Linguistic primitives: A new model for language development in robotics

Often in robotics natural language processing is used simply to improve the human-machine interaction. However, language is not only a powerful communication tool: it is deeply linked to the inner organization of the mind, and it guides its development. The aim of this paper is to take a first step towards a model of language which can be integrated with the diverse abilities of the robot, thus leading to its cognitive development, and eventually speeding up its learning capacity. To this end we propose and implement the Language primitives Model (LPM) to imitate babbling, a phase in the learning process that characterizes a few months old babies. LPM is based on the same principles dictated by the Motor Primitives model. The obtained results positively compare with experimental data and observations about children, so confirming this interest of the new model.