What constitutes good science in environmental and occupational health?

Abstract

In the last issue, I addressed the role of beliefs and suppositions in (mostly) environmental health sciences. In this issue, I explore what constitutes good science in general and in the sciences of environmental and occupational health. I do this in preparation for a major change in the editorial philosophy of this journal and as guidance for future submissions. This journal is now being taken in a new editorial direction. Archives of Environmental Health will change with the next issue. The visible manifestation of this change will be a change in name to Archives of Environmental & Occupational Health, which is intended to signal a transition in editorial management and philosophy. It is also intended to return occupational health to a position of prominence in the journal. Occupational health was the original emphasis of this journal when it was founded in the 1940s. However, the changes go deeper than the name. The peer-review process has been strengthened and the standards for scientific evaluation will return to a world standard and reflect the historical tradition of this venerable journal. The next issue will mark a transition to a new and, one hopes, increasingly valuable publication. Before making this transition, however, it i s useful to reflect on the editorial board’s expectations. What constitutes good science in environmental and occupational health? The short answer is the same qualities that make for good science in general. However, how these qualities are manifest depends on the subject matter, how readily experimental or quasi-experimental design may be imposed, and the tools available for investigation. Environmental and occupational health consists of (or, more correctly, i s artificially divided into) many subfields, including environmental epidemiology, exposure assessment, occupational epidemiology, occupational hygiene, toxicology, environmental genetics and toxicogenomics, ecotoxicology, ergonomics, demographic studies of the workforce and critical industries, environmental justice studies, environmental health policy, occupational health policy, immunology applied to issues relevant to this field, physiology, studies of psychogenic stress, ecosystem and human health, risk science (which is itself a cluster of disciplines), and all the various disciplines and subdisciplines involved in monitoring and understanding media quality. Do these fields, and others not named here, some of which have no name, have anything in common that would lead us to formulate criteria for recognizing good science? It is probably easier to recognize “junk science” than to establish criteria for good science. The reason is that good science is open-minded and good and creative scientists entertain any number of ideas that they later discard or disprove. A habit of mind of excellent scientists seems to be to consider numerous ideas, some of them wildly implausible, before settling on the ones that pan out. At the same time, excellent scientists usually do not invest too much emotion or faith in any one idea until it has been tested against sufficient evidence and has not been contradicted. However, sometimes many scientists do commit to an idea prematurely, even though they may do otherwise throughout most of their careers (j. B. Haldane’s commitment to the idea that the lung secretes carbon dioxide comes to mind’). The trick is knowing when there is sufficient evidence to turn an idea into an operating theory. Good science begins with inductive reasoning. One observes phenomena and comes up with an idea of how the world works. This is a creative act, which puts scientists on a par with artists. Indeed, during the European Renaissance, art led science in the revival of knowledge, in part because sharp observation of reality came to be newly appreciated. A newly conceived idea or theory has a bewitching grace, full of coherence and satisfying detail. Beauty in theories, as in people, is always appealing but often misleading. To paraphrase the common saying, beautiful theories usually-not always but usually-become hapless victims of brutal facts. Haldane himself once said “...the universe is not only queerer than we imagine, but queerer than we can imagine.”’ When one’s imagination is seduced by the beauty of a theory, one can no longer see the universe as it is, in all its quirkiness and “queerness.” The test of a theory has nothing to do with its aesthetic qualities or whether it seems strange. The test of a