Scientific Creativity as a Combinatorial Process: The Chance Baseline

Abstract

The goal of this chapter is to formulate a theory of creativity that uses parsimonious assumptions and logical derivations to obtain comprehensive explanations and precise predictions with respect to the most secure empirical results regarding the phenomenon. In short, the plan is to get the most with the least. The specific formulation is founded on a two-part argument. First, I argue that combinatorial models fulfill these strict requirements. That is, models based on combinatorial processes make the fewest assumptions and by logical inferences explain the widest range of established facts as well as make the most precise predictions with respect to those data. Second, I argue that even if combinatorial models are incomplete from the standpoint of one or more criteria, such models must still provide the baseline for comparing all alternative theories. That is, rival theories must account for whatever cannot be accounted for by chance alone—or what exceeds the chance baseline. This position closely parallels the concept of null hypothesis significance testing in statistics, in which researchers must demonstrate that the discovered effects, whether mean differences or correlations, exceed what could be expected by chance alone (Simonton, 2007). The rationale for this view follows from the standard scientific principle known as “Ockham's razor,” or the “law of parsimony.“ Scientists should prefer a simple explanation over a complex explanation when the former suffices to explain the data.