1.1 History of Learning Theory

Learning

the acquisition of information through experience

• need not to be conscious

memory

the retention of learned information over time

both learning & memory are forms of _________

plasticity

plasticity

a consequence of development, injury, or experience

learning and memory are different

• memory focuses on retention

• learning is the process

Plato: Nativism

the Doctrine of Anamnesis: all learning is recollection

• therefore → all information is innate

• answers How do we come to know?

• student of Socrates

• ~427BCE-~347BCE

Aristotle: Early Empiricism

• all knowledge acquired, comes from experience

• student of Plato, disagreed w/ him

• 384BCE-322BCE

the Innate Debate has taken many forms

• Doctrine of Anamnesis vs. Tabula Rasa

• Nativism vs. Empiricism

• Nature vs. Nurture

Most recent philosophical developments

Descartes (1588-1679): Dualism

• unextended substance: non-physical

  • soul(mind), free will

  • does not abide by scientific rules

• extended substance: physical

  • body

  • abides by scientific rules

• implications of Dualism

  • some parts of human behavior can be studied

    • reflexology

  • infrahuman species should be fully understandable

Materialists (e.g., Julien Offra de La Mettrie, 1709-1751)

• argued that our mind may be “machine”

• suggest we can derive the laws of the mind

Llyod Morgan’s canon (1894)

argued that “In no case may we interpret an action as the outcome of the exercise of a higher psychical faculty, if it can be interpreted as the outcome of the exercise of one which stands lower on the psychological scale.”

Aristotle’s 3 Laws of Association (and the Law of Frequency)

1. Law of Similarity

2. Law of Contrast

3. Law of Continuity

4. Law of Frequency

Law of Similarity

the experience or recall of one object will elicit the recall of things similar to that object

Law of Contrast

the experience or recall of one object will elicit the recall of opposite things

Law of Contiguity

the experience or recall of one object will elicit the recall of things that were originally experienced alongside the object

Law of Frequency

the more frequently two things are experienced together, the more likely it will be that the experience or recall of one will stimulate the recall of the second

John Locke (1632-170)

all knowledge is acquired

• Tabula Rasa - blank slate

• knowledge is built up through the association of ideas

• “Let us then suppose the mind to be, as we say, white paper, void of all characters, without any ideas: - How comes it to be furnished? Whence comes it by that vast store which the busy and boundless fancy of man has painted on it with an almost endless variety? Whence has it all the materials of reason and knowledge? To this I answer, in one word, from experience. In that all our knowledge is founded; and from that it ultimately derives itself.”

Associationists

Principles of Association

• David Hartley (1705-1757)

  • contiguity, repetition

• James Mill (1773-1836)

  • mental mechanics

  • emotional vividness (salience)

• John Stuart Mill (1806-1873)

  • mental chemistry

Charles Darwin (1809-1882) & Evolution

Natural Selection

• traits that enable a species’ survival & reproductive success become more common in a population over the generations.

  • variation exists within populations

  • variations can be heritable

  • organisms produce more offspring than can survive

  • individuals with adaptive traits are more likely to survive

  • over time, advantageous traits become more common in a population

Functionalism

• behaviors follow the rules of natural selection

• further, learning follows the rules of natural selection

• learning is hereditary as well

Structuralism

Limits of “Armchair Psychology”

• prior to the founding of psychology as as discipline, most theorists were not experimentalists

Edward Titchener (1868-1927)

• a student of Wilhem Wundt

  • “Father of Psychology”

  • systematic introspection

• all thoughts can be broken down into basic elements called sensations

  • categorized sensations

    • 4 properties: intensity, quality, duration & extent

Ivan Pavlov (1849-1936)

Nobel prize in Physiology or Medicine 1904

• “ in recognition for his work on the physiology of digestion…”

Pavlovian Conditioning

• unconditional stimulus (US)

• unconditional response (UR)

• conditional stimulus (CS)

• conditional response (CR)

Introspection by analogy (George Romanes)

“a technique for studying animal behavior by assuming that the same mental process that occurs in the observer’s mind also occurs in the animal’s mind.”

• potentially problematic

Edward Thorndike (1384-1949)

• opposed anthropomorphic interpretations proposed that animals learn through trial and error

Thorndike’s Law of Effect

“if a response in the presence of a stimulus is followed by a satisfying [appetitive] event, the association between the stimulus and the response is strengthened. If the response is followed by an annoying [aversive] event, the association is weakened.

Thorndike’s Law of Effect II

assumptions

• learning is S-R in nature

• avoids mentalistic terms

• role of reinforcement

  • S-R S-R S-R S-R S-R S-R S-R reinforcement

• learning depends on response-outcome (R-O) relationship

  • distinction between:

    • Pavlovian conditioning (S-R)

    • instrumental learning (R-O)

big takeaway of edward thorndike

early ideas of reinforcement & punishment learning

John Watson (1878-1958)

extreme empiricism

• “give me a dozen healthy infants, well-formed, and my own specified world to bring them up in, and I’ll guarantee to take any one at random and train him to become any type of specialist I might select.”

behaviorism

Watson rejected the structuralists’ method of introspection

• emphasized publicly verifiable data

  • stimuli (S) and responses (R)

  • sought to formulate behavioral laws to predict and control behavior

• posited that learning is S-R in nature

  • thus can be studied in infrahuman species (ex.rats)

• famous experiment: Little Albert

Clark Hull

Animal performance =/= learning

• behavior motivated by drives

  • food/water, etc

• learning can occur without performance in the absence of motivation

• behavior can be predicted mathematical

mathematical-deductive theory of behavior

sEr = (V*D*K*J*sHr) - (sIr + Ir +sOr + sLr)

sEr = excitatory potential

• the likelihood that an organism will respond to a stimulus

• factors increasing sEr

• factors decreasing sEr

• V = stimulus intensity dynamism

  • some stimuli will have greater influence on behavior

  • ex. conditional taste aversion

• D = drive strength

  • ex. a hungrier animal will be more motivated to perform than a satiated animal

• K = incentive motivation

  • magnitude of the goal

  • ex. an animal will be more motivated by a pile of food than by a single pelet

• J = incentive delay

  • the delay before the animal is allowed to seek the goal

  • ex. an animal will be more motivated if it has been denied the ability to seek the goal for a longer time

(sIr + Ir +sOr + sLr)

• sHr = habit strength

  • depends on contiguity and reinforcement

  • reinforcement = drive (d) reduction

  • sHr rises to asymptote

• sIr = conditioned inhibition

  • an S-R link that interferes with performing the target response

  • formed as Ir fades

• Ir = reactive inhibition

  • analogous to fatigue

  • drive state

  • fades w/ time

• sOr = random error

  • impossible to fully eliminate confounding variables

  • ex. a loud train passes by the lab where a rat is being trained. the animal hears the train whistle and freezes

• sLr = reaction threshold

  • the smallest amount of reinforcement that will produce learning

Summary from Hull

• behavior can sometimes be modeled mathematically (even if his model didn’t work well)

• performance =/= learning

Tolman Reacted Against the S-R View

drew on Gestalt Psychology

• argued that animals could form associations between neutral stimuli (Latent Learning)

• S-S relationships

• S1-R-S2 expectancies

• assumed reinforcement affects performance

• learning occurs when contiguity is present

cognitive inference

• if S1→R→S2 and S2→S-* THEN S1→R→S*

Tinklepaugh’s Experiment

comparison organisms encode the reinforcement

Tolman’s Mazes

Tolman placed rats in a maze that contained in a series of binary choice points

three groups:

• food reward