Textbooks and other popular venues commonly present science as a progressive “brick-by-brick” accumulation of knowledge and facts. Despite its hallowed history and familiar ring, this depiction is nowadays rejected by most specialists. Then why are books and articles, written by these same experts, actively promoting such a distorted characterization? The short answer is that no better alternative is available. There currently are two competing models of the scientific enterprise: reductionism and antireductionism. Neither provides an accurate depiction of the productive interaction between knowledge and ignorance, supplanting the old metaphor of the “wall” of knowledge. This book explores an original conception of the nature and advancement of science. The proposed shift brings attention to a prominent, albeit often neglected, construct—the black box—which underlies a well-oiled technique for incorporating a productive role of ignorance and failure into the acquisition of empirical knowledge. What is a black box? How does it work? How is it constructed? How does one determine what to include and what to leave out? What role do boxes play in contemporary scientific practice? By detailing some fascinating episodes in the history of biology, psychology, and economics, Nathan revisits foundational questions about causation, explanation, emergence, and progress, showing how the insights of both reductionism and antireductionism can be reconciled into a fresh and exciting approach to science.