Introduction
Recently, Thompson and Dean[1] argued (1) that their probabilist- contextualist dichotomy does a better of job of carving the field of risk studies than the positivist-relativist dichotomy proposed by me. They also claim (2) that Starr does not espouse the positivist position I attribute to him, and (3) that Wildavsky and Douglas do not hold the relativist position I say they have. On the contrary, this article shows (1) that -- although my framework differs in its accounts of Wildavsky, Douglas and lay views of risk -- Thompson's and Dean's proposed probabilist-contextualist dichotomy is nothing new. It mainly offers new labels to describe the same positivist-relativist dichotomy, including accompanying distinctions, articulated earlier by me. The essay also shows (2) that Starr defends the positivist position I attribute to him, and (3) that Wildavsky and Douglas espouse the relativist position I attribute. Finally, the article argues (4) that the Thompson and Dean account, unlike mine, is wrong to ignore the dangerous, anti-populist sentiments in the Wildavsky-Douglas position on risk.
Which label (probabilist-contextualist, positivist-relativist) better describes the risk debate? I believe that all four labels are useful; except for some warnings about anti-populist sentiments, discussed below, differences between the two frameworks are trivial. To see why there is no fundamental opposition between the two competing sets of labels, consider, first, definitions for the terms. Probabilists, say Thompson and Dean, see "risk as essentially characterized by the probability of events or their consequences."[4] Contextualists, on their account, believe "that no single attribute will be understood to characterize every instance of risk."[5] I have said, "the cultural relativists attempt to reduce risk evaluations to sociological constructs, ignoring their objective, scientific content, [whereas] the naive positivists attempt to reduce them to scientific rules, minimizing their ethical content. The cultural relativists overemphasize values in risk evaluation and assessment, whereas the naive positivists underemphasize them."[6]
Why do Thompson and Dean allege that their labels are better? They say that, "rather than suggesting two conceptions of risk in diametrical opposition," their "dichotomous terms represent opposite ends of a continuum."[7] They also claim to present a "more carefully nuanced account" of risk issues than mine. However, my account is not of two conceptions of risk in diametrical opposition; it also represents a continuum. I say explicitly that participants in the risk debate "are arrayed on a spectrum extending from pluralist or relativist views to logical-empiricist [positivist] positions."[8] I also point out that "cultural relativists overemphasize values in risk evaluation and assessment, whereas the naive positivists underemphasize them."[9] This highly nuanced account, that Thompson and Dean appear not to have read in full, distinguishes among differing relativists (e.g, soft relativists, cultural relativists) and differing positivists (e.g., early or naive positivists, later positivists) all at different spots on the values spectrum.[10] My account also warns that "most social scientists are not cultural relativists in the Wildavsky sense," and that "Wildavsky and Douglas have quite different views on relativism and on risk, with Wildavsky being more relativistic."[11] I also spend an entire chapter discussing "the probabilistic strategy" of positivist risk assessors like Starr.[12] Hence it is puzzling to see their probabilist-contextualist framework offered as if it were an alternative to my account.
Is there a difference between the two accounts? Except for several points (interpretations of Starr and Wildavsky-Douglas and of lay views of risk) to be discussed later, they diverge only in one important respect. Their labels (probabilist-contextualist) focus on the content of risk claims, whereas my labels (positivist-relativist) emphasize the epistemic reliability of risk claims. However, both sets of labels come down to the same thing: pointing out more or less emphasis on values at different ends of the risk spectrum. They note that, "as one moves from probabilist to contextualist poles, risk assessment is seen as increasingly value laden."[13] Similarly, I explain that "cultural relativists overemphasize values in risk evaluation and assessment, whereas the naive positivists underemphasize them."[14] Thompson and Dean seem to think their account is new and more nuanced only because they omit many distinctions along my spectrum and then claim those distinctions as their own. In other words, their proposed probabilist-contextualist dichotomy offers mainly new labels to describe the same positivist-relativist dichotomy articulated earlier by me.
Thompson and Dean[15] argue that Starr does not espouse the positivist position I attribute to him. I,[16] however, say that Starr is a positivist because he ignores the values component in stage two (risk estimation) of risk assessment and because he maintains that "risk assessment, at least at the stage of calculating probabilities associated with harms and estimating their effects, is completely objective, neutral, and value free."
Thompson and Dean, on the other hand, claim that "values have a role for Starr"[17] because (their only argument), in discussing "acceptability of voluntary and involuntary risks," Starr "implies that voluntariness is a subjective reaction."[18] Because they claim that Starr allows value judgments at stage three (risk evaluation) of risk assessment, they do not even make logical contact with my position. They say risk is value-laden for Starr, because he admits values play a role in determining risk acceptability. I, however, say Starr is a positivist because he denies the values component in the first and second stages of risk assessment, namely, risk identification and estimation. Moreover, I recognize that Starr accepts the value-ladenness of (stage three) risk evaluations; this knowledge is the reason that I clarified my claim about Starr's positivism in terms of the first two stages of risk assessment. Once one reads my position on Starr, it is clear that it is not at odds with the Thompson and Dean account.
throughout her other work, Douglas argues that the specific content of beliefs about purity, danger, and social taboos is arbitrary. Risk and Culture compared the arbitrary content of belief systems of several cultures to make the point that risk beliefs are part of political culture.... This view does not deny facts about risks in general, and it most certainly does not entail relativism with respect to specific well-characterized elements of risk, such as the statistical probability of events.
To defend their interpretation (just given), however, Thompson and Dean provide no quotations, no textual analysis of Wildavsky's and Douglas' statements. Instead they proclaim, ex cathedra, that Wildavsky and Douglas are not relativists. My analysis, however, derives from numerous direct quotations from Wildavsky and Douglas that reveal their relativism.
Consider, for example, the following statements of Wildavsky and Douglas that I have discussed repeatedly and analyzed:[20] "risks are social constructs";[21] "no one is to say that one [risk judgment] is better or worse" than another; [22] "each method [of risk evaluation] is biased";[23] when "thinking about how to choose between risks, subjective values must take priority";[24] people do risk evaluation "according to personal ideology";[25] "public perceptions of risk and its acceptable levels are collective constructs... like language and... aesthetic judgment... there is no reasoning with tastes and preferences";[26] "any form of life can be justified";[27] and "the selection of risk is a matter of social organization."[28]
If Thompson and Dean wish to show that Wildavsky and Douglas do not defend relativistic positions, then they need to provide analysis of Wildavsky's and Douglas' own words, such as just given. They ought not merely state their own undocumented opinion that Wildavsky and Douglas are not relativists, and expect readers to accept their begging the question. Such question begging about Wildavsky's and Douglas' relativism is all the more questionable because I have distinguished several types of relativism and explained the senses in which Wildavsky is more relativistic than Douglas.[29]
Wildavsky and Douglas also claim that laypeople "can never be satisfied" with risk reductions because they believe "there can never be enough [safety]."[33] They say laypeople are dominated by "superstitions" and by fundamentalist desires for unrealistic environmental "purity" in their views of risk.[34] They also claim that laypeople are anti-industry and anti-government and that they directly trace "chemical carcinogens in the body" to "immoral forms of economic and political power."[35] The moral of their criticisms is clear: if laypeople are irrational about risk, then it is wise to trust the evaluations of the "experts," usually the risk imposers.
It is important to note both Wildavsky's and Douglas' anti-populist claims as well as their relativism because both are dangerous to democratic social and political structures. Apart from their criticisms of lay views of risk, their relativism also contributes to a pro-industry and anti-populist bias, as the following reasoning shows: Large industrial and technological concerns tend to be imposers of potentially catastrophic risks, while members of the public tend to be potential risk victims. But if risk imposers (chemical companies, for example) are innocent until proved guilty, and if relativism precludes accurate evaluation of harm and authentic proof that someone is guilty of imposing probable harm, then risk imposers cannot be shown guilty and must be presumed innocent. In other words, stopping unjustified or dangerous risk impositions, within a purely relativistic framework, is like stopping murders within a crime context in which no one ever finds any bodies. Neither those who fight unjustified risks nor those who fight unjustified homicides can make headway without meeting certain epistemic constraints necessary for evidence. Wildavsky and Douglas -- and other relativists -- are dangerous because they deny the epistemic conditions that are necessary for the possibility of proving risk imposers are wrong.
Once one sees the anti-populist bias of Wildavsky and Douglas in Risk and Culture and of Wildavsky in books such as Searching for Safety, it is possible to see why I am so interested in clarifying the rationality of risk decisions. My theme is that experts do not have a "corner" on rationality. They do not have the last word about risk facts (how safe is safe enough) because risk decisions are also a matter of how safe is equitable enough, how safe is voluntary enough, how safe in informed enough, and how safe is compensated enough. Such considerations are a matter of neither probabilities nor contexts but instead a matter of fundamental and inalienable human rights. I describe the risk debate in terms of positivist-relativist stances because I am interested in the power behind competing risk claims. I am interested in the way that expert claims to hegemony in the risk debate ignore rightful claims of the public, claims that often focus on whether people can trust the experts. Talk about probabilities and contexts -- and not also about positivists (who claim always to be right) and about relativist (who claim no view of risk is more correct than another) -- does not capture the essential risk dispute over power. This controversy is over whether alleged risk experts have the right to ignore concerns of potential risk victims.
Thompson and Dean miss the "blood and guts" aspects of risk debates, the concerns about whose ox is getting gored. Their labels and their account completely ignore the anti-populist bias in the work of Wildavsky and Douglas. Indeed, their very words seem to indicate Thompson's and Dean's own distrust of the public. Their opening paragraph asserts that "experts who study risk or risk issues are more likely to develop well-defined, internally consistent conceptions of risk than members of the lay public."[36] Thompson and Dean, however, seem to have too much faith in the experts and too little faith in the public. Contrary to their claims, experts may not develop well-defined conceptions of risk because they often ignore important social, political, and ethical concerns of the public. These include rights to bodily security, rights to equal protection, rights to full compensation for risks, rights to free informed consent, rights to avoid risk-related violations of due process, and experts' tendencies to pass off subjective risk probabilities as "science." Experts' tendencies to define risks narrowly and quantitatively does not mean that their definitions are well defined. Nor do their reductionist emphases on a few quantitative aspects of risk mean that their accounts are more "internally consistent" than those of laypeople.[37] Prima facie, there are grounds for being suspicious of any authors who denigrate lay views of risk and who then proffer their account as "more accurately" representing the risk debate.[38] Even, and especially, risk victims have the right to be wrong. Those who bear the burdens have the right to accept or reject the load.
* Dr. Shrader-Frechette is Distinguished Research Professor at the University of South Florida. She earned degrees (Mathematics and Physics) from Xavier University and her Ph.D. (Philosophy) from the University of Notre Dame.
[1]Competing Conceptions of Risk, 7 Risk 361 (1996).
[2] 7 Risk 383.
[3] 7 Risk 377, 367.
[4] 7 Risk 369.
[5] Id.
[6] Kristin Shrader-Frechette, Risk and Rationality, 39 (1991).
[7] 7 Risk 368-369.
[8] Risk and Rationality, 7.
[9] Risk and Rationality, 39.
[10] Risk and Rationality, 39ff., 232-235.
[11] Risk and Rationality, 232.
[12] Risk and Rationality, 89-99.
[13] 7 Risk 377.
[14] Risk and Rationality, 39.
[15] 7 Risk 361-384.
[16] Risk and Rationality, 8.
[17] 7 Risk 366.
[18] Id. at 366.
[19] Id.
[20] Risk and Rationality, 31-39.
[21] Mary Douglas & Aaron Wildavsky, Risk and Culture, 186 (1982).
[22] Id. at 187, 191.
[23] Id. at 71.
[24] Id. at 73.
[25] Id. at 194.
[26] Id. at 187, 191.
[27] Id. at 187-188.
[28] Id. at 86-87, 98.
[29] Risk and Rationality, 39ff., 232-235.
[30] Risk and Culture, 5-7, 167, 182, 198.
[31] Id. at 169, 182.
[32] Id. at 190; Aaron Wildavsky, Searching for Safety, 33-34 (1988).
[33] Risk and Culture, 184.
[34] Id. at 37-39, 80-81.
[35] Id. at 46; see Wildavsky, supra note 32, at 33-34.
[36] 7 Risk 362.
[37] See Kristin Shrader-Frechette, Evaluating the Expertise of Experts, 6 Risk 115 (1995).
[38] 7 Risk 377, 367.