Word learning p1
- Created by: BKW
- Created on: 16-01-20 16:08
View mindmap
- l4 Word learning p1
- WHEN do infants start learning words?
- a lot of variability and they usually understand a word before producing them
- at 18 months we see increases in vocab
- as children get older they pick up words a lot faster - this is called a vocabulary spur
- How can we investigate this?
- lab based assessment using eye tracking to see which of the 2 images child will look at if they are told to "look at the ball"
- Bergelson&Swingley 2012 tested food and body parts as we learn these quite early
- found evidence of comprehension of concrete words at 6-9 months! this was a weak finding but 10-13 months were a lot better
- so although children know words its a lot harder to learn them because the meanings aren't obvious
- The mapping problem - Quine 1960
- picture of bunny in a field, if someone said "GAVAGAI" and points to thing in scene - new name could refer to an infinite no of aspects
- Is it the bunny itself or the ears, or they way it moves, the grass ect
- this is a really hard problem to solve but children are really goof at solving it
- Is it the bunny itself or the ears, or they way it moves, the grass ect
- picture of bunny in a field, if someone said "GAVAGAI" and points to thing in scene - new name could refer to an infinite no of aspects
- The mapping problem - Quine 1960
- so although children know words its a lot harder to learn them because the meanings aren't obvious
- found evidence of comprehension of concrete words at 6-9 months! this was a weak finding but 10-13 months were a lot better
- Bergelson&Swingley 2012 tested food and body parts as we learn these quite early
- lab based assessment using eye tracking to see which of the 2 images child will look at if they are told to "look at the ball"
- The mapping problem - Quine 1960
- picture of bunny in a field, if someone said "GAVAGAI" and points to thing in scene - new name could refer to an infinite no of aspects
- Is it the bunny itself or the ears, or they way it moves, the grass ect
- this is a really hard problem to solve but children are really goof at solving it
- Is it the bunny itself or the ears, or they way it moves, the grass ect
- picture of bunny in a field, if someone said "GAVAGAI" and points to thing in scene - new name could refer to an infinite no of aspects
- HOW do infants/children learn words?
- Word learning is special
- Markman 1990 - children innately constrained to consider only some possible word meanings and there are several constraints
- Mutual exclusivity - biased to map a new word to something we have no name for
- Whole object constraint
- taxonomic constraint - we extend words but only to the thing itself not related things
- these constraints allow fast, efficient word learning called fast mapping
- Is fast mapping specific to words?
- Markson and bloom 1997 argue not they did a study on 48 3y/o, 47 4y/o and 48 undergrads
- experimenter introduced pps to new unfamilliar object in 1/3 conditions: 1. word 2. fact 3. location
- they were tested on how quickly they learn these pieces of info immediately after, 1week later and 1 mo later
- they found children and adults fast mapped words and facts most retained for 1 mo. worse at learning location
- concluding fast mapping is not specific to just words - maybe there is no dedicated system for word learning?
- they found children and adults fast mapped words and facts most retained for 1 mo. worse at learning location
- they were tested on how quickly they learn these pieces of info immediately after, 1week later and 1 mo later
- experimenter introduced pps to new unfamilliar object in 1/3 conditions: 1. word 2. fact 3. location
- Markson and bloom 1997 argue not they did a study on 48 3y/o, 47 4y/o and 48 undergrads
- Is it specific to humans?
- No - animals can also fast map - especially doggos
- Kaminski, Cll and Fisher 2004 - Rico the german Border collie knows 200 object names
- Other dogs can do this too eg Chaser the dog
- this could be because they are social creatures bred to be comminicative
- Waxman and Booth 2000 - word learning is special for diff reasons - they are symbols and facts are not
- 4y/o in the lab introduced them to new object in 1/2 conditions 1. fact 2. word - tested on fast mapping and generalization
- Found they fast mapped words and facts the same but generalized them differently
- generalised words to objects of the same kind 100% of time but facts extended to objects of same kind less than words by 67-50% and to objects of diff kind 22-23% of the time
- Found they fast mapped words and facts the same but generalized them differently
- 4y/o in the lab introduced them to new object in 1/2 conditions 1. fact 2. word - tested on fast mapping and generalization
- Kaminski, Cll and Fisher 2004 - Rico the german Border collie knows 200 object names
- No - animals can also fast map - especially doggos
- Markman 1990 - children innately constrained to consider only some possible word meanings and there are several constraints
- Word learning is special
- Where do these biases come from
- Smith et al 2002 hypothesised word learning birning biases are built from experience with initial word-meaning associations - this is done in 4 steps with a DAM
- 1. child starts to learn concrete words
- 2. As they learn more they begin to generalize over shapes
- 3. label these features that make the shape that shape
- 4. generalize to other shapes
- can get babies who are not yet showing this to show shape biases
- 9 week longitudinal study in lab on 17month old not yet displaying whole object bias
- taught words refering to objects of same shape (trained vs untrained condition
- trained infants mapped words to objects and object properties even learning more object labels out the lab
- taught words refering to objects of same shape (trained vs untrained condition
- 9 week longitudinal study in lab on 17month old not yet displaying whole object bias
- Waxman and booth disagree with smith arguing children don't words based on perceptual info they do this based on what something is - conceptual info
- 24 3y/os native English taught novel word Dax/Rif in 1/2 conditions 1. animate 2. artifact
- shape change largely rejected in both conditions but texture was only rejected by animate codition
- conceptual info dictated meaaning
- shape change largely rejected in both conditions but texture was only rejected by animate codition
- did a second study on 12-13 /o -does conceptual info override the fact it looks animate? added eyes found adding eyes did not = rejection of texture change - more consistent acceptane of texture change than in animate condition in exp1
- shows children can look past appearance - extending words based on info about kind rather than perceptual info
- Smith et al argue this saying context of naming pushes child's attention to relevant feature but animacy cues push attention to texture
- Also argue googly eyes do not display real intimacy
- Smith et al argue this saying context of naming pushes child's attention to relevant feature but animacy cues push attention to texture
- shows children can look past appearance - extending words based on info about kind rather than perceptual info
- 24 3y/os native English taught novel word Dax/Rif in 1/2 conditions 1. animate 2. artifact
- Smith et al 2002 hypothesised word learning birning biases are built from experience with initial word-meaning associations - this is done in 4 steps with a DAM
- WHEN do infants start learning words?
Comments
No comments have yet been made