions is best described as a series of local barter markets, shaped by the ecological conditions of the society.” (Collins 2004: 158). However, as in the case of the selection effect, the interaction effects are given precedence over the effect of the morphological environment. This ecological context is considered an indispensable element of any ritual in the here-and-now, “[b]ut microprocesses are particularly important in any explanatory theory. For time, space, and numbers do not do anything; all real causal forces must come from human beings acting in some situation.” (Collins 1983: 187, original italics). Rather than being a driving force, the morphological aspects of number, space and time appear mainly as a constraint on the interactional forces that create and constitute the emotional value of the IR markets: “(…) there are numerous imperfections in the market for IRs. Many or most individuals are prevented from trying out a wide range of alternative interactions, by ecological or social barriers. There is no implication that IR markets produce social optimum or market-clearing price.” (Collins 2004: 158).These arguments from IR theory beg the question whether this restricted understanding of the social morphology, in fact, signifies a ‘return’ to the older Durkheimian notion of social morphology, as we quoted Collins above. In Mauss’ season morphological Eskimo study, the natural environment would appear to act as a real causal force, thus suggesting that collective effervescence may be shaped more directly by spatio- environmental forces than is assumed in the interaction-centered IR theory. Further developing this space morphological possibility, the article draws upon space syntax theory concerning the social effects of spatial networks (Hillier and Hanson 1984; Hillier 1996). Based on many empirical studies, space syntax suggests that the ‘global’ network pattern emerging from human built environments (e.g. the street system in cities) systematically affects patterns of movement: ‘local’ segments with a high network accessibility have a larger probability to be movement intensive, thus increasing the level of co-presence or social density in these particular spaces (Hillier and Raford 2010). Mauss (1979), Durkheim (1995), and Collins (2004) explicitly identified the degree of bodily co-presence as the triggering factor of the ritual process: “Human bodies moving into the same place starts off the ritual process.” (Collins 2004: 53). Thus, in principle, the spatial environment

160

may affect collective effervescence, for the agency of EE-seeking individuals is not the only force capable of moving human bodies into each other’s bodily co-presence. The spatial network holds a sort of ‘network agency’ (Hillier 2012), with the capacity to increase the level of movement density and co-presence in particular spaces, thus constituting these spaces as IR market sites with an effervescent atmosphere. “Being in a crowd gives some sense of being ‘where the action is’,” as Collins suggests (2004: 82), and the space syntax-inspired slant of this argument is that an effervescent atmosphere of this kind may arise regardless whether the crowd is a product of individuals acting in some situation, or an environmental effect of the spatial network itself. Collins’ restricted and Hillier’s extended understanding of spatio-environmental effects may be summarized in the following two hypotheses: H3a: The experience of collective effervescence of a group of individuals may be limited, but not constituted by the spatio-morphological environment. H3b: A group of individuals gathered in space as a result of an environmental effect of space is likelier to experience collective effervescence.

Method and data To study the causes of collective effervescence, the article applies a representative survey of festival guests attending Northern Europe’s largest music festival, the Roskilde festival 2011 (N=505). The survey was conducted in the so-called ‘warm-up days’ preceding the scheduled music program of the festival, involving approximately 100,000 festival guests living in tent camps at the festival site. These five party-intense days unfold as a ‘self-organized party’ with a ‘ritual-like structure’ that involves numerous liminal activities and effervescent experiences. These activities include visitors socializing and drinking beer in their camps, paying festive visits to the neighboring camps, and sitting on the paths to enjoy the festival spectacle (Marling and Kiib 2011; Sørensen 2009; Getz 2010; Morgan 2007). It should be emphasized that the choice of Roskilde festival as an empirical case does not reflect a specific theoretical interest in festival events. Rather, this empirical focus reflects the methodological view that festivals may be studied as a sort of ‘micro-sociological laboratory’ where we can find an extraordinarily high level of naturally occurring collective effervescence (Marling and Kiib 2011). For this reason, the festival provides an ideally suited case for differentiating between competing explanations of this collective phenomenon.

161

By applying a survey tool, the article challenges Collins’ methodological suggestion that “(…) we should engage in systematic sampling of certain kinds of microsituations. Instead of sampling individuals, as in conventional survey research, we should sample encounters.” (Collins 1983: 195; 2009: 3-7). On the one hand, the article acknowledges Collins’ view that an exclusive focus on individual-level data tends to neglect the situational process of social life, i.e. the interactional and morphological properties of the situational context in which the individuals are embedded. On the other hand, the problem of such micro-sampling strategy is that it writes off the effect of selection in advance. Testing for selection effects requires the availability of individual-level background data. So, rather than rejecting the survey tool, the article applies two methodological strategies to supplement the individual-level survey data with context- level data regarding the interactional and spatial here-and-now: First, the survey data is georeferenced. This georeferring is based on the festival’s approximately 600 predefined 100 ft² camping micro-places where visitors are asked to set up their tents. Each of these micro-places contains approximately 10-14 camps of varying size and numbers of visitor occupants. The survey was sampled in 50 randomly chosen micro-places in the camping area, from each of which an average of 10 respondents were randomly chosen to be interviewed. Data were collected with face-to-face interviews, in the period from 12.00 to 5.00 p.m., i.e. the time of day when most festival visitors were awake following the activities of the previous night and before they had started to drink alcohol again. The response rate of the survey is 96%. Second, given this geographical area sampling strategy, the data has a hierarchical structure. The individual visitors interviewed (level 1) are nested within micro-place contexts (level 2). This sampling design is tai- lored for a linear multilevel regression model that estimates how contexts affect individual outcome – in our case how the interactional and spatio- environmental contexts collectively affect the individual visitors’ feelings of emplaced effervescence. Contrary to traditional regression models that assume the independence of observations, multilevel models relax this assumption by modeling observations as being spatially dependent because of their contextual embeddedness (Kreft and Leeuw 1998: 7; Raudenbush and Sampson 1999). In technical terms, the article applies a random intercept multilevel model, allowing the intercept to vary across of contexts. That is to say, the random intercept model examines the extent to which the experienced level of effervescence varies across micro-place contexts and how micro-place context properties affect this individual outcome. The model is estimated in MLwiN (applying Stata’s

162

‘runmlwin’ module, cf. Leckie and Charlton 2013) using a Markov chain Monte Carlo (MCMC) estimation procedure. This MCMC procedure is robust towards smaller sample sizes, relaxes the parametric normality assumption and yields more precise results when modeling spatially lagged covariates, as in the case of this article (Browne 2009; Hox 2010: 273; Ward and Gleditsch 2002).

Dependent Variable As pointed out by Xygalatas et al. (2011), the literature contains very few quantitative studies with a specific interest in collective effervescence. Accordingly, there is no agreed-upon way to operationalize this condition of excitement and collective ‘dancing’, ‘animated movement’ and ‘noise-making’ (Collins 2007: 26). Given that collective effervescence is a situational intensity defined by the interplay of behavioral, emotional and acoustic elements, the article operationalizes this phenomenon using an additive index based on five items. In the survey the respondents were asked various questions on how they experienced the immediate sur- roundings of their camp: “My camp is situated in a ‘festive,’ ‘noisy,’ ‘hectic,’ ‘boring,’ and ‘calm’ place”. The response categories are: “Agree,” “Agree strongly,” “Disagree,” and “Disagree strongly.” The internal consistency of the additive index is, α=.83, suggesting adequate reliability of this collective effervescence index outcome.

Interactional Measures The article includes two variables to capture the interaction ritual chains in which the festival guests are involved. Following Collins’ (2004) and Goffman’s (1963) classic terms, these two variables distinguish between ‘focused’ and ‘unfocused’ interaction. First, the focused interaction is measured using the following question: “My camp hangs out with the neighbor camps a lot”. The response categories are: “Agree strongly,” “Agree,” “Disagree,” and “Disagree strongly.” The variable is included in the estimation as a dummy coded individual-level predictor with the reference category as the two disagreeing responses. Second, the unfocused interaction is measured as the micro-place mean for the following question: “For me the festival is about having a party.” Again, the response categories are: “Agree strongly,” “Agree,” “Disagree,” and “Disagree strongly.” Interactional unfocusedness consists in the fact that this context-level predictor does not specifically address interaction realized as face-to-face encounters, but rather captures the acoustically and quasi- visually mediated sense of being in a more or less festive party micro- context ‘where the action is’ (Collins 2009: 430; Collins 2004: 82).

163

Table 1. Summary Statistics (unstandardized) Mean SD Min Max N Dependent variable: Collective effervescence 9.04 2.87 0 15 495 Context-level predictors: Micro-place festivity 0.61 0.27 0 1.15 505 Movement density 3.50 0.47 2.30 4.19 505 Spatial lag of Coll. effervescence 9.27 2.06 4.10 13.30 505 Individual-level predictors: Interaction with neighbors 0.56 0.50 0 1 497 Male 0.44 0.50 0 1 487 Age 20.91 4.01 15 38 505 Education University bachelor/master 0.11 0.31 0 1 505 Short/medium long education 0.04 0.20 0 1 505 Vocational education 0.05 0.21 0 1 505 High School 0.45 0.50 0 1 505 Primary 0.35 0.48 0 1 505 Previous Roskilde experience 2.62 3.08 0 19 504 Party preference 0.75 0.43 0 1 503 Camp selection Plan, and realized 0.64 0.48 0 1 504 Plan, but not realized. 0.28 0.45 0 1 504 No plan 0.08 0.27 0 1 504

Spatio-Environmental Measures The article applies space syntax methods to measure the properties of the spatial environment. Space syntax makes use of mathematical graph theory to quantify the network properties of built space. Applying the space syntax tool involves the following steps. First, using the high- resolution aerial photographs taken each year by the festival organizers, we automatically generated a so-called axial map using UCL Depthmap. This axial map represents the fewest and longest lines that are needed to cover all the ways of moving around the layout and reach all spaces (Hill- ier and Raford 2010; Turner 2004). Second, based on this axial map, a segment angular choice analysis is conducted, also using UCL Depthmap. In previous space syntax studies, this particular choice measure has proven a strong predictor of actual movement in urban morphologies (Hillier et al. 1993; Hillier and Iida 2005). The choice analysis involves that the axial lines are decomposed into the segments between junctions, and the choice measure is, in turn,

164

calculated by generating the shortest paths between all segments within a given metric radius, i.e. the journey with the lowest angular ‘cost’ for each possible origin and destination pair of segments (Turner 2001). In mathematical terms, choice can be understood as a space morphological equivalent to Freeman’s (1977) ‘centrality betweenness’, as applied within social network analysis. In intuitive terms, the choice measure often identifies the ‘bridges’ that link together the network and, hence, are movement thoroughfares. Third, since space syntax methods have not been previously applied to a festival context, the choice measure is validated against count data on actual movement patterns at the festival, collected concurrently with the survey data. Following Grajewski’s (2001) guidelines, these count data were collected at 70 randomly localized gates, each of which was observed 5-6 times for 3 minutes. A Spearman’s rho suggests the existence of a highly significant and very large correlation, rS = 0.74; p = .000, between movement and a local (radius 400 meters) choice measure (note that this finding is consistent with spatial regression modeling of these data that accounts for spatial autocorrelation). This result suggests that choice may be considered a valid proxy for network-facilitated patterns of movement. Finally, converting this choice measure into a context-level predictor, the movement density of the paths adjacent to each of the 50 micro- places is calculated as the average choice value of the segments within a 65-foot radius from the center of each micro-place; that is, the immediate walkable pathway around the micro-place.

Selection Measures Five variables capture the social composition of the micro-places. First, we include three sociological background variables: age, gender, and highest level of education achieved, as measured on five levels: Universi- ty bachelor/master, Short/medium long education, High School, Voca- tional education and Primary education. Second, three variables measure the visitor’s festival habitus, i.e. the visitor’s cultural capital: (1) One variable captures the visitor’s preference for camping in a party-intense festival. This is measured with the question: “It improves my festival experience if my camp is located in an area in which people are partying inten- sively.” The response categories are: “Agree strongly,” “Agree,” “Disa- gree,” and “Disagree strongly.” The variable is included in the estimation as a dummy coded individual-level predictor with the two disagreeing responses as a reference category. (2) A continuous variable captures the visitor’s previous Roskilde festival experience with the question: “How

165

many times have you attended the Roskilde festival previous to this year.” (3) A variable captures how actively the visitors have selected themselves into their camp’s micro-place. The variable has three dummy coded levels of camp selection: in addition to the visitors who did not have a prior plan about where to deploy their camp, the variables distin- guishes between visitors who had a plan that was either carried out or not accomplished (e.g. due to competition for popular camping areas).

Control Variables As noted earlier, multilevel models are one way to account for spatial autocorrelation. However, as Chaix et al. (2005) suggested, multilevel models only provide a partial solution to spatial autocorrelation, since the model assumes that the existence of spatial autocorrelation can be reduced to within-context correlation. The multilevel model effectively treats the clusters/micro-places as socio-spatially isolated islands. First, we calculated an intra-class correlation (ICC) (on the basis of the empty Model 0, see below), which suggested that 50% of the variance of fes- tiveness experience can be explained in terms of this kind of internally correlated differences between the clusters. Second, to assess the existence of unaccounted spatial autocorrelation between the contexts, a Mo- ran’s I statistic was calculated, I = 0.74; p = .000. This result strongly suggests that micro-places with a similar level of collective effervescence have a strong tendency to be spatially clustered. Taking this spatial autocorrelation into account, the article follows Morenoff’s (2003) suggestion to include a spatially lagged version of y as a covariate in multilevel models. This spatial lag is calculated in Geoda as the weighted average of the neighboring average level of collective effervescence in the micro-places (Anselin, Syabri and Kho 2006; Ward and Gleditsch 2008).

Results Table 2 presents the estimated results from a series of multilevel models assessing how the experience of collective effervescence is conditioned by the effects of interaction, selection and/or the spatio-environmental context of the festival’s micro-places. The empty Model 0 only fits an overall constant to data, and the three subsequent models 1-4, add the interactional, spatio-environmental, selection, and control measures, respectively. Note that the Deviance Information Criterion (DIC) suggests that Model 4 has the best overall data fit (Spiegelhalter 2002). Moreover, calculation of the proportional reduction in variance suggests that the

166

final Model 4 accounts for 76% of the between-cluster variance in experienced collective effervescence. Model 1 suggests that the focused interaction with neighboring camps and unfocused festivity interaction within the micro-place both contribute significantly and positively to collective effervescence. This speaks in favor of the key IR theoretical hypothesis 1, suggesting that a group of individuals with an intense level of interaction are likelier to experience this kind of collective feeling. We should also note that while the effect size of the focused interaction variable is, by and large, unchanged by inclusion of the variables of the other models, this is not the case with regard to the unfocused interaction variable. This context-level effect reduces by half when we add the spatially lagged version of collective effervescence in M4.

Table 2. Random intercept multilevel regression analysis of collective effervescence M0 M1 M2 M3 M4 Empty Interactional Environmental Selection Control Intercept -0.01 -0.10 -0.10 -0.56*** -0.51*** Fixed effects Micro-place festivity 0.39*** 0.32*** 0.30*** 0.14** Interaction with neighbors 0.16** 0.17** 0.18** 0.18** Movement density 0.31*** 0.30*** 0.19*** Male -0.02 -0.03 Age 0.08 0.08 Education (ref: Primary) University bachelor/master 0.21 0.15 Short/medium long education 0.07 0.04 Vocational education 0.37** 0.35* High School 0.30*** 0.27*** Previous Roskilde experience -0.12*** -0.12*** Party preference 0.26*** 0.23*** Camp selection (ref: No plan) Plan, and realized 0.17 0.17 Plan, but not realized. -0.01 0.02 Spatial lag of Coll. effervescence 0.41*** Random effect Intercept variance component 0.53*** 0.36*** 0.27*** 0.25*** 0.13*** DIC 1124.11 1102.77 1101.67 1033.09 1027.94 N 495 490 490 468 468 Estimates of continuous variables are reported in Beta coefficients. *p<0.10; **p<0.05; ***p<0.01. All model estimates are obtained with 30,000 MCMC iterations.

167

The spatial lag is by far the covariate with the largest effect size, accounting for 50% of the between-context variation of collective effervescence. Significant spatial lags are often interpreted a kind of ‘spatial spillover effect’ (Anselin 2002; Ward and Gleditsch 2008). In this festival case, part of the effect size of the lagged version of collective effervescence may be ascribed to the fact that adjacent micro-places affect each other’s excitement levels. That is to say, collective effervescence develops in the form of spatio-emotional contagion, where festivity noise and activity creates self-reinforcing hotspots of collective effervescence. The halving of the effect size of the unfocused interaction variable may thus be attributed to the fact that the festivity level of one micro-place spills over from its adjacent micro-places. In addition to this interaction-based interpretation of the spatial lag it should also be noted that it is difficult to distinguish contextual effects from the effect of selection into the context (Duncan, Jones, and Moon 1998; Manski 1993). We will come back to this selection-based interpretation of the spatial lag at the end of the results section. Next, adding the movement density measure in Model 2 establishes that this context-level variable has a significant and positive effect on the experience of collective effervescence. This suggests that space affects collective effervescence by densifying movement and co-presence in certain network segments of the festival, thereby constituting these areas as crowded places ‘where the action is’. Morphological space, i.e. specifically its network properties of accessibility, would thus appear to have a more direct spatio-environmental effect on collective effervescence than was assumed by Collins or stated in hypothesis 3a. Space is not merely a constraining variable, but unfolds as a material substrate that constitutes collective effervescence as was hypothesized in 3b, in keeping with Hiller’s space syntax approach. Moreover, from the perspective of sociology, it is notable that this spatio-environmental context effect, on the whole, is unaffected by inclusion of the individual-level selection variables in Model 3. This finding underscores that the variation in collectively experienced effervescence should not be attributed simply to the aggregation of individual properties. Collective effervescence is also a contextual effect of the festival morphology as such. Model 3 establishes that the experience of collective effervescence should not be ascribed to selection effects of age, gender, and perhaps most surprisingly, to the level of the festival guests’ camp selection. On the other hand, significant selection effects include: (1) the guests’ previous Roskilde festival experience, (2) the festivity preference measure, and

168

two of the educational dummies: (3) festival guests with a high school background and (4) vocational education report a significantly higher level of collective effervescence when compared to guests with only a primary school background. These four findings may be interpreted as follows: First, the negative parameter estimate of earlier Roskilde festival experience suggests that ritual fascination and excitement wanes as guests become more familiar with the festival. This result provides evidence for the selection hypothesis 2a, suggesting that the degree of matchup between the guests’ cultural capital and ritual practice regulates the intensity of the experience. Experiencing the festival from the ‘inside’ is accompanied, so to speak, with a certain price: knowledge decreased the enchantment with the festival ritual, and this, in turn, is experienced as a decrease in the collective and effervescent force that captures the guest from the ‘outside’ (Durkheim 1995: 190ff; 2002: 9). Second, the finding that guests with a high school background tend to experience their camp as being located in a more exciting place may also be interpreted as supporting the matchup hypothesis 2a. This may reflect the fact that this is the largest educational group, representing 45% of the festival population (cf. the summery statistics, Table 1). This would suggest that the particular positive experience of this group reflects the group’s matchup – or, ‘structural homology’ (Bourdieu 1985) – with the festival as a socio-ritual space. Given the group size, it is likelier that an individual from this educational group would become involved in a type of ritual activity that matches up with that individual’s cultural capital (e.g. a distinct way of dancing or drinking). Third, the positive estimate from the vocational education dummy converges with results found in Danish alcohol studies, which have identified this educational level to be correlated with an explicit binge- drinking party culture (Østergaard et al. 2010). Thus, in the case of this particular educational group, collective effervescence, may not specifically arise from the camp context, but rather from the party-attuned individuals who have selected themselves into the micro-places. Finally, a similar interpretation pertains to the positive festivity preference measure: all other things being equal, not surprisingly, guests with a habitual taste for festivity collectively experience their camp as being located in a more exciting place. At first glance, the two latter findings seem especially to be more closely associated with the strong homology-based selection hypothesis 2a. These effects do not specifically seem to arise from the matchup with the ritual context, but from the fact that we are dealing with individuals who, regardless of the contextual properties, have a higher level of festiv-

169

ity and excitement. Here, collective effervescence emerges, so to speak, as an aggregation of individual-level excitement. Despite this suggestion, it should be emphasized that the overall effects of selection speak against the strong homophily-based selection hypothesis 2b. This conclusion is substantiated by the statistical finding that the effect sizes of the interactional and spatio-environmental context-level predictors are virtually unaffected by inclusion of the selection variables, as it may be seen in Model 3. The micro-ritual context cannot be reduced to social composition. Moreover, this picture remains unchanged after controlling for several other individual-level selection variables, such as the guests’ relationship status, their festival budget, and preferences regarding alcohol con- sumption (results not shown) However, in addition to this conclusion we need account that the large effect size of the spatial lag of y may be due to the fact that this context-level covariate captures an unobserved selection toward the festival ‘neighborhoods’ in which the micro-places are embedded (Lieber- son 1985). This kind of selection interpretation of the spatial lag would address the fact that the guests’ selective choices do not necessarily take place on the micro-spatial scale, as assumed in the analysis presented above. Instead, we may be dealing with a theoretical selection variant of what is referred to as the modifiable areal unit problem (MAUP) (Open- shaw 1984), which relates to the fact that the arbitrary partition of area units applied in most social science research may be a source of statistical bias. To determine whether the estimated results are robust in this regard, we analyze the data using a 3-level multilevel model that operationalized the neighborhood level as the 11 festival areas that follow the spatial and organizational divisions of the festival. It should be noted that while spatial lags account for spatial autocorrelation, the introduction of such higher levels of nestedness may be considered an alternative way to account for the spatial clustering in the data (Bernasco and Elffers 2010). This article will not go into the details of this 3-level data model. It is sufficient to establish that a significant portion of the effect size of selection may be identified at the neighborhood level, and that the 3-level model, on the whole, is consistent with the results of the 2-level model as presented above. That is, the 3-level models identify a number of significant effects of selection, but the fact that the interaction and spatio- environmental effects remain significant speaks against a strong homophily hypothesis. The findings in the article seem more consistent with the matchup hypothesis, and, specifically, support Collins’ key ar-

170

gument that collective effervescence is a situational property, as opposed to a simple aggregation of individual properties.

Discussion and Conclusions The findings from this research project suggest three main conclusions regarding the validity of Collins’ IR theoretical explanation of collective effervescence. Following Hedström’s typology about the three main causes of behavioral uniformities, the key explanation proposed by Col- lins can be characterized as an interactional effect, and the article has identified the ways that the effects of selection and the spatial environment have been assigned secondary explanatory importance. IR theory mainly regards selection and the spatial environment as secondary effects, which constraint but do not constitutes micro-ritual intensities. Accepting Collins’ call to test the explanatory generalizability of (IR) theoretical principles, the article’s main thrust is to ask whether the interactional bias in IR theory works at the expense of understanding the effects of selection and environment as constitutive for collective effervescence. First, our statistical analysis confirms Collins’ key hypothesis that interactional processes make it likelier that ritual participants share a collective experience of effervescence. Specifically, this effect may be attributed to focused face-to-face interactions among the festival guests as well as unfocused interactions that occur as a feeling of being at the festive epicenter ‘where the action is’ that are mediated acoustically rather than visually,. Second, the analysis supports Collins’ notion of selection processes as reflecting the degree of matchup between the participants’ cultural capital and the symbolic nature of the ritual. Matchups do not explain collective effervescence, but function by regulating the intensity of such collective feelings. This finding differs from a strong homophily-based selection hypothesis that would suggest that such emotional uniformity relates to emotional homophily. Collective effervescence is not simply a matter of the social composition of the collective of festival guests gathered in particular micro-places. At times, the interaction ritual can be an unanticipated process, and regardless of who may be participating, may emotionally captivate those individuals. In sum, with respect to the question of selection effects, the findings in the article suggest that Collins’ matchup theoretical explanation appears to be empirically valid and adequate. Regarding the phenomenon of collective effervescence, there is no immediate need for a homophily-theoretical reassessment or a stronger reformulation of selection processes within the IR framework.

171

Third, the findings of the article in relation to spatio-environmental aspects speak against the expectations of IR theory. Space does not merely restrain the interactional development of IR markets. Instead, the statistical analysis suggests that space (i.e. the network properties of built space, in particular), generates varying levels of emplaced social density or event crowding which, in turn, provides the festival guests with a sense of being at a camp site where the key festival action takes place. Overall, this conclusion suggests that the IR theory needs to permit spatial morphology to develop as a ‘real causal force’ (Collins 1983: 187), which engenders collective effervescence, and perhaps the interaction ritual events as well: “Space does not direct events, but it does shape the possibility.” (Hillier 1996: 155). At this point, we must leave it to future research to further examine the explanatory power and generalizability of such space morphological causal forces. An obvious starting point for an inquiry of this kind would be to specifically return to Mauss’ (1979) neglected Eskimo study, rather than following Collins’ vague ‘return’ toward an unspecified ‘older Durkheimian’ notion of social morphology. To my knowledge neither Collins nor the literature have conducted a comparison between IR theory and Mauss’ study, i.e. a study that presented the Durkheimian notion of collective effervescence and, in con- sonance with the findings of this article, offered a spatio-morphological explanation of this intriguing group phenomenon.

Acknowledgements. The author would like to thank the following people for their encouraging comments and helpful suggestions: Sofie Laurine Albris, Stefan Bastholm Andrade, Marie Bruvik Heinskou, Kristian Bernt Karlson, Inge Kryger Pedersen, Anders Trolle Purup, Line Vistisen, and XX anonymous reviewers.

References Alexander, J. C. (2005). The inner development of Durkheim’s sociological theory: from early writings to maturity. In J. C. Alexander, & P. Smith (Eds.), The Cambridge Companion to Durkheim. Cambridge: Cambridge Univer- sity Press. Anselin, L. (2002). Under the hood issues in the specification and interpretation of spatial regression models. Agricultural economics, 27(3), 247-267. Anselin, L., Syabri, I., & Kho, Y. (2006). GeoDa: An introduction to spatial data analysis. Geographical analysis, 38(1), 5-22. Bernasco, W., & Elffers, H. (2010). Statistical analysis of spatial crime data. In A. Piquero R., & D. Weisburd (Eds.), Handbook of Quantitative Criminology. New York: Springer.

172

Blumer, H. (1956). Sociological analysis and the variable. American Sociological Review, 21(6), 683-690. Bourdieu, P. (1985). The social space and the genesis of groups. Theory and society, 14(6), 723-744. Börjesson, U. (2013). From shadow to person: exploring roles in participant observations in an eldercare context. Qualitative Social Work, 13(2): 1-15. Brown, K. R. (2011). Interaction ritual chains and the mobilization of conscien- tious consumers. Qualitative sociology, 34(1), 121-141. Browne, W. J. (2009). MCMC estimation in MLwiN. Centre of Multilevel Mod- elling. University of Bristol. Buehler, A. (2012). The Twenty-first-century Study of Collective Effervescence: Expanding the Context of Fieldwork. Fieldwork in Religion, 7(1): 70-97. Cariton-Ford, S. L. (1992). Charisma, Ritual, Collective Effervescence, and Self- Esteem. The Sociological Quarterly, 33(3), 365-387. Chaix, B., Merlo, J., Subramanian, S. V., Lynch, J., & Chauvin, P. (2005). Com- parison of a spatial perspective with the multilevel analytical approach in neighborhood studies: the case of mental and behavioral disorders due to psychoactive substance use in Malmö, Sweden, 2001. American Journal of Epi- demiology, 162(2), 171-182. Coleman, G. (2010). The hacker conference: A ritual condensation and celebra- tion of a life world. Anthropological Quarterly, 83(1), 47-72. Collins, R. (1975). Conflict Sociology: Toward an Explanatory Science. New York: Academic Press. Collins, R. (1981). On the microfoundations of macrosociology. American Jour- nal of Sociology, 86(5), 984-1014. Collins, R. (1983). Micromethods as a Basis for Macrosociology. Journal of Con- temporary Ethnography, 12(2), 184-202. Collins, R. (1986). Is 1980s sociology in the doldrums?. American Journal of Sociol- ogy, 91(6), 1336-1355. Collins, R. (1987). Looking Forward or Looking Back?: Reply to Denzin. Ameri- can Sociological Review, 93(1): 180-184. Collins, R. (1989). Sociology: Proscience or antiscience?. American Sociological Review, 54(1), 124-139. Collins, R. (2004): Interaction Ritual Chains. Princeton: Princeton University Press. Collins, R. (2007). The classical tradition in sociology of religion. Beckford, J. A., & Demerath, J. (Eds.), The SAGE handbook of the sociology of religion. Lon- don: Sage. Collins, R. (2009). Violence: A Micro-Sociological Theory. Princeton: Princeton Uni- versity Press. Denzin, N. K. (1987). The death of sociology in the 1980s: Comment on Col- lins. American Journal of Sociology, 93(1), 175-180. Duncan, C., Jones, K., & Moon, G. (1998). Context, composition and heterogeneity: using multilevel models in health research. Social science & medicine, 46(1), 97-117. Durkheim, É (1960). Sociology and its Scientific Field. Emilie Durkheim, 1858- 1917. A Collection of Essays. With Translation and a Bibliography (Ed. Kurt. H. Wolff). Columbus: The Ohio State University Press. Durkheim, É. (1982). The Rules of Sociological Method. New York: The Free Press.

173

Durkheim É. (1994). Note on Social Morphology. In: Traugott, M. (Ed.). Emile Durkheim: On Institutional Analysis. Chicago: University of Chicago Press. Durkheim, É. (1995). The Elementary Forms of Religious Life. New York: The Free Press. Durkheim, É. (2002). Moral Education. Mineola: Dover Publication. Fine, G. A. (1991). On the macrofoundations of microsociology. The Sociological Quarterly, 32(2), 161-177. Freeman, L. (1977). A set of measures of centrality based on betweenness, Soci- ometry, 40(1), 35–41. Gans, H. J. (2002). The sociology of space: a use–centered view. City & Com- munity, 1(4), 329-339. Garfinkel, H. (1967). Studies in Ethnomethodology. Englewood Cliffs: Prentice- Hall. Getz, D. (2010). The nature and scope of festival studies. International Journal of Event Management Research, 5(1), 1-47. Goffman, E. (1963). Behavior in Public Places: Notes on the Social Organization of Gatherings. New York: The Free Press. Goffman, Erving (1967). Interaction Ritual. Garden City: Anchor Books. Grajewski T. (2001). Space Syntax Observation Manual. Bartlett School of Graduate Studies, University College London. Hallett, T. (2003). Emotional feedback and amplification in social interaction. The Sociological Quarterly, 44(4), 705-726. Hedström, P. (1998). Rational Imitation. In P. Hedström, & R. Swedberg (Eds.). Social Mechanisms: An Analytical Approach to Social Theory. Cambridge: Cambridge University Press. Hedström, P. (2005). Dissecting the Social: On the Principles of Analytical Sociology. Cambridge: Cambridge University Press. Heider, A., & Warner, R. S. (2010). Bodies in sync: interaction ritual theory applied to Sacred Harp singing. Sociology of Religion, 71(1), 76-97. Hillier, B (1996). Space is the Machine. Cambridge: Cambridge University Press. Hillier, B. (2012). The city as a socio-technical system: a spatial reformulation in the light of the levels problem and the parallel problem. Aschwanden, G., Halatsch, J., & Wonka, P. (Eds.), Digital Urban Modeling and Simulation. Berlin: Springer. Hillier, B., & Hanson, J. (1984). The Social Logic of Space. Cambridge: Cambridge University Press. Hillier, B., Penn, A., Hanson, J., Grajewski, T., & Xu, J. (1993). Natural movement-or, configuration and attraction in urban pedestrian movement. Envi- ronment and Planning B: Planning and Design, 20(1), 29-66. Hillier, B., & Iida, S. (2005). Network and psychological effects in urban movement. Spatial information theory, 3693, 475-490. Hillier, B., & Raford, N. (2010). Description and discovery in socio-spatial analysis: the case of space syntax. In: G. Walford, E. Tucker, & M. Viswanathan (Eds.), The SAGE Handbook of Measurement. London: Sage. Hox, J. (2010). Multilevel Analysis: Techniques and Applications. London & New York: Routledge. Kreft, G. G. & Leeuw, J. d (1998). Introducing Multilevel Modeling. London: Sage. Lazarsfeld, P. F., & Merton, R. K. (1954). Friendship as Social Process. In: M. Berger, T. Abel, & C. Page (Eds.), Freedom and Control in Modern Society. New York: Octagon.

174

Leckie, G. and Charlton, C. (2013). runmlwin - A Program to Run the MLwiN Multilevel Modelling Software from within Stata. Journal of Statistical Software, 52(11), 1-40. Lieberson, S. (1985). Making it Count: The Improvement of Social Research and Theory. Berkeley: University of California Press. Manski, C. F. (1993). Identification of endogenous social effects: The reflection problem. The Review of Economic Studies, 60(3), 531-542. Mariot, N. (2011). Does acclamation equal agreement? Rethinking collective effervescence through the case of the presidential “tour de France” during the twentieth century. Theory and society, 40(2), 191-221. Marling, G., & Kiib, H. (2011): Instant City @ Roskilde Festival. Aalborg: Aalborg University Press. Mauss, M. (1979). Seasonal Variations of the Eskimo: a Study in Social Morphology. London: Routledge & Kegan Poul. Mauss, M. (2007). Manual of Ethnography. New York & Oxford: Durkheim Press/Berghahn Books. McPherson, M., Smith-Lovin, L., & Cook, J. M. (2001). Birds of a feather: Ho- mophily in social networks. Annual Review of Sociology, 27, 415-444. McPherson, M., & Smith-Lovin, L. (2002). Cohesion and membership dura- tion: Linking groups, relations and individuals in an ecology of affiliation. Advances in Group Processes, 19, 1-36. Morenoff, J. D. (2003). Neighborhood mechanisms and the spatial dynamics of birth weight. American Journal of Sociology, 108(5), 976-1017. Morgan, M., (2007). Festival Spaces and the Visitor Experience. Casado-Diaz, M., Everett, S., & Wilson, J. (Eds). Social and Cultural Change: Making Space(s) for Leisure and Tourism. Eastbourn: Lesiure Studies Association. Openshaw, S. (1983). The Modifiable Areal Unit Problem. Norwick: Geo Books. Pickering, W. S. F. (1984). Durkheim’s Sociology of Religion. London & New York: Routledge. Raudenbush, S. W., & Sampson, R. J. (1999). Ecometrics: toward a science of assessing ecological settings, with application to the systematic social observation of neighborhoods. Sociological Methodology, 29(1), 1-41. Reeves, E. B., & Bylund, R. A. (1989). Social density and public ritual in non- industrial communities. The Sociological Quarterly, 30(2), 225-244. Rössel, J., & Collins, R. (2001). Conflict theory and interaction rituals: The microfoundations of conflict theory. In J. Turner (Ed.), Handbook of sociological theory. Berlin: Springer. Spiegelhalter, D.J., Best, N.G., Carlin, B.P. & van der Linde, A. (2002). Bayesi- an measures of model complexity and fit. Journal of the Royal Statistical Society: Series B, 64(4), 191-232. Summers-Effler, E. (2004). Humble saints and righteous heroes: Sustaining intense involvement in altruistic social movements. Philadelphia: Penn Press. Sørensen J K, 2009, Substance use, rituals and risk management Danish rock festivals, PhD thesis, Department of Sociology, University of Copenhagen. Turner, A. (2001). Angular analysis. Proceedings of the 3rd International Space Syntax Symposium, Atlanta, Georgia Institute of Technology. Turner, A. (2004). Depthmap: A Researcher’s Handbook. Bartlett School of Graduate Studies, University College London. Turner, J. H., & J. E. Stets (2005). The Sociology of Emotions. Cambridge: Cam- bridge University Press.

175

Ward, M. D., & Gleditsch, K. S. 2002. Location, location, location: An MCMC approach to modeling the spatial context of war and peace. Political Analysis, 10(3), 244-260. Ward, M. D., & Gleditsch, K. S. (2008). Spatial Regression Models. Vol. 155. Lon- don: Sage. Wollschleger, J. (2012). Interaction Ritual Chains and Religious Participation. Sociological Forum, 27(4), 896-912. Xygalatas, D., Konvalinka, I., Bulbulia, J., & Roepstorff, A. (2011). Quantifying collective effervescence: Heart-rate dynamics at a fire-walking ritual. Commu- nicative & integrative biology, 4(6), 735-738. Zajac, B. E. (1999). Community suicide: secularization in Catholic nuns with the manipulation of ritual and symbol, Doctoral dissertation, University of California, Riverside. Østergaard, J., Røgeskov, M., & Rasmussen, P. S. (2010): Unges rusmiddelpro- fil. In: M. Järvinen, J. Demant, & J. Østergaard (Eds.), Stoffer og natteliv. Kø- benhavn: Hans Reitzels forlag. Åberg, Y., & Hedström, P. (2011). Youth unemployment: a self-reinforcing process?. In P. Demeulenaere (Ed.), Analytical Sociology and Social Mechanisms. Cambridge: Cambridge University Press.

176

Kapitel 5 Hot Spots of Festival Co-Creativity: A Spatial Statistical Analysis

Abstract. Quantitative inquiry into the spatial setting of festivals is in its infancy. Pioneer studies have examined spatial movement patterns and, recently, how visitors’ experiences are shaped by festival settings. What is lacking in the literature are quantitative studies on the spatial behaviors of festival visitors, especially how visitors make ‘co-creative’ use of festival space. Building on the growing body of ethnographic research that has called attention to the importance of such co-creative activities, the purpose of the article is to present a spatial statistical analysis of the socio-spatial mechanisms that shape festival settings into ‘hot spots’ of co-creativity. Aerial photography data from the Roskilde Festival 2011 is applied to geocode co-creative events and micro-scale properties of festival space. Geographical data are analyzed with spatial statistical techniques and results suggest that hot spots of festival co-creativity may be attributed to effects of social selection, interaction spillovers and the micro-spatial level of event crowding.

Keywords: festival space; co-creativity; quantitative spatial modeling; Roskilde Festival

Introduction THE SPATIAL SETTING IS PART OF THE backbone of festival events. While it might be argued that placelessness reigns in so many social spheres where electronic transmissions have made physical co-presence superfluous (Virilio 2001; Castells 1996: 500), festivals are about sharing a collective experience of an emplaced ‘live’ event (Urry 2007). Few social phenomena provide a better illustration of what Boden and Molotch (1994) term people’s ‘compulsion of proximity’, and hence we should not be surprised that festival attendance has not declined with the spread of high-tech technologies. Even the most placeless of all social groups, cyber-communities, still favor attending conference gatherings where the group’s social bond is confirmed and celebrated face-to-face (Collins 2004: 57; Coleman 2010; Morgan 2007: 13). Space is a key element of festival events, yet research on the spatial properties of the festival setting has been limited by two circumstances. First, there is a general lack of studies with a specific focus on spatial

177

properties, experiences and behaviors of festival settings. This has recently been emphasized by Stevens and Shin (2012: 2), who point out that what “(…) is lacking in the current literature is any specific analytical focus on the spatial setting of festivals.” A similar call for research is found in Getz’ (2010) recent comprehensive review of the field of festival studies, and in Getz (2004) and Batty’s (2007: 322) discussions of the prospect of analyzing festival events applying a geographical perspective. Second, in looking at the relatively small number of quantitative and qualitative studies that have examined festival place, one notes a gap between these two approaches that are pursued in isolation from each other. Thus, on the one hand, a number of ethnographic studies have studied the relationship between festival space, experiences and behaviors, but have done this almost entirely isolated from the quantitative studies in the literature and their results. This isolation reflects an un- compromising, and in my assessment unfruitful, skepticism towards quantification as such. The opponents of quantification suggest that ‘measurement’ is contradicts the ‘meaning’ dimension of festival life, which they believe can only be studied by ethnographic means (Hol- loway, Brown, and Shipway 2010; Morgan 2007; Stevens and Shin 2012; Getz 2010: 22). On the other hand, quantitative studies have left themselves vulnerable to such criticism because this body of research applies a narrow focus on the movement patterns and properties of festival settings. However, as suggested by Cresswell (2006: 2ff), ‘movement’ should be thought of as ‘abstracted mobility’; that is, abstracted from the fact that moving from A to B is also experienced as something phenomenologically meaningful. The narrow aspect of movement as a ‘brute fact’, excluding the meaning content of mobility, has been examined by applying agent-based simulations of crowd dynamics (Klüpfel 2007; Bat- ty, Desyllas and Duxburt 2003), scanning participant movement patterns (Stopczynski et al. 2013) and using techniques for estimating the number of visitors attending festivals (Raybould et al. 2000). At a general level, the article is an attempt to address these two circumstances – the general lack of research and the research gap between the qualitative and quantitative contributions – which have constrained spatial inquiries into festival space in the literature. More specifically, I will attempt to do this by conducting a spatial statistical festival study against the backdrop of key insights from both qualitative and quantitative studies on festival space. By attempting to build such bridges, the article is in line with a number of recent spatial statistical studies that are unique in giving the quantitative approach to festival space a qualitative ‘slant’ toward the nature of meaningful visitor experiences. Showing that

178

the crude opposition between measurement and meaning is methodologically superfluous, these quantitative studies apply a combination of GPS data on individual movement patterns with survey data on spatial experiences (Pettersson and Zillinger 2011; Zakrisson and Zillinger 2012). Along the same line, Pettersson and Getz (2009) have triangulated ethnographic participant observations with survey data on the time-spatial nature of the event experience. These experience-sensitive quantitative studies have shown that it is possible to measure and model how festival space shapes visitors’ emotions and thought, but they only relate to a limited extent to the visitors’ actual behavior in the festival space. That is, these bridge building quantitative studies do not specifically address or operationalize what may be the most important use-centered finding across the ethnographic studies – namely, that the use of festival space is not only ‘stage-managed’ by the festival organization from above, but is also ‘co-created’ from below by festival visitors’ actions (Morgan 2007; Steven and Shin 2012; Pettersson and Getz 2009; Jamieson 2004; Samson 2012; Marling and Kiib 2011). This ethnographic finding has been pioneered by Morgan (2007), who emphasizes that extraordinary festival experiences are often closely related to the existence of co-creative actions of festival space:

(…) in terms of the use of space, the festival provides a setting where communities of enthusiasts can meet or form to co-create extraordinary experiences. They find centres, paths and domains of experiential space within the physical location. The organisers, pro- fessional performers and local businesses do not create the experience, they merely facilitate it (Morgan 2007: 15).

The specific objective of this article is to offer a quantitative analysis of the co-creative use of festival space. To my best knowledge, this article presents the first specifically quantitative examination of co-creative festival behavior. The article is divided into four sections. First, the article considers the main mechanism that may result in groups of festival visitors acting co-creatively in a similar manner. While the ethnographic studies have mainly ‘described’ the existence of co-creativity, the article addresses possible causal ‘explanations’ of this spatial phenomenon, especially whether and why co-creative activities tend to concentrate or cluster in hot spots around the festival space. Second, the article outlines its space statistical design and data basis consisting of geocoded events of co-creative actions and movement count data collected at Northern Europe’s largest music festival, the Roskilde Festival 2011. Third, results

179

are presented that suggest that co-creative hot spots are not the result of a single mechanism, but of a mixture of social interactional, selection and spatio-environmental effects. Finally, the article discusses how the ‘structural’ nature of these mechanisms challenges the actor-centered perspective of the ethnographic studies that assume that co-creativity is princi- pally the expression of the visitors’ creative impulses.

Theoretical Framework and Expectations Given lack of quantitative attempts in the literature to explain the processes shaping co-creative behavioral patterns, the article applies a broad sociological framework to explore the phenomenon, rather than a nar- rowly defined theory and hypothesis. Here, the article finds inspiration from Hedström’s (2005) analytical sociology, which proposes a heuristic topology of the main types of social mechanisms that may result in a group of individuals acting in a similar manner. According to Hedström (2005), such ‘behavioral uniformities’ can be broadly ascribed to effects of the spatial context, social composition and/or social transactions; that is, “an environmental effect is operative if we do what we do because we are where we are. A selection effect is operative if we do what we do because we are who we are. And finally, a social-interaction effect is operative if we do what we do because others do what they do.” (Åberg and Hedström 2011: 203, italics added). Applying this heuristic framework, the article raises the empirical question about which, if any, of these three processes lead visitors to partake in collective co-creative action, thus shaping festival settings into concentrated ‘hot spots’ of co-creativity. An important source for this argument is Pettersson and Getz’s (2009) identification of ‘experience hot spots’ as key festival attractions. However, instead of understanding such festival hot spots as being defined by their experience content, the article specifically examines hot spots as zones of co-creative behaviors. That is to say, the article addresses where and why co-creative hot spots are enacted, rather than how they are experienced, as it has been described in the literature’s ethnographical studies. Seen through the prism of Hedstöm’s topology, we may formulate the following general expectations regarding the mechanisms that shape hot spots of festival co-creativity. First, in terms of the festival spatial environment, a key question concerns the ratio of numbers of festival visitors present and the area volume of the festival setting, i.e. level of social density or event crowding. Spatially accessible areas of festival sites are often centers for movement, high densities and crowding (Batty, Desyl- las and Duxbury 2003; Morgan 2007). The festival event literature has

180

traditionally associated high levels of event crowding with negative experiences and spaces that visitors tend to avoid. However, more recent studies suggest that, depending on the social context and the visitors’ expectations, event crowding may also be perceived as creating attractive social atmospheres (Wickham and Kerstetter 2000; Popp 2012). This latter view is supported by the ethnographic studies on co-creativity that suggest that crowd-generated social atmospheres encourage festival visitors to engage in co-creative activities (Morgan 2007; Stevens and Shin 2012). The article considers it an empirical question which of these op- posite spatial environmental processes shapes the co-creative hot spots. Second, in terms of selection processes, co-creative hot spots may be attributed to preferences that lead visitors with a similar creative taste to select themselves into specific areas of the festival. According to Schel- ling’s (1971) discussion about the effects of such social sorting, even minor socio-spatial preferences can result in profound patterns of hot spot clustering at the aggregate level. As emphasized by Getz (2007: 324) and Pettersson & Getz (2007), selection effects are an understudied theme with regard to festival events: “An event includes many actors or target actors, and future research should segment the different groups.” (Pettersson and Getz 2007: 324). The article examines social sorting processes as segmented group preferences that crystallize into hot spot clusters of co-creative behaviors. Third, in terms of interactional processes among festival visitors, the co-creative hot spot patterns may arise when spatially proximate visitors inspire each other to partake in a given co-creative activity. In statistical terms, this kind of endogenous interaction effect is often described as a ‘spatial spillover effect’ that shapes social phenomena into a geographical hot spot pattern (Anselin 2002; Ward and Gleditsch 2008). Important in this regard is that ethnographic festival studies suggest that social interaction among festival visitors is a vital attraction for partaking in a co- creative use of space (Morgan 2007; Samson 2011; Stevens and Shin 2010). “[T]hese moments of amazement and these social interactions are as likely to be found in informal fringe events as in the main attraction.” (Morgan 2007: 14). To the extent that these processes of amazement and social interaction invite proximate visitors to engage in similar co- creative actions, the presence of co-creative hot spots should be ascribed to the effects of social interactions.

181

Data and Method We might ask why the aspect of co-creative behavior is understudied compared to the extraordinary experiences correlated with such activities? Most probably, this should be ascribed the methodological difficulties associated with micro-sampling this kind of high-resolution data about actual behaviors (Pettersson and Zillinger 2011; Collins 1983). To deal with this challenge, the article follows Pettersson and Getz’ (2007: 324) suggestion to apply aerial photographs as a mean to collect these festival event behaviors. Moreover, this micro-sampling strategy is consistent with Collins’ (1983; 2004) micro-sociological suggestion to “(…) engage in systematic sampling of certain kinds of microsituations. In- stead of sampling individuals, as in conventional survey research, we sample encounters.” (Collins 1983: 195). While the survey method (as applied in the aforementioned bridge building quantitative studies) is a suitable method to measure the visitors’ experiences, micro-sampled situational or event data allow us to examine and explain co-creative behaviors as they are taking place in situ in their natural festival contexts. In practical terms, this micro-sampling strategy applies the extraordinary high-resolution aerial photographs taken by the promoters of the Roskil- de Festival. These photographs allow visual identification and geocoding not only of a range of social and physical properties of the festival setting, but also a type of co-creative festival activity that is emblematic of the Roskilde Festival: that is, the drinking game called ‘beer bowling’ (Jensen et al. 2013; Sørensen 2009: 94; Marling and Kiib 2011). The beer bowling game uses the following convention (see Figure 1). Two competing teams, each consisting of two persons, sit side by side in camping chairs at a distance of approximately 5-7 meters. Each team has an empty bottle of beer standing in the gap between the camping chairs, and the goal of the game is to knock over this bottle with a ball, which the teams take turns bowling. The alcoholic rule of the game is that if their bottle is knocked over, a team has to drink a beer. Furthermore, this and the other rules of the game are regulated by a judge whose word is ‘law’, and any violation or lack of ‘respect’ is ‘penalized’ with beer sips – often applauded vociferously by the surrounding audience.

182

Figure 1. A beer bowling game. Note the ball being bowled by the team away from the camera. Author’s photography.

In terms of theory, the beer bowling events can be described as an ar- chetypal form of co-creative practice. It is an activity that the festival visitors orchestrate on their own creative terms and is characterized as a practice that many festival visitors consider key to their positive festival experience (Sørensen 2009). Moreover, the beer bowling phenomenon is methodologically unique because its quadratic game structure makes it possible to identify from above. Accordingly, all pathways and open areas of the festival space have been minutely examined several times and all events are geocoded in the GIS software Mapinfo. In total, 243 events are identified in camping area of the festival. Note that the article focuses solely on the camping area of the festival, thus excluding from analysis the fenced music area in which the venues are located. This highly stage- managed area contains no beer bowling events.

183

Figure 2. All 243 events geocoded on the aerial photography (left). Two events of beer bowling on the horizontal path segment (right).

Statistical Techniques: Quadrant Analysis and Spatial Regression The article applies two statistical tools to analyze the spatial distribution of the beer bowling events. First, quadrant analysis is applied to assess whether the beer bowling events are distributed randomly or clustered into spatial hot spots. The festival camping area has been partitioned into 132 equal sized quadrants, allowing for the calculation of a variance- mean ratio (VMR) by which the observed number of beer bowling events per quadrants is compared with the average number of events per quadrant. Then, a student’s t-test is applied to test whether the observed point pattern is significantly more clustered than one would expect by chance (Burt, Barber and Rigby 2009: 533ff). Second, a spatial logistic regression model is applied to examine the factors that may be causing the beer bowling events to cluster into a hot spot pattern. This logistic model is estimated in Stata applying Firth’s penalized maximum likelihood method, which is more robust when modeling data with a relatively low event rate, as in the case of this article (Zeng et al. 2013).

Measures and Definitions Measuring any sort of events in space implies a decision on what spatial unit of observation best suits the research perspective. Here, the article is inspired by Groff, Weisburd & Yang (2010) and Hillier & Raford (2010), who suggest that social phenomena benefit from being studied at the utter spatial micro-scale. Micro-spatial units of observation reduce the modifiable area unit problem (MAUP) that affects results when event- or point-based measures are aggregated into large area units (Openshaw 1984). Accordingly, I have compartmentalized festival space into the smallest meaningful areal units: 30,381 points of 4x4 meters area (in the following these areas will be referred to as ‘micro-spaces’). Note that this

184

micro-spatial area unit approximately corresponds to the spatial propor- tions of a beer bowling site, eliminating the possibility of more than one event per micro-space. Technically, this compartmentalization involves two steps: first, all walkable surfaces of the festival setting are indexed in the GIS software Mapinfo; second, the open source software UCL Depthmap is applied to automatically generate the micro-space points (Turner 2004).

Table 1. Descriptive statistics of variables used in the logistic model (unstandardized) Variables Mean Std. Dev. Min Max Beer bowling events 0.01 0.09 0 1 Dummy-coded camp proximity Several camp sites 0.23 0.42 0 1 One camp site 0.25 0.43 0 1 None camp site (ref.) 0.52 0.50 0 1 Spatial lag of beer bowling events 1.89 1.94 0 10 Movement density -1.27 0.74 -4.21 0

The dependent co-creativity variable to be included in the logistic model is measured as a binary outcome, where Y = 1 indicates an event and Y = 0 indicates that no beer bowling event is occurring in the micro-space. The regression model includes three independent variables. First, the model includes a spatially lagged version of the dependent beer bowling variable as a predictor, measuring the number of events spatially proximate from each micro-space. Spatial lags are often included in spatial regression models in order to control for spatial autocorrelation that violates the assumption of independent observations, i.e. that the values of y in one unit i are directly influenced by the values of y found in i’s neighbors (Ward and Gleditsch 2008). However, in our case, spatial lag is not included merely to control for such autocorrelation, but as a mean to explain the socio-spatial mechanisms that possibly shape the co-creative actions into spatially clustered hot spots. As we will discuss below, a significant spatial lag in beer bowling may be interpreted both as indicating a selection effect and as indicating a social-interactional effect. In technical terms, the spatial lag is measured as the number of events spatially accessible within 80 meters walking distance from each micro-place (An- selin 1992; Burt, Barber, and Rigby 2009: 570). The variable is calculated in Mapinfo, applying the place syntax tool (Ståhle, Marcus, and Karls- strøm 2005). Second, the model includes a proxy variable predicting the average density flow of visitors moving across each micro-place. This variable

185

operationalizes the spatio-environmental effect in terms of density or event crowding. In constructing this variable, the article applies the space syntax tool, ‘visibility graph analysis’ (Turner et al. 2001), which lets us measure the distance from each micro-space to the closest of the three entry gates to the music area. This reflects the assumption that the music area acts an extraordinarily strong ‘magnet’ on movement throughout festival space (Karlström and Mattssson 2009). The distance to the entry gates is measured as the minimum angular path, that is, the route that will result in the minimum change in direction as one is walking through the festival path network. The space syntax literature clearly indicates that an angular distance property of this type is far more significant than metric properties of space as individuals engage in finding their way through built environments (Hillier and Iida 2005; Turner 2009; Dalton 2003). Moreover, it makes intuitive sense that the linear artery paths that radiate directly (i.e. with a low angularity) from the entry gates will be spatial points crowded with movement. All space syntax calculations are made in UCL Depthmap. Given that space syntax tools (including visibility graph analysis) have not been previously applied to festival settings, this way of measuring movement by proxy is validated against actual movement data. Follow- ing Grajewski’s (2001) methodological suggestion, movement data were collected at 70 count points, randomly located around the Roskilde Fes- tival site. Each count point was observed 5 times for 3 minutes each time, and a count was made of all visitors passing by foot. It should be noted that movement data was collected during the 2012 festival while visibility graph analysis was conducted on the basis of aerial photography from 2011. However, this should be considered a limited source of measurement error given that the spatial layout was nearly identical for these two years. The correlation between angular distance to entry gates and movement data is examined with a linear regression model. The model includes a spatial lag of the movement outcome in order to account for autocorrelation. The software Geoda is applied to calculate the spatial lag of movement (Anselin, Syabri and Kho 2006). Estimated results are presented in the results section of the article. Third, the model includes a variable to measure the number of camping tents located in immediate proximity to each micro-space. This variable accounts for the fact that several of the pathways with the least angular distance to the entry gates (i.e. spaces with potentially high movement flow) are pathways that only have tent camps on one side or none (see Figure 2 for an example of pathways with tent camps on both sides). Without controlling for this spatial circumstance, we might conflate ef-

186

fects of spatio-environmental movement density and the effects of the festival’s spatial organization of tent areas and pathways. Further, the variable is included not merely as a control covariate, but also to capture a potential social-interactional effect. The extent to which a festival group occupies micro-space in closest proximity to their tent camp might be interpreted as indicating that the group’s interactional process spills over into the nearby festival space by means of beer bowling. This variable is included as three dummy variables, to distinguish whether the micro-space neighbors on several, one, or no tent camps (reference category).

Results This section presents the results of the three statistical models used in the article: the quadrant analysis that explores whether the co-creative beer bowling events are significantly clustered into spatial hot spots; the spatial linear regression model that validates angular distance to entries as a proxy measure for movement density; and the spatial logistic regression model that examines the mechanisms that might shape the beer bowling events into spatial hot spots.

Quadrant Analysis: Exploring the Co-Creative Hot Spots The variance mean ratio indicates that the variance is significantly greater than the mean, VMR=1.5; p<.01, suggesting that beer bowling events are more spatially clustered than one would expect by chance. This points to the existence of a social and/or spatial effect that shapes this co-creative phenomenon into a hot spot pattern. In addition, spatial clustering can also be visually verified from the aerial photography where the southern and northeastern camping areas are readily identified as zones with a low and high beer bowling frequency, respectively (see Figure 2).

Movement Analysis: Measuring Movement Density by Proxy Table 2 presents the result of the spatial regression analysis examining the relationship between the angular distance to entry gates and movement counts. Model 1 suggests that angular distance is highly significant and strongly negatively correlated with movement. Note that a partial correlation coefficient greater than .50 should be considered a large effect size (Cohen 1992). This result strongly suggests that pathways at a greater angular distance from the entry gates are less likely to be densely packed with movement flows.

187

Table 2. Spatial regression analysis of movement (N=70) M1 M2 Predictors Coef. Part Corr. Coef. Part Corr. Angular distance to entry gates -0.78 -.61 0.68(0.12)*** -.53 Spatial lag of movement (0.12)*** 0.45(0.12)*** .35

Constant 2.67(0.12)*** 2.67(0.11)*** Adjusted R-square 0.36 0.47 The angular predictor and movement outcomes are log transformed due to skewness. Results are reported as unstandardized coefficients and partial correlations. Standard errors in parenthesis. *p<0.10; **p<0.05; ***p<0.01.

Model 2 includes the spatial lag of movement as a predictor in order to control for spatial autocorrelation. Note that a Moran’s I statistics, I = .44; p < .001, suggests that movement data is characterized by a significant level of spatial clustering. After controlling for spatial autocorrelation, the unique effect of angular distance is still highly significant with a smaller, but still large effect size. The significant effect of the spatial lagged version of movement may be attributed to the fact that movement flows spill over to their neighboring micro-spaces. In sum, given the effect size and robustness of these results, the angular distance to gates may be considered as a valid movement density proxy to be included in the logistic model.

Logistic Analysis: Explaining the Co-Creative Hot Spots Table 3 presents the estimated results of the spatial logistic regression model examining the processes that shape certain spaces into hot spots of bowling co-creativity. Starting from the top, it may be established that the probability for beer bowling is substantially higher in spaces that neighbor upon one or several tent camps. This result suggests the existence of an interactional effect by which the visitors’ group transactions and collective decision to beer bowl spill into the nearest available micro- spatial site. One could object that this result could also be interpreted as an environmental effect by which it is the spatial distance to the tent camps that directly shapes the micro-spatial possibility for beer bowling events. However, following Gans (2002), it seems more reasonable in this case to specify the role of space as an ‘intervening variable’ between the visitor’s interaction and collective performance of the co-creative event; that is to say, festival space only has an indirect effect as a ‘friction of distance’, which may constrain, but does not constitute, the social interactional activity of beer bowling.

188

Table 3. Spatial logistic analysis of beer bowling events Predictors Beta coef. Odds ratio Dummy-coded camp proximity Several camp sites 2.82(0.28)*** 16.73 One camp site 1.94(0.28)*** 6.93 None camp sites (ref.) - - Spatial lag of beer bowling 0.21(0.06)*** 1.24 Movement density -0.14(0.06)** 0.87 Constant -6.878(0.25)*** Results are reported as standardized logit coefficients and odds ratios. Standard errors in parenthesis. *p<.10; **p<.05; ***p<.01.

Next, we consider the highly significant spatial lag of beer bowling. The positive estimate suggests that a micro-space has a higher probability to become a beer bowling event site if many of its adjacent micro-places are beer bowling sites. This may be interpreted as a spatial spillover effect that develops as an interactional effect: visitors in one part of the camping area get inspired by, start to imitate, or simply are invited to beer bowl against neighboring camps. Such interactional spillovers, in turn, generate a multiplier effect by which the co-creative activities feed back on the individuals who were the original co-creative source of inspiration, and who are now reminded of their enjoyment of the game and stimulated to bowl again. A self-reinforcing hot spot of festival co- creativity has been created. Cold spots follow the inverse logic, by which camping areas empty of co-creative actions generate a self-accelerating emptiness; as suggested by Girard (1987: 29), mimetic interactional processes may develop as positive or negative feedback mechanisms (cf. moreover Hillier 1996: 112; Anselin 2009). At this point, we should take note of Manski’s (1993: 531-532) argument that ‘endogenous’ interactional effects of this kind, that lead individuals to behave uniformly, are difficult to separate from the effects of ‘correlated’ selection processes and ‘exogenous’ environmental effects. Take the following example inspired by Hedström (2005: 46): when festival visitors all start to wear sunglasses it is difficult to identify if this a result of imitation of other visitors (interaction effect), a group specific preference for wearing sunglasses (selection effect) or that it simply is unpleasant to be blinded by the sharp summer sun (environment effect). Along the same lines, the significant spatial lag of beer bowling may also be interpreted as a selection effect by which visitors search out camp areas having a level and type of co-creative action that match their pref-

189

erences (McPherson et al. 2001; Collins 2004; 151). This kind of selection process might explain the almost total absence of beer bowling events in the southern camping areas of the festival, which are stage-managed by the festival as a ‘silent and clean’ zone; this leaves the social sorting mechanism to develop almost perfectly, making this an extraordinary cold spot for co-creativity. In sum, the pragmatic solution to this identification problem is to interpret the significant spatial lag of beer bowling as capturing a ‘net-effect’ (Elster 2007: 40) of interaction and selection on the co-creative hot spots rather than a separate effect. Finally, the results show that the movement density measure is significantly negatively correlated with the beer bowling events. The higher the movement density, the lower the probability that the micro-space will be a beer bowling event site. This result suggests that the festival setting itself also has a spatio-environmental effect on the co-creativity concentration process: in the movement intense pathways at a close angular distance to the entry gates, the movement congestion makes it practically inconvenient and socially unattractive to beer bowl. These movement dense micro-spaces are cold spots for co-creative beer bowling activities. In more theoretical terms, the event crowding in these micro-spaces counteracts the beer bowling activity developing as a successful and attractive ‘interaction ritual’ (Collins 2004: 48f). That is to say, the high movement density makes it difficult to sustain the mutual focus on the co-creative activity, uphold the quadratic play formation, or ensure that the effervescent mood of the beer bowling participants is not overridden by a hectic atmosphere. Faced with an environmental effect of this kind, the festival visitors search out other less crowded and micro-spaces better suited for beer bowling, thus acting as ‘social space explorers’, as defined by Hillier (1996: 155): “I suggest that all social space explorers tend to follow the same principle of occupying the most integrating lacunas available in the natural movement system.”

Discussion and Conclusions Seen through the sociological prism of Hedström’s heuristic topology regarding the three main processes (i.e. effects of interaction, selection and environment) that may lead a group of individual to act in a similar manner, the results of this research suggest that a mixture of mechanisms is at work when co-creative festival activities are shaped into hot spot pattern. First, the findings suggest that co-creative hot spots may be attributed to preferences that lead similar visitors to search out specific areas of the festival setting that match their co-creativity taste. Social

190

selection processes have not received adequate attention in festival event research and have not been previously related to the festival visitors’ co- creative use of festival space (Getz 2010: 21; Pettersson and Getz 2007: 324). Second, the article suggests that co-creative hot spots may be attributed to interactional spillover effects, by which ongoing co-creative activities invite and inspire other spatially proximate visitors to partake in the creative actions; such spillover effect, in turn, initiates a positive or negative multiplier effect that accelerated the patterning process. This sheds light on the feedback processes of interactional behaviors that condition the ‘experience hot spots’, which Pettersson and Getz (2009) described, but did not explain. Third, the article suggests that festival micro-spaces may be too crowded with movement for it to be practically possible and socially enjoyable to perform co-creative activities. The finding that movement density correlates negatively with co-creative actions differs from findings in the ethnographic literature on festival co-creativity (Morgan 2007), which suggest that event crowding constitutes an emplaced attraction that is conducive to co-creative sojourns. Here we might consider how to specifically interpret this negative association between event crowding and co-creative festival actions in the light of having operationalized co-creativity as beer bowling. The fact that this activity is performed while sitting at the center of the movement flow of a pathway may very likely have contributed to a negative feeling of ‘being in the way’, especially in the most movement dense micro- spaces. This may be illustrated by comparison to the bodily behavior of standing at the center of the movement flow, which urban sociology studies have shown to be a common practice that gives the involved pedestrians a positive sense of being where the action is (Whyte 2012: 10-16). Individuals who are standing may more easily make use of small bodily ’techniques for negotiation in motion’ (Jensen 2013: 148), that is, a slid- ing step to the side, a zigzagging move or letting someone pass through the group of individuals gathered at the center of the street. By contrast, sitting individuals are a ‘body plug’ that blocks the movement flow of the street. In addition to Mowen et al.’s (2002: 63) emphasis that there are “(…) variations in perceived crowding across different event subzones,” the above considerations suggest that the physical manner in which visitors are present in a festival micro-space may also result in variations in the perception of event crowding. On the most general level, the article suggests that co-creative use of festival space is not merely a result of the creative impulses of individuals attending the festival, but also has to be attributed to more ‘structural’

191

mechanisms (Hedström 2008: 323). This contrasts with the ethnographic festival studies, which emphasize an actor-centered explanation of co- creativity: because of the visitors’ agency initiative and creative terms, they ‘find centres, paths and domains of experimental space within the physical location’ (Morgan 2007: 25). Festival visitors are not simply ‘human props’ in a ‘carefully-staged managed performance’ (Stevens and Shin 2010; Morgan 2007), but demonstrate agency to move around freely, and to search out and thus co-create their own informal fringe events. This actor-centered perspective on co-creativity has two sources. First, the emphasis on the actors’ innate creativity is theoretically implied by the notion of ‘co-creation’. As initially proposed by Prahalad and Ramaswamy (2004), this notion accentuates the creative freedom of con- sumes as the vital force of contemporary barter markets. This seems to be the backdrop for van Limborg’s (2009: 116) structural suggestion to conduct more research “(…) not only on the consequences of co- creation, but also on the environment in which co-creation develops itself.” In a similar vein, Getz (2007: 21) has pointed out that “(…) an over-emphasis on consumer-behaviour theory and methods is limiting theoretical advancement in understanding antecedents for attending or participating in festival.” Second, I will argue that the actor-centeredness of the ethnographic studies is correlated with the inherent nature of the ethnographic methodology. While an ethnographic perspective gives us privileged knowledge about human meanings and interactions from the ‘standpoint of the participants’ (Jorgensen 1989: 12f), the method only provides limited insight into the structural ‘wholes’ – the emergent properties – of the social and physical space, which are not necessarily a product of intentional agency, nor always observable from the partial perspective of the participants. This is summarized by Hedström:

Social outcomes, like other emergent phenomena, are difficult to anticipate because the outcome depends to such a high degree on how the individual parts are interrelated. (…) small and seemingly unimportant changes in the ways the actors are interrelated can have profound consequences for the social outcomes that are likely to emerge. For this reason, social outcomes cannot simply be ‘read off’ from the properties of the individuals that generate them (Hedström 2007: 75).

Moreover, as Hedström suggests, social emergent outcomes, are better analyzed quantitatively using ‘formal analytical tools’, and the use of spa-

192

tial modeling tools in this article is in conformity with this suggestion. Quantitative tools allow analysis of the emergent hot spot pattern of co- creative behaviors, which would be difficult to ‘read’ if one had only observed the festival space ethnographically from the agent perspective of the participants. Thus, although social interactional effects are mediated by the visitors’ creative agency, the imitative spillover relations between adjacent camping areas also aggregate into an emergent pattern of co-creative hot spots. Along the same line, selection effects may unfold as purposive creative action for many festival visitors, but the aggregate result of this process is not necessarily reducible to such intentionally co-creative acts. And, even if some of the most experienced festival visitors are able to plan their camp with an eye to the environmental effects of festival space, the majority of visitors – and even, for that matter, the festival’s stage- managers (Mossberg 2007: 66) – do not have an overview of the ways micro-spatial movement distributions of the festival setting shape co- creative hot spot formation. In sum, following Schelling’s (1971: 145) emergence theoretical argument, the spatial hot spots of festival co- creativity identified by this article should, be considered as locally experienced and enacted hot spots “(…) that lead to aggregate results that the individual neither intends nor needs to be aware of, results that sometimes have no recognizable counterpart at the individual level.”

References Anselin, L. (1992). Spatial data analysis with GIS: an introduction to application in the social sciences. Technical Report 92-10, National Center for Geo- graphic Information and Analysis University of California. Anselin, L. (2002). Under the hood issues in the specification and interpretation of spatial regression models. Agricultural economics, 27(3), 247-267. Anselin, L., Syabri, I., & Kho, Y. (2006). GeoDa: An introduction to spatial data analysis. Geographical analysis, 38(1), 5-22. Batty, M. (2007). Cities and Complexity: Understanding Cities with Cellular Automata, Agent-Based Models, and Fractals. Boston: The MIT press. Batty, M., Desyllas, J., & Duxbury, E. (2003). Safety in numbers? Modelling crowds and designing control for the Notting Hill Carnival. Urban Studies, 40(8), 1573-1590. Boden, D., & Molotch, H. (1994). The Compulsion of Proximity. In R. Fried- land, & D. Boden (Eds.), Now/Here: Space, Time and Modernity. Berkeley: Uni- versity of California Press Burt, J. E., G. M. Barber, and D. L Rigby (2009). Elementary Statistics for Geogra- phers. The Guilford press. Cohen, J. (1992). A power primer. Psychological bulletin, 112(1), 155-159. Coleman, G. (2010). The hacker conference: A ritual condensation and celebra- tion of a life world. Anthropological Quarterly, 83(1), 47-72.

193

Collins, R. (1983). Micromethods as a Basis for Macrosociology. Journal of Con- temporary Ethnography, 12(2), 184-202. Collins, R. (2004). Interaction Ritual Chains. Princeton: Princeton University Press. Cresswell, T. (2006). On the Move. Mobility in the Modern Western World. London & New York: Routledge. Dalton, R. C. (2003). The Secret Is To Follow Your Nose Route Path Selection and Angularity. Environment and Behavior, 35(1), 107-131. Elster, J. (2007). Explaining Social Behavior: More Nuts and Bolts for the Social Scienc- es. Cambridge: Cambridge University Press. Gans, H. J. (2002). The sociology of space: a use–centered view. City & Com- munity, 1(4), 329-339. Getz, D. (2004). Geographic perspectives on event tourism. In A A. Lew, C. M. Hall, & A. M. Williams (Eds.). A Companion to Tourism. John Wiley & Sons. Getz, D. (2010). The nature and scope of festival studies. International Journal of Event Management Research, 5(1), 1-47. Girard, R. (1987). Things Hidden since the Foundation of the World. Palo Alto: Stan- ford University Press. Groff, E. R., Weisburd, D., & Yang, S. M. (2010). Is it important to examine crime trends at a local “micro” level?: A longitudinal analysis of street to street variability in crime trajectories. Journal of Quantitative Criminology, 26(1), 7-32. Grajewski T. (2001). Space Syntax Observation Manual. Bartlett School of Graduate Studies, University College London. Hedström, P. (2005). Dissecting the Social: On the Principles of Analytical Sociology. Cambridge: Cambridge University Press. Holloway, I., Brown, L., & Shipway, R. (2010). Meaning not measurement: Using ethnography to bring a deeper understanding to the participant experience of festivals and events. International Journal of Event and Festival Manage- ment, 1(1), 74-85. Hillier, B. (1996). Space is the Machine. Cambridge: Cambridge University Press. Hillier, B., & Iida, S. (2005). Network and psychological effects in urban movement. Spatial information theory, 3693, 475-490. Hillier, B., & Raford, N. (2010). Description and discovery in socio-spatial analysis: the case of space syntax. In: G. Walford, E. Tucker, & M. Viswanathan (Eds.), The SAGE Handbook of Measurement. London: Sage. Jamieson, K. (2004). Edinburgh. The Festival Gaze and Its Boundaries. Space and Culture, 7(1), 64-75. Jensen, A. I., Villadsen, J. S., Hansen, J. V., & Larsen, J. L (2013). This Time It’s Political? En analyse af Roskilde Festival som appliceret urbanitet, Master thesis, Roskilde Universitet. Jorgensen, D. L. (1989). Participant Observation: A Methodology for Human Studies. London: Sage. Karlström, A., & Mattsson, L. G. (2009). Place, space syntax and attraction- accessibility. Proceedings of the 7th International Space Syntax Symposium, Stockholm: KTH. Klüpfel, H. (2007). The simulation of crowd dynamics at very large events. Calibration, empirical data, and validation. In R. D. Peacock & J. D. Averill (Eds.). Pedestrian and Evacuation Dynamics. Berlin & Heidelberg: Springer.

194

Manski, C. F. (1993). Identification of endogenous social effects: The reflection problem. The Review of Economic Studies, 60(3), 531-542. Marling, G. & Kiib, H. (2011): Instant City @ Roskilde Festival. Aalborg: Aalborg University press. Morgan, M., (2007). Festival Spaces and the Visitor Experience. Casado-Diaz, M., Everett, S., & Wilson, J. (Eds). Social and Cultural Change: Making Space(s) for Leisure and Tourism. Eastbourn: Lesiure Studies Association. Mossberg, L. (2007). A marketing approach to the tourist experience. Scandina- vian Journal of Hospitality and Tourism, 7(1), 59-74. Openshaw, S. (1984). The Modifiable Areal Unit Problem. Norwich: Geo Books Popp, M. (2012). Positive and negative urban tourist crowding: Florence, Italy. Tourism Geographies, 14(1), 50-72. Pettersson, R., & Zillinger, M. (2011). Time and space in event behaviour: tracking visitors by GPS. Tourism Geographies, 13(1), 1-20. Pettersson, R., & Getz, D. (2009). Event experiences in time and space: a study of visitors to the 2007 World Alpine Ski Championships in Åre, Sweden. Scandinavian Journal of Hospitality and Tourism, 9(2-3), 308-326. Prahalad, C. K., & Ramaswamy, V. (2004). Co-creation experiences: The next practice in value creation. Journal of Interactive Marketing, 18(3), 5-14. Raybould, M., Mules, T., Fredline, E., & Tomljenovic, R. (2000). Counting the herd. Using aerial photography to estimate attendance at open events. Event Management, 6(1), 25-32. Samson, K. (2012). Den performative by - design, planlægning og oplevelser. In J. Andersen, M. Freudendal-Pedersen, L. Koefoed, & J. Larsen (Eds.), Byen i bevægelse. Roskilde: Roskilde Universitetsforlag Schelling, T. C. (1971). Dynamic models of segregation. Journal of mathematical sociology, 1(2), 143-186. Stevens, Q. & Shin, H. (2012). Urban Festivals and Local Social Spaces, Planning Practice and Research, 29(1), 1-20. Stopczynski, A., Larsen, J. E., Lehmann, S., Dynowski, L., & Fuentes, M. (2013). Participatory Bluetooth Sensing: A Method for Acquiring Spatio- Temporal Data about Participant Mobility and Interactions at Large Scale Events, International Workshop on the Impact of Human Mobility in Pervasive Sys- tems and Applications, San Diego, 18 March 2013. Ståhle, A., Marcus, L., & Karlström, A. (2005). Place Syntax: Geographic accessibility with axial lines in GIS. Proceedings of the 5th International Space Syntax Symposium, TU Delft. Sørensen J K, 2009, Substance use, rituals and risk management Danish rock festivals, PhD thesis, Department of Sociology, University of Copenhagen. Turner, A. (2004). Depthmap: A Researcher’s Handbook. Bartlett School of Graduate Studies, University College London. Turner, A., Doxa, M., O’sullivan, D., & Penn, A. (2001). From isovists to visibility graphs: a methodology for the analysis of architectural space. Environ- ment and Planning B: Planning and Design, 28(1), 103-121. Turner, A. (2009). The role of angularity in route choice. Spatial information theory. 9th International Conference, COSIT 2009 Aber Wrac’h, France. . Urry, J. (2007). Mobilities. London: Polity. van Limburg, B. (2009). Innovation in pop festivals by cocreation. Event Man- agement, 12(2), 105-117. Virilio, P. (2001). Virilio Live: Selected Interviews. London: Sage.

195

Wickham, T. D. & Kerstetter, D. L. (2000). The relationship between place attachment and crowding in an event setting. Event Management, 6(3), 167- 174. Ward, M. D & Gleditsch, K. S. (2008). Spatial Regression Models. London: Sage. Zakrisson, I., & Zillinger, M. (2012). Emotions in motion: tourist experiences in time and space. Current Issues in Tourism, 15(6), 505-523. Zeng, C., Newcomer, S. R., Glanz, J. M., Shoup, J. A., Daley, M. F., Hambidge, S. J., & Xu, S. (2013). Bias Correction of Risk Estimates in Vaccine Safety Studies With Rare Adverse Events Using a Self-controlled Case Series De- sign. American Journal of Epidemiology, 178(12):1750-9. Åberg, Y., Hedström, P (2011). Youth unemplyment: a self-reinforcing proces. In P. Demeulenaere (Ed.). Analytical Sociology and Social Mechanism. Cam- bridge: Cambridge University Press.

196

Konkluderende bemærkninger og resumé

SOM JEG ARGUMENTEREDE FOR indledningsvis, tager afhandlingen sit afsæt i et rumsociologisk spørgsmål, der blev undfanget på Dronning Louises Bro: Hvilken effekt har det fysiske mikro-rum i sig selv for skabelsen af følelsesintense mikrosituationer? Den empiriske undersøgelse af dette forudsætter en række teoretiske og metodologiske rekonstruk- tioner af det rumsociologiske felt. Teoretisk foreslår afhandlingen relevansen af Durkheim, Mauss og Co’s socialmorfologiske bidrag, der har rumsociologisk blik for, hvordan det fysiske rum betinger sociale funktioner og dynamiske densiteter, og hvordan graden af densitet er en væ- sentlig kausal komponent for den følelsesmæssige intensitet af mikro- interaktionsprocesser. Hilliers space syntax-reception af Durkheim er afgørende for videreudviklingen af dette rummorfologiske argument, idet Hillier gør det muligt at specificere rummets netværkshandlen som den morfologiske mekanisme, der fortætter, og hermed potentielt følelsesin- tensiverer, sociale interaktionssituationer. Metodologisk har afhandlingen ligeledes fundet inspiration i space syntaxen, der tilbyder det som rumsociologien og den socialmorfologiske tradition mangler: Kvantitative værktøjer til at måle det fysiske mikro-rums sociale logikker og rumlige effekter. I afhandlingens empiriske undersøgelser kombineres den mikro- durkheimianske teoritilgang med et space syntax-inspireret metodedesign til måling og modellering af mikro-rummets egenskaber og effekter.

Mikro-rumligt inviterede kropsteknikker I afhandlingens tre empiriske bidrag er bevægelsesdensitets-variablerne korreleret signifikant med undersøgelsernes respektive operationaliseringer af mikro-interaktionel intensitet. Som vi har set, har korrelationerne imidlertid forskellige retninger. Mens bevægelsesdensiteten bidrager positivt til den fænomenologiske oplevelse af bymæssig attraktionsværdi (kapitel 3) og kollektiv opbrusen i festivalkonteksten (kapitel 4), er sam- menhængen negativ, når det gælder de medkreative forsamlinger i festivalrummet (kapitel 5). Hvordan skal vi fortolke denne forskel? Fremfor at betragte dem som empirisk inkonsistent resultater, vil jeg argumentere for, at de tre resultater tilsammen giver et sammenhængende billede af samspillet mellem rum, bevægelse og følelsesmæssig intensitet. Dette viser sig, når bevægelsesdensitetens kausale effekt på den situationelle følelsesintensitet fortolkes og specificeres som en kropssociologisk mo- dereret proces. Hvorvidt bevægelsesdensiteten er negativt eller positivt korreleret med den stedslige intensitet modereres af, hvordan individer

197

befinder sig kropsligt i det bevægelsestætte mikro-rum. Bevægelsestætte mikro-rum inviterer til, at individer befinder sig i rummet på en bestemt kropsteknisk måde, mens andre kropsteknikker kun vanskeligt, og måske ligefrem ubehageligt, lader sig anvende i sådanne rum. Før dette forklares nærmere i relation til de tre empiriske resultater, vil jeg gøre mig et par teoretiske overvejelser om, hvordan individernes kropslige måde at være i mikro-rummet kan spille ind på rumpraktikkens (u)behagelighed. Med krop og kropsteknikker refererer jeg til Mauss’ (1979) begreb om “kroppens teknikker”, der påpeger, hvordan ethvert samfund har en sær- egen kropshabitus. Kropsteknikker som “at gå” eller “at sidde” læres i overensstemmelse med samfundets gældende habituelle “tradition”. Kropsteknikker er imidlertid ikke blot et spørgsmål om samfundsspeci- fikke traditioner, men også om, hvad der er “effektivt” i situationen. Hvis ikke kropsteknikkerne også er effektive til at have den effekt, som de anvendes med henblik på at realisere, mindskes (for mange, om end ikke alle, teknikkers vedkommende) sandsynligheden for, at de overleve- res gennem den samfundsmæssige tradition: “I call technique an action which is effective and traditional (…).” (Mauss 1979: 75).34 I første omgang vil jeg fokusere på effektivitetsaspektet af kropsteknikkerne, hvilket specificerer det argument, som jeg berørte ovenfor: Ikke alle kropsteknikker er lige behagelige og i denne forstand effektive at benytte sig af, når man gør kropslige ophold i bevægelsestætte mikro- rum. Det væsentlige for dette argument er, at spørgsmålet om hvorvidt det er behageligt eller ej at opholde sig fysisk i rummet, ikke alene kan tilskrives tradition og socialt tillært smag, men også er produkt af en “psykologisk mediator” (Mauss 1979: 73), der så at sige regulerer, hvorvidt måden at være kropsligt i rummet på opleves som behagelig eller ej. Mauss (1979: 85) udpeger “(…) psychological facts as connecting cogs and not as causes, except in moments of creation or reform.” Argumen- tet er altså, at anvendelsen af kropsteknikker oftest er et sociologisk spørgsmål om tillærte kropstraditioner men, at psykologien kan spille en væsentlig medierende rolle for kropstekniske udfoldelser. Her vil jeg følge denne mausske forståelse af “psychological facts as connecting cogs” samtidigt med, at det også indrømmes, at det ikke er

34 Det begreb om kropsteknisk habitus som Mauss her fremfører, er del af inspiratio- nen for Bourdieus (2002; 1984) habitusbegreb (se Ingold 2000). Bourdieus habitusbegreb synes dog at subsumere spørgsmålet om effektivitet under traditionsaspektet: Hvorvidt en kropsteknik kan anses for effektiv, er for ham et spørgsmål om sociale klassers lokale og relationelle kropstraditioner. Den særligt ranke måde som overklassen går på, er ikke essentielt set mere effektiv end andre måder at gå på, men skal tilskrives en habituel internalisering af en klassespecifik kropstradition.

198

helt klart, hvilken type psykologisk “cog-wheel”-mekanisme som Mauss refererer til. For afhandlingens argument er det afgørende, at der ikke er tale om en psykologisk mekanisme, der bevæger sig socialkonstruktioni- stisk væk fra det fysiske rums eksistens og effekt. Jeg vil således specificere, den for afhandlingens væsentlige, psykologiske mekanisme med reference til psykologen James Gibsons (1986) økologiske psykologi. Specifikt vil jeg trække på hans materialitets-sensitive begreb om “affordances”, der peger på, hvordan objekter og rumlige omgivelser i sig selv tilbyder eller inviterer individer til bestemte handlemuligheder (Gibson 1986). Affordances præger hermed den måde, hvorpå mennesker befinder sig kropsligt-psykologisk i verden. En stol inviterer til at siddes på, eller sagt med andre ord: Stolens iboende affordance gør, at den er kropsteknisk mere effektiv og behagelig at sidde på end eksempelvis at ligge på (det sociologiske spørgsmål angår omvendt, hvordan den kropslige anvendelse af stolen er socialt og normmæssigt reguleret). Denne psykologiske “cog-wheel”-mekanisme kaster lys over afhandlingens empiriske eksempler og resultater: Bevægelsestætte mikro-rum inviterer til bestemte kropsteknikker, hvis anvendelse er en betingelse for, at opholdet i rummet kan være anledning til positive frem for negative intensi- tetsfølelser. Det næste vi må overveje er, hvilke konkrete kropsteknikker, der er på spil i afhandlingens tre empiriske analyser. Dette er ikke eksplicitret i de empiriske analyser, men underforstås i analysernes respektive operationaliseringer af mikro-interaktionel følelsesintensitet. I alle tre tilfælde befinder vi os inden for det, som Mauss (1979) kalder “techniques of rest”, idet analyserne underforstår eksistensen af måder at sidde og at stå på i mikro-rummets bevægelsesfortættede strøm, samt en variant heraf, hvor individerne sanser bevægelsesstrømmen siddende eller stående på en vis rumlig afstand. Det er afgørende, at disse forskellige kropstekniske former for hvile placerer individerne på forskellige måder i rummets bevægelsesflow, og således giver anledning til mere eller mindre behagelige erfaringer. I dette lys kan forklaringen på den ene negative og de to positive korrelationer mellem bevægelsesdensitet og intensitetsmålene tilskrives det forhold, at en høj bevægelsesdensitet udgør en negativ invitation til en bestemt hvileteknik og udgør en positiv invitation til andre hvileteknikker. Hvordan dette er tilfældet i analyserne, kan bedst forklares med udgangspunkt i figur 9, der illustrerer et vejforløb med en høj netværksmæssig tilgængelighed, og som funktion heraf, en høj bevægel- sesdensitet i begge retninger:

199

Figur 9. En vej med et højt bevægelsesflow

I relation til analysen af den kollektive opbrusen (kapitel 4) befinder individerne sig kropsteknisk i kantzonen af gadens bevægelsesflow, dvs. på en af de sorte steger, der afgrænser gaden, og hvor festivalgæsternes lejre er placerede. I denne kantzone sidder og/eller står gæsterne kropsteknisk i deres lejre og sanser, i denne forstand, gadens livlige bevægel- sesflow på en vis rumlig afstand: Ved at høre og se det fortættede liv, der udspiller sig i gaderne rundt om deres lejr. Her er det vigtigt, som Løgs- trup (1995: 15) i forlængelse af Heidegger (2007) fænomenologisk har gjort opmærksom på, at sansningen er defineret ved dens “afstandsløs- hed”. Blikket er ude ved det sete. Og når man hører festivalens medrivende larm, er man ude ved det hørte. Fornemmelsen af at være der, hvor tingene sker, trænger sig således positivt og afstandsløst på i sansningen af festivalrummet uagtet, at dette sanses på fysisk afstand. En sådan sansning af festivalrummets bevægelsesflow på siddende og/eller stående afstand, realiserer således kropsteknisk en af festivalrummets iboende handlingsmuligheder for at nyde festivalens rumligt medrivende livlighed. Festivalrummet inviterer, så at sige, festivalgæsterne til at nyde festivalspektaklet ved at gøre sanseligt ophold i rummets kantzone. Den positive statistiske korrelation mellem bevægelsesdensitet og intensiteten af den kollektive opbrusen må forstås i lyset af, at festivallivet lader sig effektivt kropsteknisk nyde gennem sansning af dette liv siddende og/eller stående fra festivalgæsternes lejre. Hertil skal det tilføjes, at denne betragtning befinder sig tæt på Jonges (1967) rumsociologiske begreb om kant-effekter (“edge effects”), der fremhæver, hvordan gode opholdszoner i byer ofte befinder sig i rummets udkant eller overgangszoner. En ting, der gør sådanne kanteffekter nydelsesmæssigt effektive, er det forhold, at man derfra kan sanse byens livlige spektakel uden, at man er direkte eksponeret og/eller i vejen for byens bevægelsesflow (se også Gehl 2003: 141; Stevens 2007: 115). Lad mig nu tage den empiriske undersøgelse af den fænomenologiske relevans af Hilliers bevægelsesøkonomiske teori i betragtning (kapitel 3). I denne analyse forskydes kroppenes tekniske placering i mikro-rummet til gaden og dermed bevægelsesstrømmens centrum. Dette hænger sammen med, at undersøgelsens surveyinterviews praktisk blev gennemført stående, midt i mikro-rummenes menneskemylder. På figur 9 ovenfor

200

svarer denne kropslige placering til punktet midt mellem vejstykkets to sorte, afgrænsende streger. I denne analyse kan den positive korrelation mellem bevægelsesdensitet og den oplevelsesmæssige intensitets-attraktion af byrummene forstås således, at bevægelsestætte mikro-rum inviterer til, at man kropsteknisk gør stående ophold midt i byens bevægelses- flow for der at lade sig gribe følelsesmæssigt af byens intensive rytme. Her føles og nydes byens attraktive livlighed ikke på fysisk afstand, men snarere affektivt i form af kropslig synkronisering, fortovsdansen og høf- lig legemes-koordinering. Med dette argument befinder vi os tæt på Whytes (1988) observation, at folk i storbyen, der mødes, gør kropsligt ophold og snakker i gaderummet, ofte udviser en frivillig tendens til at “(…) gravitate to the center of the pedestrian traffic stream.” (Whyte 1988: 2). Dette identificeres som en tendens til “self-congestion”, hvilket peger på, at det er muligt at nyde gadens livlighed som midlertidigt stå- ende midt i gadens bevægelsesflow, mens mere permanente ophold nydes kropsteknisk mere effektivt ved at tage ophold i gadens kantzone (som vi så det foroven). Dette fører os til afhandlingens sidste empiriske bidrag (kapitel 5), hvor sammenhængen mellem bevægelsesdensitet og intensiteten af medkreativ aktivitet ikke er positiv, men negativ. Det kropstekniskt karakteristiske ved ølbowlingspillene er, at de gennemføres siddende midt på vejen: Selv i det omfang, at spillerne søger mod en af gadens kantzoner, gør spilformationens størrelse, at banen “rager ud i” gadens og bevægelses- strømmens centrum. Forklaringen på, at det kun er intensiteten af denne festivalaktivitet, som er negativt korreleret med bevægelsesdensiteten, skal sandsynligvis tilskrives det kropssociologiske forhold, at vi her har at gøre med en siddende kropsteknik. Mens en bevægelsestæt gade, som vi netop har set det, inviterer positivt til midlertidige stående ophold, udgør et sådant gadeflow modsat en negativ invitation til at gøre siddende ophold. Dette hænger øjensynligt sammen med, at mens et siddende ophold er kropsteknisk fastlåst, er det stående ophold fleksibelt. Som stående kan man bedre benytte sig af de små kropslige “techniques for negotiation in motion” (Jensen 2013: 148; 2010), som der er en social forventning og norm om, at fodgængerne benytter sig af (Goffman 1963). I stedet for at tage et umærkeligt skridt til siden, gøre en zigzaggende bevægelse eller lade en fodgænger passere glidende gennem en snakkende forsamling af individer, er man som siddende i vejen: Den siddende er, så at sige, en “kropsprop”, der blokerer gadens bevægelse. Mikro-rummet inviterer i en sådan situation til, at man finder et mindre bevægelsesfortættet sted at tage siddende ophold, sådan som det netop et tilfældet med den medkre-

201

ative ølbowlingaktivitet, hvor festivalgæsterne placerede sig på rumlig afstand af de bevægelseshektiske gangstier. Noget tilsvarende gør sig i øvrigt gældende på Dronning Louises Bro, hvor fortovet er så bredt, at det er muligt at sidde i en kantzone ude af bevægelsesstrømmens centrum. Hertil kunne det indvendes, at fraværet af siddende ølbowlingaktivitet i de mest bevægelsesintensive festivalrum, måske snarere skal forklares med sociologisk henvisning til, at det bryder med socialt sanktionerede normer for korrekt kropsadfærd at sidde midt i rummet. Mens en sådan sociologisk forklaring vil have gyldighed for byer, hvor der ikke er krops- habituel tradition for gøre siddende ophold midt på gågader (man ville blive skammet væk af stirrende blikke), er dette næppe den oplagte forklaring og mekanisme i Roskilde Festival-konteksten. Festivalgæsterne gør, som beskrevet etnografisk af eksempelvis Sørensen (2009), en dyd ud af at suspendere sådanne hverdagslige adfærdsnormer. I kontrast til bylivet sidder, drikker og fester festivalgæsterne bogstaveligt talt overalt i festivalrummet. Festivalen er en ekstra-hverdagslig kontekst i denne sociale henseende: De sociale sanktioner, som den siddende kropsteknik ville være blevet mødt med i byen, er, så at sige, holdt “naturligt konstant” i denne kontekst og i den statistiske analyse af denne rumkontekst. Fremfor at forklare den negative korrelation mellem bevægelsesdensitet og intensiteten af ølbowlingaktiviteten som et sociologisk spørgsmål om traditioner og normer for, hvor man må sidde, tilbyder festivalkonteksten en unik mulighed for at se og forklare fraværet af siddende aktivitet midt i bevægelsesstrømmen som en “ren” effekt af den rumlige morfologi. Der, hvor rummet har fortættet bevægelsen, er det ganske enkelt kropsteknisk vanskeligt at opholde sig og nyde den siddende spilformati- on og det fælles rituelle fokus; et sådan bevægelsesfortættet rum lader sig, som afhandlingens andre empiriske analyser viser, nyde mere kropsteknisk effektivt ved at gøre midlertidigt stående ophold i mikro-rummet eller ved at sanse bevægelsens livlighed siddende eller stående i mikro- rummets kantzone.

Begrænsninger og implikationer Det er min overbevisning, at afhandlingens teoretiske (neo-durkheimianske) og metodiske (mikro-kvantificerende) perspektiv tilbyder et væ- sentligt bidrag til det rumsociologiske forskningsfelt; et felt, der lider af teorihistorisk glemsomhed og metodologisk underudvikling. Jeg skal imidlertid være den første til at slå fast, at det sidste rumsociologiske ord hermed ikke er sagt om mikro-rums fysiske egenskaber og sociale effek-

202

ter. Andre perspektiver og værktøjer kunne og burde måske have været fremhævede, hvilket her må overlades til min eller andres fremtidige forskning. Særligt vil jeg gerne fremhæve følgende forhold, som, i lyset af afhandlingens bidrag, kalder på videre undersøgelse: For det første udgør spørgsmålet om mikro-rummets mulige kausale effekter en akilleshæl for afhandlingens gennemgående argument om, at det fysiske rum har en uafhængig effekt på bevægelsen, og som et biprodukt heraf, en effekt på rumlig brug og oplevelser. Korrelation er, som bekendt, ikke det samme som eksistensen af kausalitet, og afhandlingens rumkausale forklaringer er, i sidste instans, begrundede i et teoretisk argument, der foreslår eksistensen af en art netværkshandlen, snarere end i en eksperimentelt kontrolleret test (Lieberson 1985). Dette gør afhandlingens rumlige kausalforklaringer sårbare over for den kritik, at rummets effekt ikke er af eksogen, men af endogen karakter. Dette kunne eksempelvis være tilfældet, hvis en uobserveret selv-selektionseffekt er korreleret med den rumlige x-variabel og den sociale y-variable. Kritikken af space syntaxens manglende anvendelse af rumstatistiske metoder og håndtering af den klyngede karakter af rumlige data udgør også et sådant endogenitets-problem: Skyldes bevægelsen rummets kausale netværks- egenskaber eller rumlige spillover-effekter fra de omliggende gaderum? Til at afklare sådanne kausalitetsspørgsmål kunne afhandlingen med fordel have benyttet sig af et kvasi-eksperimentelt design, der tilnærmer sig en randomiseret og kontrolleret test af kausalitet. Et konkret bud herpå kunne være anvendelse af en instrumentvariabel-estimering. Her tilnærmes eksperimentet ved at identificere et eksogent instrument, der påvirker en central uafhængig variabel, og som kun påvirker den af- hængige variabel af interesse gennem denne centrale uafhængige variabel. Dette muliggør en kvasi-eksperimentiel test af den uafhængige variables kausale effekt (Bollen 2012). I henhold til spørgsmålet om rummets kausale effekt på bevægelse ville det eksempelvis være oplagt, og næppe metodisk umuligt, at konstruere et instrument på baggrund af et natuligt eksperiment; dette ville forudsætte identifikation af et eksogent chok af gadenettet (f.eks. i forbindelse med et omfattende vejarbejde eller et uvejr), der midlertidigt ændrer gaderummets netværksmæssige egenskaber og kun herigennem har en effekt på byens bevægelsesmønstre. Et sådan design ville i princippet gøre det muligt at teste rummets eksogent- kausal effekt på bevægelse. Mens nærværende afhandling har insisteret på, og illustreret nødvendigheden af, at space syntaxen sammentænkes med rumstatistiske teknikker, er det næste statistiske skridt at gøre dette

203

inden for et kvasi-eksperimentielt design, der muliggør en mere direkte test af rummets kausale effekter. 35 For det andet vil det utvivlsomt styrke flere af de empiriske analyser, hvis afhandlingens kvantificerende tilgang til mikro-rummet, blev trian- guleret med kvalitative metoder i et “mixed methods design”, der eksempelvis inkluderer etnografi og/eller visuelle metoder (Creswell 2013). Afhandlingens argument om, at sådanne kvalitative metoder kun i vanskelig grad kan analysere netværksrummets ikke-diskursive og ikke-lokale egenskaber, anerkender omvendt, i lighed med Goffmans (1993) blumerianske argument, at disse metoder er væsentlige i beskrivelsen og for- tolkningen af lokale oplevelser og brug af mikro-rummet. Dette understreges endvidere af, at artiklens by- og festivalbaserede analyser implicit har trukket på min egen, etnografisk “tykke” viden om disse kontekster. Afhandlingen groede, som anført, ud af etnografiske observationer, jeg gjorde mig på Dronning Louises Bro. Som københav- ner har jeg et indgående kendskab til bydelen Nørrebro, hvor dataindsamlingen er foretaget. Ydermere har jeg 15 år i træk deltaget som gæst på Roskilde Festival, som jeg følgeligt har en omfattende tavs viden om. Havde jeg ikke haft denne tavse viden om de studerede kontekster, ville gyldigheden og pålideligheden af afhandlingens kvantitative analyser u- tvivlsomt have været ringere. I udarbejdelsen af festivalsurveyen (appen- dix 2) har jeg eksempelvis kunne trække på min erfaring når det skulle afgøres, hvad der udgør gyldige indikatorer på festlighed i denne kontekst, og hvornår på dagen graden af fuldskab gør det praktisk muligt at gennemføre pålidelige interviews. Hvis jeg skulle gentage dette festival- studie, ville jeg trække mere eksplicit på denne etnografiske viden og endvidere gennemføre etnografiske feltstudier i et antal af de områder, der også kvantificeres. For det tredje skal det påpeges, at afhandlingens hovedfokus på rum- dimensionen er sket på delvis bekostning af tidsdimensionen. Afhandlin-

35 En anden kvasi-eksperimentiel måde at teste rummets kausalitet på ville være vha. et såkaldt “regression-discontinuity design” (Berk et al. 2010). Denne teknik udnytter eksistensen af en tilfældigt opstået fordeling af en lokal påvirkningseffekt rundt om et givent skæringspunkt til at vurdere kausale effekter. Dette kunne i princippet være anvendt i forbindelse med afhandlingens test af rummets effekt på den kollektive opbrusen (se kapitel 4). Der kan argumenteres for, at festivalgæsternes præcise rumlige lejrplacering er tilfældig, når denne valgproces anskues på den mindste rumlige skala. Sagt med andre ord: Individerne selekterer sig systematisk ind i et festivalnabolag, men jo længere vi zoomer skalamæssig ind, desto mere tilfældighed må der antages at være på spil i forbindelse med lejrenes præcise rumlige lokalisering. Denne tilfældighed gør det principielt muligt at undersøge, om de gæster (påvirkningsgruppen), der tilfældigvis er lokaliserede tættere på de mest bevægelsestætte mikro-rum, er kausalt påvirkede til at have et højere festlighedsniveau i sammenligning med de gæster, der tilfældigvis er lokaliseret længere væk fra de mest bevægelsestætte mikro-rum (kontrolgruppen).

204

gen efterlader det indtryk, at rummet kan studeres med tiden i analytisk parentes. Hertil kan det indvendes, at rum altid allerede har en tidslighed (ligesom tiden omvendt finder rumligt sted), og at det, som foreslået af Thrift og May (2001), ofte er mere retvisende at tale om “tidsrum”. I retrospekt burde tidsligheden ved de rumlige fænomener, som afhandlingen undersøger, muligvis være tildelt en mere fremtrædende plads. Eksempelvis kunne det være blevet nærmere undersøgt, hvad de sociale implikationer er af, at festivalens rum er midlertidigt; eller hvordan festivalens dags- og drukrytmer udgør en tidslig betingelse for, hvorvidt og hvornår mikro-rummet kan have morfologiske effekter på mikro-inter- aktionen (når folk sover, er der eksempelvis ingen bevægelses- og rumeffekter!). En væsentlig kilde til underbetoningen af tidsspørgsmålet kan spores tilbage til afhandlingens anvendelse af space syntax-metoden, hvis grafteoretiske repræsentationer og analyser af rummet netop implicerer en fastfrysning af rummets tidslighed (dvs. i netop det øjeblik, hvor det flyfoto, der anvendes som axialkortets forlæg, blev taget). O’Sullivan (2000) og Griffiths (2011) har netop kritiseret, hvordan space syntaxen har tendens til at betone rummets “struktur” på bekostning af dets tidsli- ge “proces”, og afhandlingen anvendelse af space syntaxen arver, så at sige, dette metodologiske problem. For det fjerde må det fremhæves, at afhandlingen har en vis ensidighed i retning af mikro-rummets betydning for intensive følelser i den positive ende af det følelsesmæssige register: Afhandlingen fokuserer på følelser af at være opstemt, medrevet og hensat i og af mikro-rummet. Denne positive følelsesorientering har flere kilder. Den kan spores tilbage til min etnografiske fornemmelse på Dronning Louises Bro som afhandlingens artikler kan forstås som et forsøg på at undersøge og teste generaliserbarheden af. På det teoretiske plan afspejler denne følelsesso- ciologiske orientering mit tidligere (Liebst 2009) og afhandlingens ekspli- citte forsøg på at undgå bysociologiens pessimistiske overbetoning af bylivets negative følelser af over-intensivering, fremmedgjorthed og stress (Simmel 1976; Wirth 1938; Augé 1995; Bauman 1999). Hertil kan det dog indvendes, at de færreste mikro-rum er som Dron- ning Louises Bro, og, ved at se samfund-rums relationen gennem dette eksempels prisme, har afhandlingen muligvis underbelyst relevante følel- ser i den negative ende af det følelsesmæssige spektrum. Illustrativt herfor er det forhold, at det bogstavelig talt tog mig måneder analytisk at forsone mig med bevægelsesdensitets negative korrelation med den medkreative festivalaktivitet i kapitel 5. Da jeg gennemførte analysen var det min interaktionsritualteoretiske forforståelse, at sammenhængen ville

205

være positiv. I retrospekt står det klart, at dette netop hang sammen med en overfokusering på de positivt-medrivende følelser.36 Hertil skal det tilføjes, at mit forestående postdoc-projekt netop giver anledning til at analysere mikro-rummets samspil med negative følelses- processer i forbindelse med gadevold. Dette projekt er således en mulighed for at undersøge den inverse side af afhandlingens perspektiv: Under hvilke morfologiske betingelser er mikro-rummet med til at forme volde- ligt-destruktive rum, der emmer af negative følelser af aggression og frygt?

Dansk resumé Denne artikelbaserede ph.d.-afhandling præsenterer resultaterne af et rumsociologisk forskningsprojekt om det fysiske mikro-rum som en kausal kontekst for den følelsesmæssige intensitet af stedslige mikro-interaktioner. Sociologien har gennem det sidste årti udvist en stigende interesse for rumlige fænomener. Det er centralt for denne interesse, at rummet ikke blot forstås som et passivt baggrundstæppe for sociale processer, men udpeges som et materielt objekt, der gør en forskel for det sociales konstitution. Afhandlingen har sympati for denne “rumlige vending”, men adresserer også et antal væsentlige problemer ved denne fremspi- rende rumsociologiske forskningsagenda. For det første argumenterer afhandlingen teoretisk for det rumsociologisk frugtbare ved en tilbagevenden til den durkheimianske traditions social morfologiske program som den nyere rumsociologi forholder sig tavst, afvisende og således aldrig direkte eller konstruktivt til. Denne negligering kan spores tilbage til den nutidige rumsociologis stærke inspiration fra sociologen Bruno Latours aktør-netværks teoretiske (ANT) program, hvis antidurkheimianske position i væsentlig grad baserer sig på et argument om, at Durkheim mangler blik for, hvordan rumligt-materielle objekter er medkonstituerende for det sociale. Afhandlingen anfægter denne læsning: Durkheim og den durkheimianske traditions socialmorfologiske bidrag tildeler de rumligt-materielle objekter en langt mere frem- trædende rolle, end hvad ANT-kritikerne giver udtryk for.

36 Collins’ IR-teoretiske kernebegreb om emotionel energi (EE) har muligvis også bidraget til denne følelsesteoretiske forudindtagethed. EE er grundlæggende et forsøg på at begrebsliggøre menneskets følelsesliv ud fra et kvantitativt niveau af positive følelses- impulser (Collins 1993). At forstå menneskets følelsesliv som et sådan positivt kontinuum af høj og lav EE har imidlertid, som påpeget af Barbalet (2006) og Kemper (2011), en tendens til at forbigå karakteren af og de interne forskelle mellem negative følelser. Tænk eksempelvis på forskellige følelser som frygt og depressivitet, der ikke nødven- digvis blot kan forstås som det samme lave niveau af positiv EE.

206

Dette rumligt-materielle forsvar for den durkheimianske tradition fremføres dog ikke blot af ren teorihistorisk interesse. Afhandlingen argumenterer også analytisk for, at den sociale morfologi, særligt i Marcel Mauss’, Erving Goffmans og Randall Collins’ mikro-ritualteoretiske aftap- ning, tilbyder et negligeret, men nuanceret, begrebsapparat til at studere, hvordan det sociale livs følelsesmæssigt-relationelle bånd formes i et samspil med samfundets rumligt-materielle betingelser. Dette står i kontrast til ANT-traditionen, der udskifter det sociologiske begreb om intersubjektivitet med et begreb om “interobjektivitet”, hvilket kan være berettiget i visse sammenhænge, men ikke når det, som det er afhandlingens mikro-sociologiske interesse, gælder samspillet mellem rummet og intersubjektive følelsesprocesser. I lyset af negligeringen af Durkheim, Mauss og Co’s socialmorfologi fremstår traditionen dog, i visse henseender, noget teoretisk hengemt, og afhandlingen er således et forsøg på at reformulere en rumsociologisk opdateret version af den sociale morfologi. For det andet adresserer afhandlingen den rumsociologiske litteraturs påfaldende mangel på dedikerede rumsociologiske metoder. Vi befinder os i dag i den paradoksale, og efter min vurdering, uholdbare situation, at den stadigt stigende teoretiske interesse for rummet ikke modsvares af en tilsvarende interesse for udvikling af rum-sensitive metodologier. Dette gælder mest udtalt i forhold til kvantitative metoder til at studere det fysiske rums mikro-skalaegenskaber og effekter. At afhandlingen specifikt udpeger manglen på mikro-rumlige kvantificeringsredskaber som problematisk hænger sammen med, at de kvalitative metoder, der ofte anvendes til at studere mikro-rumlige processer, må anses for at have en begrænset rum-metodologisk gyldighed: Sprogligt-baserede kvalitative metoder forbigår den ikke-diskursive regularitet ved rumlige netværk, og etnografiske, samt visuelle metoder, hypostaserer det lokale rum på bekostning af rummets globale netværksrelationer. Dette kalder på en mikro-metodologisk kvantificering af rummet, hvilket afhandlingen realiserer ved at bygge tværfagligt bro mellem rumsociologien og den rumlige netværksteori, som de analytiske arkitekter Julienne Hanson og Bill Hilli- er har udviklet under navnet space syntax-metoden. Space syntaxen er ikke tidligere forsøgt integreret i et specifikt rumsociologisk forskningsdesign. Hvad der ydermere taler for space syntaxens relevans for afhandlingens neo-durkheimianske perspektiv, er det forhold, at metoden oprindeligt blev udviklet med inspiration fra Durkheims sociale morfologi. Dette slægtskab er ej heller tidligere blevet diskuteret i den (rum)sociologiske litteratur.

207

På baggrund af ovenstående teoretiske (social morfologiske) og metodologiske (mikro-kvantificerende) overvejelser fremsætter afhandlingen en rumlig kausalforklaring på den følelsesmæssige intensitet af stedslige mikro-interaktioner, hvis generaliserbarhed herefter testes i afhandlingens tre empiriske analyser. Denne forklaring tager afsæt i Durkheim, Mauss og Co’s ritualteoretiske argument om, at mikro-gruppens følelses- mæssige intensitet er korreleret med gruppens socialmorfologiske egenskab af social densitet. Durkheim, Mauss og Co’s kausalhypotese er såle- des, at når den forsamlede gruppe bringes i kropslig kontakt, igangsættes en selvforstærkende følelsesmæssig “smitte”, der udkrystalliserer sig i en intersubjektiv oplevelse af følelsesintensitet. Med inspiration fra space syntax-perspektivet, foreslår afhandlingen en rummorfologisk radikaliseret version af denne densitet-intensitets-kausalforklaring, der reformule- res som en rum-densitet-intensitets-kausalforklaring. Fremfor at forklare gruppens forsamling og øgede sociale densitet som funktion af ritual- deltagernes formålsrettede handlinger, foreslår afhandlingen, at det fysiske rums netværksmæssige egenskaber også har handlingsmæssig, om end ikke-formålsrettet, kapacitet til at fortætte og således intensivere mikro-gruppens kollektive følelser. Situationelle følelsesintensiveringer kan således være et produkt af social handlen og netværkshandlen. Dette rummorfologiske argument skrives endvidere op imod et fæ- nomenologisk inspireret argument, der udpeger rummets overflade-æste- tiske fysiognomi som hovedvejen til, hvordan rummet kan påvirke det situerede individs følelsesliv. I dette perspektiv “stråler” rummets æsteti- ske overflader ud i rummet, hvis mellem-rum således fyldes af en følel- sesmæssigt medrivende atmosfære. Afhandlingen foreslår i stedet, at forklaringen forskubbes til det netværksmønster, som dannes mellem de rumlige overflader. Rummets følelsesmæssige effekter udgår ikke fra overflade-æstetikkens fysiognomi, men fra den rumlige netværks-morfologi, der fortætter kropsbevægelserne og således øger sandsynligheden for, at de rumligt tilstedeværende individer falder affektivt-intensivt ind i hinandens kropsrytmer. Afhandlingens empirisk-kvantitative tests af den fremsatte rum-densitet-intensitets-hypotese baserer sig på georefererede surveydata indsamlet på indre Nørrebro og på Roskilde Festival samt punkt-eventdata over mikro-rituelle forsamlinger af festivalgæster, der er georefererede ved hjælp af flyfotos over festivalen. Data estimeres statistisk med multilevel- og rumregressions-modellering, og afhandlingens kausalhypotese testes i tre empiriske delundersøgelser. I det første studie undersøges det, hvorvidt Hilliers space syntax-baserede teori om byer som bevægelsesøkonomier også har empirisk relevans for forklaringen af byens oplevelsesmæssige

208

attraktionsværdi. Hillier antyder eksistensen af en sådan sammenhæng, som imidlertid ikke tidligere er forsøgt direkte testet. Efter kontrol for individvariable peger analysen på, at byens rumlige netværk yder et positivt bidrag til konstitutionen af attraktivt-intensive urbane oplevelser. Afhandlingens andet studie tester Collins’ interaktionsritual-teoretiske forklaring af durkheimiansk kollektiv opbrusen; dvs. den intense intersubjektive oplevelse af at være del af et succesfuldt mikro-ritual. Spørgsmålet er, om Collins’ skelnen mellem effekten af interaktion som den primære forklaring af den kollektive opbrusen og effekterne af selektion og rumlig morfologi som vigtige, men sekundære forklaringer, holder til en empirisk test. Under anvendelse af surveydata bekræftes det, at effekterne af interaktion og selektion er af hhv. primær og sekundær vigtighed. Endvi- dere peger analysen på, at den rumlige morfologi har en mere direkte effekt på den kollektive opbrusen, end hvad Collins umiddelbart lægger op til. Rummet udgør ikke blot en hindring for mikro-rituel aktivitet, men er også med til at konstituere dets intensitet. Denne og den forrige analyses statistiske resultat bekræfter således i hovedsagen afhandlings rum- densitet-intensitets-hypotese. Dette billede kompliceres af afhandlingens tredje analyse, der under- søger de socio-rumlige mekanismer, der driver skabelsen af rumlige hot spots af medkreativ festivaladfærd. I henhold til effekter af selektion og interaktion konvergerer studiets resultater i hovedtræk med resultaterne i afhandlingens to øvrige analyser. Analysens resultat skiller sig imidlertid ud, når det specifikt gælder den rumlige effekt: Mens de to øvrige artikler konkluderer, at bevægelsesdensitet bidrager positivt til den stedslige intensitet, peger artiklens analyse på en negativ sammenhæng. Dette resultat peger på, at bevægelsesdensitet både kan være ophav til attraktive intensitets-fornemmelser og en stressende over-intensivering, som festivalgæ- sternes holder sig på afstand af. Disse to resultater modsiger dog ikke hinanden, men kalder snarere på en kropssociologisk forklaring: Det er, fordi den studerede medkreative aktivitet gennemføres siddende midt i gaderummet, at de mest bevægelsesintensive mikro-rum fremstår stressende over-intensiverede og således er negativt korrelerede med denne kropsaktivitet. Dette står i kropssociologisk kontrast til de to øvrigt studier, hvor det at stå midlertidigt i bevægelsesstrømmen (som i bevægel- sesøkonomianalysen i kapitel 3) eller det at sanse bevægelsesstrømmen fra en vis afstand (som i testen af den kollektive opbrusen i kapitel 4) udgør bedre kropstekniske udgangspunkter for at opleve den rumligt fortættede bevægelsesdensitet som en positiv kilde til at være der, hvor tingene sker.

209

English Summary This PhD thesis conveys the results of a micro-sociological research project on built space as a causal micro-context for the emotional intensity of emplaced micro-interactions. Over the past decade, sociology has shown growing interest in spatial phenomena, especially in emphasizing space as more than a passive backdrop for social processes: once built, the spatio-material artifact does something to social life. Although sympa- thetic to this ‘spatial turn’ of sociology, the thesis addresses a number of key problems in this emerging agenda of space sociology research. First, in terms of theory, the thesis suggests that contemporary sociology of space could benefit from a return to the Durkheimian tradition of ‘social morphology’; that is, to a tradition that the current space sociological research agenda only considers in a silent or dismissive manner, never directly or constructively. This attitude may be traced back to Bru- no Latour’s highly influential actor-network theory (ANT). Latour adopts an anti-Durkheimian position, claiming that Durkheim neglects the ways that spatio-material objects co-constitute social interactions. The thesis challenges this argument; contrary to what the ANT critics contend, Durkheim and his social morphological school do fully recognize the sociological importance of spatio-material objects. This argument is not merely a theoretical matter of historical interest: the thesis suggests that social morphology, especially as adapted by the micro-sociological tradition advanced by Marcel Mauss, Erving Goffman and Randall Collins, offers a nuanced framework for studying how the emotional-relational bonds of the social world are closely intertwined with its spatio-material substratum. In this regard, social morphology offers an intersubjectively based space sociology as an alternative to the ANT tradition, which has replaced intersubjectivity by an ontological notion of ‘interobjectivity’. Such replacement may be justified in certain cases, but not when it is applied to the interplay between space, emotions and micro-interactions, as is the case of this thesis. Given the paucity of attention in the literature to Durkheim, Mauss and Co.’s spatio-material contribution, the social morphological tradition appears antiquated in certain respects. The thesis offers a version of social morphology reassessed via space sociology. Second, in terms of method, the thesis addresses the conspicuous lack of spatially dedicated methodologies in the literature of space sociology. We find ourselves today in the paradoxical situation that the increasing interest in spatial phenomena is not matched by a corresponding interest in developing space sensitive methods. Most notable is the lack of methods for quantifying the micro-scale properties and effects of built space.

210

This methodological problem is further accentuated by the fact that the qualitative methods that are often applied at this level of spatial specificity may be of limited validity: discursively based methods miss the non- discursive regularity of spatial networks and ethnography and visual methods tends to privilege ‘local’ spaces at the expense of the ‘global’ network of spatial relations. This suggests the need for a micro-methodological quantification of space. Specifically, the thesis suggests building an ‘interdisciplinary bridge’ between space sociology and the space network theory that the analytical architects, Bill Hillier and Julienne Hanson, have developed under the heading of ‘space syntax’. Moreover, what speaks in favor of placing the space syntax approach within the neo-Durkheimian framework used in this thesis is the fact that Hillier and Hanson developed their approach against the backdrop of inspiration from Durkheim’s social morphology. The literature includes no previous attempts to systematically incorporate the space syntax approach into a space sociological research design. Based on the above theoretical (i.e. social morphology) and methodological (i.e. micro-quantification) considerations, the thesis proposes the hypothesis that the physical network properties of built micro-space shape the emotional intensity of emplaced micro-interactions. This hypothesis is developed against the backdrop of Durkheim, Mauss and Co.’s argument from ritual theory that the emotional intensity of micro- groups is causally correlated with the group’s morphological property of social density. Physically assembled individuals exercise an effect of emotional ‘contagion’ on each other. Following the space syntax, the thesis suggests a version of this density-intensity hypothesis that is radicalized by space morphology. That is, it proposes a space-density-intensity mechanism. Instead of explaining micro-ritual gatherings as a function of the purposeful actions of ritual participants, the thesis suggests that the spatial network exercises a similar, but unintentional, agency to densify (and thus intensify) micro-ritual emotions. Emplaced intensities may thus be attributable to social agency and network agency. Moreover, this morphology argument is critically related to the space phenomenological suggestion that it is the surface aesthetics or physiognomy of built space that constitutes the key mechanism by which spatiality affects human emotions. Seen in this way, aesthetic facades ‘radiate’, so to speak, into space, thus filling the empty volume of space with a particular emotional atmosphere. The thesis suggests a network morphology, rather than a surface physiognomy based explanation of such urban spatial atmospheres. Hence, it is argued that the atmospheric ‘given’ is formed by varying densities of pedestrians, compressed by the ur-

211

ban network configurations. Atmospheres emerge as a bodily affect of such spatial effects, producing a rhythmical entrainment within the dense body of pedestrians. The empirical tests for the space-density-intensity hypothesis just outlined use two kinds of data: first, geocoded survey data collected at the Roskilde festival and in the neighborhood Nørrebro, Copenhagen; second, event data on micro-ritual gatherings of festival guests, geocoded using an aerial photography of the festival site. Data is estimated with multilevel and space regression modelling tools. Applying these techniques and data, the thesis presents the results of three empirical studies. The first study examines whether Hillier’s space syntax theory of cities as movement economies provides a valid explanation of the experienced attraction value of urban life. Hillier hints at this kind of phenomenological relevance of the theory, but the literature offers no systematic attempts to empirically address this aspect of the theory. After controlling for individual selection effects, the analysis suggests that the movement economic properties of urban micro-space contribute positively to the constitution of intensive, attractive urban experiences. The second study tests Collins’ micro-sociological explanation of Durkheimian collective effervescence, i.e. the intense experience of partaking in a successful micro-ritual. More specifically, it attempts to examine whether Collins’ claim that interaction processes are the primary effect, and the effects of selection and spatial morphology are relevant, but secondary, factors in shaping collective effervescence can be statistically verified. Applying survey data, the study confirms that the effects of interaction are of primary importance and selection effects are secondary. Moreover, the analysis suggests that spatial morphology has a more direct effect on collective effervescence than Collins assumes: space acts not merely as a constraint, but also constitute the emotional intensity of micro-ritual activity. The results from this and the previous study generally confirm the key space-density-intensity hypothesis of the thesis. However, this picture is complicated by findings in the third study, which examined the socio-spatial mechanisms that constitute ‘hot spots’ of co-creative festival behavior. While the two other studies suggest that movement density contributes positively to the intensity of emplaced micro-interactions, the third study suggests a negative relationship. That is to say, movement density may be the source of attractive intensity emotions but may also be a stressful over-intensification from which the festival guests tend to keep their distance. However, these findings, although discrepant, are, not contradictory, but call, instead, for a bodily explanation. The fact that the most movement dense micro-spaces were

212

excessively intense to permit co-creative activity may be ascribed to the fact that the operationalized co-creative activity was conducted while sitting at the center of the street. And the positive results of the two other studies may be related to the existence of different bodily techniques: to stand temporarily at the center of the movement flow (as in the case of the movement economy analysis in chapter 3), or to perceive the movement flow at a certain distance (as in the case of the test of collective effervescence in chapter 4) provides better bodily conditions for a positive experience of movement dense space – it communicates the sense of being where the action is.

Bibliografi Augé, M. (1995). Non-Places. Introduction to an Anthropology of Supermodernity. New York: Verso. Barbalet, J. (2006). Emotional Payoffs of Ritual. About Randall Collins’ Interac- tion Ritual Chains (Princeton, Princeton University Press, 2005). European Journal of Sociology, 47(3), 446-451. Bauman, Z. (1999). Urban space wars: on destructive order and creative chaos. Citizenship studies, 3(2), 173-185. Berk, R., Barnes, G., Ahlman, L., & Kurtz, E. (2010). When second best is good enough: a comparison between a true experiment and a regression discontinuity quasi-experiment. Journal of Experimental Criminology, 6(2), 191-208. Bollen, K. A. (2012). Instrumental variables in sociology and the social sciences. Annual Review of Sociology, 38, 37-72. Bourdieu, P. (1984). Distinction. A social Critique of the Judgment of Taste. New York & London: Routledge. Bourdieu, P. (2002). Habitus. i: Panofsky, E. (2002). Gotik. Arkitektur, Skolastik, Habitus. Århus: Forlaget Klim. Collins, R. (1993). Emotional energy as the common denominator of rational action. Rationality and Society, 5(2), 203-230. Creswell, J. W. (2013). Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. London: Sage. Dovey, K. (1999). Framing places: Mediating Power in Built Form. London & New York: Psychology Press. Gehl, J. (2003). Livet mellem husene. København: Arkitektens forlag. Gibson, J. J. (1986). The Ecological Approach to Visual Perception. London & New York: Psychology Press. Goffman, E. (1963). Behaviour in Public Places. Notes on the Social Organisation of Gatherings. New York: Free Press. Goffman, E. (1993). An interview with Erving Goffman, 1980 (ved Verhoeven, J. C.). Research on Language and Social Interaction, 26(3), 317-348. Griffiths, S. (2011). Temporality in Hillier and Hanson's Theory of Spatial De- scription: Some Implications of Historical Research for Space Syntax. The Journal of Space Syntax, 2(1), 73-96. Heidegger, M. (2007). Væren og tid. Århus: Klim.

213

Ingold, T. (2000). The Perception of the Environment: Essays on Livelihood, Dwelling and Skill. London & New York: Taylor & Francis. Jensen, O. B. (2010). Negotiation in motion: Unpacking a geography of mobility. Space and Culture, 13(4), 389-402. Jensen, O. B. (2013). Staging Mobilities. London & New York: Routledge. Jonge, Derk de (1967). Apllied Hodology. Landscape 17. 2: 10-11. Kemper, T. D. (2011). Status, Power and Ritual Interaction: a Relational Reading of Durkheim, Goffman, and Collins. Farnham: Ashgate. Lieberson, S. (1985). Making it Count: The Improvement of Social Research and Theory. Berkeley: University of California Press. Liebst, L. (2009): Etikken i den maskerede by. Om Baumans etiske afvisning af den postmoderne bys æstetisering, Dansk sociologi, 1(20): 7-23. Løgstrup, Knud. E. (1995): Kunst og erkendelse. København: Gyldendal. Markus, T. A. (1993). Buildings and Power: Freedom and Control in the Origin of Mod- ern Building Types. London & New York: Routledge. Mauss, M. (1973). Techniques of the body. Economy and Society, 2(1), 70-88. Mauss, M. (1979). Seasonal Variations of the Eskimo: a Study in Social Morphology. London: Routledge & Kegan Poul. O’Sullivan, D. B. (2000). Graph-Based Cellular Automaton Models of Urban Spatial Processes. Ph.d.-afhandling, University of London. Simmel, G. (1976). The Metropolis and Mental Life. The Sociology of Georg Simmel. New York: Free Press. Stevens, Q. (2007). The ludic city: exploring the potential of public spaces. London & New York: Routledge. Sørensen J K, 2009, Substance use, Ritual and Risk Management: Danish rock festivals, Ph.d.-afhandling, Sociologisk Institut, Københavns Universitet Thrift, N. & May, J. (2001). Introduction. May, J., & Thrift, N. (red.). Timespace: Geographies of Temporality. London & New York: Routledge. Whyte, W. H. (2012). City: Rediscovering the Center. Philadelphia: University of Pennsylvania Press. Wirth, L. (1938): Urbanism as a way of life, American Journal of Sociology, 44(1):1- 24.

214

Appendiks #1: Nørrebro-spørgeskema

“Undskyld, må jeg spørge dig om noget? Jeg er ved at lave en eksamensopgave om oplevelsen af København som by. Det drejer sig om et par simple spørgsmål, og det tager under 2 minutter:”

 Først har jeg en række udsagn om gaden hvor vi står lige nu, som du kan erklære dig enig eller uenig i. Er du mest enig eller er du mest uenig i, at: Enig Uenig Tilbyd ikke: Ved ikke/Hverken-eller …denne gade er et spændende sted …at denne gade er hektisk …at denne gade er et rart sted at være …at denne gade er et trygt sted …at denne gade er et smukt sted …at denne gade er et råt sted

 “Dernæst har jeg et simpelt ja-nej-spørgsmål: Forestil dig, at der stod en bænk lige her hvor vi står: Kunne du have lyst til at sidde her?” JA □ NEJ □ Tilbyd ikke: Ved ikke □

 Eller alternativt, hvis der er en bænk: “Først har jeg et simpelt ja-nej-spørgsmål: Den bænk her ved siden af os: Kunne du have lyst til at sidde her?” JA □ NEJ □ Tilbyd ikke: Ved ikke: □  Næste spørgsmål er mere åbent: Hvis du skal beskrive stemningen på det sted, hvor vi befinder os, hvilke to ord falder dig så ind?

- Noter:______ Afslutningsvis vil jeg gerne spørge, hvor du er på vej hen lige nu? (arbejde, indkøb, etc.). - Notér:______ Hvad laver du til hverdag, hvad er din beskæftigelse? (murer, advokat, studerende, etc.) - Notér:______ Hvor gammel er du? ______ Bor du i København? JA □ NEJ □. Hvis ja, hvilket kvarter? (Vesterbro, Vanløse, etc.) - Notér:______

“Det var det – tak for hjælpen!” ------Noteres af interviewer, respondentens: Køn: kvinde □ Mand □ Etnisk dansk □ Anden etnisk baggrund end dansk □ Interview udført: Før □ eller efter □ frokost. Zone nr:______Port nr:______

215

#2: Festival-spørgeskema

UDFYLDES AF INTERVIEWER FØR INTERVIEW Kvadrant______Område: □ Vest □ Øst Agora: □ C □ E □ G □ J □ H □ K □ L □ P □ N □ M □ mandag □ tirsdag □ onsdag

Hej, jeg er med til at lave et sociologisk forskningsprojekt om Roskilde festivalen. Jeg har et spørge- skema, som tager cirka 10 minutter at besvare. Det vil være en stor hjælp, hvis én af jer/du har tid til at være med. Jeg skal understrege, at din besvarelse selvfølgelig er fuldstændig anonym.

(1) Hvor gammel er du?______

(2) Hvad laver du til hverdag? – læs ikke højt □ I arbejde, som hvad? (ex. tjener, advokat): ______□ Arbejdsløs □ Pensionist/efterlønner □ Barsel □ Andet:______□ Studerende

□ Folkeskole □ Gymnasial uddannelse (inkl. HHX, HTX, HF) □ Erhvervsuddannelse (ex murer, tandklinikassistent) □ Kort videregående uddannelse (ex handelsøkonom, byggeteknikker) □ Mellemlang videregående uddannelse (ex sygeplejerske, pædagog, lærer etc.) □ Bachelor-uddannelse ved universitet (prof.-bachelor tælles som Mellemlang vid. udd.) □ Lang videregående uddannelse □ Forskeruddannelse/ph.d.-studerende □ Andet:______

(3) Hvad er din højest fuldførte uddannelse? – læs ikke højt □ Folkeskole □ Gymnasial uddannelse (inkl. HHX, HTX, HF) □ Erhvervsuddannelse (ex murer, tandklinikassistent) □ Kort videregående uddannelse (ex handelsøkonom, byggeteknikker) □ Mellemlang videregående uddannelse (inkl. sygeplejerske, pædagog, sygeplejerske, etc.) □ Bachelor-grad fra universitet (prof.-bachelor tælles som Mellemlang vid. udd.) □ Lang videregående uddannelse □ Forskeruddannelse/ph.d.-grad □ Andet:______(4) Hvor i (ud)landet bor du? □ København □ Århus □ Aalborg □ Roskilde □ Odense □ Esbjerg □ Andet (notér hvilken by)______□ Udlandet (notér hvilket land)______

(5) Er du i et fast forhold? □ Ja □ Nej

(6) Hvilken dag ankom du til festivalen? □ Fredag □ Lørdag □ Søndag □ Mandag □ Tirsdag □ Onsdag

216

(7) Har du købt din billet eller arbejder du som frivillig? □ Købt billet □ Frivillig □ Andet (ex. hoppet over hegnet)

(8) Hvor mange penge regner du at bruge på festivalen, udover billetten? Kroner ______

(9) Hvor mange gange har du været på Roskilde festival før dette år (eksklusivt besøg med endagsbil- let) Antal ______

(10) Har du stået i kø ved en af de indgange, som man kunne booke hjemmefra? □ Nej Hvis ja, angiv: □ West 1 □ West 2 □ East 1 □ East 2 □ South 1

(11) Hvor mange bor I i jeres lejr? (cirka)______

(12) Har jeres lejr et hjemmebygget musik-anlæg? (inkluderer ikke boomblastere, ghettoblastere, etc.) □ Ja □ Nej

(13) Det næste spørgsmål handler om, hvad der efter din mening afgør, om en lejr er godt eller dårlig placeret på campingpladsen. Det foregår på den måde, at jeg kommer med nogle udsagn, og du fortæller, om du er mere eller mindre enig eller uenig i dem. Du har fire svarmuligheder: enig, meget enig, uenig eller meget uenig.

Meget Enig Uenig Meget (Ved ikke) enig uenig Det gør min festival-oplevelse federe, hvis min lejr ligger i et område, hvor folk fester igennem og giver den max gas. Det er gåafstanden til festivalpladsen, der er vigtigst for om en lejr er godt eller dårligt placeret. Det er vigtigt for mig at min lejr er placeret i et område, hvor jeg kan trække mig tilbage og hvor der er ikke er for meget larm om natten. Det betyder ikke så meget for mig, hvor min lejr er placeret.

(14) Var det planlagt på forhånd, hvor jeres lejr skulle ligge på camping-området? □ Ja □ Nej □ Ved ikke (gå til spørgsmål 17)

(15) Er det sted hvor jeres lejr ligger placeret cirka der, hvor I havde planlagt det? □ Ja (gå til spørgsmål 17) □ Nej

(16) Kan du markere på kortet, hvor I havde planlagt at jeres lejr skulle havde ligget. – markér på kortet, næste side:

217

218

(17) Hvor er det efter din mening det bedste sted at have en lejr på camping-pladsen? (ikke inklusiv: Get A Tent, Caravan-Camping, MC-camp eller HM Reboot Camp) - Markér med kryds □ Ved ikke

219

(18) De næste par spørgsmål handler om, hvordan du oplever det sted, hvor din lejr ligger placeret, altså området lige rundt om din lejr. Jeg kommer med nogle udsagn, og du siger i hvilken grad du er enig eller uenig. Der er igen de fire svarmuligheder: enig, meget enig, uenig eller meget uenig.

Meget Enig Uenig Meget (Ved ikke) enig uenig Min lejr ligger et festligt sted Min lejr ligger et trygt sted Min lejr ligger et larmende sted Min lejr ligger et kedeligt sted Min lejr ligger et ungdommeligt sted Min lejr ligger et roligt sted Min lejr ligger et beskidt sted Min lejr ligger et smukt sted Min lejr ligger et hektisk sted Min lejr hænger meget ud med nabolejrene

De næste par spørgsmål handler om musikken på festivalen.

(19) Hvilke to koncerter glæder du dig allermest til at høre? - Sæt to krydser (valg af kun én koncert accepteres også) - Vis helst ikke listen til respondenten, før efter denne har besluttet sig - Det accepteres, hvis en af de valgte bands allerede har spillet når respondenten adspørges

220

□ ADDISON GROOVE □ DR. JOHN AND THE □ KÚRA □ CHRISTINA ROSEN- □ THE ABYSSINIANS LOWER 911 feat. JON VINGE feat. CHRIS □ ADRIAN LUX CLEARY □ LARSEN & FURIOUS BROKAW & REFREE □ AFRICA HITECH □ DRY THE RIVER JANE □ ROVER □ ALABAMA SHAKES □ LINKOBAN □ ROYCE DA 5’9” □ THE ALAEV FAMILY □ EVIDENCE □ LITURGY □ RUM 37 □ AMADOU & MARIAM □ THE ECHO VAMPER □ LONE □ SAM AMIDON □ EL PARAISO presents □ THE LOW ANTHEM □ SAGE FRANCIS □ AMSTERDANCE feat. CAUSA SUI, EL PARAISO □ SANTIGOLD JESSE VOORN, DENNIS ENSEMBLE and PAPIR □ M83 □ MATHILDE SAVERY CHRISTOPHER and SAM □ ELEKTRO GUZZI □ MAC MILLER □ THE SHINS O’NEALL □ ELOQ □ MACHINE HEAD □ SHINY DARKLY □ ANALOGIK □ MACKLEMORE AND □ SHLOHMO □ APPARATJIK □ FIRST AID KIT RYAN LEWIS □ SLENG □ ARAABMUZIK □ LEE FIELDS & THE EX- □ MAGTENS KORRIDORER □ SLOWOLF □ ARS NOVA PRESSIONS □ MAJOVCI GROUP □ SPECKTORS □ A$AP ROCKY □ F.O.O.L. □ MALK DE KOIJN □ SPECTOR □ NILS FRAHM □ MARTYN □ SPILLEMÆNDENE □ BALOJI □ THE FREDERIK □ MASH UP INTERNA- □ SPLEEN UNITED □ BARONESS □ FRENCH FILMS TIONAL □ SPOEK MATHAMBO □ THE BARONS OF TANG □ LES FRERES SMITH □ THE MEGAPHONIC □ BRUCE SPRINGSTEEN & □ THE BEARDY DURFS □ FRIENDLY FIRES THRIFT THE E STREET BAND □ BEASTIE RESPOND □ FRITJOF & PIKANEN □ MEW □ SPRUTBASS □ BEHEMOTH □ MODESELEKTOR □ STAFF BENDA BILILI □ BELLOWHEAD □ GENTLEMAN & THE □ JANELLE MONÁE □ DJ STATIC w/ TEMU, □ BERNHOFT EVOLUTION □ R. STEVIE MOORE BLOOD SWEAT □ BIG K.R.I.T. □ GHOSTIGITAL □ MOHAMMAD REZA DRUM+BASS and □ BJÖRK □ GIRAFFAGE MORTAZAVI GUESTS □ RUBÉN BLADES □ GIRLSEEKER □ MUTINY ON THE □ SUICIDE SILENCE □ BLITZ THE AMBASSA- □ GOSSIP BOUNTY □ THE SWAY MACHINERY DOR □ GOVERNOR OF ALASKA □ JULIUS SYLVEST □ BLONDES □ GRIEVES & BUDO □ NASUM □ BOMBA ESTÉREO □ GURRUMUL □ NELSON CAN □ TALK NORMAL □ BON IVER □ NIKI & THE DOVE □ TAMIKREST □ BOWERBIRDS □ H2O □ NIKLAS □ TEAM ME □ JAKOB BRO □ HAMADCHA SUFI □ NORTHERN STRUC- □ JØRGEN TELLER & THE □ BÄDDAT FÖR TRUBBEL TARIQA OF ESSAOUIRA TURES EMPTY STAIRS feat. □ HANK3 LAZARA □ CEREBRAL BALLZY □ HELSINKI POETRY □ OF THE WAND AND THE □ TER HAAR □ CEREKLOTH □ JULIA HOLTER MOON □ TODAY IS THE DAY □ CLOCK OPERA □ HOSPITALITY □ ONEOHTRIX POINT □ THE TODDLA T SOUND □ COLD SPECKS □ SIVERT HØYEM NEVER □ TOUCHY MOB □ COMMIX □ ORQUESTA TÍPICA □ BOUBACAR TRAORÉ □ CONJUNTO ANGOLA 70 □ I GOT YOU ON TAPE FERNÁNDEZ FIERRO □ TRASH TALK □ COPENHAGEN COLLAB- □ ICONA POP □ OWINY SIGOMA BAND □ TUNE-YARDS ORATION □ INDIANS □ COPENHEAVY □ PENGUIN PRISON □ ULIGE NUMRE □ CRIOLO □ JONATHAN JOHANSSON □ PERFUME GENIUS □ MIKAL CRONIN □ DAWDA JOBARTEH □ CELSO PIÑA □ VACCINES, THE □ CROWBAR □ JUPITER & OKWESS □ KIMMO POHJONEN / □ MAÏA VIDAL □ THE CULT INTERNATIONAL SAMULI KOSMINEN and □ SOFFIE VIEMOSE □ THE CURE PROTON STRING QUAR- □ VAAGSBYGD HANDY □ KELLERMENSCH TET □ DALGLISH □ PAUL KALKBRENNER □ PRETTY LIGHTS □ WAR FOR YOUTH □ DANGERS OF THE SEA □ KESI □ PUNCH BROTHERS □ WARBRINGER □ DAUGHTER □ KHALAZER □ DANYEL WARO □ DEAD SKELETONS □ KLUMBEN & RASKE □ RANGLEKLODS □ WEEDEATER □ DEATHCRUSH PENGE □ LEE RANALDO □ JACK WHITE □ DEVILDRIVER □ SANDRA KOLSTAD □ RED FANG □ LARS WINNERBÄCK □ DIG & MIG □ KORALLREVEN □ REDWOOD HILL □ WIZ KHALIFA □ DJANGO DJANGO □ KRAFTKLUB □ REFUSED □ DOMINIQUE YOUNG □ ALISON KRAUSS & □ ROCKWELL □ YELAWOLF UNIQUE UNION STATION feat. □ THE ROOTS □ 120 Days JERRY DOUGLAS □ ANDET______□ KRIGET □ VED IKKE 221

(20) Af de to koncerter du har valgt, hvilken er så den vigtigste for dig at høre? - Angiv respondentens prioritering med “1” og “2”. (angiv også et “1”, hvis respondenten kun angiver én koncert)

(21) Så vil jeg stille et par spørgsmål om, hvad der generelt er vigtigt for din festival-oplevelse og du fortæller, i hvilken grad du er enig eller uenig. Du har igen de fire svarmuligheder: enig, meget enig, uenig eller meget uenig.

Meget Enig Uenig Meget (Ved ikke) enig uenig For mig handler festivalen om at prøve sine egne grænser af (ex. drikke meget, være ekstra udadvendt) For mig handler festivalen om at komme helt væk fra hverdagen For mig handler festivalen om at feste og fyre den maximalt af For mig handler festivalen mere om at hænge med vennerne, end at møde nye mennesker For mig handler festivalen primært om musikken og om at gå til mange koncerter For mig handler festivalen om at være mega fuld For mig handler festivalen om at få udfordret min musiksmag For mig handler festivalen om at være skæv (ex hash, andre stoffer) For mig handler festivalen om at score For mig gør det min festival-oplevelse federe, at der er mange totalt fremmede, der besøger min lejr og fester med os For mig er det vigtigere at være sammen med vennerne i min lejr end at gå til mange koncerter

De næste par spørgsmål handler om, hvordan du bevæger dig rundt på camping-pladsen:

(22) Sætter jeres lejr sig nogen gange ud på en af stierne og hænger ud? (inklusiv ophold i åbne arealer; eksklusiv ophold ved scener) □ Ja □ Nej (gå til spørgsmål 24)

(23) Kan du vise på et kort, hvor I præcist sidder? (hvis flere, seneste)- Sæt et kryds, næste side: Markér med kryds, hvor lejren sidder:

222

223

(24) Besøger du andre lejre på festivalen? □ Ja □ Nej (gå til spørgsmål 27)

(25) Hvor præcist ligger den lejr, som du besøger oftest? – markér med kryds samt “1”

(26) Er der en anden lejr, som du besøger næstmest? □ Nej - Hvis ja: markér med kryds samt “2”

224

(27) Har du nogensinde spillet ølbowling på Roskilde festivalen? (også inklusiv, hvis man har været tilskuer, dommer og andre år). □ Ja □ Nej (gå til næste side, afslut interviewet)

(28) Har du været med til at spille ølbowling i år? (inklusiv, hvis man har været tilskuer, dommer) □ Ja □ Nej (gå til næste side, afslut interviewet)

(29) Hvor faldt ølbowlingsspillet præcist sted? (hvis flere steder, seneste) - Sæt et kryds:

Så har jeg ikke flere spørgsmål – mange tak for hjælpen! Jeg skal lige huske at minde om at vi behandler din besvarelse 100 procent anonymt.

225

UDFYLDES AF INTERVIEWER EFTER INTERVIEW: (30) Informantens køn: □ Mand □ Kvinde (31) Har lejren? □ Fane(r) og/eller anden karakteristisk dekoration (ex tema-specifik) □ Hvid(e) havepavillon(er) (32) Lejrens placering i kvadranten – vælg én: □ Lejren ligger i første række, og er åben mod gaden □ Lejren ligger i første række, og er lukket mod gaden □ Lejren ligger inde i kvadranten (33) Markér hvilken VEJ, som ligger nærmest lejren, næste side:

226

#3: Segment angular analysis af Roskilde Festival

Note. Visuel repræsentation af segment angular choice-analysen af Roskilde Festival, der anvendes i kapitel 4. Farveskalaen fra rød, henover gulgrønne nuancer, til blå svarer til at segmenterne har en højere choice-værdi (dvs. er mere netværksmæssigt tilgængelige).

227

#4: Visibility graph analysis af Roskilde Festival

Note. Visuel repræsentation af synlighedsgrafanalysen af Roskilde Festival, der anvendes i kapitel 5. Analysen udregner den gradsmæssige afstand til den nærmeste af de tre indgangsporte til festivalpladsen (det røde felt til venstre). Farveskalaen fra rød, henover gulgrønne nuancer, til blå svarer til en stigende gradsmæssig distance.

228

LASSE SUONPERÄ LIEBST SUONPERÄ LASSE Department of Sociology faculty of social sciences Univeristy of Copenhagen · Denmark university of copenhagen PhD Thesis 2014 · ISBN 978-87-7611-771-9

Lasse Suonperä Liebst Staged, yet Unstaged Sociological Inquiries into Space and Micro-Interaction