A match does not make a sense: on the sufficiency of the comparator model for explaining the sense of agency

Abstract

The development of a sense of agency is indispensable for a cognitive entity (biological or artificial) to become a cognitive agent. In developmental psychology, researchers have taken inspiration from adult cognitive psychology and neuroscience literature and use the comparator model to assess the presence of a sense of agency in early infancy. Similarly, robotics researchers have taken components of the proposed mechanism in attempts to build a sense of agency into artificial systems. In this article, we identify an invalidating theoretical flaw in the reasoning underlying this conversion from adult studies to developmental science and cognitive systems research, rooted in an oversight in the conceptualization of the comparator model as currently used in experimental practice. In these experiments, the emphasis has been put solely on testing for a match between predicted and observed sensory consequences. We argue that the match by itself can exclusively generate a simple categorization or a representation of equality between predicted and observed sensory consequences, both of which are insufficient to generate the causal representations required for a sense of agency. Consequently, the comparator model, as it has been described in the context of the sense of agency and as it is commonly used in experimental designs, is insufficient to generate the sense of agency: infants and robots require more than developing the ability to match predicted and observed sensory consequences for a sense of agency. We conclude with outlining possible solutions and future directions for researchers in developmental science and artificial intelligence.

The effect of uncertainty on prediction error in the action perception loop

“…In both the comparator and predictive processing accounts, agency is associated with small prediction error, or a match between expected and actual outcomes of actions. The comparator however focuses on net retrospective prediction error and cannot account for hypothesis switches in the face of accumulating prediction error or other changes in future action based on inferences of agency (see also Zaadnoordijk, Besold, and Hunnius (2019) ). The predictive processing account positions agency in a broader theory of action and policy selection.…”

A causal view of the sense of agency

“…Nevertheless, if no evidence supports the link between sense of agency and the goal-directed behavior of the organism, this account becomes a mentalist, adultist, and non-parsimonious interpretation of infant experience. The sense of agency in newborns cannot be justified by analogy with the adult experience, because infants may lack the necessary cognitive architecture for it (Zaadnoordijk et al, 2019). Further, by stating that sense of agency is intrinsic to any goal-directed action, the classical view proposes a black-box, preformist explanation of action awareness, which is insensitive to variations of the mental implementation of behavior also in the case of adults.…”

Environment-Related and Body-Related Components of the Minimal Self

“…Empirical research has shown that a match is important between perceptual changes the agent aimed at prior to generating efferent activity (often called “goals”), and the actual perceptual feedback after the efferent activity had been emitted (for a recent overview, refer Haggard, 2017 ). If there is a match between goal and feedback, it is likely that this feedback was caused by the agent’s efferent activity ( Carruthers, 2012 ; Haggard and Chambon, 2012 ; Gallagher, 2013 ; Zaadnoordijk et al, 2019 ). If there is a mismatch, it is more likely that the postaction percept was caused by something else than efferent activity.…”