Week 9 cd Flashcards
Evidence kids know words and object mappings on the fly.
What kinds of strategies can kids use to learn words.
Proof of principle shows they did learn words will look
Mark amen study 1987
Novel word, novel object associate them. 3.5years
MUTUAL EXCLUSIVITY is what this is called. It is a heuristic or bias that leads to:
Labels for each object are musptually exclusive. 1:1 mapping object and label. Kids apply it even if not always right answer
Halberda, even 17 m olds have this thinking.
Vaby Davy
Several features vs single feature
Markman strategies for word learning and others
Heuristic biases:
- Mutual exclusivity
- Whole object /parts/feature hear a novel word associate to part feature or whole.
-using gaze, (not (association theory) first thing seen after word. )
Therefore want to act on social cues and take in people ideas. Are controlled but not realized.
Baldwin 19-20m (1993)
Use pure association , attach closest thing to what you are attending to or see after hear label.
Association vs gaze
Pay attention to gaze! Think motey
, not think of association first thing see in motey.
Looking pointing and reaching.
Problem is may not be word learning, but could be gaze learning.
To rule out only associating to gaze, ask which do you like. They choose random, ask for motey, get motey.
Conclude: stronger evidence use gaze to attach label.
Not just about gaze or association, but combo.
Yu study
Tiny cameras
When parents are labeling object they:
Wait until kid is looking at the object, to label it,
Conclude: parents are sensitive to attention and association theory,
Learning how words of togethor. Grammar syntax
Goblin copf study 17m (1987)
Cm tickled bb
Bb tickle bb
Do kids know word order.
Procedure
See both /hear 1 sentence
Results look longer at the match
Conclude: know something about syntax and grammar, doer.
Don’t know if they know meaning of tickle, but maybe know how sentence composed. Problematic early study though becauof this.
So next exp. uses made up verb.
What do kids know about abstract structure, a essence knowledge of specificity words.
Yuan et al 21 m
Scenes shown.
Two ppl Push bend over scene, guy by himself jumping jacks.
He gorps him. More likely pushing scene cuz two people
He gorps could be either one. Equal
Watch this! (Control) equal
Many possible results. Due to paths of decision.
Results are what we predicted.
Conclude: know English syntax, actiony verb heartwo nouns think two objects,
Follow up study: show two people have to be interacting in scene.
Changing scenes did not change results. They have to be interacting. Causal action, but idk could be third thing.
Kids seems to have some knowledge of sentence structure and use it to learn new words.
What kids can learn about abstract structures.
Marcus study 7m abstract structure.
Created a language, made up of sentences AAB form.
Aba wo fi wo ex.
Rest one of two languages.
Test on 7m
Use test syllable no sharing from training, know they are good at word segmentation, track frequency of occuracne of syllables.
Need to give new syllables following structure,
Differentiate between the syllables. PFERRING NOVEL ITEM.
Conclude are sensitive to structure of sentence. Form. Learning something super abstract like syntactic structure. Kids able to as shown here. Get something of oattern.
Swing,ey paper baby vaby, ray paper.
Correction,
Look more at baby,
And in fact less at raby and vaby.
In fact no distinction between vaby and raby, therefore treat similarilyt. At least in this paper. Others show dif.
Marcus study cont’d syllable combinations. Patterns sentence structures 7 m olds.
Different habituation, sae test, preference for novelty.
Can distinguish sentence structure, extract abstract form, dif, from sequence of syllables. Not syllable tracking, tracking
Used head turn preference. Method, dark room, flashing light.
Measure head turn how long. Figure out control sounds. Wow interesting method,
Marcus abstraction study, applies to any combo, is specific to language. That was their idea. Their original study did not show it was specific to language.
So could do patterns in other sensory domains. Is it specific to language. Marcus
Use tones musical notes. Timbres, animal sounds used too.
Structure same as abstraction,
Failed to distinguish, between items during test. All of them!
Does not mean did not learn. Everything about is the same. Except sounds.
So at least language is easier.
Sung syllables Marcus
It does work. As long as speechy. Also with 7 month olds.
Researches had always paired training with test. Worried about familiarity. Do they change the test trials so they
Forgo matching tests , and use animal sounds, familiarization. And then syllables!
Hypo: Can extract pattern, but can only express knowledge testing with syllables.
But did not happen! No evidence kids learning anything.
What about the reverse?
Syllable or animal sound.
If do distinguish, extract and apply to animal sound and map onto.
And they do distinguish!
Conclude,
Can learn from syllables and map onto animal sounds. Or perhaps other. Only catch on to syllable patterns.
So syllable syllable test, or syllable animal test only work.
Syllables must be in training, then can apply it, trouble learning from other patters.
So speech is special. For abstract structure according to marcus
Now onto categorization.
What is it
Group stuff together, Set list of features, specific and gen. Functions specific. Exclusion Hierarchical, Words,
TRRAT THINGS THE SAME THAT ARE DIFFERENT.
used to descriminate.
Imp.
Distinguishable but discountable at other levels.
Assignment examples show that can distinguish things, but discount things at other levels,
Make a list,
Defined by function.
Use of or intention and
Looked at triangle, chair and whale. Hard.
Definition and image,
