Implementable theory of mind, culture and activity

Why?

To implement computer models of human creativity, learning and problem solving

e.g. story generation program

To inform design of systems to support learning and knowledge working

e.g. tutoring system

Vygotsky and Cognitive Science(Frawley,1997, MIT Press)

“Why should the social and computational mind go together?”

• They often study the same problems– Context, creativity, development, consciousness, metacognition

• They complement each other– Cognitive science needs a rich theory of mind in society

– Sociocultural theory would benefit from an implementable theory of mind

• Each can be read in terms of the other– There is nothing intrinsically noncomputational about sociocultural

mind, and nothing inherently asocial about machines.

– “Vygotsky proposed a protocybernetic account of mental processes”

Sociocomputationalism

• About how the world is in the computational mind• Binding between an organism and the world affords

unified stretches of experience (‘the world in the mind’)• The world itself is not the model, but rather the

effectively computed world• The focus of sociocomputationalism is on how the

world affords effectively computed activity• Not all aspects of the world are sociocomputationally

effective. The effective aspects have a role in the language for processing.

• “Sociocomputationalism is an account of the details of the opportunistic environmental control of thinking by certain aspects of language”

Sociocomputationalism (my interpretation)

• A cybernetic not a symbolic computational theory

• About external and internal constraint, interaction, dynamic control, and how the external is internalised

• Language as mediation for (self and social) regulation

Language for mediation

VisibilityLesion visible

LocationCortical grey matterCerebral white matterPeriventricular

MarginGraded

ShapeIrregular

Interior patternUnstructured heterogeneous

SizeLarge

IntensityHyperintense

Other signsMass effect

Metarepresentation (Frawley)

• Metaconsciousness is higher-order thought but not a data-structure, not a representation of another representation, but rather an awareness of a representation as a representation, originating in the socio-cultural resources for reflexivity, chiefly speech.

• Two aspects to metarepresentation: content and bracketing

• Content: a language to describe and refer to thought

• Bracketing: a decoupling of the primary representation from reality into an imaginary situation (e.g. “this pencil is like a spacecraft”)

• Autism is bracketing without content – a language for thought (perhaps overdeveloped) with a disrupted language of thought.

Socio-computational architectures (Frawley)

• Need an architecture that “might accommodate explicit, self-reflexive focusing”

• One that gives a role to higher thinking not as problem solving per se but as its management.

• Central role to contextualised control of thought – both internal and external contextualisation

Conversation Theory(Gordon Pask)

• A domain independent theory of the processes of coming to how

• Not a theory of cognition, but of interaction, theory construction and communication

• Derived from cybernetics

• Radical constructivist epistemology

• The cybernetic theory of conversations.

• A conversation is the minimum necessary structure to enable learning

Radical constructivism

• Systems theory, concerned with the coupling of organisms and environment

• Radical constructivists “have taken seriously…the concept of knowledge as an adaptive function”

• Cognition enables us to cope with the world of experience (rather than providing an “objective” representation of an external world)

• Cognition as an evolutionary, self-organising process• Members of a species create shared ecological niches

and consensual domains of interaction and communication.

• The “models” an organism constructs as a result of adaptation are tacit.

• “Gaining knowledge” is not the storing of representations, but the process of “coming to know”

Radical constructivism and language

• Conversation is not the exchange of knowledge, but the process of becoming informed about each other’s “informings”

• The “coordination of coordinations of coordinations”

• Higher level coordinations are “tokens” for lower-level coordinations, (objects and events), which are themselves tokens for stabilities of sensori-motor activity and “structural coupling” with the environment

Pask’s contribution

• Distinction between “psychological (p-) individuals” and “mechanical (m-) individuals”

• There is no one to one correspondence between m- and p- individuals.

• Consciousness is a property of a p-individual (a system that may potentially “know with itself”) not an m-individual.

• The participants in a conversation are p-individuals. The conversation itself is a p-individual.

• A conversation is the minimum necessary system (p-individual) that can learn

Conversation TheoryG. Pask (1975) ‘Minds and media in education and entertainment: some theoretical comments illustrated by the design and operation of a system for exteriorizing and

manipulating individual theses’, in R. Trappl and G. Pask (eds.) Progress in Cybernetics and Systems Research IV, Wiley.

L language able to accommodate commands, questions, instructions

Mind organization expressed in L which, if set in motion, gives rise to thought, feeling or behaviour [very broad definition of mind! - includes e.g. theatre scripts, manifestos]

Medium processor in which a mind is realized (including brains, computing machines, machinery of theatrical production, political parties)

Individual the activity or motion permitted by an L nameable and personally pronominalized chunk of mind, such that, if it is realized and executed, its motion comprises a (logically) coherent set of beliefs. (Personalities, characters in plays, socio-political factions, if realized and executed in brains, actors, collections of brains, yield people, enactments of characters, cultures)

Pask – Distributed Cognition in Action“There is no need to see minds as neatly encapsulated in brains connected by a network of channels called ‘the media’ ... I am inviting the reader to try out a different point of view; namely the image of a pervasive medium (or media) inhabited by minds in motion. Thus, media are characterized as computing systems, albeit of a peculiar kind. In our present state of knowledge it seems prudent to regard the media as heterogeneous, and rendered modular by the existence of specially unrestricted regions (brains, for example) capable of acting as L processors (though I have a hankering to imagine these regions are ultimately determned by programmatic rather than concrete localization). It is surely true that rather powerful computerized systems greatly reduce the differentiation of the medium and coalesce the specially unrestricted nodules, so that ‘interface barriers’ are less obtrusive than they used to be.”

Organisms Mediaas channels of communication

Individuals Communcation

as program sharing and linguistic interactionbetween individuals. Medium (M) as shading

Rescher’s two-cycle model

ModelsMethodsProcedures

Application

Pragmatic correction

Metaphysical assumptionsTheoretical interpretationsConceptual systems

How?

Why?

Correction by coherence

Conversational framework (after Pask)

Facility for practical model buildingand problem solving

Teacher demonstrates or elicits models and problem solutions

Learner acts to build models and solve problems

Sets goalsAdjusts model

ActsModifies actions

How questions and responses

Teacher demonstrates or elicits models and problem solutions

Learner acts to build models and solve problems

Offers theories and ideasRe-describes theories

Offers conceptions and explanationsRe-describes conceptions

Why questions and responses

Facility for mediating agreements

The inception of a spontaneous concept can usually be traced to a face-to-face meeting with a concrete situation, while a scientific concept involves from the first a ‘mediated’ attitude towards its object.

Vygotsky, Language and Thought (cited in Laurillard)

Exteriorising agreements

“In research (and in effective tutorial operations, also) understanding must be exteriorized for observation... To exteriorize an understanding means to establish a multilevel agreement between [individuals] A, B ... with respect of a topic T. ... The concrete equipment in which A and B model derivations [agreements] ... is also a medium. The composite medium couples the media, namely brain(s), in a conversation designed to elicit understanding.”

External Representations (Pask)

“It seems clear that there must be great advantages in providing both participants with an external representation of the subject matter through which topics can be identified and discussed. In this way, explanation can be initiated by either participant.”

External Representations

An external representation is a construct that

a) mediates conversation on a topic

b) is invested with meaning [models agreements]

A full learning conversation

• Conversation about the how and why of a topic

• Conversation about the how of learning (for example discussing study skills and reflecting on experiences as a learner)

• Conversation about purposes, the why of learning, where the emphasis is on encouraging personal autonomy and accepting responsibility for one’s own learning

The complexity of coming to know(Laurillard)

For learning to success, the student must:-Apprehend the structure of the discourse-Interpret the forms of representation-Act on descriptions of the world-Adjust actions to fit the task goals-Adjust descriptions to fit the topic goal-Reflect on the cycle of goal, action, feedback

For Pask, understanding a topic means that the learner can ‘teachback’ the topic by providing both non-verbal demonstrations and verbal explanations of ‘how’ and ‘why’

Learning technology (after Pask)

Technology provides facility for practical model buildingand problem solving

Technology demonstrates or elicits models and problem solutions

Learner acts to build models and solve problems

Sets goalsAdjusts model

ActsModifies actions

How questions and responses

Technology demonstrates or elicits models and problem solutions

Learner acts to build models and solve problems

Offers theories and ideasRe-describes theories

Offers conceptions and explanationsRe-describes conceptions

Why questions and responses

Technology provides facility for mediating agreements

Implementable theory of mind, culture and activity

To inform the design of computer models of human creativity, learning and problem solving

Model of writing as creative design, informed the design of MEXICA

To inform design of systems to support learning and knowledge working

Laurillard’s framework for the design of learning technologies