Thinking
• Textbook overview– Directed versus undirected thinking– thinking and disorders of thinking
• An alternative perspective– Regulatory function of thinking
Integrating language, thinking and emotion
Groome:
A. Thinking: "Self-controlled symbolic mental activity"
Textbook overview
How can we observe thinking?
Thinking and Action
Planning
Execution
How do we know that John is thinking?
We can observe the outcome
or we can monitor his brain
Thinking and Action - areas of the brain involved in planning and execution
Left cortex Right cortexMedial cortexFront Rear Rear Front
Textbook overviewGroome:
A. Thinking: "Self-controlled symbolic mental activity"
(Goal) directed1. Problem solving
– Gestalt theory– Information processing theory
2. Reasoning
Inductive, deductive, statistical, everyday
Un-directed
Creativity, imagery, dreaming, etc.
Emotion ??
Groome (continued):
B. Disorders of Thinking: Dysexecutive Syndrome– no loss of ability, rationality, or intelligence– loss of regulatory function
Not inability to think!
Textbook overview
An alternative perspective• Regulatory function of thinking
System to represent previous experience, monitor ongoing behaviour and potential new information
• Evidence– Psychiatry (Barkley's theory of ADHD)
• Thinking subserves the executive function
• Self-regulation of behaviour and emotion
– Phylogeny (evolution of the human species)• Thinking is a substitute for action (Pinker, Darwin)
– Ontogeny (child development)• Thinking is internalized language (Vygotsky)
Vygotsky's theory of language
• Language is self-controlled symbolic mental activity used for communication and behavioural regulation of others
• Thinking is fully internalized silent inner speech• Developmental stages
– External speech– Private speech
• ego-centric language used for self-regulation
• elliptical, telegraphic
• not whispering, but link between overt and covert speech
– Inner speech– Verbal thought
Evolutionary perspective (Pinker 1997)
• Thinking is an internalized substitute for trial and error behaviour– allows testing possible future events without suffering the
consequences– allows our species to outsmart objects, plants, animals, and
each other– evolves more rapidly than defenses of other organisms
• Humans occupy a "cognitive niche" in the biological arms race
Psychiatry (Barkley, 1997)
Executive function of cognition
– biological basis is operation of inhibitory neural circuits that control overt behaviour
– behavioural inhibition makes it possible to internalize repertoirs that become mental strategies
– strategies are basis for cognitive abilities – require working memory to operate
E.g: Attention deficit / hyperactivity disorder (ADHD) is a deficit in behavioural inhibition resulting from malfunction of norepinephrine and / or dopamine circuits (but not an acquired disorder)
Barkley's model of the executive function
Behaviour repertoir
Sensing Speech Affect Play inhibition
Mental strategy
Holding events in mind
Internalization of speech
Self-regulation of affect
Analysis and synthesis of action
Cognitive ability
Bridge time Reflect upon and describe events
Motivate oneself and delay reward
Creativity, diversity, re-organization
Cognitive process Working memory
References
• Barkley, R.A. (1997) ADHD and the Nature of Self Control. Guilford Press.
• Pinker, S. (1997) The Language Instinct. Penguin Books.
• Vygotsky, L. S. (1962) Thought and Language.MIT Press.
Goal-directed thinking (Groome)
1. Problem solving"A problem exists when a person has a goal and does not
know how to reach it" (Duncker, 1945)
– Well-defined versus ill-defined problems
– Gestalt theory– Information processing theory
2. Reasoning– Inductive– Deductive– Statistical– Everyday
Problem SolvingA. Gestalt Theory
A problem exists because of the way it is represented or perceived.
– InsightRestructuring or spontaneous reorganization shows that solution
is a necessary part of the problem
– Reproductive thinkingRote application of previously rewarded responses
Novelty arises only from random trial-and-error
– Productive thinkingRe-structuring of problem situation produces novel solutions
– Obstacles to achieving insight Mental set, functional fixedness, previous experience,
over-confidence
–Problem spaceinitial state, goal state, constraints, rules, instructions, etc. that define the problem
–Well-defined problemsRepresentation of problem space
Operators to create new states
imPlementation of operators
Evaluation of current state w.r.t. goal
• Information Processing TheoryProblem solving is a search through problem space
– Analysis into parts (not Gestalt)knowledge states
operators that transform these
application of these makes paths that link states
Problem Solving
Ill-defined problems
lack one or more of these
state-action tree
• Information Processing (continued)– Obstacles
Problem definition, size of problem space, mental capacity, knowledge, time
– Strategies (ways of solving)Random trial & error
fails when many loops
Algorithmsrecipes for problem solution
never fail
Heuristicsshort-cut strategies using limited resources
success not guaranteed
Change goalfails in well-defined problems
and adversary problems
Problem Solving
Examples of well-defined problems
• Chess
– 10120 possible states
(estimated age of solar system is only 1018 seconds)
• Tower of Hanoi
– shortest solution in 2n-1 moves
2 disks 3 moves
3 disks 7 moves
5 disks 31 moves
64 disks ??? moves Shortest solution in 8,446,744,073,709,551,616 moves
Would require 584,942,417,355 years at rate of one move per second - ie. 580 billion years (estimated age of the universe is 14 billion years)
• Role of knowledge in problem solving
– analogy
– previous experience
– expertise
– mental set
– functional fixedness
• Heuristics
Problem reduction
– Means-end analysis
select sub-goals that are nearer to main goal
fails when move away from goal required
– Difference reduction (hill climbing)
fails when there are local minima or maxima
– Working backwards
create sequence of sub-goals from main goal
fails when too many paths towards goal
State-action tree: Tower of Hanoi, 3-disks
initial state
goal state
possible sub-goal
"clear largedisk"
loop
Reasoning
Thinking with rules, not content
E.g: Given that a) No students are stupid
b) No stupid people eat pizza
What do you conclude?"No students eat pizza"
Not valid
"Nothing"
Is valid
Formal logic
How people connect propositions and draw conclusions
• Pragmatic reasoning– schema bound (scripts, obligations)
– content based (derived from experience)
– incomplete (experience is limited)
– expressed in natural language
– validity equates with truth
– probabilistic conclusions (not certain)
• Formal logic– bound by normative rules
– abstract, not based on content
– complete, closed system, all information given
– expressed in formal language
– validity does not equate with truth but with logical necessity
• validity is independent of truth of the propositions
• if the reasoning is valid and the propositions are true, then the truth of the conclusion is a logical certainty
Reasoning
E.g. Inductive reasoning
E.g. Deductive reasoning
Reasoning
A. InductiveFrom particular to particular, or particular to general
Accumulation of confirmatory evidence leads to general propositions that are probably true
Empirical generalizations that are more uncertain than instances of them
B. DeductiveFrom general to general, or general to particular
Finding a conclusion that necessarily follows by rules of formal logic from what is given (premises)
– Syllogisms• Drawing conclusions about category membership of items, or
relations between items given in pairs of premises
– Conditional reasoning• Inference from propositions about antecedents and
consequences
Reasoning (continued)
C. Hypothetico-deductive reasoningCertainty by falsification
Basis of scientific method, legal argument, forensic investigation
Empirically testing truth of theories
Combines inductive with deductive logic for greater certainty• By induction, formulate a general hypothesis,or a rule • Deduce a particular proposition from this• Obtain empirical evidence
a) The hypothesis is probably true if all evidence is consistent with it
b) The hypothesis is certainly false if any contradictory evidence is found
Hypothetico-deductive (continued)
Examples of laboratory studiesWason four-card problem
Bruner concept attainment task
Many people tested show confirmation bias, fail to consider potentially falsifying evidence
Research for the way in which people tackle particular problems has led to questions about the diversity of strategies employed by people
StatisticsTaming chance to reconcile
Fortuna and Sapientia
Reasoning (continued)
D. Statistical reasoningManaging risk
Algorithms
Probability calculations make statistically independent events predictable in the long run
• use mathematical models based on all possible outcomes
HeuristicsWhen reasoning about future events we evaluate the likelihood by constructing mental models that are representative of the present. The availability of such models in terms of imaginability and retrievability does not reflect the likelihood of such scenarios in reality, and hence our estimates are biased (Kahneman & Tversky)
a) Availability• judge likelihood on basis of what is remembered
– memory is selective: significant events noted, insignificant events ignored– overestimate likelihood of significant events
Reasoning (continued)
Heuristics (continued)
b) Representativeness• judge likelihood on basis of similarity to a mental model
– typical instances are judged more likely than untypical ones– bias due to limited experience of what is typical
Example: Lottery Phenomenon Most people are irrational, because they voluntarily participate in a game with a
negative expected outcome - odds of losing are greater than odds of winning (Hill & Williamson, 1998)
• Availability: Media favour winners - millions of losers never shown
• Representativeness"Keep playing and you'll win sooner or later" (Gambler's fallacy) Long
sequence of losses believed to be untypical, but losing and winning balance only in the very long run
"Choose an even spread", "avoid adjacent numbers", "study pattern of previous draws", but random events have no pattern
B. Disorders of Thinking
Dysexecutive Syndrome– no loss of ability, rationality, or intelligence– loss of regulatory function
Textbook overview (continued)
1. Cognitive deficits
2. Working memory and forebrain
3. Supervisory attentional system
Neuropsychology of Freud's personality theory
Original 1933 diagram(colour added)
Functions mappedaccording to Solms (2004)
Disorders of thinkingResult of frontal lobe lesions1. Cognitive deficitsAny or all of the following
Attention
Abstraction
Estimation
Strategy formation
Everyday planning
• Inability to direct and sustain attention
• Distractibility
• Utilization behaviour
• Inability to suppress most salient response
• Perseveration
• Inability to utilize feedback
• Inability to suppress previously incorrect response
• Disinhibition
• Think of bizarre hypotheses, but do not disconfirm
• Generate intentions (markers) but not spontaneously reactivated later
Personality changeDeficits in emotional and social decision makingDe-regulation of affect
Frontal lobe syndrome
Disorders of thinking
2. Working memory and the forebrain – Baddeley's theorySensory information (visual, auditory) is held online
Associated actions are planned and released at appropriate time
– Goldman-RakicPre-frontal lesions in monkeys produce selective impairment of
delayed working memory task
– Object permanence in young infantsImmature frontal lobes (1 year)
Delayed-response deficits and perseveration
Disorders of thinking
Supervisory attentional system (SAS)Two cognitive modes monitored by SAS
• Routine, automatic, schema-driven
• Novel, difficult or dangerous, error monitoring, decision making
Deficits result in behaviour dominated by schemas that are triggered by irrelevant sensory input
The End
Deductive reasoning examples
• Valid SyllogismIF given: All men are mortal
and: Socrates is a man
Therefore: Socrates is mortal
• Valid conditional If P then QAffirm: P
Therefore: Q
Affirm: Not Q
Therefore: Not P
• Invalid SyllogismIF given: All men are mortal
and: All women are mortal
Therefore: All men are women
• Invalid Conditional If P then QAffirm: Not P
Therefore: Not Q
Affirm: Q
Therefore: P
Deductive reasoning: Using mental models
Consider all possible worlds. If a contrary example can be found then the conclusion cannot be a logical necessity
E.G. Constructing circle (Venn) diagrams
students stupid
eat pizza
No students are stupid
No stupid people eat pizza
Therefore no students eat pizza
All professors are pianists
All the pianists are athletes
Therefore all the professors are athletes
athletes
pianists
prof.
stupideat pizza
students
A B
No alternative is possible