Computational Creativity: Art, Natural Science, or Design Science? On the Field's Identity Crisis and the Case for Native, Falsifiable Objects

Every few years the computational creativity (CC) community returns to a question it has never fully settled: is what we do an art or a science? The recurrence is not idle — it is the surface symptom of a deeper identity crisis. This paper names four fronts of that crisis: the absence of a shared definition of the field and its goal; theoretical frameworks that remain, by their authors’ own admission, incomplete; the lack of consensual metrics for novelty, value, and surprise; and a scarcity of native, falsifiable objects. Rather than treat “art or science” as a dichotomy to be won, we argue the question is ill-posed: computational creativity is best understood as a science of the artificial (Simon) — a design discipline whose artefacts can be made with scientific rigour while their reception remains art. We propose not an absolute demarcation but a gradient of design-scientific rigour, and we illustrate, without claiming to resolve the crisis, that native, rigorously designed objects are possible, through a family of mechanisms built on a single primitive: the conceptual fracture. The claim is deliberately modest: science does not replace the human spark; it gives it a laboratory.