"The guy's obvious error is to apply this approach to the scientific realm" Which is, of course, the point. As a rule, new scientific theories must match the

"The guy's obvious error is to apply this approach to the scientific realm" Which is, of course, the point. As a rule, new scientific theories must match the data better than old ones. Over long periods of time, then, science becomes less wrong in its explanation of those things which can be scientifically investigated. The English Lit major's approach may work well with religion or philosophy, but it doesn't translate well to science. There is a minor problem with the "less wrong" approach, though: it depends upon your metric for measuring wrongness, and that choice of metric may have a philosophical basis. Consider a small set of points that nearly follow a quadratic. One might say "this appears to be quadratic with small errors"; another might say "this fits a higher-order polynomial with even smaller errors", yet the first might be conceptually/philosophically a better description of data that was quadratic but measured from a sinusoidally-vibrating platform. A poorly chosen method for measuring wrongness could actually make it more difficult to discover underlying explanations of phenomena.