Recognition Learning

1
Recognition Learning

• Psychology 3906

2
Introduction

• Basically Differential responding to something
  previously experienced
• Could be responding in some special way the first
  time a certain stimulus shows up

3
Habituation

• Decrease in responding over time
• Not too much generalization, though there is some
• More presentations the better
• Learning vs performance and habituation below
  zero
• Gilette and Bellingham (1982) showed that rats
  learn about a stimulus

4
Effects on behaviour

• Can be very broad, gill withdrawal to a decline
  in exploration
• Sherringtons Model (S-R)
• Sokolovs Model (Comparator)
• Wagners Model (SOP)

5
Perceptual Learning

• Characteristics of stimuli, not their
  relationship to each other
• Gibson and Walk (1956)
• Could be learning to associate the individual
  featerus with each other
• William James thought so!

6
Imprinting

• Lorenz said this was special as it
• Has a critical period
• Is irreversible
• Affects behaviour well into the future
• Generalizes to all members of the species

7
Conditions for imprinting

• Batesons running wheel
• Simple exposure is enough usually
• Sometimes it is certain features, say like the
  head and neck in the Burmese Red Junglefowl
• The longer the exposure the better

8
Is this just associative learning?

• Well sorta
• Shows blocking, overshadowing
• Still different though, sensory period,
  competitive exclusion
• Reversible, but not totally, depends on stimulus
• With sexual imprinting you can get combining\
• Sexual imprinting probably isnt.

9
Altruism

• I thought genes were selfish
• Well you thought right
• Kin selection (Hamilton)
• Reciprocal altruism (Trivers)
• Nicest example of the effect of genetic
  relationships is in eusocial insects

10
Eusociality

• For Eusociality we need
• Overlapping generations
• Cooperative care of the young
• Sterile castes
• In insects, euscociality occurs in
• Hymenoptera
• Ants
• Bees
• wasps

11
Eusociality

• Isoptera
• Termites
• Homoptera
• Aphids
• Some say that these two are not truly eusocial,
  but if you look at that definition, you could,
  under certain circumstances, say bees were not
  eusocial!

12
Haplodiploidy

• Eusociality has evolved 11 separate times just in
  Hymenoptera!
• Why?
• Sterile females are usually very closely related
• Basically, it does not pay to have young, it pays
  more to take care of your sisters, as they are
  more closely related to you than any potential
  offspring!

13
Haplowhatoidy?

• Daughter Son Mother Father
• .5 .5 .5 .5 female
• 1.0 0 1 0 male
• Sister Brother
• .75 .25 female
• .50 .50 male

14
Males are just clones of half of female
chromosomes
So you see, sisters are either 75 related, or
100 related or 50 related, averages out to 75
15
How are kin recognized?

• Directly? Probably not
• Indirectly seems more likely
• Cross fostering experiments
• Beldings ground squirrels, for example, treat
  familiar as kin
• Not the whole story though
• Phenotype matching
• MHC genes
• Reciprocal altruism and vampire bats

16
In conclusion

• Recognition learning has some characteristics
  that other forms of learning do
• And the opposite of that is true