Among contemporary humor theorists, the prevalent view of how a cognitive agent gets a joke is the so-called forced reinterpretation-account (henceforth FR) of incongruity-resolution (henceforth IR). Firstly, the cognitive agent interprets the set-up of the joke (or the "joke body") in a specific way. Then, the punch line violates certain expectations built into that initial interpretation. As a result, the cognitive agent is forced to reinterpret the joke body in a way that resolves the incongruity, i.e., an interpretation that allows the punch line to be congruous after all. Just as contemporary humor theorists claim that a violated expectation or incongruity triggers the problem-solving process involved in getting a joke, philosophers and historians of science assume that a (scientific) discovery process is typically set about by an “unexpected observation”, i.e., an observation that violates certain expectations built into the received view. Given this similarity between humor appreciation and scientific discovery as far as the identification of the problem is concerned, it is reasonable to expect similarities in problem-solving processes as well. However, as an account of humor appreciation, FR faces three related problems: (1) it wavers between rival conceptions of “incongruities”, (2) it exaggerates their causal role, and (3) it falls short in accounting for the actual resolution process. These problems prevent us from applying FR to scientific discovery. Moreover, by exaggerating the causal role of incongruities, FR also eclipses the creativity involved in problem-solving processes of the IR-type. As examples of both humor appreciation and scientific discovery show, IR is not simply a matter of forcing a cognitive agent to reinterpret; it also involves choices of the cognitive agent to follow through on potential solutions to the problem. What is needed then is an account of IR that provides a regimentation of the dynamic interplay of, on the one hand, the “forward chaining” involved in pursuing potential solutions, and, on the other hand, the “backward chaining” not only involved in pinpointing and quarantining the source of the initial incongruity, but also in learning from failed attempts to arrive at a solution. In turn, such an account of IR will help us to develop a theory of creativity that foregrounds “adaptive flexibility” in problem-solving processes, i.e., the ability of agile minds to alternate purposefully and swiftly between (1) following through on potential solutions and (2) redefining the problem in function of the increasing number of constraints that result from analyzing what went wrong when the initial incongruity arose and when expected solutions turned out to be unsuccessful.