PSYC 361 MT2: HIGHER LEARNING (4): Cognitive Ability In Behavioural Modulation Flashcards
“Higher” Learning: More than Stimulus-Centric Learning
Non-associative & associative learning is learning centred around nature of stimuli:
- Surprising/noxious
- Repeated/Redundant
- Predictive
- Appetitive/Aversive
Higher learning does not always follow same principles; animals learn to extract certain aspects of stimulus properties (ie. category, rule) & apply learning to new situations in which novel stimuli still trigger the proper responses
Cognition: The Basis of Higher Learning
(Tolman): among 1st to introduce concept of cognition in psychology & defined it in terms of “object-adjustments”/what could be called expectations in contemporary language
Idea that learning involves acquisition of an expectation that a given behaviour will bring the organism closer to goal object,
- cognition gives behaviour appearance of purpose & goal-direction (Tolman)
Contemporary definitions of cognition more general, defined as “the mechanisms by which animals acquire, process, store & act on info from the environment; include perception, learning, memory & decision making”
Examples of Higher Learning
Concept Learning: learning about specific property/quality of world & separate them into classes & knowledge is used for generalization, discrimination & inference
Social Learning: happens in social environment through observation & imitation of others’ behaviour; can happen even i absence of motor reproduction/direct reinforcement
Play: behaviour with no defined goals, performance improved in similar future situations; children & young animals often engage in play to interact with world
Harlow & Primate Intelligence Studies
American psychologist working with primates:
- research at university of Wisconsin-Madison
- 1 of earliest to experimentally study cognitive learning in primates— demonstrated cortex’s role for learning ability
Harlow & Primate Intelligence Studies— THE WISCONSIN GENERAL TEST APPARATUS
Department of Psychology refused request for research facility, need up taking over an abandoned building & transformed into Psychology Primate Laboratory
Developed WGTA with collegues, focused on primate learning
1 major research focus is rule-based learning & conceptual i action
Harlow & Primate Intelligence Studies— “THE FORMATION OF LEARNING SETS”
(Harlow): widely viewed as evidence of higher learning & insightful behaviour
”Learning to Learn”: transfer learned knowledge to new environment to guide behaviour
Harlow & Primate Intelligence Studies— “THE NATURE OF LOVE”
Maternal behaviour could be “motivated” such that mother’s genes can be passed onto next generation by increasing the offspring’s survival— “love” also carries biological purpose
Designed series of “surrogate mother” studies to examine “nature of love”— why does mother-infant attachment form? (Harlow)
Findings “demolished” both classic psychoanalytic view & drive reduction theory
INFANT MONKEYS SHOW PREFERENCE FOR CLOTH MOTHER
- fear induced withdrawal towards cloth mother
- attachment to cloth mother in stressful environment
Harlow & Primate Intelligence Studies— FOOD FOR THOUGHT
Not humane study; Harlow + colleagues performed experiment in which cloth mother provided unpredictable stressors (maternal rejection) for purpose of blocking the development of attachment
Early-Childhood Stress Alters a Wide Range of Behaviour
Maternal care is so important for psychological development of infants
When maternal care quality changes, wide range of behavioural changes follow:
- infants may learn through poor maternal care quality that world is not safe
- neurobiological, cognitive & behavioural traits change according to help coping with adverse environment
Comparative Approach in NRSC
Comparative approach studies similarities & differences in behavioural & biological (neural) organizations among living beings; bacteria—plants—humans
- Interested in psychological nature of humans in relation to other animals
- 1 key assumption is that cross-species comparisons can be related to phylogenetic relationship between these species
- Tinbergen’s 4 questions:
1) Function
2) Evolution
3) Causation
4) Development
Honeybee as Model for Cognitive Learning
Have sophisticated social behaviour & remarkable cognitive capacity to process environmental info; behavioural complexity supported by its “large” enough (highly differentiated brain areas of different functions) yet “small” enough (tractable size) brain
Navigational & Communication Behaviour in Honeybees
Karl Von Frisch studied honeybee behaviour particularly dancing behaviour
- discovered that when bees find nectar in flower, they fly in a “special” pattern
- perform kind of dance that conveys to other bees where to find nectar
- idea that some animal & human patterns of behaviour are innate
Systematic NRSC Research in Honeybees
Randolf Menzel:
- pioneered honeybee as model system in NRSC with respect to colour, vision, olfaction, learning & memory
- combined levels of analysis from natural behaviour to single neurons; traced perceptual & cognitive capacities to neural & cellular substrates
- 1st evidence for role of insect mushroom body in memory formation & characterized cellular & neural correlates of different phases of memory
Honeybees Exhibit All Forms of Non-Associative & Associative Learning
Conditioned Proboscis Extension Reflex (PER) is a form of Pavlovian CC
- novel odour (CS) through pairing with sugar water (US) gains predictive value of US & triggers conditioned PER (CR)
Locating the CS-US Engram in the Honeybee Brain
Engram: the (hypothetical) substrates of memory; enduring physical/cellular/biochemical changes induced by learning & changes expression of behaviour
- tractable NS of honeybee allows for identification of “engram”
Traces of Memory in the Honeybee Brain
VUMmx1 normally responds to sucrose (US) of the proboscis but does not respond to olfactory stimuli (CS); after pairing of sucrose + door, will respond to conditioned odour
The Reward Prediction Error is encoded in VUMmx1 & resembles properties of DA neurons in the mammalian VTA
- VUMmx1 is octopaminergic: contrast to role for DA in mammals
VUMmx1 activation can substitute for reinforcing effect of US
The Mushroom Body seems to act as encoding device, converting sensory info to value-based info— shares properties with mammalian hippocampus & PFC
Complex Learning in Honeybees (Symmetry)
Honeybees learn to extract bilateral symmetry/asymmetry from series of different patterns & can transfer this info to novel stimuli
Integrate symmetry & asymmetry into unique concept— after having learned 1, do not need to relearn other if it is reinforced in reversal-learning experiment
- when bees learn about symmetry they also learn about asymmetry
Showcases the Cognitive Capacity of Honeybees
Bees need not only discriminate between patterns but must extract common feature (symmetry) & learn to associate it with reward.
Bees not only able to perform this generalization but can learn reciprocal task much more quickly; demonstrating certain capacity for abstraction
None of existing models for pattern recognition can explain this performance
Must assume insects’ ability to perceive patterns is much more powerful & flexibly than we had hitherto believed
What Other Complex Learning Can Bees Do?— RULE EXTRACTION: DELAY MATCHING-TO-SAMPLE TASK
Not only can bees extract rule & apply to new situations within same sensory modality (vision) but also to different sensory modality
This experiment: bees appeared to transfer rule learned from a visual task to an olfactory task + vice versa
Rule Extraction Summary
Results similar between delayed matching-to-non-sample task
- learn to choose stimulus different from sample
Because bees continue to choose appropriate matching (/non-matching) stimulus even in new situations, concluded that they can form & use concept of sameness & difference in making their choices
Ability both broad & robust— can be transferred across domains & modalities
Prospects for Future Research
Cognitive capacity for concept learning considered cornerstone of human cognition, demonstrated in honeybees
Minimal neural architectures capable of extracting regularities underlying concept formation; this capability also achieved in absence of PFC with simpler & smaller neural structures; may/may not be implemented by same neural principles compared to cognitive abilities in humans
Little is known about how bee NS supports this cognitive ability; future: identify neural circuits that mediate these forms of learning
Limitations with Honeybees
Often studied as wildly caught bees; ecological validity BUT:
1) Seasonal Effects: seasonal changes in performance (ie. different reversal learning in summer vs. Winter); may reflect differences in food availability (motivation?)
2) Cannot control previous experience: behaviour may be altered by prior learning
3) Cannot raise them outside colony & can only survive several months if colony is removed from environment
4) Genetic manipulation possible but very inefficient; comes with risk of contaminating gene pools for wild bees, leading to detrimental effects to environment of much larger scale
Summary of Higher Learning
Cognitive learning ability is less stimulus-centric, rather learning of the categories, concepts, rules etc, is transferred to new situations to achieve behavioural goals
Harlow’s pioneering work showed that some forms of learning is incremental but others are insightful
Early childhood stress can affect a wide range of behaviour
Comparative approach in NRSC research allows for cross-species comparisons to understand behaviour & reduce research questions to tractable size, facilitated by use of model organisms
Honeybees demonstrate complex behaviour & remarkable cognitive capacity; its powerful yet tractable NS, complex learning such as **concept learning* & rule extraction can be achieved