An evaluation strategy, described in [Traub, FPCA 89], under which all redexes are evaluated in parallel except inside the arms of conditionals and inside lambda abstractions. This is an example of an eager evaluation strategy.