A Classification of Scientific Evidence in Geology

Abstract

All facts are not equal, nor are they equally theory-laden. Facts can be separated into a three-level, five-part classification. The three levels are first-, second-, and third-order evidences, going from most certain to least certain evidence. First-order evidence may be either a simple fact (one based directly on a measurement of length, time, or mass, or on an expert's identification) or a derivative fact (an induction from simple facts). Second-order evidence may be a simple inference (from insufficient first-order evidence) or a derivative inference (from the cumulative indication of simple facts, derivative facts, and simple inference). Third-order evidences are hypothetical explanations, known as theories to their proponents. Scientific examples of the five classes are as follows: Elevations, clast sizes, and rock types employed by J Harlen Breli to explain the Channeled Scablands are simple facts. The understanding of the Basin and Range of western North America as an extensional terrane is a derivative fact. The ocean bottom topography on maps by Heezen and Tharp are simple inferences. The thicknesses of the crust under Lake Mead or Lake Bonneville are derivative inferences. Many theories are third-order evidences, although successful theories can graduate to become first-order derivative facts, with sufficient first-order support. This classification developed from close readings of Oreskes's The Rejection of Continental Drift and Victor Baker's essay-review of that book in Earth Sciences History.