Abstract
Abstract Philosophical discussions of experiment usually focus exclusively on testing predictions. In this paper I compare G. Morpurgo's experimental test of the Gell?Mann/ Zweig quark hypothesis with two neglected uses of experiment: constructing representations of new phenomena and inventing the instruments that produce such phenomena. These roles are illustrated by J. B. Biot's 1821 observations of electromagnetism and by Michael Faraday's invention of the first electromagnetic motor, also in 1821. The comparison identifies similarities between observation and experiment, showing how both observation and experiment actively engage the natural world and how each engagement shapes representation and subsequent empirical work. This challenges the post?empiricist assumption of the sufficiency of knowing only the outcomes of experiments. I conclude that traditional views of observational access have looked in the wrong place for empirical constraints on theorizing. The active character of observation implies that a realist interpretation of experimenters? discourse should be grounded in the fine structure of experimental practice rather than the supposedly decisive, golden events favoured by hypothetico?deductive methodology