Reading Risks

A Proposal to NCGIA Initiative 17: Collaborative Spatial Decision-Making

Seymour J. Mandelbaum

May 11, 1995

This brief account of the paper I am writing describes two complementary focii. The first places the relationship between geographic information systems and collaborative group dynamics within the general framework of a process through which planning tools are made and broken. The second explores that framework using the representation of risk as an analytic probe. The first focus dominates this note; the second will dominate the paper.

The call for the conference is grounded in a archetypal story. In this stylized account, a group is engaged by an issue that appears (at least to some in the circle) to call for collective action. The conversation begins and then, at one point or another, the discussants are presented with a spatially organized set of measurements that I will call, for simplicity's sake, a "map." The group processes continue -- sometimes ignoring the map and sometimes using it to shape the issue and the group's decisions -- until (with the authority of the narrator) the story is brought to a close.

The conference and the research program to follow seek -- as I interpret the call -- to understand and improve the processes of creating and reading maps -- in all their variety. "How," we ask within the epistemic community of map-makers, "does the representational form, timing, social provenance and authority of a map -- or better, a set of maps -- influence its uses? How do the characteristics of issues and of groups shape the dynamics of reading?"

The goal of the enterprise is to increase the ability of a group to articulate and evaluate alternative "solutions to spatial problems," and to resolve "spatial conflicts." The political framework within which these aspirations are set is a general commitment to "collaboration." The term is ambiguous -- masters and slaves may, after all, be said to "collaborate" -- but I assume that it here implies a high level of uncoerced consensus within the group, resolving conflicts in a way that is not only legitimate but satisfactory. In the fashionable term of the mediator's art, a collaborative sensibility seeks "win/win" solutions.

Within this context -- engaged in the service of the collaborative sensibility -- maps and the processes of creating and reading them must capture and discipline the issues that occupy the collaborators privately. Absent that engagement, maps are rejected (either tacitly or explicitly) as irrelevant or pernicious; are controlling only by the use of coercive mandates.

That is, I suspect, a familiar caution. If the level of aggregation in a geographic information system is too large to capture the moral concerns of the collaborators, it is difficult for them to express and compare their worlds. Maps confuse and contaminate the conversation when boundaries are "wrong," "distance" measures don't capture "location," social conflicts are misread in the language of space or protean processes are deceptively frozen in time.

Familiarity is often, however, only small comfort. Warnings may be ritually repeated with little impact on the deep structures that make them difficult to follow. The measures and the representational protocols that are brought to bear in "group-based" processes are characteristically generated in complex polities and depend upon both radical simplifications and legitimate coercion. Like the U.S. Census, they are trenches designed to constrain and manage group processes by providing a public conception of their external environments. Inevitably, however, the external measures colonize the representation of the internal environment: we are bound sometimes to describe ourselves as if we were figures in the Census or the reporting forms attached to funding requests. (In this way, we adopt one of the prescribed ethnic identities or measures of "need.") Uncoerced collaborations make it difficult, however, to repress identities and secret concerns. When they "out," they threaten the discipline of an imposed public order (or map).

The perils of the external information system are complemented by internal difficulties. A collaborative process often changes the Mind and archetypal stakeholders of a group, altering the balance of open and strategic speech and shifting preferences and conceptual frames. The stages of deliberation alter the demands for information. Boundaries, dimensions and levels of aggregation that capture and discipline speech in the opening moves of a collaborative process may be too simple for the middle game. As a group works towards convergence, however, the complexity of the middle may retard closure if it is not shed. A geographic information system to "support" uncoerced decisions must be appropriate to its moment. It is, however, virtually impossible to create a map that changes with the group if the map-makers see themselves as creating a stable, foundational order of "fact" and if the process of reading is not robust enough to support a conflictual conversation of incomplete intimations.

The often faltering and flawed adjustment of geographic information systems to the dynamics of collaborative groups exemplifies a ubiquitous process of making and breaking planning tools. The making of tools may start with the social construction of a privileged image of a "problem" and the subsequent crafting of an instrument appropriate to their construction: "Since this is a spatial problem, we need a map." The flow may, however, proceed in the other direction: "We have a map so let's think of this as spatial problem."

In a symmetrical fashion, the breaking of tools may be initiated either by the denial of the cogency of a problem construction or the relevance and effectiveness of the instrument. The symmetry is not, however, complete. It does not take a tool to break a tool. A technological consensus may be eroded without a replacement arising instantly in its place. (In the interregnum, both savvy professionals and ordinary men and women may grieve for lost certainties.) The rhetoric of making and breaking may also be different: one confident and romantic; the other skeptical, angry and even tragic. ("What becomes of the precious notion of the general welfare," we wonder sadly, "if the technical bases of a utilitarian calculus no longer command our respect?")

There is a great deal of similarity between the making of planning tools and the (much more familiar) crafting of technological innovations in manufacturing. Some manufacturing innovations begin on the shop floor and spread initially from one set of users to another without the mediation of cosmpolitan scientists or engineers. Others, originate in research laboratories and are first expressed in abstract rhetorical forms before they are "applied" to "practical" affairs. In the same way, some planning innovations originate in the dense world of collaborative groups and others in the necessarily simplified rhetoric and authoritative claims of large polities. Wherever the tale begins, innovations in manufacturing and planning only earn a consensual authority -- only, in effect, appear as "tools" -- when they have both been honed in the practices of particular places and explicated in the simplified rhetoric of every place.

The process of making tools at two different scales and in two different rhetorical styles, creates complementary disciplines. Collaborative group processes test both the claims to authority characteristic of large polities and the requirements for simplification associated with the management of diverse units. Hierarchical processes in large polities sustain public orders that support inter-communal relations and make it possible to distinguish the array of "distinctive" places in which a tool is apposite.

The call to the conference quite sensibly suggests that experience matters; that some problems in the design of geographic information systems are hidden when we design for every place and in the abstract language of the map-makers themselves. Only in the practices of collaborative groups do we finally appreciate how unfriendly are our designs or how much they distort the worlds in which users struggle with collective choices. (A similar observation applies, of course, at the other end of the spectrum: experience matters in the appreciation of the limits of collaborative processes.)

Experience is not, however, our only teacher. Many of the qualities in information systems that frustrate collaboration are apparent in the arguments phrased at a high level of abstraction in the corridors of power and formal science.

Consider the representation of risk.

Some geographic information systems purport to map explicitly the incidence of risk: here is a cancer belt, there the aquifer carries industrial pollution from brownfields to rural pastures. There is, however, a larger sense in which every spatial display -- even the most benign -- evokes a reading of risk. Presented with a map, we all sensibly wonder: who and what is at risk in either trusting the display or in distrusting it? what sorts of risks are attached to the spatial pattern elaborated in the map? The ways in which we answer those questions express our essential images of time and planning. If a collaborative process does not provide an opportunity for those readings of risk to be expressed, they will (like the water-borne pollution) surface in strange places.

Tools for the representation of risk are grounded into one of two images that are formally distinctive and even conflicting though in the muddied and opportunistic language of planning they often appear together. In the first, risk appears as an expected consequence of enterprising behavior so that we talk of ourselves as responsible agents "taking" risks. We may prudently act to mitigate future dangers but we cannot expect to eliminate them. We should, indeed, be wary of devoting ourselves so fully to mitigation that we foreclose valuable opportunities; of being so risk aversive that we accomplish nothing. The tools we construct within this image seek to decompose bundled risks and opportunities into discrete elements and to create a common metric that allows us to compare and aggregate prospective gains and losses. The tool-makers, acting within this image, seek in the words of a recent NSF-EPA request for proposals, to develop a "systematic and compelling valuation mode."

In the second image, we speak of ourselves as victims whose rights are violated when we are put "at risk" from the behavior of others or our own ignorance or akrasia. The risk -- always prospective -- is represented as if it were a tort -- an unjustified injury in the past -- that must (rightfully) be rectified. The tools of representation engage the allocation of blame and the assignment of responsibility for rectification. Seymour J. Mandelbaum
Professor of Urban History
Department of City and Regional Planning
University of Pennyslvania
Philadelphia, Pennsylvania 19104-6311

Telephone: (215) 898-6492
Fax: (215) 898-5731
E-mail: mandel@pobox.upenn.edu

This proposal follows upon a resolve to increase contact between the communities of interpretive planning theoreticians and of designers crafting new planning technologies. That resolve is announced in an essay -- "The Technological Sensibility" -- scheduled for publication in the April issue of the Town Planning Review and in my introductory essay -- "The Talk of the Community" -- at the beginning of Explorations in Planning Theory, scheduled for publication in December. Dick Klosterman, Judy Innes and I have arranged two sessions on the theory/technology nexus at the Detroit ACSP meeting in October.

The tack I have taken -- framing the making and breaking of planning tools within the dynamics of epistemic and practice communities -- continues a theme that runs through virtually all my work: from Boss Tweed's New York (1965) and Community and Communications (1972) to a new book, Open Moral Communities, that will, I hope, be out next Spring. The title I have given this proposal emphasizes this continuity by echoing an earlier essay on "Reading Plans," Journal of the American Planning Association, 56 (1990), 350-356.

In addition to the project on tools, I have joined a recent proposal to NTIA to use Libertynet -- the regional "freenet" -- as a complement to the "collaborative" development activities of the Philadelphia empowerment zones.