Draft Paper for Project Varenius Specialist Meeting on Empowerment, Marginalization, and Public Participation GIS

Draft as of September 30 1998

J.B. Krygier
State University of New York at Buffalo
Department of Geography
105 Wilkeson Quad
Buffalo NY 14261

 jkrygier@geog.buffalo.edu

 http://www.geog.buffalo.edu/~jkrygie r/

 Home/Office Mailing Address / Contact Academic Year 1998-99:

 231 Crestview Rd
Columbus OH 43202
614-261-0336

 The Ohio State University
Department of Geography
1036 Derby Hall
154 North Oval Mall
Columbus OH 43210-1361

email: krygier.3@osu.edu

For additional information see the WWW site for the research described in this proposal which includes additional information, the text of a Master's Project on the topic, and a prototype PPGIS / PPVis site:

http://www.geog.buffalo.edu/~jkrygier/krygier_html/lws/ppviz.html

Introduction

 Many of the vital issues in geographic information systems today are not primarily technical issues. As GIS develops and plays an expanding role in the way we manage, analyze, and understand spatial phenomena, the societal consequences of GIS come to the forefront of research in GIS. Research on "GIS and Society" is important not only as a means of understanding the impact of existing technologies on society, but in also imagining and engineering new technologies - for scientists, researchers, and the diverse "public" who increasingly have access to GIS and its analytical capabilities. Research on GIS and Society, then, can play an important role in the future development, implementation, and evaluation of GIS and related geographic methods.

 My work on Public Participation GIS (PPGIS) and what I am calling Public Participation Visualization (PPVis), as reported in this paper, consists of two interrelated parts. First, I discuss a prototype PPGIS / PPVis WWW site developed with ESRI's Map Objects and Internet Map Server. This WWW site was developed for a low-income, inner-city neighborhood in Buffalo, NY. My second, and more general goal is to begin to delineate a praxis - a theorized practice - of PPVis and PPGIS. This involves an explicit awareness of concepts and theories of information, its representation, of people, social relations, and power, and how these shape and are shaped by socially-infused technologies such as PPVis and PPGIS. The results of this research are preliminary, but point the way to a series of future research projects.

 1. Public Participation Visualization, GIS, and the WWW

 "Public Participation GIS" (PPGIS) have been conceived broadly as an integrative and inclusive process-based set of methods and technologies amenable to public participation, multiple viewpoints, and diverse forms of information. I define Public Participation Visualization as an important component of PPGIS. Geographic visualization (GVis) has been conceptualized as predominantly private types of map use involving high human-map interaction wedded to exploratory analyses (MacEachren 1994). Such advanced 'visual analysis' is increasingly linked to the analytical component of Geographic Information Systems. In other words, maps and other visual representations on not merely the 'output' or end result of a GIS analysis, but are integral to the analysis itself - the boundaries between GIS, visualization, and mapping are becoming blurred. GVis Research has focused on particular users and applications: highly skilled scientists engaged in scientific research using advanced computing technologies. However, rapid advances in technology, and the World Wide Web (WWW) in particular, are allowing a much broader array of non-scientific users to engage in visualization-type map use. Developments in WWW-based programming languages are making advanced, highly interactive GVis and GIS applications available to anyone with a modem and internet browser. Users can not only access existing geographic information, but can interactively explore 'what if' scenarios and amend and add information to publicly accessible WWW sites. Users can 'make' and 'un-make' information and thus shape and reshape the way they understand their neighborhood, region, county, and the world. This is an active process of "sense-making" by diverse people, using geographic information from a variety of sources, represented in many different ways (maps, images, text, sounds).

 An initial step in my research on PPGIS and PPVis was to evaluate available software for providing GIS, mapping, and visualization functions via the WWW.

 2. PPVis and PPGIS in Application: Buffalo NY Case Study

 The practical issues surrounding the design and implementation of a PPVis and PPGIS WWW site occupied my attention in during parts of 1997 and early 1998. A small grant allowed me to fund a graduate student (Keng-Pin Chang) to develop a prototype PPVis / PPGIS site. The prototype site was developed for and in cooperation with an inner-city Buffalo community development center. The design and implementation of the site has been documented in a Master's Project by K. Chang (Chang 1997) and in the WWW site associated with this project. Documentation from the project has served as the basis of an initial evaluation of the software and technology: what skills are needed to create such applications? What map and GIS functions are available? What hardware is necessary? How much time is involved? Is the approach taken viable and worth pursuing beyond the prototype stage?

 2a. Buffalo's Lower West Side Community

 I chose an inner-city neighborhood on Buffalo's Lower West Side as the geographic basis of the prototype WWW application. Work began in the Summer of 1997 in cooperation with Buffalo's Lower West Side Development Corporation (LWSDC) and its Director, Mark Kubeniec. Earlier contact with Kubeniec and other LWSDC employees (who lived in the neighborhood) suggested the Lower West Side community would be a good location for the prototype and, possibly, a more sophisticated future project. The Lower West Side Community is diverse; dominated by Hispanics and recent Latin American immigrants, it is also home to a significant number of Asian Americans, African Americans, and Whites. While nearly 50% of the residents have incomes below the poverty level, the eastern edges of the community overlaps the 'fashionable' Allentown area - a historic neighborhood dominated by middle and upper income whites and their refurbished, Victorian-era homes.

 2b. Choosing an Appropriate Technology

 There are many 'technologies' available for PPGIS and PPVis. On one hand, paper maps and colored pencils are a cheap and relatively effective technology already used in many communities. Digital alternatives are also diverse and have their own benefits and problems. The primary alternative to WWW-based mapping and GIS is traditional microcomputer based software and hardware installations. Providing mapping and GIS functions on microcomputers in urban community centers has been investigated, but such physically located resources may be difficult (if not impossible) for certain individuals to access (Ghose 1994). Delivery of such mapping and GIS via the WWW is vital as it could maximize public access to mapping and GIS, and may be the most cost effective means of providing people (and particularly those in marginalized communities and regions) with analytical tools which would not otherwise be affordable. Further, familiarity with the user interface provided by the WWW may enhance useablity. Users could focus on learning about the substantive use of mapping and GIS to solve problems, rather than struggling with an unfamiliar computer interface. Finally, the WWW provides easy access to excessive additional on-line information which may supplement and enhance particular applications of WWW-based mapping and GIS.

 Several methods exist for providing mapping and GIS capabilities on the WWW. It is possible to make spatial data and analytical software available over the WWW via a spatial data library. The user, then, must perform their own analysis on their own computers after acquiring the data and software. Another method is for a human to undertake a GIS analysis and generate maps independent of the WWW - possibly in response to a query from an interested user - and post the results on the WWW. This process can be automated with the use of a map generator. Users set the parameters of a map or GIS analysis on a WWW-based form. This form is passed to a map or GIS server, which generates a map or series of maps, then posts the results on the WWW page. The U.S. Census Bureau's Tiger Mapping Service is a good example of this type of technology. Real-time map browsers, such as ESRI's Map Objects and Internet Map Server provide similar functionality in a package explicitly aimed at component- and WWW-based GIS developers. Early in the research it was decided to use real-time map browser technology for the Buffalo project, as this approach provided more sophisticated, real-time GIS and mapping capabilities than spatial data libraries or the pre-generated map/analysis approach. Real-time map browser technology also provided the best support for the CGI scripts and basic GIS components the project required. Two grants provided the funds necessary to purchase necessary computer equipment and pay for Keng-Pin Chang to program the prototype. The prototype was developed Summer and Fall of 1997, with some updates made during the Spring and Summer of 1998.

 2c. Developing the Prototype WWW Site

 At the same time that Chang and I were reviewing our technology choices, discussions with Kubeniec and others from the Lower West Side Development Corporation resulted in preliminary foci for the prototype PPVis/PPGIS site. First, the site should focus on housing issues. One of the primary goals of the LWSDC was to confront the numerous problems caused by absentee landlords in the community, and to subsidize home sales to community members. To this end, the site should have at least two map scales with associated databases: a neighborhood scale (with streets, lots, and building outlines) and a more generalized city scale. Users of the site could view information about housing in their neighborhood, then compare themselves to the city as a whole. The city scale map existed in a compatible (Arc View) format, but the neighborhood map had to be digitized from paper maps. It was determined that the WWW site must also provide a series of basic mapping and GIS functions: zoom in, out, and pan, search and find (eg., a particular lot or building owner, or a particular kind of housing violation), identify (eg., get the name of an owner by clicking on a particular lot or building of interest), hyperlink (eg., click on a city owned property links to the Buffalo City WWW pages relevant to that property), a comment function (eg., a user can post a comment on a particular property - that it seems to have been abandoned, or drug activity was observed), and finally, a more restricted function that lets a 'master user' (eg., Kubeniec) access and change or add to the database via the WWW (eg., change ownership status or data on housing violations associated with a property). These latter functions were incorporated to empower members of the community, allowing them to assess and amend information about their neighborhood. Again, the purpose of the prototype was not to offer an exhaustive set of functions; instead I wanted evidence that the software could be programmed to allow users to access data and to add their own information and local knowledge to the database via a map-based interface.

 Second, it was decided that a prototype site should be created prior to extensive discussions with the community at large. The prototype would allow us to assess the software capabilities and technology costs - in other words, is the technology appropriate in the given context? Further, few community members were familiar with GIS or computer mapping; the prototype, once finished, would give community members a real sense of what the technology could do, and would hopefully spur community involvement in developing a more sophisticated system. Kubeniec and others had many additional ideas which were not incorporated in the initial site. For example, they wanted the site to provide a balanced view of the housing issues, to focus on the positive (potential home ownership, quality of life, community activities) as well as the negative (problem properties, drug issues, crime, etc.). Such ideas, along with more sophisticated GIS and visualization functions, would be incorporated into the site in the future.

 2d. Preliminary Evaluation of Prototype PPGIS / PPVis Site

 The project to create the basic prototype PPGIS / PPVis site was, in general, successful. Details are provided in Chang 1997, but in general, given particular software and hardware and a moderate amount of programming time, the chosen development platform (ESRI's Map Objects and Internet Map Server) were able to provide us with the basic set of WWW-based mappng and GIS functions we desired. The project used a computer with an Intel 486DX 33 MHz chip, and required Microsoft Windows NT and at least 8 mb of memory and 10 mb of disk space (the machine actually had 64 mb memory and 3 gb disk space). Software and programming skills required included those necessary to use Map Objects, the Map Object Internet Server, Arc View (primarily for digitizing the neighborhood map), Visual Basic, CGI, and HTML. While Chang was familiar with Visual Basic and a limited amount of CGI and HTML, the rest of the software and skills for the project were learned in the process of creating the prototype. Total time spent on the project - including digitizing and all programming - was 240 hours. This would have been less for a programmer with more familiarity with Map Objects and the Internet Map Server.

 Additional (and not unsubstantial) costs to consider include Krygier's time in guiding the project, Kubeniec's time in helping Krygier and providing maps, data, and guidance from the Lower West Side Development Corporation. In addition, upgrades of Map Objects, the Internet Server, and Visual Basic in the Spring of 1998 required an additional 20 hours of modifications to the prototype. Any PPGIS / PPVis site requires a programmer familiar with Map Objects to maintain and update the system. Indeed, the prototype site occasionally crashes, and it requires someone to check that it is working properly on a daily basis. While volunteers can help with data gathering, planning, and some computer components of a WWW-based GIS project, funds would be required to maintain a programmer on a part time basis. The map server itself is situated in the SUNY - Buffalo computer network and thus there were no internet access costs for this project. If a private provider of internet access was needed these costs would have to be figured into the project. Computer hardware and software were paid for by grant monies. The largest investment is ESRI's Internet Map Server. Fortunately, the Geography Department at the University of Buffalo has an educational licence for the map server software, otherwise the costs are substantial. Finally, access and use of the site requires at least some computers in the community with internet access. Some access exists: in libraries, schools, and in some homes. Other access would have to be developed, possibly in community centers. In any case, the costs of PPGIS and PPVIs delivered via the WWW are not insubstantial. While the benefits seem potentially significant, it will always be difficult to find sufficient resources in communities with limited sources of money and the skills required to undertake and maintain such a project. Universities and academics are certainly an important means of providing such resources and skills to marginalized groups and places.

 A serious issue with any kind of PPGIS / PPVis project has to do with the complexities of communities and the vagueries of funding for community projects from year to year. In late Spring of 1998 state funding for the Lower West Side Development Corporation (and other similar agencies in the state) was cut by the Governor of New York. While the Corporation still exists on paper, it is down to a minimal staff and is currently unable to support further development of the Lower West Side site. Existing community groups in the Lower West Side are not particularly cohesive (split along income and racial lines) and thus it has been difficult to develop a new home for the Lower West Side PPGIS / PPVis project. Unfortunately, in most cases it will be those communities that are more stable, wealthy, and less vulnerable that can support the development of PPGIS and PPVis sites on the WWW. Such insight, while not particularly surprising, is one of the primary results of this research project.

 Regardless of the ultimate fate of PPGIS and PPVis in Buffalo's Lower West Side Community, the project has served to focus many theoretical and conceptual issues. To reiterate, I am concerned with developing a theoretically informed practice of PPVis and PPGIS: wedding conceptual and theoretical ideas to actual implementation of a site in a community. A review of some of the most important conceptual issues surrounding the praxis of PPGIS and PPVis are described in the final section of this paper.

 3. Conceptual Issues in the Praxis of PPGIS and PPVis

 A praxis or theorized practice of PPVis and PPGIS consists of an explicit awareness of the concepts and theories of information, its representation, of people, social relations, power, and how these shape and are shaped by socially-infused technologies such as PPVis and PPGIS. Simultaneously, such awareness must be brought to bear on real applications used by real people. A focus on only the practical (technical, software, hardware) or the conceptual (social, information, or cognitive theories) may limit the development of actual PPGIS and PPVis technologies in ways that substantially benefit real people in real communities. I discuss a few important issues in the praxis of PPVis and PPGIS below, and more exist (other issues are discussed at the WWW site for this project). My descriptions of these issues are brief and I plan to explore them in more detail, as research projects, in the future.

 3a. The Geography Behind PPGIS and PPVis

 The prototype WWW site developed for Buffalo's Lower West Side is based on very simple geography which allows users to interact with where particular phenomena (property owners, housing violations) are. Geographers (and others) have developed many sophisticated methods for analyzing and understanding geographic phenomena. For example: many concepts and models and methods of analyzing economic data exist and are used by geographers, planners, and regional analysts. The technology for providing such geographic methods of analysis via the WWW exists or will exist soon. To what extent can these more sophisticated means be incorporated in PPGIS and PPVis applications? What are the potential problems and benefits of the 'general public' having access to such geographic methods and models? What do users of such applications need to know to use and understand such methods? How can we build an educational component into PPGIS and PPVis applications? This component of PPGIS and PPVis may be guided by existing literature on the design and implementation of educational multimedia and other pedagogic materials (discussed in Krygier et al. 1997a). The importance of geographic education in the context of PPGIS and PPVis cannot be underestimated. A future PPGIS and PPVis site will incorporate several levels of geographic analysis tools, from the simple (as in the prototype) to the more sophisticated (modeling, statistics, etc.). The design and evaluation of such sophisticated analytical tools, and what users need to know to effectively use the tools, are discussed below.

 3b. The Medium and Site Content: Representation, the Visual Forms, Hypermedia and Intertextuality

 PPGIS and PPVis are not only maps and GIS, but images, video, text, and sound (Krygier 1994). The way these interrelated representations are linked together - the intellectual design of PPGIS and PPVis - must be carefully considered (Krygier 1999). This intellectual design is guided by cognitive, social, and geographic theories and may (should?) be open to modification by the users of the site. This component of my research focuses on the manner in which current concepts and theories in human geography relate to certain fundamental aspects of visualization and PPVis - the significance of interconnected representational forms (Cosgrove 1984, Krygier 1997b), the spatiality of the map, linked to the development of spatial components in social theory (Sayer 1992, Krygier 1995, 1996), and hypermedia, linked to hypertextual theory (Bolter 1991, Landow 1992, Krygier 1995, 1996). This aspect of my research, then, links issues of representation back to the concepts and theories of geography discussed in the previous section (3a).

 3c. Public Participation and Information Technology

 Enhancing public participation with the use of information technology consists of more than just making the technology available to people. For example, one can have access to tools which provide a sophisticated geographical analysis of environmental data for an area, but not actually understand the analysis itself. Of particular interest to me and of some importance to PPVis is the idea of " non-threatening graphics" - simply defined, graphics that encourage rather than discourage participation. I first heard of this term from academic planner John Felleman (SUNY ESF, Syracuse NY). Felleman told me that public participation in traditional settings (eg., public meetings) can be diminished if the graphics used to present information about alternatives look too polished, professional, and finished. Felleman suggested that sketchy and less finished looking graphics tend to encourage public participation: the graphics look like the proposal is still in the "sketchy" and undecided stages. The same phenomena was noted by Alan MacEachren in his book How Maps Work (p. 456). What do 'non-threatening graphics' look like? What constitutes a threat? How can we evaluate if certain designs are threatening? The issue of 'non-threatening graphics' is broader than only graphics, and includes all aspects of the design of a PPGIS and PPVis application in order to insure effective use by the public. Some possible characteristics of 'non-threatening graphics' for PPVis include:

 - use of game- and role-playing metaphors

 - allow people to explore issue at home (rather than only in public meeting)

 - intermediaries used in public meetings to do what people ask

 - the look of the graphics: sketchy/scruffy vs. finished

 - hinge graphics: ex) panorama as hinge between situated and map view

 - design to move people through increasing levels of complexity

 - design to make people critical (different perspectives on same issue)

 - include on-line encyclopedia of basic concepts needed to participate

 3d. Evaluation

 Evaluation is a complex and important issue: how do we assess the impact and consequences of the use of PPVis and PPGIS? I have modified a broad approach to evaluation described in an earlier paper (Krygier et al. 1997a) to fit the context of PPVis and PPGIS (described in Krygier 1999). To summarize, evaluation can play a role through an entire project - helping to shape and reshape the design in the process of its development and implementation. Further, evaluation should be conceived of broadly, consisting of four (often interrelated) functions: goal refinement, documentation, formative evaluation, and impact evaluation. Goal refinement consists of a detailed plan of action and set of goals; documentation is, simply, documenting what is actually done in the process of creating the site; formative evaluation consists of the systematic collection of information during the process of creating the site in order to get preliminary feedback on its viability; and impact evaluation consists of the evaluation of the final site. Each of these evaluation functions can be facilitated with a range of evaluation methods, including interviews, focus groups, questionnaires, observations, ratings assessment, expert review, and achievement tests (a range of both qualitative and quantitative methods). An important component of impact evaluation for PPVis and PPGIS may be Brenda Dervin's Sense-Making approach (Dervin 1998, Gluck 1998). For practical purposes, Sense-Making has well tested methods and numerous applications in many fields. Sense-Making, then, is particularly viable as a means of understanding and evaluating the complex interactions between users and PPVis applications.

 A major advantage of Sense-Making is that it is based on the same conceptual and theoretical ideas which infuse contemporary human geography and social science. Sense-Making is based on a conception of humans moving through complex time/space contexts that is strikingly similar to Hagarstrand's Time Geography and Giddens structuration theory. Dervin brings these important theories into the realm of information design by arguing that all information is designed: "humans make it and un-make it" (Dervin 1998). In other words, Sense-Making provides both theory and methodology which help guide the development of systems which not only deliver information to people, but which allow people to modify, change, and adapt the systems and information. "Sense making ... explicitly privileges the ordinary person as necessarily a theorist involved in the development of ideas that provide guidance not only for understanding personal worlds but necessarily for understanding collective, historical, and social worlds as well" (Dervin 1998). This is the goal of PPGIS and PPVis - to empower users rather than only provide them with existing information. Sense-Making can be a vital element of the praxis of PPVis: an explicit theoretically informed approach to information design which, as Dervin argues, assists "humans in the making and unmaking of their own informations, their own sense" (Dervin 1998).

 Conclusions and Future Directions

 This paper has discussed both technical and conceptual issues related to Public Participation GIS and visualization. My research suggests that the most vital issues for PPGIS and PPVis are not technical issues. The successful development of a prototype WWW-based site, as described in the first section of this paper, demonstrates that existing software can provide the kind of functionality necessary for PPGIS and PPVis applications. While the technical development of such sites is certainly not easy, I believe that it is currently possible, given sufficient resources, to design, produce, and implement a sophisticated PPGIS / PPVis application. The most pressing issue concerning the technology itself is the costs involved - of hardware, software, and programmer time. Universities and academics have an important role to play in providing such resources to marginalized people and places.

 More vital are the concepts and ideas which shape the development of actual PPGIS and PPVIs applications. This paper has reviewed a few of those issues, and all will play a role in the future development, implementation, and evaluation of such applications. A theorized practice or praxis of PPGIS and PPVis is vital. What concepts and theories of geography underpin the analytical capabilities we provide to PPGIS and PPVis users? How do we structure and build the GIS and visualization tools to enhance the user's understanding of, and participation in shaping, the content and analysis available? How do we design PPGIS and PPVis sites so that they encourage, rather than discourage participation? How do we evaluate the impact of such tools given the complexities of the publics who use them? I have offered some preliminary ideas, theories, and research directions which may contribute to answering such questions. I am encouraged by the manner in which PPGIS and PPVis encourage substantive links between geographic methods (GIS, visualization) and geography, and between geographers and people outside of the academy.

Bibliography

 Bolter, J. 1991. Writing Space. Hillsdale NJ: Lawrence Erlbaum.

 Chang, K. 1997. "The Design of a Web-based Geographic Information System for Community Participation." Master's Project, State University of New York at Buffalo, Department of Geography. (copy available at http://www.geog.buffalo.edu/~jkrygier/)

 Cosgrove, D. 1984. Social Formation and Symbolic Landscape. Totawa NJ: Barnes and Noble.

 Dervin, B. 1998. "Chaos, Order, and Sense-Making: A Proposed Theory for Information Design. (http://edfu.lis.uiuc.edu/allerton/95/s5/dervin.draft.html). [a revised version of this is to be published in a book called Information Design edited by R. Jacobson].

 Egenhofer, M and D. Mark. 1995. "Naive Geography." In: A. Frank. Spatial Information Theory. Berlin: Springer-Verlag.

 Ghose, R. 1994. "Use of Information Technology for Community Empowerment: Transforming GIS into Community Information Systems." Paper presented at Association of American Geographers Annual Meeting, April 1994.

 Gluck, M. 1998. "Interpreting Geospatial Representations: Social Semiotics for Cartographic Analysis." Paper presented at the 1998 Association of American Geographers Meeting, Boston MA.

 Krygier, J. 1994. "Sound and Cartographic Visualization." In: D. Taylor and A. MacEachren, eds. Visualization in Modern Cartography. Oxford: Pergamon Press.

 Krygier, J. 1995. Visualization, Geography, and Landscape. PhD Dissertation, Pennsylvania State University, Department of Geography.

 Krygier, J. 1996. "Geography and Cartographic Design." In: C. Wood and C. Keller. Cartographic Design: Theoretical and Practical Perspectives. Chichester: Wiley. pp. 19-34.

 Krygier, J., C. Reeves, D. DiBiase, and J. Cupp. 1997a. "Design, Implementation and Evaluation of Multimedia Resources for Geography and Earth Science Education." Journal of Geography in Higher Education. 21:1. pp. 17-39.

 Krygier, J. 1997b. "Envisioning the American West: Maps, The Representational Barrage of 19th Century Expedition Reports, and the Production of Scientific Knowledge." Cartography and Geographical Information Systems 24:1. pp. 27-50.

 Krygier, J. 1998a. "Public Participation Visualization." Paper presented at the 1998 Association of American Geographers Meeting, Boston MA. (copy available at http://www.geog.buffalo.edu/~jkrygier).

 Krygier, J. 1998b. "Project Ketch: Project Plowshare in Pennsylvania." Ecumene. (forthcoming)

 Krygier, J. 1999. "Cartographic Multimedia and Praxis in Human Geography and the Social Sciences." Forthcoming in: W. Cartwright, M. Peterson, and G. Gartner, Multimedia Cartography. Berlin: Springer-Verlag. (forthcoming)

 Landow, G. 1992. Hypertext: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins University Press.

 MacEachren, A. 1994. "Visualization in Modern Cartography: Setting the Agenda." In: A. MacEachren and D. Taylor. Visualization in Modern Cartography. Oxford: Pergamon. pp. 1-12.

 MacEachren, A. 1995. How Maps Work. New York: Guilford.

 Sayer, A. 1992. Method in Social Science: A Realist Approach. London: Routledge.