Osiurak and Badets (2016) examined the validity of the manipulation-based versus the reasoning-based approaches to tool use in light of studies in experimental psychology and neuropsychology. They concluded that the reasoning-based approach seems to be more promising than the manipulation-based approach for understanding the current literature. Buxbaum (2017) questioned this conclusion and raised certain theoretical limitations with regard to the reasoning-based approach. She also suggested that this approach is not well-equipped to integrate the existing psychological and neuroanatomical data in the tool use domain. In this context, she presented a neurocognitive model—the “Two Action Systems Plus” (2AS+) framework—deeply anchored in the embodied cognition approach. In this reply, we address the key points raised by Buxbaum, leading us to draw 2 new conclusions. The first is that the reasoning-based approach integrates the existing psychological and neuroanatomical data not only in the tool use domain, but also in the motor control domain. As a matter of fact, it is even better equipped than the 2AS+ to account for recent neuroscience data. The second is that the 2AS+ suffers from epistemological and theoretical limitations, generating confusion as to what manipulation knowledge—a core concept in this model—precisely is. To sum up, 2AS+ illustrates potential misuse of embodied cognition, viewing tool use mainly as a matter of manipulation and not of understanding mechanical actions between tools and objects.