In the common law tradition, criminal conviction depends on both actus reus (a harmful consequence and mens rea (the intent to harm). Young et al. (PDF here) set up an experimental test using different small stories to demonstrate that a subject's belief that he/she has caused intentional harm causes a larger increase in the activation of a region of the right temporal parietal junction (RTPJ in the figure.) than attempted harm, unknowing harm, and neutral effect.) The study suggests that moral judgments depend on the cognitive processes mediated by the RTPJ, previously associated with belief attribution, and, to a lesser extent, the PC, LTPJ, and MPFC, which compose a network of brain regions implicated in theory of mind.
The discussion of the paper is well worth reading. Here is one clip:
The current results also reveal an asymmetry between moral judgments of incompetent criminals (whose false beliefs prevent intended harm from occurring) and unlucky innocents (whose false beliefs lead them to cause unintended harms. Judgments of incompetent criminals were harsh, made on the basis of beliefs alone, and associated with enhanced recruitment of circuitry involved in belief attribution. By contrast, unlucky innocents were not entirely exculpated for causing harm on the basis of their false beliefs. Instead of showing an increased response in brain regions associated with belief attribution, whole-brain analyses revealed recruitment of brain regions associated with cognitive conflict: right inferior parietal cortex, PC, bilateral middle frontal gyrus, and bilateral anterior cingulate sulcus. All of these regions have been implicated in cognitive conflict associated with moral dilemmas, specifically where subjects endorse emotionally salient harmful acts to prevent greater harm. Here subjects had to override judgments against harm in favor of utilitarian considerations (e.g., the greatest good for the greatest number). Analogously, in the context of unknowing harm, subjects may partially override judgments against harm to exculpate agents on the basis of their false beliefs. Moral judgment may therefore represent the product of two distinct and at times competing processes, one responsible for representing harmful outcomes and another for representing beliefs and intentions.