Thursday, 26th July 2018 - Social Learning Flashcards
In today’s lecture:
- Introduction.
- Processes involved in social learning.
– Enhancement in rats, bumblebees.
- Social learning transmission.
- Discussion of culture.
– Japanese macaques.
• Imitation example.
– ‘Do as I do’ procedure.
Learning outcomes:
- Be able to describe and explain the main processes involved in social learning, using examples.
- Demonstrate understanding of who animals learn from (the various possibilities).
- Understand the concept of culture, including what defines it.
- Have knowledge of imitation, and the example:
Do As I Do procedure.
Social Learning We often think about individual learning.
• Individual experiences, trial and error learning.
Social learning = changes in an individual’s behaviour resulting from attending to another’s behaviour or its products.
(van Schaik, 2010, and references therein.)
- Social species.
- Maybe those with long parent-offspring associations.
Various behavioural processes:
• Enhancement.
– Stimulus enhancement.
– Local enhancement.
• Observational conditioning.
• Goal-directed emulation.
• Imitation.
Not always necessary for animals to observe/understand the precise actions performed by another for social learning to take place.
• Attention to a location/object/individual is required.
Enhancement:
- A naïve animal learns something new via the presence/behaviour of a conspecific.
- Stimulus enhancement = another animal’s presence/behaviour near an item, attracts the naïve animal.
– E.g., mates, carcass.
– E.g., pointing experiments
(Research Seminar lecture).
• Local enhancement = another animal attracts the naïve individual to a location.
– E.g., feeding areas.
• This learning can be indirect (non-observational).
– It isn’t necessary for naïve animals to attend to the demonstrators’ actions.
– E.g., signs of foraging but not direct observation.
Rats given strongly flavoured foods and put to sleep, put next to observer rats and later observer rats had preference for food sleepy rats had eaten despite being a generalist species.
Perhaps like how a human is triggered to crave a certain food once being exposed to the smell stimulus of it.
Social learning has been studied with rats.
Example: Heyes et al. (2000) examined rats’ attraction to levers depending on conspecifics’ interactions with the levers.
- Demonstrator rats: left / right levers. (100 R+.)
- Observer rats: screen / no screen to dis/allow viewing of demonstrator.
• Observers were then placed in chamber and responses on both levers were reinforced.
• Observers were then placed in chamber and responses on both levers were reinforced.
DEMONSTRATOR RAT PRESSING LEFT
What they found was the observers who could actually see the demonstrator rat and what side they were pressing on, when they got the chance to go in and press the leaver, they responded more on the same side as the demonstrator rat.
No screen = high proportion of responding to left (same as demonstrator)
Screen =
Great example of stimulus enhancement because when observers could see exactly where their demonstrators were pressing, observers too showed a very strong biased to pressing in the same location.
You would expect a rat working 100x to press a lever to leave odour cues that conspecifics would be able to pick up on. The non-observer rats who had the screen couldn’t see the demonstrators so they should have still gone to the leaver that the demonstrator were working busilly on behind the screen but it was what they could see that influenced the observer behaviour mostly which is why the observers went mostly to the lever they could see.
Example: Worden & Papaj (2005) examined flower choice copying in bumblebees.
• Observer bees watched: – Live bee models: demonstrator bees trained to forage on either orange/green ‘flowers’.
– Artificial bee models: model bees on orange/green ‘flowers’.
Used bumble bees (social animal)
Trained obsever bees to either forage on green or orange flowers by controlling demonstrator model bees.
Both demonstrator bees (model and real) had an affect on where the observer bees went next.
Example: Worden & Papaj (2005).
- Model bee (live/artificial) colour choice significantly affected choice of observer bees.
- May have implications for how colonies monitor changes in food resources – through copying non-nest mates to exploit new food sources.
*Conspecific responding to demonstrator actions (prey animals/threats)
Elliciting fear responses in other animals. Reese’s monkeys born in laboratory and not exposed to natural predators in the same manner as wild born monkeys, they might show no fear response initially to snakes but when they observe wild born monkeys seeing snakes and seeing their fear response they too will rapidly learn to respond in the same way to that danger stimulus.
Predisposition to biologically significant things such as a major threat to survival such as a predator. An inherint ability or tendancy to behave in that way by observing others behave that way they are able to learn that response and adopt it in a similar situation.
ADVANTAGE OF THIS:
Avoiding injury/death
More likely to survive
Goal-directed emulation:
• Naïve animal attempts to achieve the same goal
– but produces own behaviour pattern in doing so.
– Not the same topographical behaviour.
– E.g., chimps using a rake to retrieve food.
Imitation: Naïve animal replicates another animal’s behaviour to achieve goal.
Observational social learning is considered more cognitively demanding than non-observational.
Adaptive Value of Social Learning Is there adaptive value to social learning?
• Reduced costs.
– Time taken for individual learning.
– Energy expenditure, risk of predation during foraging, etc.
Who do animals learn from?
• Vertical/oblique social learning.
– Vertical (offspring from parents)
• E.g., infants following and observing mothers foraging.
– Oblique (juveniles from other adults)
– If there is a choice of demonstrators, there can be a bias towards more skilled (and observation-tolerant) individuals.
• E.g., nut-cracking capuchins.
• Horizontal social learning (Across peers)
A group of animals young or old foraging at the same time, conflict at times with members of group. Information produced by conflict, learn something about pack members such as weaker and stronger animals.
- E.g., foraging areas.
- E.g., individuals eavesdropping on the outcomes of conflict.
A group of animals possess particular behavioural traits which are transmitted via social learning. • Innovative behaviour, learned via other individuals. • Transmission can be vertical or horizontal. • AKA traditions. – Debate over definition. – Some: transmission should only be via imitation (or teaching).
Culture
A group of animals possess particular behavioural traits which are transmitted via social learning.
- Innovative behaviour, learned via other individuals.
- Transmission can be vertical or horizontal.
- AKA traditions.
– Debate over definition.
– Some: transmission should only be via imitation (or teaching).
Example of potato-washing and wheat-in-water, Japanese macaques.
Some contention over actual culture.
• Is it transmitted via social learning (definition of culture)?
– E.g.,speed of learning.
If its social learning you would expect it to transmit through the group very fast because your going to see some individuals perfoming it observed by a group of observers which would lead to more demonstrators.
• Question: what are some alternative explanations, how else could other memebers observers naive animals in this group be learning this behaviour?
-Individuals might just give it a go
–Thirsty when they eat them/makes them softer?
-Oppertunisticly near the water and accidently dropped them in
-Vertical learning - parent to offspring
-Horizontal learning - Pear to pear
-Immitation learning from human caretakers
-Training by care takers to promote tourism (Positive reinforcement training)
Imitation – Do As I Do
The ‘Do as I do’ (DAID) training paradigm takes advantage of social learning in dogs.
- It has been used to examine imitation abilities of many species, including great apes, parrots, dolphins, and dogs.
- Training is initiated by training dog to perform a small number of simple tasks via operant conditioning, on command “do it”.
- Then novel behaviours are demonstrated to the dog, and they’re instructed “do it”.
- See Topal et al. (2006) for full details.
http: //www.youtube.com/watch?v=n013vP3xuEs&index=10&list=UU40ktPZovMjkswsXxVB7NWg
Imitation – Do As I Do
Fugazza & Miklosi (2014) compared DAID and shaping.
30 dog-owner dyads:
– 15 experienced with shaping.
– 15 experienced with DAID.
Owners required to train three new behaviours in 15 min each (breaks allowed): simple, complex, sequence.
Results:
– Simple
– no significant differences.
– Complex
– DAID dogs had significantly shorter learning latencies.
– Sequences
– significantly more DAID dogs learned the tasks.
Authors note preliminary training for both methods wasn’t accounted for, and comparison of learning rates is difficult.
Overall, DAID may be a useful addition to current popular techniques, esp. for complex tasks.
https://www.youtube.com/watch?v=bOkj7lJpeoc
Reading – Social Learning
Cornell et al. (2011). Social learning spreads knowledge about dangerous humans among American crows.
