Klaus Gärtner

In recent years, the debate about cognition culminated in a clash of two oppositional attitudes: on the one side, some think that cognition has to be thought of in representationalist terms (e.g. computationalism). On the other side, embodied, dynamical, action-orientated, and enactive programs begin to challenge this view. Their most radical versions explicitly claim that the relation between the mind and brain, body and the world has to be spelt out in an active world engaging way, and that we have to get rid of representations in the process. Some of the proponents take this even a step further and argue that by purging representations a revolutionary alternative to classical views of cognition can be created. To show its full potential, they are eager to ‘radicalize’ what can be considered enactivist friendly views, including predictive processing. In this talk, I will argue that one should be careful about being too radical. To do so, I will first introduce the strategy of this radical approach. After this, I will ask what kind of explanation of cognition we really want. In this context, I will consider one of the most prominent theories about cognition, namely predictive processing. Here, I will concentrate on the idea that whether representations form part of a cognitive economy, should be based on mechanistical considerations, but what we see right now, however, is a content-debate blocking our view. Finally and most importantly, I will focus on a philosophical analysis of mechanisms and discuss how explaining content cannot play the critical role it is supposed to. I will conclude that we should be laid-back about representations in cognitive science and, although their job description might have changed, continue to take advantage of them in explanations.