Discovery, Generalization, and Assimilation: DGA as Transformations of Scientific Knowledge

抄録

Scientific knowledge does not become part of humanity’s durable capabilities merely by being discovered. This paper distinguishes three transformations of scientific knowledge: Discovery, Generalization, and Assimilation (DGA). Discovery increases the resolution of human knowledge; Generalization extends existing knowledge non-obviously beyond its previously established or ordinarily implied scope so that it becomes functional for a new problem; and Assimilation incorporates knowledge into humanity’s established repertoire of capabilities. The framework is reconstructed from three research histories: SARS-CoV-2 genomic phylogenetics, Liberality as a quantitative description of cellular state, and electrical measurement of cells. In the SARS-CoV-2 case, genomic phylogenetics, already used to reconstruct origin and transmission history, was connected to the socially salient problem of border-control effectiveness, allowing recurrent independent introductions to generate a conclusion about intervention feasibility. DGA is therefore not a taxonomy of research outputs or a sequence of maturity stages, but a framework for describing historically situated transformations of knowledge. The paper further proposes that Generalization can alter the search space for subsequent Discoveries and may increase the probability of Assimilation, and that large language models may reduce the search cost for candidate Generalizations.