Urban Nuclei

8/12/2019 Urban Nuclei

1/26

Article: Mehaffy M, Porta S, Rof Y, Salingaros N, 2010,Urban nuclei and the geometry of streets: Theemergent neighborhoods model, Urban DesignInternational, 15-1, p.22-46.

Disclaimer:

This paper not necessarily reflects the final definitivepublication: it might be a pre-copy-editing or a post-print author-produced .pdf or in any case a differentversion of that. Therefore the reader is advised torefer to the publishing houses archive system for theoriginal authenticated version of this paper.


2/26

Original Article

Urban nuclei and the geometry of streets: The emergentneighborhoods model

Michael Mehaffy a,b, *, Sergio Porta b, *, Yodan Rofe c, w and Nikos Salingaros d, w

a Structura Naturalis Inc., 900 Cornell Street, Lake Oswego, Oregon 97034, USA. bUDSU Urban Design Studies Unit, University of Strathclyde, 131 Rottenrow, Glasgow G4 0NG, UK.E-mail: [email protected] Architecture and Urban Planning Unit, The Jacob Blaustein Institute for Desert Research,Ben-Gurion University of the Negev, Sede Boqer Campus, 84990 Midreshet Ben-Gurion, Israel.d University of Texas at San Antonio, One UTSA Circle, San Antonio, Texas 78249, USA.

*Corresponding author.

wAll authors are members of the Environmental Structure Research Group.

Abstract A controversy remains among planners and urban designers about the proper location of the non-residential core (nucleus) of a neighborhood in relation to thoroughfares. One school of thought suggests thatthe nucleus should be located along the busiest thoroughfares; a second school holds that it must be somedistance away from them which, because of their disruptiveness, should form the edge of the neighborhood;and a third school proposes that it should be somewhere between the two as an eccentric nucleus. The threeschools may be overlooking the underlying variables that govern this problem under different conditions, andso we propose a model for establishing the best location and distribution of urban nuclei as these conditionsvary. This requires firstly, a redefinition of the neighborhood as distinguished from a pedestrian shed. Weargue that a neighborhood can either emerge within a sanctuary area between thoroughfares, or span across both sanctuary areas and thoroughfares, if the latter are properly designed; a pedestrian shed, by contrast,can overlap with neighborhoods and with other pedestrian sheds. We propose a 400 meter rule, a surprisinglysmall maximum spacing of main thoroughfares that empirical observation shows that traditional, pedestrian-governed urban fabric has always tended to obey, for reasons that are likely to have to do with the self-organizing logic of pedestrian movement and social activity. In so doing, we advance a more fine-grained,permeable, potentially lower-carbon model and illustrate its advantages with several historic and modernexamples.URBAN DESIGN International (2010) 15, 2246. doi:10.1057/udi.2009.26

Keywords: urban nucleus; thoroughfare; neighborhood; urban morphology; urban model

Introduction

This article takes up one of the major issuesunder debate in New Urbanism and other reformmovements in urban planning theory and prac-tice. Should the commercial nuclei of neighbor-hoods be placed to straddle major thoroughfares,or should they be planned to locate away fromthem in order to allow for a safe and tranquilpedestrian neighborhood environment? We inter-pret this issue as a composite of two separate, if interrelated, questions: (1) the need for pedestrian

access to everyday needs and city wide andregional public transit (the pedestrian shed); and(2) the importance of neighborhood communityfor urban life.

We begin by showing that the urban structuretypical of post World War II development was brought about by the application of two moder-nist city planning concepts: the neighbor-hood unit, and the hierarchical classification of urban streets. These two concepts developedindependently in the early years of the twentiethcentury and were joined together, after World

r 2010 Macmillan Publishers Ltd. 1357-5317 URBAN DESIGN International Vol. 15, 1, 2246www.palgrave-journals.com/udi/


3/26

War II, to form the contemporary orthodoxy of planning for urban areas. As many critics havenoted, this orthodoxy is the cause of much of thesegregation of activities and lack of life in urbanstreets so typical of modern planned develop-

ment. The New Urbanist movement has donemuch to expose the damage to the city caused by the hierarchical classification of streets, andto bring back (at least at the local level) thetraditional urban network of streets. But it haskept on the modernist idea of the neighborhood,surrounding a distinct center, as a structuringelement of the urban region. The debate withinthe movement only varies (often vociferously) onthe appropriate location of the neighborhoodscommercial center: should it be within theneighborhood, within walking distance from allthe homes in it, and removed from through going

traffic on the major arterials surrounding theneighborhood, or should it be at the edge of theneighborhood, and enjoy the accessibility andvisibility that such a location provides? If at theedge, what happens in the very busy thorough-fares between centers?

We offer here a way to break the impasse of thedebate, by shifting to a deeper examination of neighborhood structure. To do so we developideas based on the notion of urban configurationas a network phenomenon. We rely on insightsof Christopher Alexander and his colleagues(Alexander et al, 1977; Alexander, 20012005),the New Urbanists (Krier, 1998; Duany et al,2000) as well as previous work on urban form andspatial networks (Hillier and Hanson, 1984; Rofe `,1995; Hillier, 1996; Porta & Renne, 2005; Salingaros,2005; Porta et al, 2006; Steil et al, 2008; Porta et al,2009). We introduce an abstract model for thelocation of urban nuclei, relative to the neighbor-hood and the movement network, and use it toclassify the various positions in the debate, as wellas variants that can be found in urban regionstoday. The analysis of those combinations aseither conducive or deleterious to urban commu-nity formation leads us to propose what we termthe emergent neighborhood model. In this model,neighborhoods are not fixed-designed entities, but mayemerge historically and socially from inhabitantspatterns of movement and activity, as well as fromadministrative decisions and actions. For neighbor-hoods to emerge, however, they need the basis of a well-formed urban structure. This structure is based on urban nuclei that shape pedestrian sheds,and whose size is dependent on their nature,density and characteristic uses. These nuclei span

or abut the main thoroughfares of the citysmovement network, while forming an edge tosanctuary areas protected from through movement.The thoroughfares, in turn, are smaller and morepermeable, allowing greater transversal connec-

tivity across them and a greater range of uses(including residential and institutional) alongtheir length.

Based on our observation of historic cities fromdifferent cultures, and the work carried out by theItalian morphologists (Caniggia & Maffei, 2001)we propose that the maximum edge length for asanctuary area (the area between major thorough-fares) is governed by a surprisingly small 400-meter rule. An extensive case-study research ispresently being carried out by the authors thatwill provide empirical support to the 400-meterrule idea. The study is basically a morphological

comparison of many urban layouts at the samescale taken form Google Earth. Cases includespontaneous and planned urban fabrics char-acterized by the most different historical, geo-graphic and broadly cultural (including religious)conditions. Here we anticipate the broadest resultof the study, the existence of the rule itself; wethen explain why this finding is so relevant in thehistory of urban form and the theory of urbandesign, and finally we propose a model to explainits formation in practice as the outcome of apiecemeal, largely self-organized process of evo-lution in time.

But neighborhoods may emerge in one sanc-tuary area, or span between two or more of them.Different people, or even one person at differenttimes in his/her life cycle, may come to relate todifferent levels of the neighborhood, but from ourstudies and that of other authors it seems that thenatural limitations of pedestrian movement intime (manifested in the 5 min walk) are funda-mental to community formation.

A key feature in our model is its ability toexplain how normal forces of urbanization willdrive one configuration to mutate into another,often resulting in a sudden reversal of urbanmorphology.

This model can be used to design new urbanfabric, or to upgrade existing inhuman urbanregions into genuine communities. We discussseveral contemporary urban pathologies, such ascities with neither pedestrian sheds nor urbannuclei, and describe some built examples thatembody those pathologies. We then show ourmodels extension to the regional scale, andhighlight how it resolves the aforementioned

Urban nuclei and the geometry of streets

23r 2010 Macmillan Publishers Ltd. 1357-5317 URBAN DESIGN International Vol. 15, 1, 2246


4/26

debate within reformist urbanism. We concludewith some thoughts on the importance of emergent neighborhoods in todays globalizingworld, and their importance in preventingthe social exclusion and complete isolation from

any functional urban life, for those without themeans or the capacity to connect with largerscale city and global networks. Reiterating therole of urban design in facilitating this complexurban growth over time, we propose seedingthose neighborhoods through the optimal organi-zation of main streets, urban nuclei and sanctuaryareas.

The Modernist Destruction of UrbanStructure

Urban fabric built today is pathological becausethe integrated pedestrian system typical of thetraditional city has been transformed by designinto a fragmented and dispersed car-orientedenvironment. The two main reasons for thispathology are: the hierarchical functional classifi-cation of urban streets, and the way in which itwas married to the neighborhood unit idea (Perry,1929).

The neighborhood unit idea stems from a socialand functional critique of the nineteenth centuryindustrial city, where for the first time in humanhistory many cities grew beyond the size thatallowed easy access to the open space surround-ing them, and where the social control of a closecommunity of kin and neighbors, typical of villages and small towns was absent. The conceptattracted universal attention after its application by Clarence Perry for the First Regional Plan of New York in 1929 and has become an acceptedmode of practice after the Second World War(Ben-Joseph, 2005). This despite early criticism onthe grounds that it fosters segregation, andstudies that showed that the neighborhood unitis not perceived as a community by its residents,and that peoples expectations with regard totheir community are very different (Banerjee andBaer, 1984 is perhaps the most explicit. See alsoBrower, 2000 and Biddulph, 2000).

On the other hand, city planners addressedmovement across larger areas with a thoroughfare(arterial) system, whereas access was relegatedto so-called collector and local streets. Theresult was that the thoroughfares were designedwith speed as an overriding consideration, whichwas most efficiently addressed with relatively

few, larger thoroughfares or arterials. By separ-ating movement from accessibility, and by closingmajor urban streets to their immediate surround-ings, their role as meeting places and urbaneconomic generators is thwarted. Because of this

separation of movement from accessibility, pre-sent-day streets fail to become social spaces,and thus fail to coalesce into a living city (Gehl,2002). This compromises pedestrian safety, and,ultimately, neighborhood integrity. Moreover, itcreates an area between retail-supportive andtransit-served streets that is simply too large.

Because the rationale behind Perrys neighbor-hood unit was to create an area with sufficientpopulation for a primary school, so that youngchildren will not have to cross major streets ontheir way to school or open space, the amalgama-tion of both paradigms was natural, and occurred

sometime after World War II. The combination of both paradigms, enshrined as they were in cityplanning and transportation manuals and guide-lines is the reason for the marked difference inurban structure between pre-war and post-wardevelopment in all countries of the developedworld, as well as in communist-controlled coun-tries. Its outcome is the lack of civic and streetlife in cities and city parts developed throughoutthe world a lack of life that has laid cities bareto the development of out-of-town shoppingcenters and malls which obviously tend todevelop as close as they can to the movementnetwork.

In fact, it is the movement network and the wayit structures daily movement patterns that drivespeoples social awareness and sense of commu-nity, as masterfully described by Jacobs (1961) andformally shown by Hillier (1996).

Empirical studies of peoples cognition of thecity reveal that major streets, their organizationand their relationship to surrounding areas arethe primary elements used for cognition andrepresentation of the urban environment (Conroy-Dalton, 2003). Thoroughfares that are still con-nected to the city fabric (those that have not beenturned into expressways) are the primary ele-ments for localization and orientation in aneighborhood. Usually, they are not perceived as boundaries between neighborhoods. Peoplesimages of their city also reflect their neighbor-hoods access to the larger movement and socialnetworks of the entire city, that is, how well theirlocal area functions as an entry point to theregional transportation and social networks (Rofe `,1995; Talen, 2000).

Mehaffy et al

24 r 2010 Macmillan Publishers Ltd. 1357-5317 URBAN DESIGN International Vol. 15, 1, 2246


5/26

Certain routes will be favored over others foreveryday use, creating social patterns of interac-tion through shared paths. Nevertheless, commu-nity (as a shared system of spatial recognition andawareness) is shaped both by the integrators of

the urban grid and by the barriers. Local breaks inthe grid, such as institutional buildings, gardensand so on, help to shape these gradients. By theirvery presence away from the busiest thorough-fares, small squares or parks constrain movementand reduce centrality, and yet they are a destina-tion that draws people by virtue of their nature anduse. The same goes for schools and other institutions.

The lesson for city planning is that all streets mustprovide for both movement and access. That meansthat thoroughfares (not expressways) must allowtransversal permeability by pedestrians and allother kinds of road users (Philibert-Petit, 2006); it

also means that intersection and direct exchangemust connect thoroughfares and local streets with-out other intermediate classes of streets in-between.In a recently proposed taxonomy by Marshall(2005), any subjection to access constraint (that isthe rigidly hierarchical succession of intersectingstreets) is abandoned. This is a precondition to re-establishing a traditional conjoint type of streetlayout in city planning after decades of anideological shift towards dendritic or mosaicmodels. In this way, thoroughfares can be properlyre-framed as actual main streets (Jacobs, 1993).

Within the New Urbanist movement, and in theurban design literature there is an acknowledg-ment that the functional hierarchy of streets has been damaging to the city (Murrain, 2002;Hebbert, 2005; Marshall, 2005). Even some trans-portation engineers have begun to realize that itwas not even an efficient transportation network(Kulash, 1990; US-EPA, 2004). Others, have con-tested the safety benefits of the hierarchicalclassification (Ernst and McCann, 2002) andparticularly for low-income groups (Loukaitou-Sideris et al, 2007). Questions are also arisingregarding the overall public health benefits of hierarchical classification if it leads to higherreliance on car travel (Cervero and Duncan, 2003;Badland and Schofield, 2005; Ewing and Kreutzer,2006; Frank et al, 2006; Boarnet et al, 2008; deNazelle and Rodr guez, 2009; Forsyth et al, 2009;Townshend and Lake, forthcoming).

However, the neighborhood idea still persists,as an essential part of the tools to create urbanism.This can be seen in the chapter devoted to it in theCharter of New Urbanism, in various diagramsproduced by Duany Plater-Zyberk (DPZ 1990) as

examples of good urbanism, in the diagramsproduced by Calthorpe (1993) as part of his ideasfor Transit Oriented Development, (see alsoBernick & Cervero, 1997), in the literature onurban villages (Franklin, 2002; Brindley, 2003;

Sucher, 2003) and most recently in the diagrams of sustainable urbanism produced by Farr (2008).Although these ideas have been contestedstrongly within the New Urbanist movement mostly by Jones and Kauffman (2009) andMurrain (2009) who developed different modelsof district development the debate so far has been raging about the location of neighborhoodcenters and their location relative to the move-ment network. Here we attempt to analyze thisdebate, and offer our resolution of it, by redefin-ing the relationship between urban nucleus, itspedestrian shed and the neighborhood.

Framing the Debate: Urban Nucleus,Pedestrian Shed, Neighborhood andMain Thoroughfares

Let us sketch out the current debate in planningand urban design literature and practice, in termsof how one defines a neighborhood or urbanquarter, and where its core is to be located. Theurban nucleus contains the highest density of each function (residential, commercial and so on)as well as the densest mixture of all distinctfunctions. There are several choices, but some of them represent dysfunctional city form.

We propose here a four-way classification based upon combinations of the two independentfactors of position and connectivity (Figure 1):protagonists of these combinations are urbannuclei (red-to-yellow smaller circles), urban neigh-borhoods (pink larger circle), pedestrian sheds(dotted red circumference), thoroughfares (thick black lines, dotted when tamed for pedestriancompatibility) and local streets (thin dashed blacklines). Every urban situation can be described assome binary combination, simple or complex, of these factors.

1. CENTERED NUCLEUS. The urban nucleus islocated somewhere around the geographiccenter of the neighborhood.

2. EDGED NUCLEUS. The urban nucleus islocated somewhere at the edge of the neighbor-hood.

3. EXPOSED NUCLEUS. The urban nucleus ismainly considered a result of the movement




6/26

network and is therefore located in closeproximity to urban thoroughfares.

4. SHIELDED NUCLEUS. The urban nucleus isconceived as the outcome of far more complexdynamics, where the movement network is justone factor among many others, thus thenucleus is considered suitable for quietersanctuary street realms.

Some urbanists are in favor of a CENTERED/SHIELDED option, where centered urban nucleiserving the local community are not locatednecessarily in combination with urban thorough-fares (Figure 2). This model places activities at thecore of a neighborhood, where the neighborhoodsare enclosed by but are detached from thorough-fares. Neighborhoods and pedestrian sheds there-fore overlap, which ensures a good efficiency forthe latter. The limitation with this model is thatneighborhoods cannot grow beyond the limitimposed by surrounding thoroughfares, and theurban space remains fragmented in disconnectedpatches. More seriously, major mixed-use hubsinevitably gravitate around prominent crossings

of motorways or major urban arterials in acar-oriented environment, which undoes theCENTERED/SHIELDED typology. This processseparates the local neighborhood community levelof mixed-use locations from the district-to-metro-politan level (around thoroughfares) in differentareas of the city, thus creating unresolved forcesthat lead to morphological instability.

Other urbanists are more supportive of anEDGED/EXPOSED option, where edged urbannuclei are this time tightly connected in spacewith thoroughfares that bring with them theso-called movement economy (Figure 3). Thisrepresents social and commercial hubs traversedand served by major channels of information/human/goods flows. An example is the shoppingmall or high-rise office cluster at the intersectionof thoroughfares. In this case, neighborhoods are bounded by thoroughfares but are kept awayfrom each other, from the thoroughfares, and fromurban nuclei alike. Even when the traffic iscalmed, transversal connections are spatiallydistant. Although in a traditional urban fabricthe neighborhood might possibly benefit from a

Figure 1: Urban nucleus and edge conditions.Note: colours described in the text can be seen in the online version of this article at www.palgrave-journals.com/UDI.

Mehaffy et al



7/26

direct walkable connection to urban nucleiat intersections, in many later instances it isdisconnected, and access is roundabout and bythoroughfare only. Posed in this way, the opposi-tion between the CENTERED/SHIELDED versusEDGED/EXPOSED standpoints could not besharper: it is in fact irreconcilable.

The alternative CENTERED/EXPOSED core(Figure 4) places urban nuclei close to thegeographical center of the neighborhood in strictand direct proximity to trafficked roads in orderto take advantage of the movement economy. Thisis the traditional compact urban scenario, aremarkable scheme that has always been a modelfor traditional urbanization processes before themotor age. It survived until main streets begansevering rather than joining urban communitiesdue to an increased level of traffic, and separatingthrough movement from access, as mandated bytransportation engineers. Here, neighborhood boundaries are no longer created by streets, but by some other type of natural or artificial barrieror discontinuity. The model becomes problematicwith the disturbance of fast moving automotivetransportation on major roads through the urban

centers. Unresolved forces generated throughincreasing traffic flow create an instability that canpush the CENTERED/EXPOSED core (Figure 4)to transform into the EDGED/EXPOSED model(Figure 3). To prevent this, the whole scheme ismade practicable when thoroughfares are traffic-calmed locally into main streets, which therefore become supportive of the denser social dynamicsthat give life to urban nuclei.

The opposite alternative is EDGED/SHIELDED(Figure 5), in which urban nuclei are placed at theedges of neighborhoods, or actually removedfrom them, but are also disconnected fromthoroughfares. This model combines all possiblenegative aspects of severed communities, sec-luded social/commercial hubs and inefficientpedestrian sheds. Abstract though it is, this modelhas nevertheless been widely practiced for theurbanization of western societies since afterWorld War II.

Neighborhoods and Pedestrian Sheds

A key to understanding the resolution we offer of this debate is the distinction we make between the

Figure 2: Centered/Shielded relationship of pedestrian shed, neighborhood nucleus and movement network.




8/26

notions of neighborhood and pedestrian shedin relation to the urban nucleus. There is anassumed relation between these two notionsimplicit in the debate and in most urban designliterature, which we seek to dissolve.

An urban nucleus is by definition the center of its own pedestrian shed, which may or may not be identifiable with a neighborhood. Urbannuclei should be reachable within some 5 minwalk from most residences and non-industrialworkplaces. This accessibility is necessary inorder to achieve several essential goals of thesocial and environmental sustainability agenda(Newman and Kenworthy, 1999). Urban nucleiwill in most cases be located somewhere near thegeographic center of the pedestrian shed, becauseany pedestrian shed is by definition determined by the location of the urban nucleus itself.

On the other hand, there are many waysto define neighborhoods. As a result, parish boundaries could come from centuries of history;other boundaries may be defined by the servicesheds (some larger, some narrower) of differentcategories of services; we may have different

systems of administrative boundaries, eitherhierarchical or partially overlapping. In histor-ical territories, past administrative boundarieswere in use in medieval times and typicallyenclosed main streets as central spines or back- bones of neighborhoods (quite the opposite of most administrative boundaries currently in use,which coincide with main streets). Finally, wehave the fuzzy boundaries of ethnic groupconcentrations, as well as other boundaries thatshow up when other demographic factors (suchas age or income) are mapped. When interview-ing people about their neighborhoods we comeup with many other different boundaries, some-times consistent and sometimes divergent, butoften different from the formal ones used bypublic officials.

At the local scale, the interplay and overlapping between countless personal and collective formsof social and cultural relationships (which we callcommunity) finds temporary spatial configura-tion. This process occurs at every step of theurban evolution into emerging aggregations.Neighborhoods are layered (fractal) formations

Figure 3: Edged/Exposed relationship of pedestrian shed, neighborhood nucleus and movement network.

Mehaffy et al



9/26

of overlapping sub-formations, each of which isstill layered in itself: they should not be conceivedas fixed and delimited geographical entities. Aneighborhood is not identical with the geo-graphic and socio-cultural extension of an urbancommunity. Spatial configurations influence evol-ving and borderless neighborhoods specific tosite, social groups and individuals. Places andcommunities interact precisely because they areinherently different in nature.

Urban designers have always been conceivingneighborhoods (during the entire history of ourdiscipline) as fixed and stable community unitsaggregated around a center of basic services thatshould ideally be accessible on foot. This imagethus unites in the same concept both a socio-cultural meaning related to the term com-munity, and a geographic one related to theterm accessibility. We, however, stress a moredynamic idea of the urban community as a socialand cultural construct, especially in contemporarysocieties (Castells, 2000; Talen, 2000; Salingaros,2005).

In short, neighborhoods are only definable in aspace characterized by variable geographies thatdepend on who uses this structure. Moreover, thisvariable geography of use is dynamic not just inspace, but in time as well. It endlessly changesand evolves along with society and values overtime. That is why we distinguish very clearly theneighborhood as a social, cultural and historicalentity (complex of entities, to be sure) from thepedestrian shed. A pedestrian shed is basicallyspatial proximity characterized by some sort of distance to be covered by citizens towards theirdaily destinations to satisfy ordinary needs.

Predominantly residential areas, creating areasof less through movement, but hopefully morelocal movement, can help to generate a sense of community among people sharing them. Bythemselves, however, these areas are not enoughto create a neighborhood. In the final analysis,neighborhoods are created by the social networks,strong and weak, that arise in such places, andthese networks are only partially dependent uponthe spatial structure. Therefore, we dont believe

Figure 4: Centered/Exposed relationship of pedestrian shed, neighborhood nucleus and movement network.




10/26

neighborhoods should or can be designed: to placeanything, including urban nuclei, as the center of aneighborhood is not within our domain as urbandesigners. Managing public policies of transportand land-use such that most residences are locatedat walkable distance from transit access, retail and basic services (in other words, shaping pedestriansheds) is within our domain.

Summing up, our proposal implies that pedes-trian shed and neighborhood need not be con-gruent. This leads to many possible resolutions of the disputes about urban nuclei and thorough-fares: the most important is that neighborhoodsmay have (but are not required to have) a quietcenter with civic space, a corner grocery ordestination retail, and perhaps a bus stop (buteven that is not needed) that should not necessa-rily coincide with either the center of pedestriansheds or urban nuclei. A new scheme is thereforederived from the EDGED/EXPOSED compacturban model previously illustrated by relaxingthe geographical conditions of neighborhoodformation. It may be termed the EMERGENTNEIGHBORHOOD MODEL (Figure 6), and isdeveloped in more detail in the next sections.

The Solution: Well-connected Nuclei WithDiffuse Internal Flow, or the EmergentNeighborhood Model

What brings people into a certain space more thaninto another? This is the great problem; yet,considering what drives urban dynamics in time,consistent evidence indicates that centrality is thekey. Central places tend to evolve, all other thingsequal, into popular places with many passersby,which makes such places attractive for retail andservices, which in turn makes them even moreattractive for additional passersby, and so forth ina self-reinforcing cycle. Having many passersbydoes not necessarily mean congestion of cars andtrucks, however. It just means many people, and if space is limited, then preference should be givento the modes that use space more efficiently:pedestrians, bicycles and public transit, whilelimiting access to other modes either by spaceallocation or by time allocation to off-peakperiods.

We must place urban nuclei in the bestconditions in which they can root in and growin the long run. That is feasible when: (1) they

Figure 5: Edged/Shielded relationship of pedestrian shed, neighborhood nucleus and movement network.

Mehaffy et al



11/26

form out of urban clusters in close proximity tohighly central places, especially intersections between popular thoroughfares; and (2) theymatch access points to major public transporta-tion systems, thus establishing a convenientpedestrian shed both in terms of accessibility tolocal retail and services and to the region as awhole through the transit system.

All this is pictured in the proposed EMER-GENT NEIGHBORHOOD MODEL (Figure 6).This model is an evolution of both the traditionalCENTERED/EXPOSED scheme (Figure 4) andits contemporary reformed EDGED/EXPOSEDversion (Figure 3) illustrated above, with twoimportant differences: the pedestrian shed is heredistinguished from the neighborhood. The neighborhood also has lost the rigidity of the one-circleshape to take the form of a variable-geographycomplex of layered areas. These areas gravitatetowards urban nuclei but are not centered onthem necessarily.

The problem is the level of conflict that ensueswhen a node of activity is spatially centered on amajor urban road, due to the severance effect of motor vehicle traffic and, in addition, the width of the right-of-way. This is not an insurmountableproblem. It has been shown that a well-designed boulevard can carry a very high flow of trafficwithout bisecting the pedestrian shed. Boulevardslike the Diagonal in Barcelona, for example, cancarry around 100 000 vehicles per day (Jacobs et al,2002). Traffic-calming and street design techni-ques make it possible to solve the conflict betweenthoroughfares and dense activity centers, even incases when the former take the form of heavilytrafficked roads if one is willing to accept reducedspeed (which does not necessarily mean muchreduced capacity).

In what sense then does the proposed EMER-GENT NEIGHBORHOOD MODEL (Figure 6)reduce the schism between the standpoints repre-sented in Figures 2 and 3? Neighborhoods can be

Figure 6: The proposed Emergent Neighborhood Model. Like in both the Centered/Exposed and Edged/Exposed options, of whichthis represents an evolution, urban nuclei are on thoroughfares. However, neighborhoods are no longer constrained into a fixedgeography, but are represented as the outcome of the dynamic overlapping of many layers that vary in space and time throughprocesses of self-organization. That allows including in this model the emergence of local focuses of mixed-use activities within areasof higher neighborhood potential detached from thoroughfares.




12/26

quieter areas framed by local streets and never-theless converge around some sort of communityarea placed more or less in the center of theneighborhood itself. Some services, like specialkinds of retail shops or the lower grades of

educational services, may take advantage of urban environments that put some distance fromthe hustle and bustle of urban nuclei. Theproposed scheme presents ripples of denserneighborhood layers at the interstices of pedes-trian sheds and away from dense activity centerson thoroughfares.

In this model, neighborhoods can still bequieter areas, but activities strongly need to stayclose to channels of the movement economy, thatis thoroughfares. On the one hand, neighbor-hoods do not rigidly respect the structure of centers and thoroughfares but flow in different

ways over the substrate of this structure. Onthe other hand, pedestrian sheds abandon anynecessary reference to neighborhoods and remainanchored to the structure of urban nuclei andthoroughfares.

The solution to the placement and structure of the urban nucleus incorporates some sophisti-cated notions learned from biology. The centerof our nucleus accommodates urban activitythrough flow, but only up to a certain carefullymonitored threshold. This is analogous to havingdifferent systems of internal body flow that arenecessary for the organism to function. There isalways a very delicate balance, however, betweenthe structural wholeness of the organs themselvesand the requirements of the adjoining transporta-tion system that is required to feed them. Inaddition, much of internal body flow is in factcapillary: that is diffusion rather than medium orfast flow. Our conception of human-scale urban-ism shares this notion of capillarity. Humanpedestrian movement being the most delicateform of movement (compared with vehicles)corresponds to capillary flow in the urban fabric(Salingaros, 2005, Chapter 6).

Urban Nuclei, Main Streets and SanctuaryAreas: The 400-meter Rule

Distinguishing between the concepts of pedes-trian shed and neighborhood, as was discussedabove, re-defines the concept of neighborhood.The conclusion that we draw is radical: we cannotdesign neighborhoods, because neighborhoodsself-organize beyond any hierarchical top-down

control. What then remains for urban designers tohandle after the disappearance of designableneighborhoods? Basically two things: (1) pedes-trian sheds, and (2) main streets and sanctuary areas.

Conventionally, pedestrian sheds are deter-

mined by the location of mixed-use and transpor-tation hubs (urban nuclei), and the type of streetlayout. Pedestrian sheds are simply a spatialrange of accessibility to facilities. The pedestriansheds form and extension depends upon thethreshold of accessibility that we consider accep-table for that category of facility, and also on thetypes of facilities that we select for essentialaccess. We can assume a threshold distance of the 5-minute walk, roughly equal to 400 m (1/4mile), for basic services and light rail transit. Thisthreshold can be less for local transit services like buses, more for underground metro and still

higher for regional transit services like railways.Facilities that include mixed-use commerce/service activities and access to transit constitutewhat we term an urban nucleus.

The type of street layout determines actualpedestrian sheds, with highly interconnectedgrids leading to higher permeability of the urbanspace, thus better approximating the virtualpedestrian shed circle. At the schematic level,however, we can represent local pedestrian shedsvery roughly in the form of circles of 200 m radius(or, say, 400m radius for longer-reach transit)surrounding the urban nuclei.

A pedestrian shed, therefore, has a fairlydefinite geographical shape so that we can indeeddesign it. Pedestrian sheds become a key tool inthe hands of urban designers; their impact spansover major roads and local streets, residentialquiet areas and intense activity centers, primaryurban functions and ecological reserves likegardens and parks. The new Emergent Neighbor-hood Model addresses this complex relationshiparound two interrelated urban notions: mainstreets and sanctuary areas. This approach leadsto an understanding of urbanization as a timelessprocess based on the 400-meter rule; it alsoprovides a conceptual framework for interpretingthe greater part of urban fabrics, which are neitherdense/active places (that is urban nuclei), norappendages of thoroughfares.

With neighborhoods seen as flowing, evolving behavioral and socio-cultural formations ratherthan spatial entities, we use the term sanctuaryarea to define those quieter urban areas, mostlyresidential, here and there punctuated by servicesor local retail, where most of us live. Sanctuary

Mehaffy et al



13/26

area refers to Donald Appleyards distinction between busy streets full of people and activitiesfrom calm, local, mostly residential realms that hetermed sanctuary streets (Appleyard, 1981). Heconceived these two distinct domains as being

essential elements constituting the good city. Wewill explain how sanctuary areas are related tourban nuclei and thoroughfares.

A comparative study of many cities withdifferent historical, cultural, geographic, climaticand economic characteristics reveals that sanctu-ary areas (that is those urban areas made of sanctuary streets) are one of the two building blocks of human cities across all times andplaces; the second is main streets. We can see thisdynamic in action in the paradigmatic case of the historical center of Bologna, Italy (Figure 7,plate a). The same dynamic applies to a case as

culturally and historically different as Al Hofuf,Saudi Arabia (Figure 7, plate b). We observe thatmain streets in all living cities are those connect-ing the global to the local scales (in historicinstances, they connect the city center with theancient city gates), whereas sanctuary areas arequieter spatial entities bordered and defined bymain streets. Such sanctuary areas are roughlyquadrangular (or triangular in the case of con-verging streets); their size is determined by oneedge that measures between 350450 m. Thismeasure of around 400m on the side is surpris-ingly constant all over the world and across urbanhistory. The rationale behind this 400-meter ruleis embedded in the general mechanism of theurban fabrics evolution that has been especiallyinvestigated by the Italian school of urban mor-phology (Caniggia & Maffei, 2001). We can extendthose findings with a new focus on the interplay between main streets and sanctuary areas.

The process begins (Figure 8, plate a) with adensification around one major street at one pointthat is for some reason prominent (for example,an intersection with another important street).Buildings and lots directly abutting on the twomain streets organize from both sides (here werepresent just one side). A few small streets beginto grow from the two main streets inwards,showing a dead-end configuration; these streetsare the seeds for the future development of theinner street network that will frame the sanctuaryarea at the end of the cycle. In a second step(Figure 8, plate b) a new main street stemsorthogonally from the generator path somewherearound 400 m from the previous intersection, beginning here from both sides. What we have

constitutes a first merged urban fabric, still notvery dense, towards the original core of the city. Itis framed by a complex but still not clearlyhierarchical network of local service streets.

In a third step (Figure 8, plate c), the first

sanctuary area finally gets closed by anothermain street emerging parallel to the generatorpath. As the internal street network becomesmore complex, a few routes emerge that providedirect shortcuts to the higher main streets net-work. Local service streets that evolve intothrough routes connecting main streets becomelocal main streets; they assume some of thecharacteristics of main streets, that is density in built fronts and location of non-residential activ-ities (especially at ground floor), though at alower intensity. This type of street offers the bestconditions for the spill-over effect to happen, by

which the mixed-use system of commerce andservice gradually expands from main streets ontolocal main streets, and thus into sanctuary areas.The depth of this penetration dynamic dependsupon local conditions, particularly on the grade of attractiveness exerted by the outer main street atits intersection with the local main street.

The dynamic relationship between sanctuaryareas and main streets that we call the 400-meterrule is a timeless pattern in city evolution thatemerges whenever the urban fabric is the out-come of a historical self-organized process of formation. This rule has to do with severaluniversal characteristics of the human body inrelation to the environment, and correctly bal-ances the need to navigate urban space effectivelyon foot, with the competing need to maximize theuse of space for developing buildings and activi-ties at a proper density. Interestingly enough, wefind this rule not only in the spontaneousformation of self-organized cities in history, butalso everywhere in the rationally planned newfoundation cities before the industrial age, fromancient Roman colonies like Augusta Taurinorum(now Turin) to Renaissance ideal cities likePalmanova. It is only with modernism that cityplanning divorced itself from the 400m rule of evolving cities applied during all preceding times, by introducing the concept of neighborhood as ageographic entity based around the population of an elementary school, and disconnecting adjacentunits by high-speed inaccessible arterials (see ourearlier discussion, above).

We conclude that sanctuary areas as building blocks of cities in history have always beensensibly smaller than the neighborhood as we




14/26

know it in literature and practice. This leads to aprofound structural diversity in the scale of urbanfabrics. The half-mile (800 m) Jeffersonian grid,a measure that has persistently framed theneighborhood literature in our modern discipli-nary tradition (including some New Urbanistschemes), is simply out-of-scale. Other examplesare Chandigarh in India and Beer-Sheva in

Israel: both planned according to modernist cityplanning dogmas and based on an 800 1200mgrid, each cell of which was somehow meant to bean autonomous neighborhood and is in factseparated from all others by major roads. Thisscheme is an abysmal failure based upon acomplete misunderstanding of human natureand urban community.

Figure 7: The timeless pattern of main streets (thick solid red) and sanctuary areas (on the background, defined by main streets) hereat work in the cities of Bologna, Italy (plate a), and Al Hofuf, Saudi Arabia (plate b). Local main streets (dashed orange) emergeregularly within sanctuary areas as denser shortcuts connecting the higher network of main streets.

Mehaffy et al



15/26

Figure 8: The 400-meter rule as an outcome of the process of self-organized urban accretion that is conducive to the timeless patternof main streets and sanctuary areas.




16/26

Given the human limits of pedestrian move-ment, all main streets must maintain connectivityon the order of 400 m maximum spacing, and thedistance between transit-served and retail-sup-portive streets (as determined by their catch-

ments) must be on the order of 400 m maximumseparation. The 400 m rule establishes a diffuse,connected street structure containing the mini-mum spacing for continuous walkability, whichensures a close relationship between the san-ctuary areas and the urban nuclei: close enoughthat you can actually walk to four differenturban places from anywhere within the sanctuaryareas.

The Emergent Neighborhood Model in theRegion: The Case for Urban Seeding

A taxonomic investigation (Figure 1) of therelational position of urban nuclei, pedestriansheds and neighborhoods in the urban quarterspace resulted in a formation that we termed theEmergent Neighborhood Model (Figure 6). In ourproposal, traditional compact urban developmentfinds a new synthesis based on the distinction between pedestrian shed and neighborhood, andthe consequent reframing of neighborhoods asdynamic, social-spatial formations. Introducingthe term sanctuary areas and the 400-meter ruleas a metric for the self-organization of urban

fabrics, we argued that this is a timeless metric of cities throughout urban history, regardless of culture, until the codification of the neighborhoodconcept in the early twentieth century. We nowextend our proposed conceptual model (Figure 6)to a new vision that embeds the spatial dimen-sion: the Emergent Neighborhood Model in urbanand regional space (Figures 9 and 10 (a)(e)).

At the urban scale (Figure 9) the model offers afiner hierarchy of nuclei and transit systems,which also implies an equally finer hierarchy of pedestrian sheds. The main points of this visionillustrate:

K The formation of denser developments inproximity to transit service of a higher rank(that is infrastructure-dependent mass transit).

K The generation of mixed-use corridors as anoutcome of contiguous urban nuclei mergingalong such systems.

K The aggregation of local foci along local mainstreets within sanctuary areas as a result of thespill-over effect from main streets.

K The continuity of a connected ecologicalnetwork of linear and planar green elementssuch as parks, tree lines, boulevards or riversall over the urban fabric, with arrangements atintersections with main streets that maintain

the critical connectivity of the system.Compared with similar visions in the history of urban design, including recent New Urbanistschemes (Duany & Plater-Zyberk, 2002; Farr,2008), the proposed scheme introduces severalnovel features:

K The lack of a tree-like hierarchy (in the sense of the large-scale dominating the small-scale) of neighborhoods around one neighborhood ornucleus of higher rank. That type of hierarchyof communities makes the small-scale depen-dent upon the large-scale, and would in fact betoo rigid to allow any kind of self-organizationand emergence of neighborhood to take place.

K The relinquishing of the neighborhood as aplanning unit: the building block of urbanfabric is no longer the fixed neighborhood unit, but is instead the network of pedestrian shedsconnected by transit stops within urban nuclei.

K The 400 m rule defines quieter predominantlyresidential sanctuary areas, instead of regionsof 800 m or more typical of twentieth centuryneighborhood unit-based models. Our propo-sal shows how the 400 m rule matches well

with both the hierarchy of urban nuclei andthat of transit systems.

As a result, our model stresses a continuous andpermeable urban space, framed by an accessibletransit infrastructure located within urban nuclei.The urban hierarchies are flexible and areconstantly being reframed through a self-organiz-ing process acting on a spatial substrate of places.

These characteristics are also legible in theextension of our vision to the larger regional space(Figure 10 (a)). Overlapping social and geographicunits show up as emergent neighborhoods andpedestrian sheds of different sizes structuredaround urban nuclei. This evolution in turnreverses the top-down hierarchical structuresproposed on the basis of a geographic notion of neighborhood in all post-traditional urban-ism from the Garden City to the modern city(Alexander, 1965), down to the latest New Urban-ism and Place-making schemes. A complex inter-lacing of social, cultural and economic layersemerges and evolves over a spatial substrate thatincludes features of mobility (private and public)

Mehaffy et al



17/26

Figure 9: The proposed Emergent Neighborhood Model in urban space.




18/26

Figure 10: : The proposed Emergent Neighborhood Model in the regional space: ( a) the complete picture; ( b ) the network of privatemobility; ( c) the network of public mobility; ( d) the network of the built environment; ( e) the network of the natural environment.

Mehaffy et al



19/26

and the environment (built and natural). Althoughsocial, cultural and economic phenomenaare rapidly changing, the spatial substrate (Figures10 (b) (e)) is relatively permanent in the sensethat it changes at a far slower pace. It lies indeed

within the domain of urban designers, in the sensethat its structure should be the matter of arenovated discipline of sustainable, human, adap-tive urban design or, in short, of urban seeding.

In the proposed model, main streets (Figure10 (b)) are the backbone of the ecological network.They provide the main linear natural features thatmake the whole ecological network connected.Nevertheless, main streets must not be conceivedas parkways in a large green strip; in that case,they would split the built fabric of sanctuary areas(Figure 10 (c)) into visually and spatially dis-connected regions. That would inhibit the self-

organization of dynamic neighborhoods, pushingthe system back again into a hierarchicaltop-down structure as in Chandigarh, Brasilia,Cambernauld, Milton Keynes and countless othermodernist and post-modernist planned estates.This is because, as Kevin Lynch stated, the goodcity is one in which the continuity of this complexecology is maintained while progressive change ispermitted (Lynch, 1981, p. 116).

Linear natural features (Figure 10 (e)) shouldinstead be interpreted as tree-lines, green storm-water medians, water streams, shrubs or hedgesand the like, which do not fragment the builtfabric into separate parts and do not interrupt thecontinuity of the citys complex ecology withspatial vacuums and barriers. Inserting thesefeatures as part of the design of main streets is amatter of innovation and research, creating anew urban component of the sustainable city of tomorrow. Local linear greens in the form of smallpublic or private gardens, ponds and the like(what in short constitutes the ecologicalpatrimony of a good residential realm) are to beconsidered part of the sanctuary areas structure,directly connected with the urban ecologicalnetwork.

The overall structure of the main streets needsto be coordinated, designed and built at the scaleof the city as a whole, or in the case of majorthoroughfares, at the urban regional scale. How-ever, the street and built-up fabric of sanctuaryareas (Figure 10 (d)) can and should be structuredsuch that a continuous process of adaptation andevolution takes place from the bottom up,through the personal initiatives of residentsand citizens, outside of any formal process of

participation. This form of grassroots contributionto the constitution of the urban fabric has always been the original and powerful form of citizensinvolvement in the construction of their livingurban environment, with the greatest advantage

for the common good. In order for this participa-tory dynamic to emerge and flourish again, theproper spatial, proprietary and procedural pre-requisites should be devised and established atthe local scale.

There are by necessity discontinuities in theurban realm (Salingaros, 2005). These may belarge elements of infrastructure, wildlife corridorsand ecologically important areas: areas of heavyindustry, or warehousing or areas of extremeconcentrations of particular uses such as down-town commercial centers. How are these disconti-nuities to be negotiated within the new paradigm

we are proposing?1

In fact, there is a very limited number of highlyspecialized activities that need proximity toexpressways. The vast majority of economicactivities and services, including all those thatdirectly sustain the daily life of urban commu-nities, need the kind of centrality that is based onhow streets are positioned and mutually inter-connected. This is pretty much independent frommotor traffic flows. It is therefore crucial to avoidany confusion about the word thoroughfare,which in this article means a major channel of information/human/goods movement, but notnecessarily anything like expressway or evenarterial road. In our model thoroughfares arere-considered in a system of main streets. In allcases where urban nuclei are to be centered onactual major traffic roads, an intensive use of traffic-calming techniques, including turningarterial roads into boulevards, are needed toresolve the conflicts.

Urban Pathologies Interpreted Within theUrban Nucleus Framework

In the following three subsections, we describethree known urban pathologies in terms of thegeneral urban nucleus framework introducedabove. The first two cases describe conditions inwhich urban nuclei are either non-existent orextremely weak, mainly due to the lack of pedestrian sheds, or their misplacement relativeto major movement corridors. The third caserepresents suburban sprawls response to urbannuclei: a simulated nucleus that is divorced from




20/26

any pedestrian shed. The three pathologiesare (1) strip development, characteristic of theearly phases of suburbanization, (2) plannedunit developments and (3) edge city and variouscampus-like isolated developments, such as

enclosed shopping malls, or their recent reincar-nations as lifestyle centers.

Strip development

Of all suburban typologies, strip malls are perhapsthe easiest to deal with. Their typology is EDGED/EXPOSED, but because of the very low densitiesof development, the pedestrian shed is relativelyempty, and therefore the commercial life along themajor arterial is completely dependent upon caraccess. Early strip malls may have developed inareas where the street grid was fairly dense, butthe enforced segregation of uses and the com-pletely different scale of the commercial develop-ment along the arterial, relative to the residentialareas on streets perpendicular to it, separates theexposed world of the commercial street from theresidential areas surrounding it. In later stripdevelopments, this separation is exacerbated bythe drastic reduction in the number of streetsallowed to intersect with the arterial. Otherfactors that hinder the development of nucleiinclude: extensive parking lots fronting and some-times completely surrounding the properties; the

design of the street for capacity and speed of cars;a lack of shade; a lack of adequate sidewalks; andfrequent curb cuts. All of these make pedestriancrossing and access difficult.

The strip mall configuration could be trans-formed rather easily into the Emergent Neighbor-hood Model by increasing residential densitiesaround selected nuclei; improving public trans-port along the corridor; reducing speeds; improv-ing the pedestrian environment and crossingability; and removing front parking lots to ashielded position. The pedestrian scale connectionof such nuclei with their residential hinterland

has to be implemented according to the 400 m rulein conjunction with such improvement.

Planned unit development: Cities without anyurban nuclei

The second widespread urban pathology we findtoday results in the complete absence of anysignificant urban nuclei in expansive areas of residential suburbs. In merging the neighborhood

unit idea together with the functional hierarchy of streets, neighborhood centers were envisagedcombining commercial and civic facilities in thecenter of neighborhoods. This is the CENTERED/SHIELDED nucleus type described above (Figure 2).

However, strict zoning codes, low densities andthe lack of movement through the nucleusprevent an urban nucleus from actually forminginside a suburban pedestrian shed, even if thosewere designed, and so in recent years suchattempts were abandoned (Steil et al, 2008). ManyNew Urbanist developments have tried to redressthose problems by increasing density (particularlyaround the nucleus), improving internal walk-ability and inviting urban design. The lack of through movement and exposure to the outside,however, has meant that these neighborhoodcenters have difficulty in succeeding commer-

cially, and in generating more complex uses.When they do succeed, it is often throughbranding, where they become an attractionto drive to, even if not located on the mainthoroughfares of the city.

Edge cities, shopping malls, employmentcenters, and gated communities: Isolatedconcentration

Even areas of suburban sprawl need commercialand employment centers, and civic functions.The commercial edge city we now see amongdormitory suburbs is both spatially fragmentedand not really dense. Each commercial node isisolated from the pedestrian shed, but it is alsoisolated from every other neighboring commercialnode. These nodes exemplify the EDGED/SHIELDED nucleus option (Figure 5). As notedabove, it would at first appear that this combina-tion should never be experienced on the ground(but it is!), as it encourages severed communities,peripheral urban nuclei and inefficient pedestriansheds.

In this category, we find high-density mono-functional clusters that are built in peripheralareas and are thus poorly connected to any urbannucleus. Significantly, they themselves do notdefine a very good urban nucleus. Even in exten-sive industrial estates, educational campuses ormilitary bases that are isolated, attempts to mimicnormal urban life are never entirely successful because of the lack of urban variety and restrictedconnectivity with the outside world. Character-istic examples include light industry that was

Mehaffy et al



21/26

removed from the urban nucleus; high-rise apart-ment blocks built in the periphery and verypoorly connected by roads to the rest of the city;schools and shopping malls built in the middleof nowhere and connected only by one road. In

short, we see the typology of a highly concen-trated mono-functional urban node, usuallyaccessible by a single feeder road. Industrial esta-tes and prison camps were also of this typology:these concentrated urban structures are intention-ally isolated from other urban regions by difficultaccess (Kargon & Molella, 2008).

Many of these examples are further isolated bya perimeter fence (for security reasons), thuspreventing diffuse connections from all sides.That is why they are nowhere inside a pedestrianshed. Entry and exit is usually via a singlecontrolled road: not a through road but a dead-

end feeder road. This restricted access is foundalso in residential developments of suburbansprawl, yet those do not follow exactly the samepathological pattern because, while also mono-functional, they are low-density. The constrainedtraffic flow is the same, however, as the feeder roadto a residential subdivision in suburban sprawl becomes the only access to a very large area.

These various pathologies described abovemade temporary economic sense in the era of cheap oil, during a time while their externalenvironmental and health costs could be ignored.The new economics increasingly makes thesetypologies unfeasible. We draw the analogy witha physical system that is artificially kept fromnucleation (that is condensing into an orderedstructure that includes a central core) by havingconstant energy input. That is precisely what ittakes to prop up suburban sprawl, which isinherently unsustainable. Indeed, economistshave argued persuasively that the faulty, unsus-tainable sprawl model was a major driver of the global financial crisis that began in 2008(Leinberger, 2008). The economic dimension of this issue, though beyond the scope of this article,nonetheless parallels and reinforces these othertrends. Thus it is critical to recognize the need forreform of sprawl-supporting economic processes just as much as correcting design issues.

Balancing the Movement Economy With theQualitative Factors of a Sanctuary Area

The model offered here is not a standard, but atype. There is a crucial difference between the

two. A standard provides rigid outlines withinwhich all variations must occur. A type, on theother hand, offers a set of relationships that can bevaried greatly according to local conditions. It is amore generic version of what Alexander and

colleagues have called a pattern (Alexander et al

,1977). In our proposed type we offer the condi-tions under which the location of the nucleus willvary in relation to the thoroughfare. Moreover, wedescribe how variations in this spatial substrateoccur so that the evolution of neighborhoods can be influenced to proceed in a more desirabledirection. Some existing arterial roads may beimpossible to change in the early stages, but can be planned to convert to boulevards over time.This is particularly true in the USA and othercountries where auto-dominated planning is onlyslowly giving way to a more efficient multi-mode

system.Our Emergent Neighborhood Model accountsfor the qualitative experience of a neighborhood:its beauty and vitality, and the kind of access itoffers to various kinds of enriching urbanexperience such as parks and civic spaces. Thesewere key motivations for the original CityBeautiful concept that is one of the inspirationsfor New Urbanism.

Under the 400m rule, the small scale of thesanctuary areas allows people living in them to bein a relatively quiet area, in which both pedestrianand car movement are restricted. This area mayinclude neighborhood parks and small (notnecessarily local) social institutions. Reducedmovement is not a function of social limitationor lack of permeability, but of the reduced degreeof centrality. Streets forming the edges of thesanctuary areas form active boundaries, not barriers, with denser functions and more mixeduse, and of varying scale and intensity. Some may be through streets, but with relatively littlecommerce; some may be commercial streets;others may be residential avenues; some may beeven wider urban parks, or larger urban squaresor institutions. The details depend on density, thelocation of the sanctuary area within the city andregion, its relationship to natural features andmajor urban uses.

One example of a planned urban area wherethese relationships hold is the northern part of central Tel-Aviv, known as the Geddes Plan(Figure 11). This area designed by Patrick Geddesfor the Jewish Agency in 1926 guided the growthof Tel-Aviv in the 1930s, as Jews fleeing theascendancy of Nazism and Fascism in Europe




22/26

immigrated to British-controlled Palestine (Rofe `,2008). Geddes developed the idea of the home- block that is similar in intention to our sanctuaryarea. An area of roughly 200 300 m, whichincluded an internal public square and publicinstitutions, is surrounded by through streetsforming a larger scale grid. Each home blockwas adapted to its urban and natural context. This

plan successfully negotiated the explosive growthof Tel-Aviv, and adapted to a density muchhigher than the garden suburb originally envi-saged by Geddes. Even after becoming the centralarea of a thriving metropolis of 2.3 million people,its quiet residential streets within the sanctuaryareas, with their little gardens and paths stillallow for a quiet and at times even reclusive

Figure 11: The Geddes Plan of Tel-Aviv, showing the grid of main streets and home-block sanctuary areas.

Mehaffy et al



23/26

residential experience, only steps away from thehustle and bustle of the major streets (Rofe ` &Schwartz, 2007).

Other examples stem from attempting tointegrate natural beauty into the neighborhood,

using patterns like eccentric nucleus

, as part of alarger methodology for repairing sprawl (Steilet al, 2008). The integration coincides with, but issomewhat distinct from, the integration of natureper se, as wildlife corridors. As it happens, we believe the two can work together at the interface between main streets and sanctuary areas.

For example, a principal boulevard can sur-round a park-like natural area, with commercialand mixed use flanking the sides opposite the

park. The connection across the boulevards to theopen space occurs principally at the major nodes,where everything can be stopped and pedestrianscan be given priority. Other connections couldalso be made with a series of bridges or traffic-

calmed transverse crossings. Several real-worldexamples of similar structures, which are in fact beautiful and livable areas, can be found inPortland, Oregon and along Hyde Park in London(Figure 12).

Note that the neighborhoods illustrated hereseem to have strong quiet centers along local mainstreets. In the case of Portland, there is a schooland its play area, doubling as a public park. Thisis a very successful local neighborhood focus,

Figure 12: A principal boulevard can surround a park-like natural area, with commercial and mixed use flanking the sides oppositethe park. Examples can be seen in Portland, Oregon, in the Northwest neighborhood and in London, UK, in the area surroundingHyde Park.




24/26

though not so far from the edge that it becomes adisturbing segregation.

Once again, the optimum spacing of theseconnective nodes is surprisingly small: only about400 m, which is about half of common twentieth

century practice. This spacing would then defineadjacent cells of the size of small pedestriansheds, which would then also be about 400 macross (1/4 mile diameter), and can easily accom-modate about 1500 dwelling units. All houseswould be no more than 400m or 1/4 mile, from aconnecting node.

These estimates show that the sanctuary areasare getting close to about one-half of ChristopherAlexanders theoretical pattern of 7000 people fora COMMUNITY OF 7000 (Alexander et al, 1977).As neighborhoods flow across sanctuary areas, allwe need is two such cells to satisfy Alexanders

criterion. Note that we need not preoccupyourselves with a rigid definition of either thepedestrian sheds or the neighborhoods. Our400 m spacing is used as a guide to make surethat activity nodes are within optimum walkingdistance. In this way, the different points of viewin this debate are reconciled. Areas of retail notdirectly on main streets can still function, becauseof their proximity to the movement economygenerated by the main streets. This is how themovement economy works on retail and services:the closer you are to people flows, the more likelyis your business to succeed; it is this increasedlikelihood that makes all the difference in thelong run.

Regardless of technology or urban context,study examples consistently remind us that thereis one fundamental biological unit at work in anysanctuary area: the ability of a person to walkconveniently to many nearby destinations. This iswhat mathematicians term a structural attractor(an essential and consistent geometrical solutionto the problem) for the coherence of any sanctuaryarea. A person will need to walk to all destina-tions within or bordering the sanctuary area, notonly public transit, or retail or any other singlefunction. So it is very useful to start with thispedestrian shed and work with the other struc-tures retail, transport and so on as overlaypatterns that may vary based upon conditions.Then the question becomes, how do the otheroverlaid patterns work in relation to such afundamental biological unit? There is not a one-size-fits-all spatial configuration, but there is aspatial measure. This is one of the key points of confusion and controversy on this issue. We argue

here that there is one scale for this fundamentalunit, and it is smaller than usually assumed. This hasimportant implications for the required connectiv-ity of any scheme.

Those who think that a neighborhood needs an

edge, and those who think that, more importantly,it needs close connectivity to the next neighbor-hood, are both likely to find a new synthesis inour proposed framework. Sanctuary areas, whichfollow the 400m rule, between main streets, arelikely to result in a permeable urban spatialcontinuum, over which social and cultural com-munities such as neighborhoods can layer, aggre-gate and change.

Yet another distinction comes from those whohave focused upon transit-oriented development,as opposed to those who are more focused uponcreating complete communities whose public

transit connections to other communities may belimited. The former include urban infill planners,whereas the latter include planners of moreperipheral and rural settlements. The debate overcenter versus edge has persisted, with someauthors making dire warnings about putting toomuch at the center, and others arguing that atleast some things (the corner grocery, a transitstop, perhaps more nodes) can and should go tothe center. We suggest that in either case, the issueis not transit-oriented development per se, but theuse of integrated multiple modes, starting withpedestrian movement.

Conclusion

The Emergent Neighborhood Model (Figures 6, 9and 10 (a)(e)) proposes that urban nuclei should be located in close proximity to urban thorough-fares in order to benefit from the movementeconomy (Porta et al, 2009). Such thoroughfaresshould be conceived as a network of main streetsshaped by a 400-meter rule, so that a spatialdistance of roughly 350450m is left between twoneighboring major intersections. The location of urban nuclei on or immediately adjacent to such anetwork of main streets ensures the best potentialfor retail and services to stabilize and grow, and toserve urban communities in the long run. Theurban nuclei by definition shape the geographyof pedestrian sheds, which are simply theareas surrounding them under the critical 400mdistance.

The urban fabric is, therefore, an interplay between main streets and quieter, mostly residential,

Mehaffy et al



25/26

sanctuary areas. Neighborhoods, however, areinherently self-organizing and evolving socialand cultural entities, whose dynamics of formationand change are simply too complex to be shapedtop-down by spatial design. Pedestrian sheds

and a complex of main streets and sanctuary areascreate the geography from which neighborhoodsmay emerge from the interplay between thesubstrate of space and the social life of itsinhabitants.

A high grade of attractiveness of the urbannucleus can enact dynamics of retail and servicespillover towards the inner areas of the neighbor-hood. That regularly happens on local mainstreets, that is, shortcuts across sanctuary areasthat connect to the higher network of urban mainstreets. Therefore, we can evaluate the potential of places to attract and stabilize a flourishing local

retail and service system. People make places fitfor running a business, and the more a place iscentral the more it is likely to be popular andused. The densest and most diverse variety of retail commerce and services is very likely toemerge and grow in central places. In order to beaccessible to pedestrians, those central places fullof shops and services must be reachable withinwalking distance, say about a 400 m radius.Finally, if in such central places, located atwalking distance from home, one is able to jumponto a transit service, this will heavily enhancethe popularity of the place. At the same time, easyaccess to transit ensures that a condition of equityis restored that frees residents from requiredownership and extensive use of the privateautomobile, to enjoy the best of what their cityoffers.

Note that our discussion covered topics of urban design, social structure, economics, themathematics of generativity and other topics. We believe this kind of inter-disciplinary investiga-tion offers a key avenue for future research.Indeed, the lesson may be that the notion of urban design itself must undergo a transition,from a simple linear application of standards,schema or parti, to a complex process of facilitating beneficial pattern growth. This is aparadigm shift in the management of complexurban growth, towards a discipline that may bethought of as urban seeding.

We recognize, however, that this is a shift thateven traditional urban designers/reformers mayfind hard to make. Yet, we believe the loomingchallenges of resource depletion, habitat destruc-tion, massive informal/illegal urbanization, cli-

mate change, social equity, political instability andother crises make such a transition mandatoryand urgent. It is in urban settlements, and intheir integration and their ultimate quality, thatthe principal struggles of the future will be won

or lost.

Note

1 A notable example is in our colleague Paul Murrains designfor Harlow New Town, an urban extension of one of theUKs struggling modernist towns built north of London afterthe war. The extension is connected over a wildlife corridor by only a single bridge flanked by retail, in a configurationnot unlike the Ponte Vecchio in Florence. Yet, as Murrain hasargued, the connectivity remains strong (Council Report III,CNU).

References

Alexander, C. (1965) A city is not a tree. Architectural Forum122(1): 5862, (Part I); and 122(2): 5862, (Part II).

Alexander, C. (20012005) The Nature of Order, Books 14Berkeley, CA: Center for Environmental Structure, Book 1:The Phenomenon of Life, 2001; Book 2: The Process of CreatingLife, 2002; Book 3: A Vision of a Living World, 2005 and Book 4:The Luminous Ground, 2004..

Alexander, C., Ishikawa, S., Silverstein, M., Jacobson, M.,Fiksdahl-King, I. and Angel, S. (1977) A Pattern Language.New York: Oxford University Press.

Appleyard, D. (1981) Livable Streets. Berkeley, CA: University of California Press.

Badland, H. and Schofield, G. (2005) Transport, urban design,and physical activity: An evidence-based update. Transpor-tation Research Part D: Transport and Environment 10(3) (5):177196.

Banerjee, T. and Baer, W.C. (1984) Beyond the NeighborhoodUnit: Residential Environments and Public Policy. New York:Plenum Press.

Ben-Joseph, E. (2005) The Code of the City: Standards and the Hidden Language of Place Making. Cambridge, MA: MITPress.

Bernick, M.S. and Cervero, R. (1997) Transit Villages in the 21stCentury. New York: McGraw-Hill.

Biddulph, M. (2000) Villages dont make a city. Journal of URBAN DESIGN 5(1): 6582.

Boarnet, M.G., Greenwald, M. and McMillan, T.E. (2008)Walking, urban design, and health: Toward a cost-benefitanalysis framework. Journal of Planning Education andResearch 27(3): 341358.

Brindley, T. (2003) The social dimension of the urban village: Acomparison of models for sustainable urban development.URBAN DESIGN International 8(12): 5365.

Brower, S. (2000) Good Neighborhoods: A Study of In-Town andSuburban Residential Environments. Westport, Connecticut:Greenwood Publishing Group, Incorporated.

Calthorpe, P. (1993) The Next American Metropolis. New York:Princeton Architectural Press.




26/26

Caniggia, G. and Maffei, G.L. (2001) Architectural Compo-sition and Building Typology: Interpreting Basic Building.Firenze, Italy: Alinea Editrice, First published in Italianin 1979.

Castells, M. (2000) The Information Age: Economy, Societyand Culture. The Rise of the Network Society. Malden, MA:Wiley-Blackwell.

Cervero, R. and Duncan, M. (2003) Walking, bicyclingand the urban landscape: Evidence from the SanFrancisco Bay Area. American Journal of Public Health 93(9):14781483.

Conroy-Dalton, R. (2003) The secret is to follow your nose:Route path selection and angularity. Environment andBehavior 35(1): 107131.

de Nazelle, A. and Rodr guez, D.A. (2009) Tradeoffs inincremental changes towards pedestrian-friendly environ-ments: Physical activity and pollution exposure. Transporta-tion Research Part D: Transport and Environment 14(4) (6):255263.

Duany, A. and Plater-Zyberk, E. (2002) Lexicon of the NewUrbanism (version 3.2). Miami: Duany Plater-Zyberk &Company, http://www.dpz.com/research.aspx.

Duany, A., Plater-Zyberk, E. and Speck, J. (2000) SuburbanNation. New York: North Point Press.

Ernst, M. and McCann, B. (2002) Mean streets 2002, surfacetransportation policy project report. http://www.transact.org.

Ewing, R. and Kreutzer, R. (2006) Understanding the relation-ship between public health and the built environment.Report prepared for the LEED-ND Core Committee. U.S.Green Building Council.

Farr, D. (2008) Sustainable Urbanism, URBAN DESIGN withNature. Hoboken, NJ: John Wiley and Sons.

Forsyth, A.J., Oakes, M., Lee, B. and Schmitz, K.H. (2009) The built environment, walking, and physical activity: Is theenvironment more important to some people than others?Transportation Research Part D: Transport and Environment

14(1) (1): 4249.Frank, L.D., Sallis, J.F., Conway, T.L., Chapman, J.E., Saelens,

B.E. and Bachman, W. (2006) Many pathways from landuse to health: Associations between neighborhood walk-ability and active transportation, body mass index, andair quality. Journal of the American Planning Association72(1): 75.

Franklin, B. (2002) Constructing an image: The urban villageconcept in the UK. Planning Theory 1(3): 250272.

Gehl, J. (2002) Life Between Buildings: Using Public Space.Copenhagen, Denmark: Danish Architectural Press.

Hebbert, M. (2005) Engineering, urbanism and the struggle forstreet design. Journal of URBAN DESIGN 10(1): 3959.

Hillier, B. (1996) Cities as movement economy. URBAN DESIGN International 1(1): 4160.

Hillier, B. and Hanson, J. (1984) The Social Logic of Space.Cambridge, UK: Cambridge University Press. Jacobs, A. (1993) Great Streets. Cambridge, MA: MIT Press. Jacobs, A.B., Macdonald, E. and Rofe`, Y. (2002) The Boulevard

Book. Cambridge, MA: MIT Press. Jacobs, J. (1961) The Death and Life of Great American Cities. New

York: Vintage Books. Jones, E. and Kauffman, C. (2009) Integrated urban structuring

in Australia. Paper presented at Optimizing Urban Struc-ture session, CNU 17; June, Denver, CO.

Kargon, R.H. and Molella, A.P. (2008) Invented Edens: Techno-Cities of the Twentieth Century. Cambridge, MA: MIT Press.

Krier, L. (1998) Architecture: Choice or Fate. Windsor: AndreasPapadakis.

Kulash, W. (1990) Will the traffic work? 11th Annual PedestrianConference; October, Bellevue WA.

Leinberger, C. (2008) Study from the metropolitan programof the Brookings Institution. http://blog.islandpress.org/author/Christopher.

Loukaitou-Sideris, A., Liggett, R. and Sung, H. (2007) Death onthe crosswalk: A study of pedestrian-automobile collisionsin Los Angeles. Journal of Planning Education and Research 26:338351.

Lynch, K. (1981) A Theory of Good City Form. Cambridge, MA:MIT Press.

Marshall, S. (2005) Streets and Patterns. London: Spon Press.Murrain, P. (2002) Understand urbanism and get off its back.

URBAN DESIGN International 7: 131142.Murrain, P. (2009) Beyond Perry: Learning from several

thousand years of the social logic of the structure of settlements. Paper presented at Optimizing Urban Structuresession, CNU 17; June, Denver, CO.

Newman, P. and Kenworthy, J. (1999) Sustainability and Cities.Washington DC: Island Press.

Perry, C. (1929) The Neighborhood Unit. Regional New York and itsEnvirons. New York: Regional Plan Association.

Philibert-Petit, E. (2006) Connectivity-oriented urban projects.PhD Thesis available from Delft University of Technology,to be published as a monograph by Techne Press,Amsterdam, 2010.

Porta, S., Crucitti, P. and Latora, V. (2006) The network analysisof urban streets: A primal approach. Environment andPlanning B: Planning and Design 33: 705725.

Porta, S. et al (2009) Street centrality and densities of retail andservices in Bologna, Italy. Environment and Planning B:Planning and Design 56: 450465.

Porta, S. and Renne, J. (2005) Linking urban design tosustainability: Formal indicators of social urban sustain-ability field research in Perth. URBAN DESIGN International

10: 5164.Rofe, Y. (1995) Space and community: The spatial foundations of

urban neighborhoods. Berkeley Planning Journal 10: 107125.Rofe, Y. (2008) The white city of Tel-Aviv. Urbanistica Journal

of the Italian Institute of Planning 136: 95102.Rofe, Y. and Schwartz, H. (2007) Vision, implementation and

evolution of Patrick Geddes urban block in Tel-Aviv.Proceedings of the Conference on Regional Architecturein the Euro-Mediterranean Area: IIIrd InternationalConference on Architecture and Building Technologies.Second University of Naples, Ischia, Italy.

Salingaros, N.A. (2005) Principles of Urban Structure.Amsterdam, The Netherlands: Techne Press.

Steil, L., Salingaros, N. and Mehaffy, M. (2008) Growingsustainable suburbs: An incremental strategy for recon-

structing sprawl. In: T. Haas (ed.) New Urbanism and Beyond:Designing Cities for the Future. New York: Rizzoli Inter-national, pp. 262274.

Sucher, D. (2003) City Comforts. Seattle, WA: City Comforts.Talen, E. (2000) The problem with community in planning.

Journal of Planning Literature 15(2): 171183.Townshend, T. and Lake, A.A. (2009) Obesogenic urban form:

Theory, policy and practice. Health & Place 15(4): 909916.US Environmental Protection Agency (US-EPA). (2004) Char-

acteristics and performance of regional transportationsystems. Washington DC. http://www.epa.gov/smartgrowth.

Mehaffy et al

Documents

Urban Nuclei