I’ll introduce a simple but general notion of a ‘translationally FLC’ space of (Euclidean) patterns. This covers standard cases of interest, including translational hulls of FLC tilings and uniformly discrete (but possibly non-relatively dense) point sets, and more. Extending familiar structures from hulls generated by a substitution rule, we say that such a space $X$ is ‘substitutive’ if $LX$ maps to $X$, surjectively, under some local derivation map. We show that, even without minimality, we may extend recognisability results of Solomyak to this setting, and even classify the patterns with unique pre-images in terms of non-periodicity (for an appropriate power of substitution). In the minimal case, such structures, up to MLD, arise as hulls of primitive substitution patterns. This viewpoint adds some clarity when asking questions such as ‘which cut and project (c&p) sets are substitutive’? Time permitting, I will explain recent joint work of Edmund Harriss and Henna Koivusalo which answers this, in simple terms of the c&p data, for schemes with Euclidean total space. We get a particularly simple necessary and sufficient condition when the acceptance window is also chosen to be polytopal.