Chapter 6

Non-Associative Learning:

Learning about Repeated Events

6.1

Behavioral Processes

3

6.1 Behavioral Processes

• Learning about Repeated Stimuli

• Learning and Memory in Everyday Life— Sex on the Beach

• Perceptual Learning

• Models of Non-Associative Learning

4

Behavioral Processes

• Non-associative learning—learning involving only one stimulus at a time In contrast, associative learning is learning to

associate one stimulus with another or to a new response.

5

Learning about Repeated Stimuli

• Habituation—lack of response to originally noticeable stimuli. A widespread, basic form of learning

Automatic, reflexive

• Examples: Learning to ignore the hum of the air conditioner.

Learning to ignore traffic sounds.

Getting used to wearing glasses.

Your dog getting used to your whistling.

6

The Process of Habituation

• For study, researcher use the simple, controllable acoustic startle reflex. Defensive response to a loud, unexpected noise

• Researchers also use orienting response. Innate response to a novel stimulus

Examples:Infants turn their head and gaze at unfamiliar visual

stimuli.

Dogs cock their head in response to novel stimuli.

7

Habituation

(a) Adapted from Davis, 1980; (b) adapted from Malcuit et al., 1996.

8

Factors Influencing Rate and Duration of Habituation

• Massed exposures to stimuli facilitate habituation. After a period of no stimulus presentation,

spontaneous recovery (response reappearance) can occur.

• Dishabituation—novel stimulus can renew reflexive response. e.g., you make a new sound and your dog cocks

his head again.

9

Learning and Memory in Everyday Life—Sex on the Beach

• Sexually explicit photos and recordings elicit sexual arousal in male undergraduates.

• If the stimuli are repeated, sexual habituation occurs, even without conscious awareness.

• Dishabituation may be facilitated with novel stimuli, locations or lovemaking techniques.

10

Sexual Habituation and Dishabituation

Adapted from Koukounas and Over, 2001.

11

The Process of Sensitization

• Sensitization—a startling stimulus leads to a strong response to a later stimulus. New stimulus might otherwise evoke a weaker response. e.g., strong electric shock increases rats’ startle response

to future loud noise for 10–15 minutes.

• Humans can show sensitization of their startle reflexes. Skin conductance response (SCR)—change in skin’s

electrical conductivity; response to emotion.

12

Sensitization of the Rat Acoustic Startle Reflex

Adapted from Davis, 1989.

13

Priming

• Priming—prior exposure to a stimulus can improve an organism’s recognition later.

• In humans, priming often studied with word-stem completion task. Fill in the blank:

MOT_____ / SUP_____

Examples:

motel, motor, moth / suppose, supper, supreme

14

Priming in Blue Jays

• In blue jays, recent observations of one kind of moth “primed” them for later recognition.

Adapted from Bond and Kamil, 1999.

15

Perceptual Learning

• Perceptual learning—prior experience with a set of stimuli make those stimuli easier to distinguish from each other. Increased ability to make fine distinctions among

highly similar stimuli.

• Examples: Chicken-sexers

Medical diagnosticians

Dog show judges

16

Mere Exposure Learning

• Mere exposure learning—occurs with only exposure to stimuli, no explicit prompting.

• In studies: Rats pre-exposed to shapes learned to differentiate

shapes more quickly than rats with no pre-exposure.

Over time, humans learned to distinguish a target scribble without feedback on accuracy.

• Related to latent learning—original learning is not observed until a later time.

17

Mere Exposure Learning in Humans

Adapted from J. J. Gibson and Gibson, 1955.

18

Discrimination Training

• Expert perceptual training includes: Distinguishing among examples.

Receiving feedback on accuracy.

Mere exposure is part of this process.

• In studies: Participants improved perception and discrimination of

tactile stimuli with training.

Perceptual learning has learning specificity (does not transfer automatically to discrimination of other stimuli).

19

Learning Specificity in Humans

Adapted from Fiorentini and Berardi, 1981.

20

Spatial Learning

• Learning one’s environment or surroundings (much is latent learning).

• Tolman studies: rats ran maze over 22 days. Group 1 received food for reaching food box on every

trial. Over time, learned maze with few errors.

Group 2 received no food for ten days; on 11th day, began to receive food. Exposure-first rats showed rapid learning; ultimately showed fewer maze errors than group 1.

21

Rat Learning by Exploration

Adapted from Tolman and Honzik, 1930.

22

Spatial Learning in Wasps

• Tinbergen studies: Encircle wasp burrows

with pinecones.

After orientation flights, wasps leave to find food.

Next, researchers move pinecone circles.

Returning wasps search moved pinecone circles for burrows.

Adapted from Tinbergen and Kruyt, 1972 (1938).

23

Models of Non-Associative Learning

• What processes might be involved? Dual Process Theory

Comparator Models

Differentiation Theory

• Each theory may explain certain features of non-associative learning.

24

Dual Process Theory

• Dual process theory—habituation and sensitization are separate but parallel processes. Operate in the same manner; do so

independently.

Similar to neural circuits in cat spinal cords.

25

Dual Process Theory

(a) Adapted from Groves and Thompson, 1970.

26

Comparator Models

• Comparator models—habituation is a special form of perceptual learning. Brain experiences a stimulus and develops neural

representation.

Compares to existing representations of previously experienced stimuli.

If no match, respond is triggered (e.g., an orienting response).

If match, behavioral response is suppressed (i.e., habituated).

27

Differentiation Theory

• Differentiation theory—brain develops representations over time; incorporates new details each time stimulus is presented. Brain has limited processing capabilities.

Develops a vague mental representation from first stimulus exposure.

Mental representations become more detailed with each subsequent exposure.

28

6.1 Interim Summary

• Habituation = decreased strength or frequency of behavior after repeated exposure to the behavior-producing stimulus. In spontaneous recovery, a behavior may reappear

at original level if stimulus is presented again after a delay.

Behavior decreased through habituation can also be renewed (dishabituated) by a novel stimulus.

Habituation is stimulus-specific.

29

6.1 Interim Summary

• Sensitization = increased response to a stimulus. Exposure to a threatening or highly attractive stimulus

causes a heightened response to any stimuli that follow.

Sensitization is not stimulus-specific.

• In priming, prior exposure to a stimulus improves the organism’s ability to recognize that stimulus later.

30

6.1 Interim Summary

• In perceptual learning, experience with a set of stimuli improves the organism’s ability to distinguish those stimuli. In mere exposure learning, simply being exposed to the

stimuli results in perceptual learning. Related term = latent learning (learning without

corresponding changes in performance).

Perceptual learning can also occur through discrimination training.

Organism learns to distinguish stimuli via feedback about stimulus class.

31

6.1 Interim Summary

• Many kinds of spatial learning take the form of perceptual learning. Is often latent learning that results from mere

exposure as the organism explores its world.

• Comparator models suggest that habituation is a special case of perceptual learning.

32

6.1 Interim Summary

• In dual process theory, changes in behavioral response after repeated stimulus exposure reflect combined effects of habituation and sensitization. Habituation decreasing responses; sensitization

increasing responses.

• In differentiation theory, perceptual learning results from new details being added to existing stimulus representations.

6.2

Brain Substrates

34

6.2 Brain Substrates

• Invertebrate Model Systems

• Perceptual Learning and Cortical Plasticity

• Unsolved Mysteries—Why Did Cerebral Cortex Evolve?

• The Hippocampus and Spatial Learning

35

Invertebrate Model Systems

• Aplysia—sea slugs with simple nervous systems, large neurons Gill-withdrawal reflex

Simple sensory-glutamate-motor reflex closes gill when siphon is touched.

• Aplysia serve as a simple model for non-associative learning.

• Dual process theory provides best explanation for this example.

36

Aplysia Californica

Da

vid

Wro

be

l/V

isu

als

Un

lim

ite

d

37

Neural Circuits in Gill-Withdrawl Reflex

38

Habituation in Sea Slugs

• Occurs rapidly, can endure for 10–15 minutes.

• Associated with decreased glutamate release. Synaptic depression—less presynaptic terminals

on sensory neurons and eliminated synapses

• Homosynaptic—involves only synapses activated during the habituating event. Changes in a siphon sensory neuron will not affect

sensory neurons in the tail or mantle.

39

Sensitization in Sea Slugs

• Mild tail shock (T) sensitizes response to stimulation of siphon and upper mantle.

• Heterosynaptic—involves changes across several synapses (including those not activated by the sensitizing event). T’s interneuron releases a neuromodulator, (e.g. serotonin);

arouses other sensory neurons (siphon and upper mantle).

• Repeated touch to the siphon will elicit a stronger gill closure.

40

Sensitization in Aplysia

41

Perceptual Learning and Cortical Plasticity

• Within sensory cortices, different neurons respond to different properties of stimulus.

• Receptive field—the range of physical properties to which a neuron responds. e.g., an auditory neuron might have a range of

800–900 Hz.

Field can change with experience, contributing to cortical plasticity—change in cortical organization.

42

Receptive Field in Guinea Pig Auditory Cortex

Adapted from Weinberger, 2004.

43

Cortical Changes after Mere Exposure

• In one study: After 3 hours finger tip stimulation, participants

showed temporary discrimination learning.

fMRI shows greater activation in the left somatosensory cortex.

• In a similar study: Magnetoencephalographic (MEG) recording showed

positive association between changes in cortical activity and quality of tactile discrimination.

44

Cortical Reorganization after Finger-Tip Stimulation

Fig

ure

s c

ou

rte

sy

of

Be

n G

od

de

, J

ac

ob

s C

en

ter

for

Lif

elo

ng

Le

arn

ing

, J

ac

ob

s U

niv

ers

ity

Bre

me

n,

Ge

rma

ny

45

Cortical Changes after Training

• Cortical areas may increase or decrease after discrimination training.

• Decreases may reflect fine tuning within the cortex. Quality over quantity of neural firing

• Changes correspond to the comparator models of non-associative learning.

46

Plasticity during Development

• Neuroimaging studies found visual association cortical areas more activated in blind people when engaged in Braille reading and other tactile tasks. Compared to sighted people.

47

Cortical Reorganization of Opossums

• Different cortical mapping was observed in opossums blinded at birth. Blinded opossums

showed unique multimodal areas, responding to combination of auditory and tactile stimuli.

Adapted from Kahn and Krubitzer, 2002.

48

Hebbian Learning

• What is the neurological mechanism underlying cortical plasticity?

49

Hebbian Learning

• Donald Hebb (1949) proposed that neurons that fire continuously strengthen their connections to each other.

• Hebbian learning—forming firing patterns among contiguous neurons, speeds and strengthens behavioral responses. Neurons that fire together, wire together.

Corresponds to long-term potentiation (LTP).

50

Unsolved Mysteries—Why Did the Cerebral Cortex Evolve?

• The simplest multicellular organisms with a nervous system have sensory and motor neurons.

• A cortex is not necessary for complex learning and memory. As seen in the octopus

• Cortex may have evolved to give the brain its ability to reorganize neural connects. A mechanism for learning

51

The Hippocampus and Spatial Learning

• Rats with hippocampal lesions have difficulty learning spatial tasks. e.g., radial maze

• Humans with medial temporal amnesia also may have problems with spatial tasks.

52

The Hippocampus and Spatial Learning

• Some neurons in rats’ hippocampal region fire only when rats move into specific external locations. O’Keefe calls neurons place cells.

• Place fields—external locations associated with place cells’ maximum response.

53

Identifying Places in Rats

• Place cell firing may be mediated by visual input (Cues or landmarks).

• Visual stimuli change andplace cell location may also change. Place cell location may rotate accordingly.

Ad

ap

ted

fro

m L

en

ck-S

an

tini e

t a

l., 2

00

1.

54

Place Fields Are Not Maps

• Place cell-to-place field associations may develop during learning, but cells may be recycled to learn new spatial locations.

• Rats’ place cells describe specific locations; do not form two-dimensional map. Beyond place cells, information is needed to form a

cognitive map.Researchers still looking

Possibly from cortical areas

55

6.2 Interim Summary

• In marine invertebrates (e.g., Aplysia), habituation = a form of synaptic depression in circuits that link a stimulus (sensory neuron) to a reflexive response (motor neuron), as proposed by dual process theory. Habituation in Aplysia is homosynaptic; changes in one

sensory neuron do not affect other sensory neurons.

Sensitization in Aplysia is heterosynaptic and reflects increases in synaptic transmission.

56

6.2 Interim Summary

• Cortical plasticity = cortical networks ability to adapt to internal or environmental changes. During perceptual learning, cortical changes parallel

improvements in discrimination abilities.Includes refinement of neurons’ receptive fields in

response to sensory inputs.

Can lead to widespread changes in cortical map. In extreme cases (e.g., a form of sensory input is absent

from birth), cortical map may reorganize so that active inputs take over the areas normally devoted to processing the missing inputs.

57

6.2 Interim Summary

• Hebbian learning, based on the principle that neurons that fire together, wire together. Repeated exposure can strengthen associations

within particular subsets of cortical neurons.

Subsets then provide an increasingly reliable basis for discriminating the stimuli that activate them.

A mechanism for cortical plasticity.

58

6.2 Interim Summary

• Place cells = hippocampal neurons; become most active when animal is at a particular location (the place field for that neuron). Unclear how information from different place cells is linked

together to form a useful spatial map for environmental navigation.

• Place fields change with learning; if place cells are disrupted, so is spatial navigation. With environmental familiarity, corresponding place cells

become more selective, responding to increasingly precise locations in environment.

6.3

Clinical Perspectives

60

6.3 Clinical Perspectives

• Landmark Agnosia

• Rehabilitation after Stroke

• Man–Machine Interfaces

61

Landmark Agnosia

• Stroke may cause lesions in parahippocampal region (cortex near hippocampus). Extend into the visual cortex.

Adapted from Takahashi and Kawamura, 2002.

62

Landmark Agnosia

• Patients may: Become disoriented in known or new locations.

Have problems recognizing familiar locations in photos.

Have problems recognizing faces (prosopagnosia).

63

Rehabilitation after Stroke

• Stroke may cause significant loss in perceptual function (e.g., desensitized arm). Leads to the patient favoring the unimpaired arm.

• To counter learned non-use: Place mitt (constraint) on unimpaired arm for most

of the day; forces patient to use desensitized arm.

May increase desensitized arm use in some patients.

64

Overcoming Learned Non-Use

Adapted from Taub, Uswatte, and Elbert, 2002.

65

Man–Machine Interfaces

• Sensory prostheses—mechanical devices; interface with neurons to produce sensation.

• Example: cochlear implant Electrically stimulates the auditory nerve.

Need training to interpret these “virtual sounds” (perceptual learning).

Training yields initial rapid improvement with slower gains in discrimination learning over time.

Recency of hearing loss is a factor in implant success.

66

Cochlear Implant

Adapted from Clarke, 2002.

67

6.3 Interim Summary

• Landmark agnosia = inability to identify familiar buildings and landscapes. Condition often results from damage to the

parahippocampal region of the cortex.

68

6.3 Interim Summary

• After stroke, many patients experience large losses in perceptual and motor function. May suffer from learned non-use.

A functional limb takes over the role of a limb that still has motor function but has lost sensation.

Learned non-use can be overcome by restraint therapyIndividual is forced to use desensitized limb.

Recovery of function in stroke patients is thought to result from cortical plasticity.

69

6.3 Interim Summary

• Sensory prostheses = electronic devices that interface directly with neurons or sensory receptors. Designed to provide sensory processing capabilities

individuals would not otherwise have.

• Cochlear implant: Most extensively developed sensory prosthesis; used to

treat profound deafness.

Training leads to perceptual learning; improves ability to discriminate simulated speech sounds.