O'Brien, Braine, and Yang argue that the mental model theory of propositional reasoning is easy to refute, and they report 3 experiments that they believe falsify the theory. In contrast, Bonatti argues that the model theory is too flexible to be falsified. We show that O'Brien et al.'s experiments do not refute the model theory and that Bonatti's claims are ill founded. Formal rule theories of propositional reasoning have 3 major weaknesses in comparison with the model theory: (a) They have no decision procedure; (b) they lack predictive power, providing no account of several robust phenomena (e.g., erroneous conclusions tend to be consistent with the premises); and (c) as a class of theories, they are difficult to refute experimentally.