Consensus in Traditional Cognitive Science

1
Consensus in Traditional Cognitive Science

• Unit of Analysis
• Laboratory
• The brain is the cognitive system.
• Control the environment to isolate the brain.

2
Are there limitations?

• Question Can laboratory experiments and results
  subtend and describe the full set of cognitive
  phenomena?
• Answer Most likely not.
• Move unit of analysis beyond skin, so that our
  cognitive system is
• Physically distributed
• Socially distributed
• Temporally distributed

3
Distributed Cognition as a Framework

• Goal Still to explain the cognitive phenomena of
  a cognitive system memory, attention, inference,
  reasoning
• BUT, with the approach that these processes are
  distributed through time and culture, as well as
  space (and space beyond the brain).

4
Distributed Cognition

• Edwin Hutchins, UCSD
• Navy Bridge navigation
• Flight deck navigation
• D-COG makes 2 commitments

5
1) Unit of Analysis

• Extends beyond the skin. But how far?
• Bateson systemic wisdom, leave nothing
  inexplicable
• Look for cognitive processes on the basis of the
  functional relationships of elements that
  participate together in the process.

6
2) Range of Mechanisms in System

• Cognitive events not encompassed by the
  manipulation of symbols in the brain.
• Cognitive events involve a rich array of
  manipulations between internal and external
  elements (eg, artifacts).
• Information propagates across media.

7
What do we get from D-COG?

• We can see how cognitive processes may be
  socially distributed.
• We can see how cognitive processes may be
  temporally distributed.
• Cognitive processes may involve the coordination
  of internal and external.

8
Important

• D-COG is a framework for examining the cognitive
  processes of cognitive systems as they exist in
  the world (in the wild).
• Not a theory of mind.

9
Other Frameworks

• Situated/Contextual Cognition is not a new idea.
• A few other frameworks are worth mentioning.
• Some more clearly state their goal as theory of
  mind.

10
Activity Theory (Vygotsky, Rubinstein, Leontjev,
Lurija)

• From USSR school of cultural-historical
  psychology
• 1) human activity is directed towards material or
  ideal objects
• 2) activity is mediated by tools or artifacts
• 3) activity occurs within a culture

11
Situated Action (Lave, Wenger)

• Environment dictates the activity
• An emerging, evolving interaction between human
  and environment
• Moment-to-moment interactions situated within a
  setting (itself part of an arena).

12
Related Topics/Readings

• Cultural Models (see Shore)
• Cultural blending (see Fauconnier)
• Books to get started with
• Mind in Society - Vygotsky
• Cognition in the Wild - Hutchins
• Cognition in Practice - Lave

13
Cockpit Cognition
14
Cockpit Cognition

• The cockpit as a whole is taken to be one
  cognitive system.
• Our unit of analysis is the cockpit and the
  elements (humans, artifacts, culture, etc)
• We examine cognitive processes as functional
  relationships between humans and artifacts.
• http//hci.ucsd.edu/10/Week7/

15
Analysis grounded in meaning

• Culture plays an important role in D-COG.
• We study the meanings of a culture to ground our
  analysis of the cognitive system.
• Actions and language are culturally meaningful.

16
Intersubjectivity

• I know that you know that I know.
• Shared understanding lt-gt Intersubjectivity.
• Expectations, and their importance in socially
  distributed systems.

17
Trajectories of Information

• Propagation of representational states across
  representational media.
• Expectations and intersubjectivity play a role in
  the trajectory.
• The value and function of particular media (its
  representation) is critical.

18
Trajectories of Information
19
Computational Artifacts

• Artifacts, when used, can perform a computation.
• The representation of the artifact scaffolds the
  task required to do the computation.
• Hutchins notes humans are good visual processors.
  Instruments aboard a ship bridge or cockpit are
  highly visual instruments (dials, meters, scales,
  linear alignment).

20
A common navigation problem

• A ship travels 1500 yards in three minutes. What
  is the speed of the ship in nautical miles per
  hour?

21
Solution 1 Paper and Pencil

• D R x T
• 1 nm 2000 yards
• 60 min 1 hour

22
Solution 2 Calculator

• D R x T
• 1 nm 2000 yards
• 60 min 1 hour

23
Solution 3 Three Scale Nomogram
24
Solution 4 Three minute rule

• (demonstration)

25
Artifacts as computation

• The task needed to solve the problem is related
  to the structure of the artifact used to solve
  the problem.

26
Solution 1 and 2 (pen/paper/calculator)

• First change equation to
• R D / T
• Then, convert units
• yards to nautical miles (0.75 nm)
• minutes to hours (1/20 hours)
• Then, perform division
• Answer 15 nautical miles / hour (knots)
• The computation is an analytical task.

27
Solution 3 Nomogram

• The same computation is now a perceptual task.
• You only need to draw a line across two points,
  the third point is the answer.

28
Solution 4 Three minute rule

• The same computation is now a conceptual task.
• Also, no need to understand how it works to use
  it.
• And, no need to design it to discover it.

1500 yards
15 knots
29
Computational Artifacts (again)

• The structure of the artifact defines the task.
• Cognitive systems are coordinated systems of
  elements (artifacts, humans, culture, etc) that
  represent information in such a way (the type of
  task) that some computation may be done.

30
Other Properties of Cognitive Systems

• Memory (redundancy, for example)
• Error checking
• Problem solving
• Distribution of cognitive load

31
Conclusions

• We expand the unit of analysis beyond the human
  skin, earning a more systemic view of cognition
  processes.
• Immaterial and material artifacts (culture and
  tools and humans) exist as elements of a
  cognitive system.
• The interaction of internal and external creates
  trajectories of information. These trajectories
  are cognitive processes (and may perform
  computation).