It ain't necessarily so-or is it?

Scholars who have couched their propositions in the form of necessary conditions owe a great debt to Gary Goertz for his recent work evaluating the status of claims to necessity and developing methods for testing them. In his most recent essay, Goertz makes a powerful case that necessary conditions hypotheses can reasonably be understood as probabilistic, not deterministic. Goertz suggests two benefits from this reconceptualization: necessary conditions hypotheses are more likely to get a fair hearing from social scientists who would otherwise dismiss them as deterministic, and those hypotheses will not be held hostage to single counter-examples that would unfairly be taken as disconfirming when in fact they likely represent measurement error.