Bioethics Research and the Language of Methodological Uncertainty

The editors of The American Journal of Bioethics are to be congratulated. By publishing such an overtly flawed paper and recruiting experts in law, medicine, and the social sciences to critique it, they were either so clever as to foresee that it would evoke discussions about analysis and argument in bioethics, or bold enough to judge that such a set of discussions would increase readership and thereby perhaps, draw renewed attention to the content and conduct of scholarly work in the field. I read William P. Cheshire's paper (2004) as a call to bioethicists to be mindful of rigor and quality in their representations as much as a story about a "skewed linguistic distribution" in journalism concerning the moral status of embryos. Bioethics turned a corner a few years ago with the growing acceptance of empirical work. For those of us who choose to conduct empirical research, it is important to advocate that it be done well. Such studies are certainly not the only way to provide insight into complex moral dilemmas, but if done properly they can be powerful tools with which to shed light on actual practices and relationships in science and medicine and their ethical and social implications.

My fellow commentators have done an excellent job in pointing out the methodological and conceptual problems with Cheshire's paper and the study it reports. I need not rehearse them here, other than to underscore the key points that the assumptions on which the study was based were faulty, the lack of context meant that terms could be multiply interpreted, and the absence of explanation about methods and choice of sources leaves the author open to criticism. A number of scholarly works have been written on the role of media in public disputes, as have studies of the degree of effect of media coverage on public understanding of scientific controversies or on decision making (e.g., Nelkin 1995; Friedman, Dunwoody, and Rogers 1999; Priest 2001). These could have guided the author in his understanding about the media and their use of terminology and framing. However, none were referenced.

Most disconcerting, however, is the way studies such as this become persuasive by virtue of the way in which the information is represented—in this case appearing to be valid simply because it is quantified. As social scientists have recognized for many years, representations are social facts—that is, knowledge claims and the use of statistics, images, and other analytical and narrative devices take on an air of authority and reality when circulated in certain ways. In the stem cell debates biomedical ethicists have been placed in a position of authority relative to possible alternative sources of expertise, such as social scientists, advocacy groups, religious authorities of various persuasions, and others. As such, it is critical that commentators on issues of immediate social, scientific, economic, and moral concern take care in the way they tell their stories, collect their "facts," and construct representations to each other and to public audiences. Cheshire has a right to express his views that human embryos should have a special status and that they should not be used as a source of stem cells. Yet he should have stated his position explicitly from the beginning, as it is common in biomedical ethics and in social research to do. To attempt to veil one's stance in pseudoneutrality as he did by creating a study that illustrates his opinion, is disingenuous and shows a disrespect for readers. More important, as Green (2004) points out, it is pernicious to disseminate his opinion in the form of pseudoscientific "data." Quantitative measures of various sorts are often used to convince readers that a study is well-documented. "Four out of five dentists recommend" sugarless gum is an old example of a believable but unfounded "fact."

A great deal is at stake. The way science, medical therapies, regulatory guidelines...