Some discussions in moral philosophy revolve around asking what the discussants’ intuitions tell how the questions should be answered. For instance, in a discussion about whether it is wrong to directly take the life of a terminally ill patient in pain as opposed to letting the patient die, some report that their intuitions are that the direct killing is wrong while others report their intuitions are contrary. The interminable “trolley cases” fall into this pattern. What’s the pattern? A case is invented wherein there is a forced choice between violating a standing moral rule such as: Don’t kill! Don’t lie, and bringing about an outcome with less suffering by violating the rule.

These discussions go on-and-on because many of us have intuitions supporting both sides. What are these intuitions? These inuitions are based on our innate knowledge of how to use moral language – how to think morally. I call this knowledge “semantics of moral language.” In the semantics of moral language the direct violation of a moral law is to proclaim that some harm ought to be brought about. This is the thought of retribution. So thinking of violation a moral law is to think that there ought to be some harm. Thinking that some harm ought to be brought about is prima facie not a proper way to think.. On the other hand, in the semantics of moral language thinking that some preventable suffering need not be prevennted is prima facie not a proper way to think. The invented examples do not allow for going beyond prima facie thought. The semantics of moral language will not settle the case. So these discussions continue throughout the ages.

Here my goal is not to provide a recipe for how to settle these cases. My purpose is to diagnose these continuing discussions as based on the semantics of moral language and most importantly:the notion of retribution is in the semantics of moral language.