In practice, the question is what sort of lens cognitive scientists and neuroscientists use for studies. When conducting a scientific experiment, you have to m

In practice, the question is what sort of lens cognitive scientists and neuroscientists use for studies. When conducting a scientific experiment, you have to make some assumptions about what system you're studying, how you're modeling it, and how you're connecting model to hypothesis to evidence via experimental design. My favorite argument by Andy Clark specifically is his take on the Outfielder Problem: if you study the brain with a "disembodied", action-free theoretical lens, you have a really hard time parsimoniously explaining how an outfielder runs to catch a fly-ball. If you take the "embodied", "action-oriented" view, you see that solving cognitive problems using bodily and environmental resources via continually-updated motor planning is not only the theoretically elegant solution (the one that makes most precise use of the available sensory data, has the fewest parameters, is cheapest to compute), but simply resembles what we observe people actually doing (linear optical acceleration cancellation).