Tracking objects, Tracking agents
Résumé
Animals and humans have to keep track of individuals in their environment, both in perception (sensorimotor tracking) and in cognition (e.g., spatio-temporal localization and linguistic reference via memory, communication and reasoning). Items that are typical targets for tracking are things such as stationary physical objects (e.g., rocks, plants, trees, buildings, or attached artifacts), moving physical objects (e.g., animals, certain artifacts) and human beings. All such items are located in a spatial environment, persist over time, and are – or at least closely related to, in the case of conspecifics' bodies – physical objects that respect non trivial objecthood criteria such as having cohesive parts, following continuous spatio-temporal paths, or possessing causal powers and dispositions. Perceptually tracking these objects through space and over time requires possessing sensorimotor systems (e.g., the oculomotor system, the visuo-haptic system or the auditory system) that can anchor into and smoothly pursue objects' properties. Nonetheless, one may suspect that tracking intentional agents (i.e. creatures to whom it is natural to attribute intentional states such as beliefs, plans, desires, and who may even participate in shared intentionality), as opposed to physical objects without mental states (i.e. objects which are not intentional agents) exploits or requires further abilities and strategies. In particular, at least for humans, tracking conspecifics amounts to tracking intentional agents. This raises the question of how the perceptual tracking of non-intentional objects relates to the keeping track of intentional agents. Here, we propose an extension and augmentation of recent work on object-tracking to the tracking of intentional agents. Based on the examination of the elementary procedures available for pursing agency, our principal suggestions are as follows. First, identifying intentional agents is significantly dependent upon the perceptual abilities of physical-object tracking, and might therefore be explained by the ‘object-file' hypothesis (this hypothesis is explained in section 3), which we suggest to use in the study of the tracking of intentional agents (section 4). In the most elementary case, humans track intentional agents as physical objects: they track such agents by tracking their bodies (section 5). Even though this kind of tracking is insufficient for keeping track of human individuals ‘as' intentional agents (and explaining their behavior with an intentional stance), it may suffice to explain a number of interacting and situated behaviors in social contexts with intentional agents. Second, however, tracking intentional agents ‘as' intentional agents requires additional capacities for detecting and understanding intentional states and certain further properties which creatures with such states can exhibit (section 6) – e.g., (ir/)rationality, and the capacity to participate in shared intentionality. We note, however, that tracking a human individual as an intentional agent may require an appeal to specific perceptual cues and may even recruit basic sensorimotor skills – such as the detection of biological motions – whose tracking might be independent of the understanding of conspecifics' mental states. For reasons of parsimony and computational economy, unless we have reason to think that there is a separate system devoted to tracking intentional agents, we should suppose that the same mechanism used to track physical objects is recruited for tracking intentional agents.
Domaines
Loading...