The project seeks to understand whether conscious knowledge can obscure the influence of unconscious representations on behavior and decision-making. To this end, we use the context of implicit learning, through which humans acquire complex regularities in their environment, without being fully aware of what they learn (e.g., consider the way we learn the grammar rules). Employing novel adaptations of standard implicit learning tasks (e.g., Artificial Grammar Learning), together with motion capture in sports, we test the hypothesis that conscious representations are not always direct reflections of internal states, but are often only partial, inaccurate,  reconstructions.