Pretty-Big-Step Semantics

In spite of the popularity of small-step semantics, big-step semantics remain used by many researchers. However, big-step semantics suffer from a serious duplication problem, which appears as soon as the semantics account for exceptions and/or divergence. In particular, many premises need to be copy-pasted across several evaluation rules. This duplication problem, which is particularly visible when scaling up to full-blown languages, results in formal definitions growing far bigger than necessary. Moreover, it leads to unsatisfactory redundancy in proofs. In this paper, we address the problem by introducing pretty-bigstep semantics. Pretty-big-step semantics preserve the spirit of big-step semantics, in the sense that terms are directly related to their results, but they eliminate the duplication associated with big-step semantics.