Nuclear Questions, Unclear Answers


Nuclear Power and Public Policy: The Social and Ethical Problems of Fission Technology, by K. S. Shrader-Frechette, Boston: D. Reidel, 1980, 176 pp., $19.95/$10.50.

Working under grants from the Kentucky Humanities Council and the federal Environmental Protection Agency, Ms. Shrader-Frechette says she discovered "critical lacunae in allegedly comprehensive analyses of various energy technologies." Specifically, she is concerned about nuclear power. Declaring an attempt to avoid "overly emotive and question-begging" negative assessments and "unsubstantiated assertions" by many environmentalists, she nevertheless argues that nuclear fission has been promoted using faulty scientific methodology and logical fallacies and concludes that it violates due process and equal protection rights.

"This is applied philosophy at its very best," says another philosopher, Michael Ruse, on the dust jacket. Unfortunately, this may be true. Shrader-Frechette is significantly guilty of begging the question herself when she maintains that "it is generally admitted that any amount of radiation increases the probability of cancer and genetic damage." She presents the same presumption, the heart of the so-called linear hypothesis, at least eight more times in her book without ever mentioning the alternative "threshold" hypothesis, which posits that there is no harm below a certain level (see my article, "Is Any Radiation Too Much Radiation?" REASON, July 1980, and letters and reply, Oct. 1980).

Did she bother to consult one medical article in her research? She does not cite any in her entire book. In fact, there is an ongoing debate on this issue. Moreover, even if there were not, since she observes in a footnote that "dose-response coefficients for low-level radiation are obtained by hypothesis from data on high levels of exposure" (emphasis added), she abdicates the philosopher's role by failing to examine carefully the foundations (or lack of them) for such scientific hypotheses in the philosophy of knowledge.

After begging this important question, she goes on to ask, "According to what moral principles…ought society to determine when the carcinogenic and mutagenic costs of low-level radiation have outweighed its technological benefits? What number of cancers and genetic deaths ought society to trade for the economic advantages of not requiring tighter radiation controls?" Assuming the truth of the linear hypothesis, she correctly argues that government policymakers are engaged in an unjust "utilitarian" analysis of maximizing the greatest good for the greatest number at the expense of some who pay in the form of health costs.

This is certainly true (given the linear hypothesis). She fails to note, though, that such a utilitarian policy is neither new nor unique to nuclear energy policies, since it has been used since at least the 1880s to justify diminishing due process of law in court cases ranging from property taxation to air pollution from coal-burning plants.

She then maintains that the health costs are borne inequitably, which violates equal protection rights, since the government sets up the same radiation exposure limit for everyone. But, she correctly notes, some persons are at higher risk during normal nuclear plant operations (again, assuming the linear hypothesis)—namely, children, those living closest to the plant, and those with previous medical X-ray exposures.

It may well be true that this argument can be made out, but she does not do so, since, in particular, she does not deal with the usual legal interpretation of "equal protection." As Black's Law Dictionary quotes a leading California case, the constitutional guarantee of "equal protection of the laws" means that "no person or class of persons shall be denied the same protection of the laws which is employed by other persons or other classes in like circumstances," (emphasis added). Thus, again, she fails to provide philosophical argument for why her interpretation of "equal protection" is more moral or rational than the one used in law today.

Shrader-Frechette's "arguments" about knowledge are even more philosophically undernourished than her arguments about ethics. Her primary epistemological argument is that, since the Price-Anderson Act requires a person to prove that his injury was due to a catastrophic accident, "it is likely that some injured persons will be unable to prove this and hence likely that they will remain uncompensated." While neglecting to mention that this is true throughout the law, she goes on to maintain that "in principle, there is no way to prove whether or not radiation from a particular plant has caused a cancer." Hence, placing the burden of proof on the accuser "forces…the courts to employ…an argument from ignorance. They will have to conclude that an accident did not cause a particular cancer whenever the victim is unable to prove that it did."

Since she never says what the "argument from ignorance" is, I will spell it out here. It is a logical fallacy held to be committed by someone who maintains (1) that a statement is true simply because it has not yet been shown to be false or (2) that a statement is false simply because it has not yet been shown to be true.

But as Irving M. Copi correctly notes in his widely used Introduction to Logic, (2) is not held to be a fallacy in a court of law. Since a person is presumed to be innocent until proven guilty, the burden of proof is placed solely on the accuser. Unless the plaintiff can show that the defendant harmed him, the court presumes that the statement that he did harm him is false, simply because the plaintiff has failed to show otherwise.

It is simply not true that this places an impossible burden on the plaintiff and thus violates due process and equal protection. Given that Shrader-Frechette's notion of "proof" is not the standard philosopher's call for omniscience, some cases of injury can be proved, as she herself notes ("proof of injury is not difficult for the more serious and immediate types of radiation and deaths and injuries").

But what about exposures to lower doses, which can have effects with latency periods of up to 20 or 30 years? I grant that this is a difficult problem, but I also maintain that the "more impossible" task would be an owner attempting to prove that his nuclear plant did not cause cancer in all of the people who contracted cancer during the lifetime of his plant and 20-30 years thereafter.

Moreover, "differential diagnoses" can be and are made all the time (for example, injuries from medical X-ray overdoses, or probability estimates for lung cancer due to smoking). And since the types of radiation emitted from nuclear plants affect specific organs, their effects can be distinguished, for instance, from sunlight, which causes only skin cancers. (I also think further "reasonable" distinctions can be made, and "reasonably certain" causes arrived at.)

Since she does not recognize such possible objections, she fails to begin to provide the necessary epistemological argument that would require us to give up a basic legal principle like presuming innocence, which has been adopted in every civilized country since early Roman law (but which has increasingly come to be ignored by government agencies, including the IRS, FTC, SEC, FDA, EPA, and OSHA). The alternative to giving up the presumption of innocence is to decide what "statistical" evidence counts as "proof" and perhaps allowing class action suits and compensation based on probabilities of harm, all the while keeping the initial and ultimate burden of proving harm with the plaintiff.

Unfortunately, the objections and possible alternatives I have raised do not point so much to the failings of this specific work as to the impoverished state of philosophy in general (and epistemology in particular). If this is, indeed, "applied philosophy at its very best," then philosophy has not even begun to deal with the legal and scientific concerns of today.

Bruce K. Bell has a B.S. in physics and an M.A. in the history and philosophy of science. He is currently working on a book on chemicals.