Semantic Mapping

Question 1

Boobaa- Kiki phenomenon

Answer 1

• Non-arbitrary association of sounds with visual shapes
• Round shape = boobaa
• Angular shape = kiki
• Aspect of human perception
• Robust effect
• Observed in non-neurotypical + neurotypical populations
• Link between evolution of lang + innate perception of shape

Question 2

Maurer et al (2006) - boobaa + kiki

Answer 2

• Effect is robust across multiple stimuli
• Systematic bias (most over 50%) even in children with dev disorders
• People associated certain sounds with spikey/round shapes
• E.g. Baa-moo vs Kuh-Tay
• Goo-ga vs Tee-tay

Question 3

Explaining Kiki/Boobaa effect

Answer 3

• Ramachandran and Hubbard (2001) speculate that it arrises from cortical connections among contiguous brain areas that unite:
  1) Visual perception of a nonsense shape
  2) Feel of movement of lips + tongue when say word
• Happens almost instantaneously in brain

Question 4

General principles of semantic learning

Answer 4

• Markman + Wachtel (198)
• Mutual exclusivity priciple: object can only have one label (difficult to learn 2 lang)
• Whole object assumption: a word refers to whole object rather than features (key milestone - requires holisitic understanding of object

Question 5

Mechanism of semantic mapping - Fast mapping

Answer 5

• Semantic learning occurs quickly (Carey + Bartlett 1978) almost instant following a learnign experience
  -retention is short-lived (Horst + Samuelson, 2008)
• Retention varys depending on type of word (colour, shape words poorly retained)

Question 6

Mechanisms of semantic mapping - Probabilistic

Answer 6

• Progressively learns to map to correct level (whole object not indiv parts)
• More space learning - exposed to it over time
• Slow but better retention
• Trial + error

Question 7

Horst + Samuelson (2008) study

Answer 7

• Referent selection trial - object given name (known) + target (unknown)
• With no training shown 3 images + asked to retrieve target
• Feedback
• Good at picking target - process of elimination
• Retention trials
• Extention trials same objects but slightly different examples of them (test if it extended)
• Poor on retention + extention trials - evidence of fast mapping mechanism
• Fast mapping of novel names just as good as known
• Retention + extension are around chance
• Retention for fast-mapping is short

Question 8

Challenges during mapping

Answer 8

• How do we find correct reference level
• Superordinate level (animal)
• Basic level (dog)
• Subordinate level (chihauaha)
• Default level used by speakers is basic
• Optimal in commuinication

Question 9

Semantic errors - over extensions

Answer 9

• Car = any vehicle that moves on road
• Blue = broad range of colours
• Not immediately noticeable
• 2 stages: true over-extention followed by deliberate semantic misuse

Question 10

Semantic errors - under extensions

Answer 10

• Car = only familys volvo
• Blue = only colour of sky not sea
• Not incorrect per se
• Under-recorded

Question 11

Pereira et al (2014) study

Answer 11

• Aim = see