Inductive generalization is the primary way by which human beings arrive at the construction of knowledge. Usually, it is assumed that it operates in a linear manner-each new feature becomes "piled up" in the inductive accumulation of evidence. We question this view, and otherwise claim that inductive generalization is essentially a non-linear dynamic process that fits the theoretical premises of the Dynamic Systems Theory. In our study, we explore the ability that young infants have when making inductive generalizations -previous studies show the existence of this capacity not earlier than at the age of 14 months. These studies have been cross-sectional in nature, but they do not offer an answer to the question of emergence of cognitive capabilities, therefore, a short-term longitudinal study is needed. Based on 3 case studies carried out longitudinally in infants ranging from 9 to 14 months, we demonstrate how the process of inductive generalization occurs from a conceptualization of nonlinear dynamic systems. We use Min - Max and State Space techniques, which allow us to show how the infant uses diverse pathways of actions with everyday objects to facilitate inductive generalization. The identified paths are not the same, they present differential and common moments that confirm the dynamic nature of development, and provide empirical evidence on the emergence of non-linear, non-sequential or inductive generalization.