The Epistemic Benefit of Transient Diversity

Because of the structure of inductive reasoning, error is an ever-present worry in science. Excluding errors that result from a lack of creativity (where better alternatives remain unconceived), scientific error can be divided into three broad categories. Some occur because of, widely termed, sociological factors. Powerful political parties, cultural biases, or scientific alliances may cloud the judgment (or publication) of scientific results, preventing a superior theory from being accepted. Other errors are the result of scientists’ misconduct; data is invented, hidden, or manipulated and as a result others are convinced to accept an inferior theory. Finally, errors may be the result of simple, good faith mistakes on the part of scientists. In this case, everyone did everything right, but as a result of some random occurrence or another, the data was misleading. These three types of errors can be extremely damaging, especially if an available, superior alternative is dismissed as the result of a scientific error. A discarded theory might never again be revisited; future scientists may remember that this theory was abandoned by their predecessors, and avoid revisiting the decision. A better theory