Chengzhi Peng. School of Architecture, University of Sheffield, Arts Tower, Western Bank, Sheffield S10 2TN, UK
Based on a previously Web-based multi-tiered virtual city server, this paper reprts on a further study into how online virtual city models can be turner into digital canvas for urban designers to engage in 3D concept design modeling. This research design environment, called Townscaping, provides a set of tools for users to instantiate and edit 3D graphic elements that can be positioned directly onto the virtual city model retrieved from the Web server. Because of the visual immersion and interaction afforded by virtual world technology, early concept designs could be developed as though they are worked out by designers working directly on the site, hence in-situ 3D concept design.
Keywords: 3D concept design, modeling, virtual reality, system design, design education
In many areas of design studies, concept (or, schematic) design often refers to the early stages of design processes in which designers are motivated to generate and play out initial/outline ideas in response to the problems, issues or situations arising from the project briefs. In architecture and urban design, concept design is particularly associated with contextual thinking – the forming of strategies that could position new designs in some dialogue with existing city sites. How to foster design students’ ability to engage in contextual thinking is therefore one of the key objectives in design education. In the past few years, the School of Architecture at the University of Sheffield has undertaken an educational research programme called Sheffield Urban study, in which students worked in teams to unravel the history of the city following a common research method. In retrospect, the programme has delivered a notable pedagogical result: the students have learned a method and various skills of how to do urban contextual study that could inform new designs (Blundell Jones et al., 1999). A secondary result from the urban study is the acquisition and accumulation of urban contextual datasets of the city of Sheffield that have never been systematically assembled before. The historical maps, physical models and other associated information about the historical urban fabric collected through the study have thus formed a library of Sheffield Urban Study accessible to students and researchers for future consultations.
Based on the physical datasets assembled through the Sheffield Urban Study, a separate project was developed to investigate how the physical datasets could be best converted into digital format allowing for wider access. Funded by the UK Arts and Humanities Research Contextual Databank system has been designed and implemented as a dynamic virtual city server. It provides novel Web-based system architecture and user interfaces for depositing as well as retrieving the multidimensional urban datasets including 3D virtual city models. The key results of SUCoD system design and how the system has been applied in an undergraduate urban design computing course were reported earlier in Design Studies (Peng et al., 2002b; Peng and Blundell Jones, 2004).
Following out precious research and system development, we continue to explore new online facilities that could link dynamic urban contextual modeling more directly with the actual processes of urban and architectural design modeling. Although users can now explore how their proposed designs could best fit into the urban contexts through the uploading facility, it is assumed that the actual design of urban proposals such as MicroStation or 3D Studio Max and then converting contextual models. It seems to us a logical step to investigate if SUCoD can be further extended to provide yet another new facility that could allow the accrual design of urban proposals to take place directly over the city’s contextual models, at least at the early stage of concept design. In doing so, the virtual city models can become designers’ 3D online working models for developing early schematic designs when designers are mostly concerned with basic issues of urban designs such as volume, location, orientation, viewpoint, scale and proportion etc.
By experimenting with the Java 3D technology on the SUCoD server platform, we find that the virtual city models can be turned into 3D digital canvas for designers to engage in concept design. Because of the visual immersion and interaction afforded by the virtual city modeling, early stage concept designs could be developed as though the designer is working directly on the site, hence in-situ 3D concept design. The current design and implementation of the research design environment, called Townscaping, provides a set of tools for users to create and edit 3D graphic elements to be positioned directly onto the user-specified virtual city models. One of the key features of Townscaping is to enable concept design while in navigation: designers can perform concept design and gain immediate feedback continuously while navigating the 3D virtual city context to any viewpoint at any moment.
The remainder of the paper will first give a brief introduction to the ongoing development of a dynamic virtual city server system – Sheffield Urban Contextual Databank (SUCoD). A detailed account of the current design and implementation of the Townscaping environment follows, explaining the key system features aimed at supporting in-situ 3D concept design. Two working examples developed by the author are then presented to demonstrate how in-situ 3D concept design modeling can be performed on Townscaping. In comparison with other related system designs, the findings from the current study are discussed in the final section together with a list of pointers to further investigating.
1 SUCoD: ongoing development of an experimental virtual city server
The Sheffield Urban Contextual Databank (SUCoD) system was developed through previous research that looked into how the usability and reusability of urban models and associated datasets could be significantly improved by adopting a web-based multi-tiered framework (Peng et al., 2002a). Central to the design of SUCoD is a set of facilities for allowing users to freely navigate and retrieve 3D virtual city models, images of historical maps, and web links to other related historical resources according to spatial-temporal attributes specified by the users. In so doing, the restrictions on end-user access and retrieval, as seen per-determined by conventional model developers when building the urban models and other datasets in the first place, can be removed completely.
As demonstrated through the Sheffield Urban Study, 3D reconstruction of the city fabric remains a powerful means of telling a city’s story. In building SUCOD, we were particularly interested in how users could access the interactive 3D urban according to their own spatial. temporal and thematic interests. In search for the technical solutions, we were finally able to formulate the Multi-tiered Extensible platform for Dynamic Interactive Urban Modelling (MEDIUM) framework that could fulfill simultaneously several objectives set out in the requirement study (Peng et al.,2002c). In our view, MEDIUM is a generic framework that can be applied to building urban contextual datasets and SUCoD’s functionality for use interaction is open to continuous extension and adaptation over time.
Figure I shows a screen shot of the latest implementation of SUCoD. Based on the Java platform, an applet was built to display an interactive city map of historical Sheffield containing multiple layers of selectable graphic objects depicting the city’s terrain, streets and buildings of different ages and usages. Secondly, interactive functions were built into the Java applet for users to retrieve the urban contextual resources stored on the server including historical maps and 3D virtual city models created in the VRML format. Two other user interaction functions were also attempted: Upload VRML was designed to allow users to submit their own VRML worlds that can be further combined with the city models they retrieve from SUCoD; while List VRML was designed for the user to generate a table in real-time that shows the history of city model retrieval and / or user model upload. Listing can be invoked at anytime during a single live session, and the history of user interaction is formatted as a simple Web page containing checkboxes for all VRML models registered through either retrieving (Get VRML) or uploading. With records generated by the listing function, users can go on selecting any of the VRML models from the list and creating their own personal virtual worlds, which could well be syntheses of urban contexts and user proposed 3D designs.
As an experimental virtual city server, SUCoD has been developed to explore how 3D city modeling could migrate from developer-centre to user-driven. We learned that the usability of 3D city models and other urban datasets could not be significantly improved if users were not given interaction routes at a level leading to the generation of personal virtual worlds to their own purposes. We have since considered SUCoD an open extensible platform for experimenting further interactive functions beyond those of retrieving, uploading and listing. In the next section, I describe our development of the Townscaping 3D concept design environment on the basis of SUCoD.
2 Townscaping: an in-situ 3D concept design environment based on SUCoD
Townscaping was developed in some way along Gordon Cullen’s principle of ‘serial vision’ or ‘sequential study’ as demonstrated in his seminal work on Townscape (Cullen, 1996; Gosling and Maitland, 1984). Speculatively, we consider that Cullen’s principle could be translated to something like sketching / designing while navigating in the realm of 3D digital modeling and design. This suggests that architects and urban designers could be provided with an intuitive 3D contextual concept design environment in which developing new designs could always go hand in hand with designers’ cogitating the city or town context. Our pursuing this hypothesis on the ground of SUCoD means to explore the benefits of designing while in navigation if designers are enabled to freely navigate to any position in a virtual city model and, at the same time. To perform 3D concept design activities without lesving the virtual contextual model.
We therefore set out to develop and test the Townscaping environment on the basis of SUCoD. Conceptually, Townscaping is composed of the following component functions.
Select, access and define. As seen in SUCoD, interactive maps of a city are provided for users to select an area of interest and access the virtual city models of that area. Users define 3D city contexts in terms of spatial boundary and thematic content that they consider relevant to their tasks at hand.
Navigation. Once a 3D urban context is defined and generated as a virtual world, the user can freely navigate the world through various modes such as Walk, Fly, Pan, Look etc. as commonly seen in VRML or X3D based viewers such as ParallelGraphics’ Cortona or Bitmanagement Software’s Contact VRML / X3D.
Viewpoints. Similar to the function of book-marking Web pages when people find their favorite websites, viewpoints can be set up by users freely during navigation. These may be the viewing points where the users consider the scenes captured are of special interests or importance that can be quickly revisited if needed.
Generation. Instances form a set of basic 3D graphic elements can be specified and generated into the virtual city model. Depending on the computer graphic language chosen in the implementation, the types of graphic elements and the scopes of their attributes may vary from one language to another.
3D editing. Graphic objects generated in the virtual reality model can be manipulated individually (or, as a group defined by the users) to enable changes in spatial positions, dimensions, appearances, or even behaviours.
World save. At anyone time, users may decide to leave Townscaping and save whatever has been produced during the session including the urban contexts specified. Users can reload their saved worlds later to resume design editing.
Unlike conventional CAD systems, the Townscaping environment was envisaged to support early conceptual design by (a) aiding visual understanding of the urban design context, (b) making a set of 3D graphic editing tools applicable directly onto the design context as represented by a 3D virtual city model, (c) switching between navigation and concept design more easily without going through complicated mode change on the user interface, and (d) enabling immediate feedback on concept design via virtual world based navigation and interaction.
3 In-situ 3D concept design with Townscaping: some working examples
The current Townscaping prototype is implemented as a Java TM Web Start application that can be accessed online from the SUCoD Web server . Once set up properly, Townscaping can be launched directly from a user’s desktop shortcut. Java, java 3D and Perl are used to implement the key functionalities for accessing urban contextual models and designing while in navigation (Crawford, 2004). In particular, the VRML97Loader class provided by Java 3D is used to load the current Sheffield urban contextual models, and this opens up the use of a great deal of Java 3D capabilities in Townscaping such as allowing users’ adding new viewpoints into the contextual models. When accessing Townscaping, the city’s index map is first displayed for the user to specify a general area of interest by selecting one or more 200-m square tiles, and then the Get IVL Map for Townscaping function can be activated for building the final 3D urban context (Figure 2). The green-shaded area on the index map indicates the current scope of the city’s VRML models currently available from the SUCoD server, which will be extended as more contextual datasets are built into the system.
As soon as the Get IVL Map for Townscaping function is activated, the user is shown another interactive city map in a separate window (Figure 3). Created in the IVL format, this is a much more detailed map containing 20 layers that depict the 2D layouts of urban features gathered through an urban study survey. With this interactive map, the user can further define a more exact area of interest by fine-tuning the content selection down to individual buildings and street blocks if considered appropriate.
When the user is satisfied with their content selection on the IVL map and activates the SUCoD 3D Townscaping function, a virtual city model is displayed in a new window with the Townscaping toolset provided on the left (Figure 4). The 3D city model is initially shown at the viewpoint (‘view 0’ by default) that is at the top of the Views list, which contains all the viewpoints preset by the original city model developers. On the Navigation panel, the user can move around the model via MOVE, PAN or LOOK with Align to reset the sky-ground of the city model into a horizontal position. Direction of navigation or where to look is controlled by the user’s moving and pressing the mouse cursor into a position relative to the centre of the field of vision. As mentioned earlier, the user can add new viewpoints at anytime by navigating to the intended locations, adjusting the views and then entering names of the new viewpoints through the Views panel (Figure 5). The marking up of new viewpoints can be a valuable exercise in setting up how an emerging concept design may be developed and reviewed in sequence.
On the Designer panel, the user is provided with a set of graphic elements including Cube, Cone, Cylinder and Sphere (Figure 6). This should be seen merely as the initial toolset, which will be extended to afford other types of 3D elements / tools over time Selecting the Cube tool, for instance, a Cube-Properties pad is shown prompting the designer to enter the name, dimensions and colours of a cube element. As the Apply button is pressed, an instance of Cube is generated into the virtual world at a preset location and orientation. The cube object can then be manipulated through the set of spatial editing operations provided on the Buildings until a desirable status is reached.
As a worked example to illustrate Townscaping in action (Figure 4-6), a hypothetical street performance stage is conceived in relation to a city monument site, the Jubilee Monument in Barkers Pool back in 1900. Instances of Cube and Cylinder of various properties were generated and manipulated into the intended positions and orientations. It is worth mentioning that defining new viewpoints can be useful during concept design process because it allows the designer to navigate around those bookmarked views quickly while developing the design composition.
Figure 7 and 8 show another example of in-situ 3D concept design where an attempt was made to carry out a study of block structure for a chosen city site. The volumes, heights, shapes of a hypothetical office-residential complex were played out in relation to the existing historical buildings and street blocks. During the concept design, the setting up of a number of new viewpoints around the site before inserting any new shapes was found particularly useful; designers could form strategies of concept design through the process of view-adding to acquire some kind of 3D contextual modeling frameworks. It will be interesting to observe how different designer may set up their viewpoints for the same city site when creating their conceptual proposals. These user-specified views could reveal something informative about how the contextual characteristics of the city site were appreciated by the designers, and how best their resultant concept designs should be looked at from those specifically chosen locations and orientations.
作者：Chenzhi Peng. 来源：0142-694X S – see front matter Design Studies 27 (2006) 439 – 455
在设计研究的许多领域，概念（或图解）设计，通常指的是设计过程的早期。这时，设计者被激发产生以及展示最初的构想，作为对工程概念中所产生的问题，争端以及情况的回应。在建筑学和城市设计中，概念设计跟前后关系思维有特别地联系——策略的形成，在和现存城市景点对话的基础上，安置新的设计。因此，怎么去培养设计专业学生致力于前后关系思维能力，是设计教育中主要的目标之一。在过去的一些年，在谢菲尔德大学的一些建筑学院已经进行了一次教育研究项目，被称为谢菲尔德城市研究。在这次研究中，学生根据普通的研究方法，运用团队合作来阐明城市的历史。回想起来，这个项目发表了一个著名的教育成果：学生们已经学会了怎么样处理城市前后关系研究的一种方法和几种技巧，这将会新生出新的设计（Blundell Jones et al.,1999）。从这次城市研究所取得的第二层次的结果是，关于谢菲尔德城市前后关系数据的采集和积累，而这之前从来没有被系统地采集过。历史地图，物质模型和其他关于城市建筑方法的有关信息，通过这次研究，被收集起来，因此形成了谢菲尔德城市研究的一个图书馆。这个图书馆为学生和研究人员进行进一步的咨询提供了条件。
在谢菲尔德城市研究中所汇集起来的物理数据的基础上，另外一个独立的工程被开发起来去研究。这些物理数据怎么样才能最好地被转化成数字格式，允许进行更广泛的（数据、信息的）选取。由于英国艺术和人文学研究委员会拥有超过两年的资金资助（2000~2002），谢菲尔德城市前后关系数据库（SUCoD,http://sucod.shef.ac.uk）已经被研究成动态的城市虚拟服务，并使之生效执行。它提供最新的以网页为基础的系统建筑和使用者界面，用于存取和检索多维城市数据，包括3D虚拟城市模型。SUCoD系统设计的重要结论以及系统怎么样被应用于一位大学生的城市设计计算机课程中，早已经在《设计研究》中进行过报道（Peng et al.,2002b;Peng and Blundell Jones,2004）。
依照我们先前的研究和系统发展，我们继续探索新的在线设施，这些设施可以直接把动态的城市前后关系模化成型和真实的城市和建筑摸化成型过程聚集起来。虽然使用者现在可以探索怎么样通过下载设施，使得所计划的设计可以最适合城市前后关系。我们猜想，城市计划的真实设计将会在SUCoD之外发生，使用常用的CAD密封装置，例如MicroStation或3D Studio Max，然后把这些设计转换成VRML格式，使得和城市的3D前后关系模型所兼容，研究SUCoD是否可以更进一步扩大到提供另外一种新的设施，它可以允许城市设计的真实设计，可以直接在城市前后关系模型之上发生，至少在概念设计的早期阶段发生，对我们来说，这看起来是逻辑上的一个步骤。通过的这些，虚拟城市模型可以成设计者的3D在线工作模型，用于发展早期图解设计。早期的图解设计者更关心城市设计的基本问题，比如大小、位置、朝向、视角、比例尺、比例等。
通过在SUCoD服务器平台上用Java 3D技术进行实验，我们发现虚拟城市模型可以被转换成3D数字屏幕，用于设计者参与进行概念设计。由于虚拟城市摸化成型的视觉参与相互作用，概念设计的早期阶段可以发展演化到设计者仿佛直接在现场工作一样，因此被称为in-situ 3D概念设计。现在的研究设计环境的设计和履行，被称为城镇风景化，它为使用者提供了一整套工具，为了创造和编辑3D图表关系，这些元素被直接安置于使用者所指定的虚拟城市模型中。城镇风景化更重要的一个特点就是在导航的时候进行概念设计：设计者可以在任何时候以任何视之航行于3D虚拟城市，当这个时候，设计者可以实施概念设计并且取得即刻的回馈。
谢菲尔德前后关系数据库（SUCoD）系统通过先前的研究而发展起来，先前的研究是来调查城市模型的使用和循环使用以及相关数据怎么样通过采用以网络为基础的多等级结构恒有意义地被发展起来的（Peng et al.,2002a）。SUCoD的主要设计是一系列设施，允许使用者自由地航行以及恢复3D虚拟城市模型，图片和历史地图。根据由使用者所指定的暂时的空间属性，网络链接到其他相关的城市资源。通过做这些，终端用户的接入和检索限制，就可以完全地清除。这些终端用户的接入和检索的限制就是我们所见的传统的模型开发者在修建城市模型和其他数据第一场地时所预见的。
城镇风景化，在某种程度上是沿着Gordon Cullen的“连续视觉”和“连续研究”原则，这些原则在他的著作《城镇风光》（Cullen,1996;Gosling and Maitland,1984）中所阐述过。推测而来，我们认为Cullen的原则可以翻译成那些类似于航行与3D模化成型和3D的设计领域中画示意图和设计之类的东西，这就建议我们提供建筑师和城市设计者一个直观的3D前后关系概念设计环境，这里所开发出来的新设计总是和设计师所谋划的城市和城镇内容同步。我们所追求的这个假设是以SUCoD为基础的，意味着在探索设计过程的益处，当航行的时候，如果设计者被促使自由地航行于虚拟城市模型和任何方向，同时，在没有虚拟城市前后关系模型的情况下，可以执行3D概念设计活动。
航行。一旦3D城市环境背景被下定义以及作为一个虚拟世界产生，使用者可以自由地通过各种方式航行于这个世界，比如，行走、飞行、拍摄全景、观看等。就如我们所常见的从观看者为出发点的VRML或X3D，例如ParallelGraphics’Cortona或Bitmanagement.Similar’s Contact VRML/X3D。
当前的城镇风景化原型，是作为一个Java网页开始应用而生效的，这些应用可以在SUCoD网页服务器上在线寸取（http://sucod.shef.ac.uk/townscaping）。一旦合理地建立，城镇风景化可以直接从一个使用者的桌面上简化操作，接着被装入运行。Java，Java 3D和被用于执行关键的功能，用于航行时接入城市关系模型和设计（Crawford,2004）。尤其是，由Java所提供的VRML97程序被用于加载当前的谢菲尔德城市前后关系模型，这些开发了Java 3D在城镇风景化中的大量使用功能，例如允许使用者在前后关模型中加入新的观点。当接入城镇风景化时，城市的索引地图首先显示给使用者，通过选择一个或者大于200平方米平铺显示，来指定感兴趣的大致范围，然后那个城镇风景化功能的Get IVL地图可以被激活去建立最终的3D城市背景（图2）。那个在索引地图上的绿色阴影区域暗示了当前城市的VRML模型的范围，通常适用于SUCoD服务器，这些作为建立到系统中的数据，将会得到扩展。
当使用者对他们在IVL地图上的内容选择感到满意以及激活SUCoD 3D城市风景化功能时，一个虚拟的城市模型在一个新的窗口中展示出来，城市风景化的工具箱将会被提供在左边（图4）。3D城市模型最初在视点那出现（view 0默认），那是在视图清单的最上面，它包括了所有被原始城市模型开发者预先装置的视点。
Confronting the Challenges in Reconnecting Urban Planning and Public Health
Although public health and urban planning emerged with the common goal of preventing urban outbreaks of infectious disease, there is little overlap between the fields today. The separation of the fields has contributed to uncoordinated efforts to address the health of urban populations and a general failure to recognize the links between,for example, the built environment and health disparities facing low-income populations and people of color.
I review the historic connections and lack thereof between urban planning and public health, highlight some challenges facing efforts to recouple the fields, and suggest that insights from ecosocial theory and environmental justice offer a preliminary framework for reconnecting the fields around a social justice agenda. (Am J Public Health. 2004;94:541–546)
DESPITE THE COMMON
historical origins and interests of urban planning and public health, only minor overlaps between the 2 fields exist today. One result of this “disconnect” is an uncoordinated approach to eliminating the glaring health inequalities facing the urban poor and people of color. While public health is increasingly concentrating on biomedical factors that might contribute to different morbidity and mortality rates between the well off and least well off, the field is just beginning to seriously investigate the role of land use decisions and how the built environment influences population health. At the same time, urban planning practice shows few signs of returning to one of its original missions of addressing the health of the least well off. The result is that work in the 2 fields is largely disconnected, and both areas are failing to meaningfully account for the economic, social, and political factors that contribute to public health disparities. However, the public health significance of the disconnect between planning and public health has not gone unnoticed.
A series of recent reports have emphasized the importance of reconnecting planning and public health. For example, a 2001 Institute of Medicine report titled Rebuilding the Unity of Health and the Environment emphasized that the “environment” should be understood as the interplay between ecological (biological), physical (natural and built), social, political, aesthetic, and economic environments. The National Center for Environmental Health of the Centers for Disease Control and Prevention, in its 2000 report Creating a Healthy Environment: The Impact of the Built Environment on Public Health, argued for the reintegration of land use planning and public health, explicitly linking transportation and land use planning to public health outcomes such as increased obesity, asthma, and mental health.A 1999 report published by the World Health Organization, Healthy Cities and the City Planning Process, emphasized the importance of developing a model of “healthy urban planning” to ensure the health of the world’s increasing urban and poor populations. Finally, Healthy People 2010 lists eliminating health disparities as 1 of its 2 top priorities and acknowledges that only an interdisciplinary approach to health promotion will accomplish this goal.
While these reports are important steps toward reuniting planning and public health, what is missing is an articulation of the challenges each field must confront in any reconnection effort and a theory or framework articulating why and for whom the fields should be reconnected.10 This article highlights some of these challenges and offers a framework by drawing on insights from ecosocial epidemiology and environmental justice. I suggest that ecosocial epidemiology and environmental justice are useful paradigms because the former provides an explicit framework that attempts to explain health disparities across populations and how social relations can be pathogenic, biologically “embodied,” and expressed as health inequalities, while the latter outlines a democratic research and public decisionmaking agenda that is attentive to the distributive, procedural, and corrective justice concerns of people of color.
THE DISCONNECT BETWEEN PLANNING AND PUBLIC HEALTH
Public health, city planning, and civil engineering in the United States evolved together as a consequence of late-19thcentury efforts to reduce the harmful effects of rapid industrialization and urbanization, particularly infectious diseases.8,18,19
Reformers recognized that poor housing conditions, inadequate sanitation and ventilation, and dangerous working conditions helped cause devastating outbreaks of cholera and typhoid.18 Planning and public health were regularly affiliated during this era of miasma and contagion, and engineering-based sanitary reforms, largely influenced by the Chadwick report in Britain, were instituted to limit hazardous exposures through such measures as sewerage, garbage collection, and rodent control.20–22 Planners also used the power of the state to separate out populations suspected of causing disease. Yet,both miasma and contagion failed to explain certain aspects of population health, such as why epidemics occurred only sporadically, even with the seeming ubiquitous filth present in many urban areas, and how diseases traveled.
By the end of the 19th century, the driving ideology in public health had shifted to germ theory, and this shift continued through the first half of the 20th century. According to germ theory, there are specific agents of infectious disease, in particular microbes, and these agents relate in a one-to-one manner to specific diseases.20 This conceptual shift was accompanied by shifts in public health and planning practice. Public health research shifted from investigating ways to improve urban infrastructure to laboratory investigations of microbes and interventions focused on specific immunization plans,with physicians, not planners, emerging as the new class of public health professionals.12,19
In urban planning, the Germaninspired “Haussman model” of zoning began to take hold in the United States during this same period. 23 This model focused on functionality and a hierarchical ordering of land use that tended to separate residential areas from other land uses, particularly those involving industry.24 At the core of the Haussman model was the idea of dividing up functions within the economy (e.g., zoning), isolating those functions deemed unhealthy (e.g., industry), and placing strict regulations on the kind of contact occurring between people and land use functions. 24 Zoning was aimed at “immunizing” urban populations from the undesirable externalities of the economy, such as industrial pollution.
As clinicians increasingly implemented public health measures in the mid-to-latter half of the 20th century, the field shifted toward addressing the “hosts” (e.g., individuals) of disease, because the “environment” (e.g., the world outside of microorganisms) was harder for physicians to influence. 20 During this era, public health largely ignored the social dimensions of disease and emphasized modifying individual “risk factors” reflected in one’s lifestyle, such as diet, exercise, and smoking.25 Planning, searching for an identity in postwar America, turned to promoting economic development through large infrastructure and transportation projects.26 Planning shifted from attempting to restrain harmful “spillovers” from private market activities in urban areas to promoting suburban economic development.27 Models of economic efficiency were used in planning new towns, regional planning authorities were established to provide inexpensive and reliable resources to these areas, and an era of urban divestment and residential segregation took hold.26,27
By the latter half of the 20th century, the biomedical model of disease, which attributes morbidity and mortality to molecularlevel pathogens brought about by individual lifestyles, behaviors, hereditary biology, or genetics,25 was firmly entrenched as the dominant paradigm in epidemiology. Yet, the biomedical model was oriented toward explaining molecular-level pathogenesis rather than explaining the distribution of disease among populations or disease incidence or distribution at a societal level.20,25
Urban planning underwent an analogous shift in its orientation toward environmental health by adopting the environmental impact assessment (EIA) process. The EIA process, institutionalized after passage of the National Environmental Policy Act of 1969, ushered in the use of the environmental impact statement (EIS) for analyzing the ecological and human health effects of plans, projects, programs, and policies.28 The EIA process is generally accompanied by a quantitative risk assessment in which human health effects are considered.29 Risk assessment was institutionalized as part of the EIA process in almost all site-specific analyses of human health after the 1980 Supreme Court decision supporting the use of risk assessment in the regulation of benzene.30
Yet, both the EIS and quantitative risk assessment have been widely criticized as methods for assessing population health because they tend to overemphasize carcinogenesis at the expense of other chronic diseases,31,32 treat all populations as similarly susceptible while ignoring the disproportionate hazardous exposures experienced by certain populations,33 restrict analyses to quantitative data while minimizing or ignoring other kinds of information, 34 and limit the discourse and practice to experts, which can undermine the democratic character of the process by determining who is empowered to frame analyses and who will be excluded, deemed inarticulate, irrelevant, or incompetent.29,34,35 Thus, wholesale adoption of practices such as EIS and risk assessment leads to planning becoming disconnected from environmental health.