Who is the logic for

A subject the papers are about. The loosest grouping, and the one to reach for last.

Four designs for a logic, each settled by naming who will use it. Robinson drops the requirement that an inference step be graspable "in a single intellectual act" because the agent is a machine; Gries and Schneider make every function total but underspecified because "tools should be as simple as possible, if they are to be used reliably"; Makarov prefers separation logic to linear logic because a student already understands heaps; Demri and colleagues refuse labels and nominals for an exhaustive account of expressive power, with "no claim for practical applications". The audiences are a program, a practitioner, a student and a theorist, and each choice looks wrong from the other three.