Abstract

Contemporary accounts of laws of nature face two recurrent problems. On the one hand, some accounts capture what is distinctive about laws---their nomic necessity and their support of counterfactuals. However, they do this at the cost of making obscure how we are ever able to justify a claim that something is a law. On the other hand, there are accounts that seem to give us epistemic access to laws, but these appear unable to capture their distinctive characteristics. ;I resolve this dilemma by providing a new account of laws. The natural way to express the modal character of laws, including the counterfactuals that they support, is to employ possible worlds semantics. On all available ways of understanding this semantics, however, the dilemma arises. I propose a new solution that treats possible worlds in the Carnapian spirit as maximal sets of propositions. What is distinctive about my account is the relationship I draw between modal, higher-order, and first-order facts. Various kinds of modal facts are generated by higher-order constraints on the sets of first-order propositions that are possibly true. Nomically necessary truths hold in virtue of a particular type of higher-order fact: the relations among properties; and the nomically possible worlds consist of all those combinations of first-order propositions that are compatible with the constraining relations among properties. ;By understanding nomically necessary truths as relations among properties, it becomes clear how the routine practices of scientists---observation and experiment---can lead to knowledge of laws. These methods allow scientists to explore the relations among properties by altering and investigating what combinations of properties are present in various samples. Scientists, in effect, create and explore the nomic possibilities. In so doing, they are able to discover the nomically necessary truths and, hence, the laws. We thus achieve an account of laws on which the distinctive modal characteristics can be naturally understood, and at the same time clarify how we can justify our beliefs about them by just the sort of methods scientists have always employed. Perhaps God did not play dice with the universe, but we can