Hossein Jorjani
First publication: 2026 – 07 – 18
David Hume’s critique of induction is widely regarded as one of the decisive turning points in the philosophy of science. Hume did not deny that human beings routinely generalize from experience, nor did he question the practical success of scientific prediction. His challenge was more fundamental: he argued that no purely logical argument can justify the inference from observed cases to unobserved ones. From the fact that an event has always occurred in the past, it does not logically follow that it must occur in the future.
This conclusion did not eliminate induction from scientific practice. Scientists continued to formulate generalizations, predict future events, and rely upon regularities observed in nature. What disappeared was the hope that such reasoning could be justified by deduction alone. The problem of induction therefore survived precisely because scientific practice survived.
Modern philosophy of science may, in part, be understood as a prolonged series of attempts to respond to Hume’s challenge. Kant sought to secure scientific knowledge through the a priori structures of cognition. John Stuart Mill attempted to codify inductive reasoning into systematic methods. Later logical empiricists, probabilistic philosophers, and Bayesian approaches each proposed new foundations for inductive inference. None, however, restored induction to the self-evident logical status it had enjoyed before Hume.
Meanwhile, scientific methodology gradually developed along a different path. Rather than deriving theories directly from observations, scientists increasingly formulated hypotheses, deduced observable consequences from them, and compared those consequences with experiment. In the hypothetico-deductive approach, observations primarily serve to test explanatory models rather than to generate them. This development did not solve Hume’s problem so much as reduce its centrality to scientific practice.
Consequently, the modern philosophy of science contains two parallel histories. One concerns the continuing search for a philosophical justification of induction. The other concerns the evolution of scientific methodology itself, which has increasingly relied upon hypotheses, models, experimentation, and prediction without requiring such a justification. Understanding this distinction helps explain why philosophers have continued to debate induction long after scientists largely shifted their attention to the testing and refinement of theoretical models.
Hossein Jorjani 