With very few exceptions philosophers of science have tended to shy away from questions of science and values, science and social responsibility, science and its role in the larger society. So it is good to have a book like John Forge's The Responsible Scientist, especially since this book is forcefully argued, engagingly written, and replete with interesting and timely examples. Indeed, Forge fills a gap even in the slender neck of the woods in which he finds himself. For he sets out to explore the extent and limits of scientists’ responsibility for the outcomes of their research, whether those outcomes are products of technology or the culture-shaping effects of science as a “body of ideas.” But those who have written on related topics—for example, feminist philosophers of science such as Helen Longino and Sandra Harding or philosophers of biology such as Philip Kitcher and John Dupré—have tended to assume rather than argue that scientists are responsible for the outcomes of their research. They were not remiss in this, of course. The examples they dealt with were ones in which the responsibility of scientists for their research outcomes seemed obvious. And so these philosophers moved quickly to the changes in science that needed to be made as a result. Forge's book, by contrast, focuses on physical science rather than social and biological/biomedical science, including “basic” or “pure” physical science. Here the argument for the neutrality of science vis-à-vis research outcomes is far more plausible and the argument for the responsibility of scientists for those outcomes correspondingly more arduous. Forge's goal, nonetheless, is “a unified account of the responsibilities of all scientists engaged in what has been called the ‘research system,’ the system that comprises all forms of scientific research” (4).

The unified account Forge offers is what he calls a “wide” account of scientists’ responsibility: scientists are responsible for the intended outcomes of their research as well as the unintended outcomes they either foresee or should have foreseen. And this holds of all scientists, those engaged in “pure” research as well as those whose research is applied. Indeed, “what distinguishes pure from applied research,” for Forge, “is not primarily the work's character or ‘content,’ but rather where and why it is carried out” (33). Pure research is done “for its own sake,” and its “proper” product is a scholarly scientific paper that attempts to describe the world. By contrast, applied research aims for a technological product beyond the research context. Thus, Forge tells us, the contributions of such scientists as Otto Frisch, Rudolf Peierls, and Enrico Fermi to the Manhattan Project were “applied” since they were aimed at developing an atomic bomb but would have been “pure” had they been done “for their own sake,” out of curiosity. Regardless of whether research is pure or applied, however, it will have outcomes—technological outcomes or ideas that shape our lives—outcomes for which (if they are foreseeable) the scientists doing the research will be responsible. In fact, these outcomes can be thought of as part of the research itself. Frisch and Peierls, for example, can equally well be described as trying to determine certain critical nuclear parameters in early 1940 or as trying to develop an atomic bomb.

So all scientists are responsible for the (foreseeable) outcomes of their research. Two kinds of responsibility are involved here that derive from a “basic and uncontroversial understanding of morality” (136). Scientists’ first—“backward-looking” or “negative”—kind of responsibility is a prohibition: scientists should not, through their research, cause harm. It derives from the moral rule that “at the very least people should not gratuitously harm others, harm considered gratuitous if it cannot be justified” (136). Scientists’ second—“forward-looking” or “positive”—kind of responsibility, on the other hand, derives from the moral ideal that people should strive to prevent harm. Unlike the first kind of responsibility, it is not a requirement but rather something to be encouraged—that scientists should, through their research, prevent harm—a commitment that may involve “a great deal of sacrifice in terms of career and status” (147). Scientists thus have a “two-tiered” set of responsibilities: “scientists are required not to do wrong by causing harm and are encouraged to do good by preventing it” (148).

What follows from this general account of scientists’ responsibilities? Forge emphasizes that his book is “not (or not just) an erudite exercise in philosophy but also something engaged with scientists themselves and scientific practice” (117). So it is entirely reasonable to expect practical results. The most detailed one Forge offers concerns weapons research: “research conducted with the intention of designing or improving weapons systems or the means for carrying out activities associated with the use of weapons, such as command, control, and communications” (153). Indeed, weapons research—especially the research that surrounded the Manhattan Project—is the issue that dominates Forge's book. The conclusion Forge comes to is that no scientist should take part in weapons research “under any circumstances” (153). “My view is that all those involved in the design and manufacture of an artifact are committed to its primary purpose and that this entails responsibility for instances of its use for this purpose; this is surely obvious—an artifact's primary purpose is what it is supposed to do, what its creators intend that it do” (157). But the “primary purpose” of a weapon is to do harm. So scientists’ backward-looking responsibility precludes weapons research.

Realizing that this call for a “complete moratorium on weapons research” (153) might cause surprise, Forge spends a great deal of time developing his case. Thus, he argues against research on “precision” (“inherently more discriminating”) weapons because, he says, such weapons can be used to increase civilian causalities (terror bombing) as well as decrease them. He argues against research on “inherently defensive” weapons because, he says, there are no such weapons: any candidate for an “inherently defensive” weapon can be, and be interpreted by adversaries as, part of the means for an offensive first-strike capability. (Forge reminds us here that the Soviets interpreted President Ronald Reagan's “Star Wars” initiative as a plan to develop a defensive shield to take care of most of the surviving retaliatory missiles after a U.S. first strike on Soviet offensive systems.) He argues against research on weapons for “deterrent” purposes because, he says, any such weapon can be used for offensive as well as deterrent purposes (the “primary purpose” of any weapon, remember, is to harm, and the use of weapons for deterrence depends on this primary purpose). He argues against research on weapons to be used solely by democratic states that espouse human rights, freedom, and other desirable political characteristics because, he says, the historical record shows that even such states cause unjustified harm (the Manhattan Project and its aftermath, the bombing of Hiroshima and Nagasaki—against the wishes of many of the scientists who had signed on to the project in order to deter the Nazis—is only the most dramatic illustration of this). And so on.

Still, Forge's call for a complete moratorium on weapons research leaves something to be desired. After all, by his reasoning there should be a complete moratorium on weapons research for police forces as well as military forces. If the “primary purpose” of a weapon is to harm and there are no such things as “inherently defensive” weapons, weapons used solely for deterrence, and so on, then research to develop or improve on such products as tear gas, pepper sprays, and stun guns should be unacceptable to Forge—just as unacceptable as research to develop or improve on any other kind of weapon. Clearly something is awry, but what? Forge says that the primary purpose of a weapon is to harm—which is simply part of the meaning of the term weapon—and hence that purely defensive weapons, deterrent weapons, and so on do not exist. But arguing in this way begs the question against such weapons, whose primary purpose (according to the meaning of defensive weapon, deterrent weapon, etc.) would be to defend and protect rather than harm. Indeed, Forge says that “an artifact's primary purpose is what it is supposed to do, what its creators intend that it do” (157; quoted previously). But weapons scientists could intend (surely many have intended) that their creations be used as defense weapons or as deterrents, and more important, these weapons might actually be used over time in just these ways. Similarly (to take another of Forge's examples), “precision” weapons could be designed—have been designed—to decrease civilian casualties, and this is also the way they have been used. Hence, this has been their “primary purpose” even though they could have been used to increase civilian casualties. Of course, using them (or tear gas or the rest) or doing the scientific research that makes this possible surely ends up doing harm. But the relevant question is whether the harm is justifiable—for example, whether the harm is done in self-defense or to prevent far greater harm, whether it represents the least amount of harm that will suffice, and so on.

The upshot: we simply cannot say, without a wealth of additional information regarding particular weapons in particular contexts, that scientists should not take part in weapons research—Forge's main case study in The Responsible Scientist. No matter. Forge has provided a framework with which such questions can now more fruitfully be pursued.