Thanks for your highly informative response. I’m glad that you recognise some element of ideology in your belief system. Until there are convincing arguments, most of our actions will have some kind of ideological basis. I’d like to highlight this ideology a bit by discussing the articles you cited and explaining why I don’t find them entirely convincing.
I agree unreservedly that there is an abundance of literature showing the value of interaction in learning. However, it must be recognised that whole swathes of research have been carried out on ideologically motivated grounds. Both the Bredo (1994) and the Packer and Giocoechea (2000) articles centre on the contested notion of nonduality (Dilley, 2004). Bredo’s position is that symbol-processing fails because (to simplify somewhat) there can be no mapping of the external world onto an internal symbol. From this, he assumes that symbol-processing is a form of dualism, from which he develops his situated view of cognition as an answer to the questions in learning that dualism cannot answer. I find this sequence flawed.
Stating that Theory A is flawed does not mean that Theory B is correct. It may well be that Theory B is correct, but logically it is a non-sequitur to premise B’s accuracy on A’s faults. The best that Bredo can do is compare and contrast. His description of symbol-processing’s language-and-reality weaknesses is very good. However, I find his ‘Mind and Body’ section poor. His premise here is that because AI systems and robots “take an inordinately long time to act” (p. 27) humans’ cognitive architecture is similarly disposed when made to learn passively. This may be true in reality, but implicitly, he’s saying that if AI can’t do it, neither can humans. Later, Bredo discusses the development of AI systems that have a situated component and argues from that to the similarities in humans’ perception, “minding is a tangibly physical matter” (p. 31).
Both examples so far can be reframed. As an epistemology of understanding human perception, modelling on robots is a useful method of testing researchers’ measurement capacities. So far, so good. It is a categorical error to extend this epistemology to claim that if AI can’t do it, there’s the possibility that there’s nothing to measure. Or if AI can do it, the mind-body duality is questionable. The former ‘if’ is flawed because it is a measurement issue. The latter ‘if’ is a genuine question. However, (as of 1994 when Bredo was writing) the lack of measurement instruments, i.e. good robotics, does not mean that there is no mind-body nonduality.
This post is getting too long, so I have to stop. I would like to respond to Packer and Giocoechea and explain my stance on the duality question. However, I’ll finish with Dilley’s observation that;
“Steven Pinker, after explaining for nearly 600 pages how the mind supposedly works, admits that there are still six enigmas that continue to baffle cognitive scientists: consciousness, the self, free will, meaning, knowledge, and morality” (Dilley, 2004, p. 136).
Bredo, E. (1994). Reconstructing educational psychology: Situated cognition and Deweyian pragmatism. Educational Psychologist. doi:10.1207/s15326985ep2901_3
Dilley, F. B. (2004). Taking consciousness seriously: A defense of Cartesian dualism. International Journal for Philosophy of Religion, 55(3), 135–153. doi:10.1023/B:RELI.0000034594.88195.23
Packer, M. J., & Goicoechea, J. (2000). Sociocultural and Constructivist Theories of Learning: Ontology, Not Just Epistemology. Educational Psychologist, 35(4), 227–241. doi:10.1207/S15326985EP3504_02