Belief, reality, and self-deception in environmental and occupational health.

“The first rule of science is that you must not fool yourself-and you are the easiest person to fool.”Richard Feynman’ The desire to make a fundamental contribution drives all scholars-scientists more than most-but the great motivating desire to discover and innovate also renders the originator of a new theory the worst judge of its validity. Self-deception is the easiest error in science. One looks at one fact, matches it with another, links to something relevant that someone wrote several years ago, and voiM: A new theory of the universe. It all makes compelling sense to the creator, and it is very hard for the believer to abandon even if it is proven wrong. However, this used to be the way that early scientists thought until the invention of rigorous, disciplined experimentation in the late Middle Ages and the widespread adoption of hypothesis testing and reproducibility of findings as the norms by which scientists conduct studies. In Europe, in the 18th century, this systematic way of knowing crowded out the many personal theories and traditional explanations that contemporaries called, and historians of science today call, ”systems.”2 Medicine came late to the scientific method, perhaps because it developed early as a compendium of astute empirical observations united by various idiosyncratic theories. Medicine was dominated by systems until the rise of scientific medicine in the German universities of the 19th century and their subsequent influence when the scientific method began to bring reason and validity and gave birth to modern biomedical science.2 The belief in spontaneous generation gave way to the continuity of life through the experiments of Pa~teur,~ and the patchwork of therapeutic systems gave way to a rational understanding of how drugs work, which also indirectly led to modern toxi~ology.~ Even so, medicine remains plagued with untestable explanations, hypotheses, and nonscientific systems, as demonstrated by the long persistence through the 20th century of psychoanalysis and the continuing popularity of homeopathy. Science progresses in a methodical way by setting up hypotheses that can be found false whether the discipline is inherently experimental, such as chemistry and toxicology, or inherently observational, such as astrophysics and epidemiology. The test of a science rests on whether a discipline involves making predictions based on a theory and testing those predictions against observations. This is the doctrine of “falsification,” as elaborated by Popper and other students of ~c ience.~ A theory is not a science nor does a science prove a theory by working backward from an observation to string together a mechanism to explain it. Science also does not prove a theory by searching out facts that appear compatible. In truth, scientific theories cannot be proven at all, only disproven. A science does not rest on the plausibility of its argument, but is valid because the body of knowledge and i ts theory of explanation can sustain rigorous testing. In other words, a scientist cannot persuade other scientists or find objective truth simply by compiling observations and making an argument that they are connected. That way leads to error because it is far too easy for scientists to pick and choose among the literature and the data to bolster their preconceived notions and to ignore or fail to search for the evidence that does not. The literature of environmental and occupational health is filled with examples, ranging from the miasma theory of disease that continued into the late 19th century to the notion that the introduction of method tertiary butyl ether (MTBE) caused the current worldwide rise in the incidence of asthma because it happened to coincide with the first documented rise in Philadelphia, a belief recently expressed to me in an informal conversation as “fact.” In every generation there are those who believe in an alternative science of wishful thinking. At times, stories seem so persuasive to some people that they conclude that the stories must be true. The presentation of selected, misleading, and biased data as evidence, the piling on of extraneous and often irrelevant detail, and the recreation of forms that resemble scientific papers but, in reality, have no scientific design, all obscure the absence of real data and misapplication of the scientific method, resulting in the accretion of extraneous detail mistaken for proof. Science does not recognize proof; it recognizes only that a theory has not been proven false and is, therefore, a convenient working explanation.