This is what's wrong with philosophy in cognitive science - lots of empty rhetoric which completely ignores the fact that all of the simulations/models of human cognition which most closely replicate empirical data of human behaviour require internal representations of some kind.
Despite the admonition about "vacuous scholasticism", I think the idea of recognizing that representations are "a label for whatever best characterizes and explains cognition" is probably a good one. The real issue is that we want a good model of cognition at various levels of organization, so we still have to (eventually) understand what exactly "representation" or "content" corresponds to in our model/theory.
From an outsider's perspective, I have to say that the talk of "Mordor" and "defeating" sounds kind of histrionic. Does anyone really care whether there are "representations" or "content" when we know that the way we conceptualize these things is going to have to change a lot to tie them to real physical processes? A lot of Philosophy of Mind feels like a big category error sometimes.
It is surprising to see how a focus on action can bring to disregard representations. Actions do not exist by themselves. Agents implement actions for given reasons, to satisfy some constraints. And constraint satisfaction processes use meaningful representations which come before action implementation. We bring a spoon of porridge to our mouth because our organism needs energy in order to stay alive. The representation of the hand guiding the spoon is constraint satisfaction driven before being action oriented. The action exists because the agent has a constraint to satisfy. Constraints to be satisfied produce meaningful representations that will guide actions. The constraint satisfaction process basically leads to defining meanings and meaningful representations with intrinsic/derived status. Short paper: http://cogprints.org/6279/2/MGS.pdf The Meaning Generator System: http://www.mdpi.org/entropy/papers/e5020193.pdf Meaningful Representations: http://crmenant.free.fr/2009BookChapter/C.Menant.211009.pdf
This is what's wrong with philosophy in cognitive science - lots of empty rhetoric which completely ignores the fact that all of the simulations/models of human cognition which most closely replicate empirical data of human behaviour require internal representations of some kind.
ReplyDeleteDespite the admonition about "vacuous scholasticism", I think the idea of recognizing that representations are "a label for whatever best characterizes and explains cognition" is probably a good one. The real issue is that we want a good model of cognition at various levels of organization, so we still have to (eventually) understand what exactly "representation" or "content" corresponds to in our model/theory.
ReplyDeleteFrom an outsider's perspective, I have to say that the talk of "Mordor" and "defeating" sounds kind of histrionic. Does anyone really care whether there are "representations" or "content" when we know that the way we conceptualize these things is going to have to change a lot to tie them to real physical processes? A lot of Philosophy of Mind feels like a big category error sometimes.
"A lot of Philosophy of Mind feels like a big category error sometimes."
ReplyDeleteMy thoughts exactly. David, I agree with everything you wrote.
http://robots.net/article/3502.html
ReplyDeleteIt is surprising to see how a focus on action can bring to disregard representations.
ReplyDeleteActions do not exist by themselves. Agents implement actions for given reasons, to satisfy some constraints. And constraint satisfaction processes use meaningful representations which come before action implementation.
We bring a spoon of porridge to our mouth because our organism needs energy in order to stay alive. The representation of the hand guiding the spoon is constraint satisfaction driven before being action oriented. The action exists because the agent has a constraint to satisfy. Constraints to be satisfied produce meaningful representations that will guide actions.
The constraint satisfaction process basically leads to defining meanings and meaningful representations with intrinsic/derived status.
Short paper: http://cogprints.org/6279/2/MGS.pdf
The Meaning Generator System: http://www.mdpi.org/entropy/papers/e5020193.pdf
Meaningful Representations: http://crmenant.free.fr/2009BookChapter/C.Menant.211009.pdf