IEAM Spotlight: Environmental Scientists as Philosophers?
Brian Church, Windward Environmental LLC
Philosophical issues don’t often warrant serious discussion in scientific circles, but in an upcoming Brief Communication in Integrated Environmental Assessment and Management (IEAM) entitled “Pragmatism: A Practical Philosophy for Environmental Scientists,” Suter and Cormier provide a brief lesson in scientific epistemology to shed light on the fact that, whether we accept it or not, environmental scientists are philosophers too. Not only are we philosophers, but as environmental scientists we ascribe to one particular school of thought: Pragmatism. This lesson, and the authors’ discussion that stems from it, is intended as a refutation of a comment made previously by an IEAM editorial board member. It was said that, “science is not a matter of voting,” which Suter and Cormier think is less than spot-on; they pull philosophy into the fray to explain why.
As a general rule, the pragmatist theory of truth states that truth and knowledge are only valid if they apply tangibly to reality. The crux of pragmatism is that it requires concepts to have an application, such that experience informs knowledge and not the other way around; that is, we can’t determine what is true while sitting in our armchairs. William James points out in Pragmatism that “truth” doesn’t necessarily imply “true under all circumstances;” a true statement simply explains to the best of our ability what is observed. Pragmatists and scientist alike argue, somewhat obviously, that if you can’t test it, you can’t disprove it. This is the scientist’s need for a practical application. Therefore, if you are a practicing scientist, you are also most likely a practicing pragmatist. Better get used to it.
While that seems good and obvious to us, it isn’t the whole issue. The definition of truth described by Suter and Cormier also implies that truth can’t be arrived simply by conferring with other people. This caveat seems to be the source of confusion. Charles Sanders Peirce and Oliver Wendell Holmes describe scientific truth not just as a convergence of ideas but as the competition between hypotheses and the ultimate synthesis or emergence of lasting and testable theories. Although we may “vote” for the best explanation, we do not do so without examining all available evidence. In this way, we vote on what is “true,” according to the pragmatists’ definition.
The essential tenants of pragmatism are simple and, as the authors reluctantly summarize, somewhat “shallow” (Suter and Cormier, in press). The simplicity of this approach is quite striking compared to other, more nebulous epistemologies, which is perhaps what makes pragmatism so appealing to the modern scientist. An example can be seen in William James’ concept of pluralism, which posits that more than one approach to a problem may result in various practical solutions. This approach is certainly useful when weighing multiple lines of evidence. The authors cite anthropogenic global warming as one such issue for which this approach should be used to arrive at truth and to inform decision makers. The whole idea of a National Academy of Sciences, and other forums and councils akin to it, arise from the need to compare notes before arriving at the “whole truth.” But, as environmental scientists, is this philosophy all that we need or use?
Under many circumstances we do apply these methods, but we do not use pragmatism for all problems (or at least not only pragmatism). It is clear from the articles accepted (but yet unpublished) from Stahl and Cimorelli and Forbes and Calow that we practice other philosophies as well, whether we are cognizant of them or not.
Not all situations call for the pragmatists’ joyously Taoist “whatever works” mentality. Some environmental problems require more discussion and deconstruction, and others are downright “wicked” (Stahl and Cimorelli, in press). Whereas Suter and Cormier advocate openness, transparency and consensus within the scientific community as precursors to decision-making, they fail to mention the key role that stakeholders play in the decision-making process.
Involvement of the public is touched on in two upcoming articles from Stahl and Cimorelli, and Forbes and Calow. Each presents a slightly different approach to informing decisions, but both draw on another school of philosophy called Utilitarianism. Utilitarianism deals with maximizing what is good based on expected consequences, whatever that may mean (Bentham, 1907; Mill, 1909), and in the case of environmental issues there may be a variety of positive ends for which to strive.
At first glance, modern science and utilitarianism appear to be at odds. Many would argue that it is utterly incompatible with science because the introduction of subjective bias (e.g., what is good) is something that scientists have pooh-poohed since time immemorial. Of course, subjective bias is inherent to the utilitarian approach since the consequences of any available decision are necessarily deemed desirable, undesirable or some compromise gray-area between, based on a subjective value system. The application of values to results might not seem scientific, but it can be found applied in numerous ways. In particular, many decision frameworks exist that embrace this shift away from the objectivist paradigm in light of new approaches that better tackle wicked problems, those with social, economic and political implications in addition to environmental. Wicked problems are deeply subjective, and many different viewpoints must be considered to appropriately manage the problems.
Stahl and Cimorelli discuss a decision tool developed by the United States Environmental Protection Agency (USEPA) called the multi-criteria integrated resource assessment (MIRA) decision analytic approach. MIRA has been developed under a variety of circumstances for a decade or more. Its purpose is not only to provide a scientific basis for decision-making but also to involve stakeholders throughout the analytical process. This is a step beyond what Suter and Cormier imply, consensus of scientists first and then education of the masses; it is instead a two-way learning process by which scientists and modelers inform stakeholders of the outcomes of multiple model scenarios and by which stakeholders inform scientists on what matters by assigning numerical values to model parameters. The authors aptly refer to the process as “clumsy” although “necessary” (Stahl and Cimorelli) due to the non-linear, iterative and possibly contentious nature of assigning systems of valuation.
The MIRA approach involves a great deal of modeling (i.e., fate and transport models, spatial models and benefit analysis), which allows scientists to first introduce stakeholders to the issues, their expert opinion of value (i.e., “indexing”), and likely scenarios and solutions. Then stakeholders, by assigning their value set (i.e., “preferencing”) allow for value-relevant model outputs, as well as model back-casting to desired outcomes. Using this approach, scientists and stakeholders use real world data and reliable models with two subjective value sets to arrive at mutually acceptable determinations of risk and benefit. If the result is not mutually acceptable, back-casting allows for the reevaluation of what really matters and adaptation of the value set. The information can then be used by managers to make decisions that “maximize the good.”
The MIRA framework can also be used within a closed scientific circle by assigning a value set based on the literature. Stahl and Cimorelli provide the example of air quality in a large area of the northeastern United States. They model how management of various air pollutants affect localized changes in 2.5 μm particulates, nitrogen oxides and various volatile organic compounds, all parameters with significant environmental implications. All the parameters are given different values based on stakeholder inputs, best professional judgment and/or inhalation reference concentrations. The model then spits out relative values that rank management options according to their overall value to stakeholders and the environment.
Forbes and Calow present a similar approach to informing decision makers formed by synthesizing two seemingly complementary approaches, Ecological Risk Assessment (ERA) and Ecosystem Services (EsS). The authors present two value-relevant programs that have taken this approach. Although the authors note the EsS philosophy as the basis for these programs, it would be equally as pertinent to discuss the utilitarian philosophy from which it most certainly emerged. The first program described is the update of the European Food Safety Authority’s (EFSA) technical guidance documents for pesticide risk assessment. The program provides regulators with insight into what receptors matter, where and on what scale. These insights are paramount to a successful risk assessment since they form the basis of the questions that the entire assessment aims to answer. The difference between this approach and typical risk assessments is that instead of making decisions from a toxicological standpoint (and often at the scale of a single individual), the technical guidance documents attempt to inform decisions from an EsS standpoint. That is to say that it poses questions like, what is the net benefit of such-and-such decision, and more to the point, how can we maximize the good.
The second program that the authors present is the USEPA framework for Comprehensive Environmental Response, Compensation and Liability Act (CERCLA), and Natural Resource Damage Assessments (NRDA). This synthesis of approaches is similar to what was attempted by the EFSA, in that ERA is integral to CERCLA and ES is integral to NRDA. The problem with the current USEPA approach is that ERA (as part of CERCLA) and ES (as part of NRDA) are separate and may proceed in two disjointed steps. Specifically, the endpoints evaluated as part of an ERA are likely to differ from those evaluated as part of NRDA.
Forbes and Calow provide potential solutions to joining the two steps--a series of models that link toxicological data of varying scale to services rendered. Mechanistic effects models are the first; they include a broad range of specialized uses that link standard test endpoints to impacts on service-providing “units” (SPUs) (e.g., populations). Different models deal with different SPUs, scales or life stages and can incorporate other complicating factors such as density dependence and stochasticity. The second group of models is ecological production functions that link SPUs to the delivery of services, the practical benefit of managing an SPU. Measurable population parameters can thus be equated to an ES value.
By combining the predictive power of mechanistic effects models with ecological production functions, the data typically evaluated within an ERA context can be directly transferred to the NRDA process. The outstanding issue with the CERCLA/NRDA framework (aside from the nascence of the aforementioned models) as well as the EFSA and MIRA approaches is that value sets are neither static nor universal. There is not yet a unified “valuation method” (Forbes and Calow) that can be applied to all situations. The authors point out that such a framework would be advantageous in the marriage of CERCLA ERA to NRDA as a unified, value-relevant approach.
Taking these studies into consideration, is it correct to say that environmental scientists are practicing pragmatists? Do they become practicing utilitarians under different circumstances, like when dealing with wicked problems? Whether we tackle an issue collaboratively within the scientific community using only objective data and our wits, or engage the public with value-driven discourse in order to reach mutually acceptable management decisions, what we consider our professional philosophy might not ultimately matter. “Whatever works.”
- Bentham, J. An Introduction to the Principles of Morals and Legislation. Oxford: Clarendon Press. 1907. Library of Economics and Liberty [Online] available from econlib.org/library/Bentham/bnthPML.html accessed 26 February 2013.
- Forbes, Valerie E, Calow Peter. Use of the ecosystem services concept in ecological risk assessment of chemicals. Integr Environ Assess Manage. DOI 10.1002/ieam.1368.
- Mill, JS. Principles of Political Economy with some of the Applications to Social Philosophy. Ashley, W.J., ed. London; Longmans, Green and Co. 1909. Library of Economics and Liberty [Online] available from econlib.org/library/Mill/mlP.html; accessed 26 February 2013.
- Stahl, Cynthia, Cimorelli, Alan. A demonstration of the necessity and feasibility of using a clumsy decision analytic approach on wicked environmental problems. Integr Environ Assess Manage. DOI 10.1002/ieam.1356.
- Suter, Glenn W II, Cormier, Susan M. Pragmatism: a practical philosophy for environmental scientists. Integr Environ Assess Manage. DOI 10.1002/ieam.1382.
Author’s contact information: firstname.lastname@example.org
Return to the Globe