L15 - Spatial Learning Flashcards
What is spatial navigation?
Ability to use space to reach a goal:
- forging for food
- finding a mate
- locating a home
Geometric Information
Encoding scalar (magnitude) and vector (mag and direction) properties e.g. distance, direction and length
Allows you to map out shape of the env
Featural Information
Encoding surface marking and form
E.g. colour, pattern and shape
Position land marks on a cognitive map
Spatial Cues, geometric and featural information experiment
Cheng 1986
3 black walls one white wall
- Rat in the area
- Food in the top right corner
- Can encode location of the food in terms of geometrical and featural info
○ Geometrical - in the corner where long wall is on the left and short wall that is on the right,
○ Featural - located in the corner that is black and white
- In a rectangular area you have 2 geometrically equivalent corners
- Also have 2 featural corners
Cheng 1986
What happens if you disorientate the rat?
Use the cues that are more salient to reorientate themselves
○ Geometric cues are more salient, when you put them back into the arena half will go to correct corner, the other half will show geometric confusion and go to the other corner
- Confusing as they are geometrically symmetrical but featurally distinct
□ After one trial they only learnt geometric cues not featural
□ When disorientated they rely on geometric cues
What did Hermer and Spelke (1996) find?
Children at 24 months confuse geometrically symmetrical corners
What is allocentric spatial language?
Using featural information to break the geometric symmetry
- being able to describe things to an objective framework
Why might geometry prevail over featural reorientation?
There may be a (separate) geometric module in the brain:
- exclusively encode shape information
- Stored separately from any landmark information
Evolutionary explanation as to why geometry prevails
Not testable
When searching for food and shelter, they always stay the same, colour of a forest can change
How is a geometric module different from a cognitive map?
GM - idea of a region of the brain purely for geometric info
Cognitive map- forming a mental representation of an env as you go out in life
Roberts and Pearce (1999)
Water maze experiment (can’t see what is below water)
Design:
Cues around the pool so they want to get to the platform and stay dry
- When they find it on the first trial they learn where it is for the next trial and use cues to find it again (1 trial)
- In the blocked group we a training rats to learn using a beacon, landmark conditioned at the goal (8 trials)
Roberts and Pearce (1999)
Removing beacon: Findings
Shorter amount of time means you have learnt it better.
Time spent in quadrant is very low in blocked group - can’t find it well
control groups can use the cue to find the goal
Landmarks can block other landmarks - competing CSs
Rules governing space link to rule of associative learning
Pearce et al (2001)
Geometric module testing
Design
Phase one block group learn beacon relationship
phase two: all groups introduced to beacon and corner shape
Corner + beacon, with the corner cue you learn platform is located
Pearce et al:
Get rid of beacon, keep corner Findings
No blocking!
- They have learnt in second phase, phase 1 training has failed
- The lack of blocking isn’t due to lack of learning in phase one
○ They do have good learning, can find it based on the beacon
○ Not sufficient to block the shape
Support for geometric module? - learning about features isn’t able to block geometry - separate!
Alternative explanation to Pearce findings
Geometric cues are more salient to finding a platform
Just like taste for food
Using parallelogram and obtuse angles can lead to blocking
