psych 421
Chapter 6.2:
Applying dynamical systems: Two examples from child development
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Applying DST to child development
- Time-sensitivity of dynamical systems theory potentially useful for studying development of motor and cognitive examples
Motor [infant walking]
Cognitive [A, not-B error]
- Both illustrate subtle dependencies on (apparently) cognitively irrelevant variables – can be modeled using DST
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Computational model of motor control
• Planning movements (e.g. reaching) begins with CNS calculating position of target and position of hand
• coordinating input from vision and proprioception
Planning movement requires (a) calculating trajectory, (b) working out a series of muscle movements that will take the hand along that trajectory
Executing movement requires calculating changes in muscle movement to accommodate visual/proprioceptive feedback
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
A typical computational model
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
DST alternative
- Walking is not a planned activity involving a specific set of motor commands “programming” limbs to move in certain ways
- Walking emerges out of complex interactions between muscles, limbs and different environmental features
• coupled system
• multiple feedback loops
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Learning to walk
- Infants standardly show a U-shaped developmental trajectory
Infants capable of making stepping movements during first 2 months of life
Ability disappears during non-stepping window [2-8 months]
Reappears when walking begins
- Standard explanation neural – depends upon maturation of areas of cortex responsible for executive control
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Manipulating the trajectory
Esther Thelen and Linda Smith showed that
(1) Stepping can be induced during the non- stepping window by holding the baby upright in warm water
(2) Non-stepping infants will make stepping motions on a mechanical treadmill (compensating when legs move at different speeds)
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Thelen and Smith model
- Changes in bodily and environmental factors can significantly alter walking behavior
• leg fat
• muscle strength
• gravity and inertia
• task environment
- Can be modeled by dynamical equations in which these are parameters
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Another modeling task: Object permanance
Object permanence = infants’ understanding that objects continue to exist when unperceived
• Important stage in development of child’s naïve physics
• Studied by Swiss psychologist Jean Piaget in terms of search for hidden objects
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Note that this is also discussed in chapter 9.
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
A - not-B error
- Occurs between ages of 7 and 12 months
- Children reach to the original hiding place, even after seeing the object moved from container A to container B
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive explanations
- Piaget’s explanation
Children not able to form abstract representations before 12 months
Before that children’s knowledge is in the form of sensori-motor routines – takes time to adapt to new location
- Other explanations
• development of prefrontal cortex required to inhibit reaching response
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Manipulating infant performance
• effect disappears if no delay between hiding and search
• effect disappears if posture is changed
• performance significantly improved by drawing infant’s attention to right side of table
• most reliable predictor is number of times the infant reaches for toy in A trials
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Thelen and Smith’s dynamical field model
• represents space in front of infant
• infant reaches when activation at a point exceeds threshold
• activation in one trial becomes input into next trial
• predicts time- and task-sensitivity of performance
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
*
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Dynamical field model
3 different types of input (all coded in the same way - as locations in the visual field)
(1) Environmental (distance to containers, salience of target. . .)
(2) Task demands (experimenter drawing attention to target)
(3) Memory input (from previous reaching trials)
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Key features
- Thelen and Smith’s model captures some features of infant reaching behavior
- It does not explicitly appeal to standard CogSci factors
representational states
emergence of executive control
Cortical maturation
- Highly sensitive to initial conditions
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Assessing the DSH
- DS models certainly give powerful tools for studying the evolution of cognition and behavior
More complex than standard accounts (bring a wider range of factors into play)
Simpler than standard accounts (because they do not invoke representations and computation)
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
An alternative to traditional CogSci?
- The analogy with the Watt governor does not carry over to cognition
• Watt governor and computational governor are mutually exclusive
• but DS models are compatible with information-processing models of cognition
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
The benefits
- DS models allow cognitive scientists to abstract away from details of information-processing mechanisms in order to study how systems evolve over time
- But we still need an account of what makes it possible for the system to evolve in those ways
Seems likely that this will be an information-processing model
How are we to interpret the activation variable in Thelen and Smith’s model?
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
An analogy
- Physicists have constructed models of traffic interpreting traffic jams as the result of interactions between single particles in many-particle systems
• good at predicting stop-and-go traffic and fact that traffic jams occur before road’s capacity has been reached
• But this doesn’t show that we can understand traffic without thinking about internal combustion engines. . .
• “Abstracting away from” ≠ “replacing”
Cognitive Science José Luis Bermúdez / Cambridge University Press 2020
*