Tuesday, June 9, 2015

Is a hybrid radical embodied-representational cognition model a theoretical option?

A recent paper by Beckes et al. attempts to outline "a radically embodied neuroscience of attachment and relationships."  The content here doesn't matter much.  What I want to consider is the conclusion: 

we entertain the thought to explain such individual differences without mediating representations or computations, but in the end propose a hybrid model of radical embodiment and internal representations.  
Is a hybrid model a viable theoretical option? Radical embodied cognition denies computation and representation, so such a model violates the basic tenets of that framework. And I would argue that since a "hybrid model" includes representations, then this is just traditional cognitive psychology.  In a "hybrid" model, representations still exist in the mind/brain and they must be processed. This is the foundation of the standard cognitive, computational, or information processing model of the mind.  What is embodied is that the "hybrid model" talks more about the body's contribution, which is useful.  But traditional cognitive psychologists don't deny that the body plays an important role.  How else could you get retinal disparity as an important cue to depth or interaural level differences as an important cue to sound localization?  In this sense ALL traditional cognitive/computational/information processing models are hybrid embodied-representational models.

So while I welcome thoughtful discussions of the contribution of the body and the environment to cognition, calling a model a hybrid of radical embodied cognition and representational approaches is just terminological gobbledygook.

2 comments:

Unknown said...

What makes this model hybrid is that it proposes different systems linked to behavioral control in different (e.g., predictable vs. unpredictable) environments, some of those systems being representational and others radically embodied.

Unknown said...

To add to what Mattie says, if you take a hardline radical embodiment viewpoint you may be right that any "hybrids" accept some representational component and therefore must be a traditional cognitive/computational model. I think this a red herring though, and more semantic than substantive. The entire conceptual point of radical embodiment is that the nervous system doesn't represent the outside world, but rather functions to transduce signals, keeping the content of those signals relatively unmodified. Thus the debate is really about the degree to which the underlying processing structure modifies the input in any substantial way that makes it representational. If one imagines the neural structure of a snail, one can easily imagine the entire system being non-representational in nature. Over evolutionary time many new species may emerge from that snail species and its decedents. Perhaps at some point one of those species had an adaptation that included a new neural structure or set of structures that organized in such a way that incoming signals were sufficiently modified to no longer be radically embodied, but are now representational in nature. Does that necessarily imply that the original architecture that was radically embodied no longer exists? What if the new system is at least somewhat functionally distinct from the phylogenetically older system despite obvious interactions? If this occurred in the course of evolution of our species or our ancestor species doesn't the manner in which each system deals with incoming information merit a distinction? One that is currently best captured by a hybrid model concept? If you throw out radical embodiment in this case, you may be ignoring a potentially useful conceptual framework because it doesn't adhere to a rigid philosophical standard. I'm not sure that is the right way to think about these issues.