Formal Epistemology in a Tropical Savanna
As a philosopher who also practices ecology and has a small laboratory, conservation biology and disease epidemiology have been two areas that have afforded opportunities for testing the relevance of philosophical methods in the field, particularly decision theory as reconstructed within formal epistemology'. What follows is the story' of one such episode, an attempt to use decision analysis to support a complex multi-criteria, multi-agent decision to be made concerning land use in a remote region of Indonesia. This example shows both the promise of these methods and the practical difficulties faced in their attempted application.
I begin with some personal remarks explaining my' involvement in this project. In 2003, when Branden Fitelson and I began organizing the Formal Epistemology Workshops (FEW), our initial concern was mainly with reinvigorating formal work in fields such as confirmation and theory choice which had once concerned key figures such as Keynes, Ramsey, Carnap, Reichenbach, and others but had come to be neglected in the recent philosophy of science.
As we worked to develop formal epistemology, I initially felt a tension between its embrace of abstract examples divorced from science in practice and the work that my laboratory did. We were involved in applied work in disease ecology and conservation science. In both cases we worked extensively in fieldbased projects that involved working with local communities with multiple stakeholders. However, it soon became apparent that, beyond purely ecological scientific analysis, we were necessarily placing ourselves in policy contexts through our work, where the best feasible choice had to be made between feasible but imperfect alternatives.
Consequently, we began to see that our work was essentially falling within the domain of decision analysis—that part of formal epistemology' that deals with the rationality of choice and action. As will be seen below, we were dealing with particularly complex decision problems, sometimes technically termed “wicked problems,” because even the problem formulation is open to dispute. For conservation problems, we developed a framework that relied heavily on multi-criteria analysis. The example below discusses a case where this theoretical framework was translated into practice. For us, this experience has demonstrated the mutual value of interaction between philosophical theory and practical policy, but it has also led to a cautionary strategic attitude about developing collaborative partnerships with actors from outside academia, in particular with large non-governmental organizations (NGOs) and corporations.
To start at the origins of my interest in biodiversity and conservation science: when conservation biology first emerged as an organized discipline in the United States in the late 1980s, two publications were particularly important for many of us who had long been politically involved in attempts to conserve biodiversity. The first was Michael Soule’s manifesto “What is conservation biology?,” which appeared in BioScience (Soule 1985) a few months after the founding of the Society for Conservation Biology' (Soule 1987). The second was Dan Janzen’s exhortation to ecologists, “The future of tropical ecology” (Janzen 1986). Janzen’s rhetoric was powerful. In Costa Rica, where he mostly worked, as well as elsewhere in the tropics, he witnessed natural habitats disappearing under pressure from powerful socio-economic forces. If biologists wanted to continue having tropical nature to “biologize” about, Janzen argued, they must undertake the political activism necessary for conservation. They must do so by embedding themselves and propagating their conservationist imperative within local cultures. There was no alternative. It was a demanding and compelling vision.
To a philosophy' graduate student, as I then was, the contrast between Janzen’s and Soule’s pieces could not have been greater. Soule, like Janzen, also conveyed a sense of urgency by' declaring conservation biology' to be a “crisis discipline,” but he then proceeded with great pomp to produce hyper-theoretical “postulates.” He had two types of postulates, functional ones describing the state of the world and normative ones explicating an ethic of conservation. The functional postulates were relatively innocuous claims about ecology, how co-evolution has taken place, how ecological processes may enter chaotical, dynamical regimes beyond some parameter values, and so on. None were either very original or controversial. What held my attention were Soule’s normative postulates. Though Soule invoked the hallowed name of the philosopher Arne Naess, who had once been part of the Vienna Circle, his normative postulates were ex cathedra pronouncements with little supporting argument.
Ecological complexity is good, Soule declared. So is evolution. But, I remember wondering, evolution consisted of three main processes: adaptation, diversification, and extinction. If evolution is good, presumably all these processes should also be good. But wouldn’t finding extinction good be awkward within a conservationist ethic? Unless, perhaps, extinction was supposed to pave the road for further evolution. From Soule’s account, there was no way to tell: there was no further discussion. More controversially, Soule simply announced that biological diversity had intrinsic value. He did not even seem to understand that, here, he was treading into well-worked and controversial philosophical territory, and that claims like this should not be made without adequate defense. Perhaps even more troubling was that Soule’s ethic had no place for human values and interests. Like most other Northern conservation biologists of that decade, he harped on the biodiversity that was at risk in the global South. But he did not once acknowledge a need to engage those Southern populations that would bear the brunt of his proposed policy measures.
In contrast to Janzen’s inclusive vision, Soule’s normative program was imbued by the values of just one faction of North American environmentalism. This faction styled itself deep ecology, with the honorific “deep” presumably intended to indicate that adherents were privy to some profound insights inaccessible to ordinary ecologists and environmentalists who were concerned, e.g., with resource depletion or pollution. Deep ecologists often gloated on their distaste for human values and interests, putting what they regarded as Nature above mere humans. In 1989, this ideology was subjected to a blistering attack— “Radical American environmentalism and wilderness preservation: A Third World critique”—by Ramachandra Guha, an Indian anthropologist visiting the Yale School of Forestry and Environmental Studies. In my assessment, very little of Soule’s conservationist ethic survived Guha’s attack, which became one of the most anthologized contributions in the emerging discipline of environmental philosophy. My rejection of Soule’s vision of conservation biology paved the way to eventually viewing conservation problems as decision problems rather than ecological problems.
Even though I taught, at McGill University in 1995, what was one of the earliest undergraduate courses in the philosophy of conservation biology, I did not anticipate doing work in conservation biology. However, biodiversity protection remained a guiding political goal in my environmental activism from that period. That changed when I spent the 1996—1997 academic year at the Wissen-schaftskolleg zu Berlin (WiKo for short). The Kolleg encouraged free-ranging discussion and collaboration among its Fellows and the time spent there was intellectually rewarding. While there were about 40 resident Fellows every year, many others came for shorter periods including those from earlier years. Two such Fellows who visited WiKo in Spring 1997 were Guha and Chris Margules, then working at the Commonwealth Scientific and Industrial Research Organisation (CSIRO) in Australia. Thanks to their presence I turned from ongoing work on the history of evolutionary genetics to problems of biodiversity conservation.
With Guha’s encouragement, I completed a philosophical critique of wilderness preservationism that contrasted it with biodiversity conservation (Sarkar 1999). My interactions with Margules were more significant. During the 1993-1994 academic year at WiKo, Margules had been part of a Working Group, “Biodiversity Reserve Selection Methods,” that had brought together experts from Australia, South Africa, and the United Kingdom. Together, this group began to develop a framework for biodiversity conservation that came to be called “systematic conservation planning” (Sarkar and Margules 2014). Margules and Pressey (2000) went on the publish the foundational document for the new approach shortly afterwards.
The central problem of systematic conservation planning was the selection of a set of areas earmarked for biodiversity conservation. Originally, these areas were presumed to be reserves but, as a few of us soon emphasized, reservation is not the only valid approach to conserving biodiversity. Moreover, as the philosophical critics of wilderness preservationism had been pointing out for over a decade, creating reserves was associated in many cases with involuntary displacement of resident peoples. Consequently, the term “conservation area” came to replace “reserve” (Sarkar 2005). Moreover, since it was also recognized that successful conservation required engagement with the entire spatial matrix of the region, and would be futile if restricted only to selected conservation areas, the term “prioritization” came to replace “selection.” The prioritization of conservation areas in a landscape or seascape remains central to systematic conservation planning today.
In Berlin, Margules and I realized that conservation area prioritization is a decision problem. We had a set of feasible alternatives, the potential sets of conservation areas. Here, “feasible” means that each such set satisfies constraints such as the total budget for conservation measures. The criterion by which performance was to be measured is the representation of biodiversity features. Because of the large size of the data sets typically required for such decisions (e.g., hundreds of thousands of habitat patches and thousands of species), computer-based methods were needed. These, in turn, required the development and use of software decision support tools. Margules and other Australian researchers, as well as Paul Williams from the Natural History Museum in London, had already begun developing such software but there was much more that could be done. I began to see that my background in decision theory as a philosopher (as well as experience in computer science) could be useful in systematic conservation planning. The context was set for a potentially fruitful interaction between formal epistemology' and conservation science in the field.