I think that Norm Friesen is right to say that we need to look beyond the simple accounts of interaction offered by the early theorists of computation. But I do not agree that we should go so far as to say, with Friesen, that "computers cannot enter the world of obligation and concern that ultimately gives language its meaning." Pulling out the phenomenalism of Merleau-Ponty doesn't establish the case. When Merleau-Ponty says that emotions such as love and anger are "directly manifest" to us, rather than interpreted, he is, in my view, wrong: there isn't some sort of direct person to person interaction that does not go through the media of communication and interpretation. But where Merleau-Ponty is right - and where a proper criticism of Friesen's thesis ought to begin - is in the assertion that not all communication is accomplished through the mediation of signs or symbols representing a certain state, idea or emotion. When we perceive a tiger, we do not (necessarily) create a 'tiger' token in our brains; we infer immediately from sensation to perception via properties of the sensation itself, not through an intermediary language. Today's computers do require a translation to signs, and so are not able to emulate this human capacity. But there is no reason why the cognitivist paradigm ought to prevail forever in computational theory, certainly, no more reason than it would prevail in the psychology of perception.
Today: 9 Total: 32 [Direct link] [Share]
