The Language of Mens Rea

To be guilty of a crime, generally one must commit a bad act while in a culpable state of mind. But the language used to define, partition, and communicate the variety of culpable mental states (in Latin, mens rea) is crucially important. For depending on the mental state that juries attribute to him, a defendant can be convicted-for the very same act and the very same consequence-of different crimes, each with different sentences. The influential Model Penal Code ("MPC") of 1962 divided culpable mental states into four now-familiar kinds: purposeful, knowing, reckless, and negligent.' Both before the MPC and since, scholars in criminal law and philosophy have actively debated how best to define and apply the mens rea categories.2 Yet few empirical studies have explored the actual relationships between specific mens rea formulations and legally relevant outcomes. A 2011 article coauthored by several of us, Sorting Guilty Minds, presented experiments that tested the MPC's twin assumptions that: (1) ordinary people naturally do-or at least can, when instructed- distinguish these four categories of mental states with reasonable reliability; and (2) holding the act and harm constant, the average person would punish acts reflecting these four mental states in the manner corresponding to the MPC hierarchy-that is, they would punish purposeful conduct the most severely and negligent conduct the least. Those experiments found that these assumptions held, for the most part. But an interesting and important exception emerged at the boundary between knowing and reckless conduct: in sorting the mental states and in assigning punishment, subjects were much less able to differentiate between knowing and reckless conduct. On the basis of those findings, the article outlined several possible reforms-assuming the results were validated in future studies. To validate and extend those results, we have conducted a series of additional experiments, reported here, with more than 1,600 additional subjects.