6th Avant Workshop: Between computation and coordination: Distributed perspectives on language, cognition and text-based artefacts
On a distributed perspective language and cognition arise as dialogical human activity is shaped on various timescales by biological and cultural environments. This perspective challenges the assumption that language and cognition are constituted by discrete text-like symbols stored in individual memory and governed by formal rules. Accordingly, the distributed framework is incompatible with classical organism-centered views of mind and cognition. Further, if the language of thought and a representational theory of mind are seen as essential to computational models, computationalism too is incompatible with the distributed perspective. In recent decades, however, cognitive science has extricated computation from organism-centered representational models of language and thought. For example, connectionism, certain neuroscientific theories and wide functionalism adopt computational approaches and, crucially, allow that thinking can take place in extended systems. Further, there ceases to be any reason to ascribe it to either an individual mind or the use of lingua-form symbols.
The computational theory of mind is rooted in the model of an external artefact which is usually understood as a von Neumann computer. However, the idea of digital computer derives from more basic cultural practices that consist in writing, re-writing, reading and transforming symbolizations as a literate person draws on external tools such as paper and pencil. Therefore, the heart of computational approach lies in appeal to interaction between a person, the environment and artefacts that can be used to make systematic use of symbolizations.
Why does it matter? The distributed perspective stresses that the multi-scalar dynamics of talk and text-making differ between cognitive domains. Accordingly, many insights can be gained from contrasting the bodily dynamics of, for example, talking, writing and computing. We explore the idea that, in contrast to talk, writing and computing use transformations of artefacts that can encode connotation reduced (or free) meaning. This opens up a new kind of cognitive domain – one centred on text-genres and, latterly, programs. Further, if human minded behaviour is a product of bio-cultural co-evolution, the “machine” for recognizing these artefacts (and, indeed, for drawing on bodies to enact face-to-face talk) has to be installed in a society that relies on coordination by embrained bodies. Therefore the domain of computations can be re-framed as a special case of social practices with text-based artifacts. Accordingly, the concept of computation can be broadened to include interactive dynamics that link people with both the perceived world and ways of translating their experience into a systematic use of artefacts.
These threads open up general questions: Is it desirable to weave together computational and distributed approaches to language, cognition and artefacts? How they can enlighten each other? What are the opportunities and threats posed by such a synthesis? These problems generate more specific issues:
1. What can we gain from framing computations as linking formal transformations with a domain of social rules and interactions?
2. How can the computations we normally ascribe to individual minds be implemented in interaction?
3. How do the multi-scalar dynamics of operating on text-based artefacts differ from mental computations?
4. What is the role of interactions with symbolizations in the performance and acquisition of algorithmic skills?
5. Does a computational approach need a code model of language?
6. How far we can go in explaining language or languaging without the concept of symbolization?
7. Are computations in the head less real or essentially different than computations on paper?