Wheeler (2005) begins his book with two explicit assumptions:
The first task Wheeler sets himself is to demonstrate that orthodox cognitive science is generally Cartesian in character, which in my opinion he does successfully. He then wants to show that, in contrast, more recent developments in cognitive science (such as the dynamical systems approach and evolutionary robotics) are generally Heideggerian in character.
I outline some of the difficulties that Wheeler is facing with this second task, and trace them to his characterization of scientific knowledge as observer-independent and context-free (thus providing the foundation for his first assumption). This view of science has fundamentally Cartesian consequences, and I provide some examples of Heideggerian cognitive science which shows that it therefore fails to satisfy Wheeler's own second assumption. Finally, I will briefly outline a characterization of scientific knowledge (Maturana 1988) which avoids these difficulties, and therefore paves the way for a more consistent Heideggerian paradigm in cognitive science.