Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Language and kids Linguistics lecture #8 November 21, 2006.

Similar presentations

Presentation on theme: "1 Language and kids Linguistics lecture #8 November 21, 2006."— Presentation transcript:

1 1 Language and kids Linguistics lecture #8 November 21, 2006

2 2 Overview Why study children’s language? Evidence for innateness Modeling language acquisition

3 3 Why study children’s language? Innateness can be studied by finding out how much kids are born “knowing” Modularity can be studied by finding out how language development relates to general cognitive development Also, it’s cool! - Kids develop language so easily, without any explicit teaching, that linguists call it language acquisition (acquiring), not language learning

4 4 A critical period? Why is it so easy for babies and kids to acquire a language, and so hard for adults? Maybe human brains develop by a biological clock, so there is a critical period of time when the brain can easily acquire a language Some birds have a similar critical period for acquiring their songs

5 5 Innateness in speech perception? The hardest thing for an adult to learn is phonology, including speech perception. Does this imply innateness? One way to find out is to test little babies: Do they also show categorical perception? If they do, they probably did not have time to learn it, so maybe they were born “knowing” how to do it.

6 6 But how can we “ask” them? Babies like to hear new things, so if they recognize a change in speech sounds, they suck faster, as measured by computer:

7 7 Babies & categorical perception Even newborn babies (just a few hours old) show categorical perception They perceive categories that may not be used in their parents’ language - Spanish babies perceive [pa] vs. [p h a] even though Spanish adults only use [ba] vs. [pa] They still must learn to focus attention on the sounds important for their language

8 8 What about words? Of course individual words are not innate, and must be learned. But still, something must be innate about the process of learning words. Do you remember why? It’s because of the gavagai problem: people seem to already “know” that names refer to whole objects.

9 9 Kids acquire grammar, not language Babies and children don’t just memorize phonological and syntactic atoms, but also learn grammatical rules For example, kids are good at this test: This is a wug.Here are two more of them. These are two ______ wugswug[z]

10 10 Is the “rule” concept innate? Of course kids have to learn the particular rules in their parents’ grammar But do they have to learn the concept of “rules”, or are they born expecting rules? In fact, kids seem to love rules, so much that they use them even when their parents don’t This is shown by U-shaped development

11 11 U-shaped development Kids begin by imitating words: go - went Then they start to guess rules, and sometimes go too far: go - goed 2 years 3 years 4 years... 100% correct go - went eat - ate talk - talked go - goed eat - eated talk - talked go - went eat - eated talk - talked

12 12 Is there a biological clock? American and English kids start to overuse rules like past tense (goed) around the age of 3 Deaf kids acquiring American Sign Language start to overuse verb marker rules and other rules around the age of 3 Taiwanese kids start to overuse classifier rules ( 兩個狗 ) around the age of 3

13 13 Do parents teach grammar? Nobody teaches kids to say “goed”! Most parents don’t know enough linguistics to teach grammar (could you teach your kids the tone system in Taiwanese?) Even when some parents do try to teach grammar, the kids just ignore them!

14 14 Want other one spoon, Daddy. You mean, you want the other spoon. Yes, I want other one spoon, please, Daddy. Can you say “the other spoon”? Other... one... spoon. Say “other.” Other. “Spoon.” Spoon. “Other... spoon.” Now give me other one spoon?Other … spoon.

15 15 Can cognitive development explain language acquisition? Does language develop as an independent module? Language development also requires: - Theory of mind: the belief that other people have minds like yours - Object permanence: the belief that objects continue to exist even when you don’t perceive them

16 16 Are these abilities also innate? At first, babies don’t act as if they have a real theory of mind, but later they do…. But chimpanzees never seem to develop much of a theory of mind at all! Piaget ( 皮亞傑 ) thought that object permanence takes a long time to develop... But more recent experiments show that very young babies already have it!

17 17 Modeling language acquisition Some models focus on innateness and modularity, e.g. learnability theory Other models focus on learning from experience and interaction, e.g. connectionism

18 18 Learnability theory This theory uses math to study innateness The focus is on how kids try to figure out their parents’ grammar This task faces the problem of induction: figuring out a general statement from a finite set of data How do kids take a finite set of sentences and figure out an infinitely generative grammar?

19 19 A formal example You’re a baby and you hear these sentences: a … aa … aa … a … aa Suppose that you are born with no innate expectations about language. Then you would have to consider both of these possible grammars: G  = ({a}, {S}, {S  a, S  aS}) G 2 = only a, aa are grammatical

20 20 That won’t work! Suppose the correct grammar is G 2, but you guess G . Can you ever learn the right grammar? No. Both a and aa are grammatical by G , so why change to G 2 ? What if G  is correct but you guess G 2 ? If you hear aaa, just change to G 3 ; if you hear aaaa, just change to G 4 ; etc. How could you ever conclude that the grammar was really G  ?

21 21 A very different approach Some researchers argue against this standard innateness-and-modularity view of language acquisition They use connectionism to provide an interactionist model of language acquisition in which very little must be innate

22 22 Rules or memory? Connectionism models memory and “rules” the same way (remember 繩子  條 ?) So do kids really need to learn a “+ed rule” (or a “ 個 rule”)? Maybe they just generalize from memorized examples.

23 23 Grammar = memorizing atoms? If so, the pattern of word learning should match the pattern of syntactic acquisition

24 24 What about the biological clock? Maybe the critical period exists because the brain “gets too full” to learn a second language perfectly nothing’s innatelearn Chinesetry to learn English

25 25 Summary Experiments show that babies are born “knowing” some things about language Kids acquire grammar; they don’t “learn language” by simple imitation Learnability theory highlights the induction problem, supporting innateness Connectionist models imply that innateness and modularity may really be less important

Download ppt "1 Language and kids Linguistics lecture #8 November 21, 2006."

Similar presentations

Ads by Google