Abstract
Historical contingency is commonly associated with unpredictability and outcome variability. As such, it can be seen as an undesirable aspect of experimental investigations. Many might agree that experimental methodologies that include enough control help to by-pass this problem and thereby make for more secure knowledge. Against this received view, we argue that, for at least some historically contingent processes, an over-emphasis on control might mislead by obscuring the very object of investigation or by preventing fruitful discoveries. In discussing cases from evolutionary biology, developmental biology, and geochemistry/astrophysics, we show how investigating through approaches that don’t prioritize environmental control, while allowing for greater variability of outcomes, better respects the object/environment entanglement of these systems. Finally, we defend the idea that, despite the lower level of control, these types of experiments do not have a lower epistemic value than more highly controlled experiments.