Abstract
There has been a reappraisal of phylogenetic issues in cognitive science, as reconstructing cognitive phylogenies has been considered a key for unveiling the cognitive novelties that set the stage for what makes humans special. In our opinion, the studies made until now have approached cognitive phylogenies in a non-optimal way, and we wish to both highlight their problems, drawing on recent considerations in philosophy of biology. The inadequacy of current visions on cognitive phylogenies stems from the influence of the traditional "linear cladograms," according to which every seemingly new or more sophisticated feature of a cognitive mechanism, viewed as a novelty, is represented as a node on top of the old and shared elements. We claim that this kind of cladograms does not succeed in depicting the complexity with which traits are distributed across species and, furthermore, that the labels of the nodes of these traditional representational systems fail to capture the "tinkering" nature of evolution. We argue that if we are to conceive of cognitive mechanisms in a multi-dimensional, bottom-up perspective, in accordance with the Darwinian logic of descent, we should rather focus on decomposing these mechanisms into lower-level, generic functions, which have the additional advantage of being implementable in neural matter, which ultimately produces cognition. Doing so renders current constructions of cognitive phylogenies otiose.