LECTURE 4

Question 1

MULTI STORE MEMORY MODEL

Answer 1

Environmental input -> Sensory memory (if not forgotten) –ATTENTION-> Short term memory (if not forgotten and continuously rehearsed) –CONSOLIDATION -> Long Term Memory (if not forgotten) –RETRIEVAL -> Short term memory (and the process continues)

Question 2

MODULARITY (FOR REF.)

Answer 2

How the different components can be separated and recombined

Question 3

LOCAL

Answer 3

Concepts are stored and represented as one coherent unit. “Grandmother cell” – one neuron (or local group of
neurons) responds when seeing your grandmother (and nothing else)

Question 4

DISTRIBUTION

Answer 4

Concepts are defined by their pattern of activation across many smaller units.

Question 5

SERIAL PROCESSING

Answer 5

A serial process occurs in a linear order, with the next operation occurring after the previous has finished. Things happen one at a time

Question 6

PARALLEL PROCESSING

Answer 6

A parallel process occurs simultaneously, with all operations occurring at the same time. Things happen
simultaneously

Question 7

HYBRID PROCESSING

Answer 7

Things enter/exit system in serial but everything within
system processed in parallel

Question 8

ALGORITHMS

Answer 8

Strict step-by-step rules. Solution is guaranteed. Might not always be the “best” method, in terms of processing power and speed

Question 9

HEURISTICS

Answer 9

Rules of thumb. Solution is likely. Typically more efficient than algorithms. Humans tend to prefer heuristics over algorithms (efficiency, effort, flexibility)

Question 10

CONNECTIONISIM

Answer 10

“Neural Network”
* Inspired by neuroanatomy of the cortex
* Nodes – units of info – analogous to neurons
* Connections between nodes – excitatory or inhibitory
* Concepts are built from distributed representations
of smaller units activated in parallel
* Networks learn to associate input with desired (or
taught) output
* Moving away from “classic” computer metaphor
(but still computational)

Question 11

EMBODIED COGNITION

Answer 11

Abstract concepts and processes are understood through sensory and bodily experience

Question 12

EMBEDDED COGNITION

Answer 12

What happens inside the head is only one piece of the
story – environment is part of cognition (calculators)

Question 13

DYNAMIC SYSTEMS

Answer 13

Emphasis on time, context, and interacting subsystems
* Subsystems including things like body and environment

Question 14

PANDEMONIUM MODEL

Answer 14

STUDY THIS MODEL

Question 15

What are connectionist models good at?

Answer 15

• Learning
• Pair input with desired output – supervised learning
• Reinforce connections that lead to desired output –
  reinforcement learning (basically operant conditioning)
• Generalization and discrimination
• Network can treat similar inputs as the same, and different inputs as different, based on some sort of “similar” and “different” criteria
• Learn to generalize between dog breeds, and discriminate from cats
• Can do this with previously unseen stimuli
• Sometimes make over-generalization errors…but so do humans!

Question 16

ENVATTED BRAIN AKA “BRAIN IN A JAR” MODEL

Answer 16

Information processing models consider all cognition to be localized inside the brain

• Perception is encoded as symbolic input and action is based on symbolic output. Perception, action, the body,
  and the environment are not a part” of the black box of
  cognition; just provide input and produce output

Question 17

Problem with Brain in a Jar System

Answer 17

People are not brains in jars –we are embodied, and reside in a dynamic perception- action loop
* Perceive the environment, act on it, perceive consequences of actions, act on them…etc.
* We can extend cognitive processing using our bodies and environment
* Embodied cognition
* Extended cognition

Question 18

Standard information processing models

Answer 18

Perception -> Symbolic Representation -> Symbolic
Processing -> Action
* Box and arrow models (like Atkinson-Shiffrin)
* Envatted brain

Question 19

Embodied and extended models

Answer 19

Perception -> Action -> Perception -> Action -> Perception….(perception-action loop)
* No symbolic representation = not a Turing machine
* This structure allows for tighter interplay between mind, body, and environment