Abstract
The commentaries on my target article can be clustered into three topics. First, I discuss the broader theoretical background underlying the notion of cognitive boundary, and argue that this notion should be taken to refer to the boundary by which cognitive autonomous agents actively distinguish themselves from their environment. Subsequently, I address the difference between the notions of Markov blanket and operational boundary, pointing out that the two notions of boundary are compatible in some, but not all cases. In the last section I deal with some concrete examples of extended mind and argue that the proposed account does justice to instances of distributed assembly.