LING 101 Lecture notes W Apr 10 Today s topics: Acquisition of morphology, syntax UG in language acquisition (1) Background reading: CL Ch 9, 4, 5 CL Ch 9, 6 (review) 1
1. Review L1 acquisition key ideas Why do linguists like to use the term acquisition rather than learning for children s language development? Why is child language acquisition much more connected to descriptive grammar / mental grammar than it is to prescriptive grammar? 2
1. Review L1 acquisition key ideas A child who is in the process of acquiring his/her target (adult) language goes through different stages of development - These stages reflect intermediate mental grammars on the way to the adult grammar A child often shows variable behavior - A rule may be applied only some of the time - Multiple versions of a rule may be in use But we can still find a great deal of systematicity in children s language behavior 3
2. Morphological development Two strong sources of evidence that children are constructing a mental grammar as they acquire their language come from morphology: - overgeneralization (also known as overregularization) - productive use of morphology (wug-tests) 4
2.1 Overgeneralization Example of morphological overgeneralization (also called overregularization) Stage 1 show showed go went Stage 2 show showed go goed Stage 3 show showed go went 5
2.1 Overgeneralization Does a child hear forms like goed or mans in the speech community? - Why does the child produce such forms, often after a stage with the correct forms? 6
2.1 Overgeneralization Does a child hear forms like goed or mans in the speech community? - Why does the child produce such forms, often after a stage with the correct forms? This is evidence for morphological rules - At first, the child stores each form (pres/ past, sing/plural) separately in the lexicon - Then, the child learns a morphological rule, and sometimes applies it even to forms that are irregular and should be lexically listed as exceptions to that rule 7
2.1 Overgeneralization How we analyze what the child is doing Stage 1 show showed go went lexically listed lexically listed lexically listed lexically listed show lexically listed show lexically listed Stage 2 showed go formed by rule lexically listed Stage 3 like adults showed go formed by rule lexically listed goed formed by rule went lexically listed 8
2.2 Productive use of morphology Children perform quite well at tasks like these: - This is a wug. Now there is another one. There are two of them! There are two. - What would we call someone who crushes things? Someone who crushes things is a. Children who can complete these tasks have the relevant inflectional and derivational morphological rules in their mental grammar See the original wug-test article (very accessible): Berko [Gleason], Jean. 1958. The child s learning of English morphology. Word 14: 150-177. [PDF file] 9
3. Content vs. function morphemes A useful distinction in morphology (and syntax): Content morphemes (also called lexical morphemes) have real-world meaning - N, V, A - Derivational affixes Function morphemes (also called nonlexical or grammatical morphemes) have grammar-related meaning - Det, P, auxiliary verbs,... - Inflectional affixes 10
3. Content vs. function morphemes The first morphemes acquired are typically content morphemes - Why might this be? 11
3. Content vs. function morphemes Function morphemes often have a typical developmental sequence in a given language - Why? Where does this sequence come from? 12
3. Content vs. function morphemes Typical child developmental sequence for English function morphemes (CL, Table 9.12, p 365) 1. -ing 5. past tense -ed 2. plural -s 6. 3rd person singular -s 3. possessive s 7. auxiliary be 4. the, a Compare: Typical relative frequency of function morphemes in parent speech (CL, Table 9.13, p 366) 1. the, a 5. possessive s 2. -ing 6. 3rd person singular -s 3. plural -s 7. past tense -ed 4. auxiliary be 13
3. Content vs. function morphemes Frequency of function morphemes in adult speech does not predict how early they will be acquired! What factors do seem to predict early acquisition? - Occurs frequently at the end of the utterance - Forms a syllable on its own - Not a homophone - Behavior is regular few exceptions - Allomorphic invariance (one sound shape) - Has a clearly discernable semantic function 14
4. Syntactic development Syntactic development also proceeds in stages Examples: - Stages in utterance length - Stages in development of transformations 15
4.1 Development of utterances The one-word stage (12 to 18 months) One-word utterances are used to express the meaning of a whole sentence Some examples from A.: More ( I want more milk ) Foot ( My foot is stuck ) Leaf ( That s a leaf / I see a leaf ) Mama ( Mama should do it ) 16
4.1 Development of utterances The two-word stage (a few months later) Words very often lack inflection at this stage - Sometimes, children treat adult phrases as words in this stage (A. had V-it for transitive verbs) Some examples from A.: More crackers ( I want more crackers ) That bicycle ( That s a bicycle ) Papa eat-it ( Papa should eat it ) Duck head ( I have a duck on my head ) Mama up ( Mama should pick me up ) 17
4.1 Development of utterances The two-word stage Do children have syntactic categories in the two-word stage? - How could we test this? Can we tell? Word order mostly matches adult language - But children may learn word order verb by verb at first (before using a general X'-schema) 18
4.1 Development of utterances The telegraphic stage (approx. age 2) What morpheme types are missing here? - From CL, p 370 Chair broken. Man ride bus today. Car make noise. - From A. Eat-it orange fork mouth. Mama draw big blue O. Is there evidence for X' structure here? Once the telegraphic stage begins, further development is very rapid (see CL, Table 9.19, pp 371-2) 19
4.1 Development of utterances One fact about individual children that is often reported in research on child language is the child s MLU, or mean length of utterance - This can be measured in words or morphemes Why might MLU be more useful than chronological age in comparing children? Can MLU help indicate when a child has left the telegraphic stage and become more adult-like? (What should happen to MLU at this point?) 20
4.2 Later development: Movement The Inversion rule: How does this develop for English-acquiring children? a. Questions signaled by intonation only b. A relatively rare pattern: Can he can look? What rule does this child s grammar have? c. Adult-like application of Inversion Some children pass through a stage where they can apply Inversion, except when they have to apply Wh Movement too 21
4.2 Later development: Movement Draw a tree and apply the appropriate rules for this wh question: What do you think is in the box? - Hint: How many TPs/CPs do we have here? Suppose we wanted to study the syntax of this type of wh question in child language. How might we collect data? - Naturalistic vs. experiment studies (what are the pros and cons?) Here is a video of a wh-question study 22
4.2 Later development: Movement Consider the syntax of the child in the video: What do you think what is in the box? - What does this child s current Wh Movement rule seem to be? 23
5. What makes L1 acquisition possible? Next we will consider: What role do the following types of parent/child interaction play in a child s language development? - Imitation - Feedback / correction by adults (Next lecture:) - Caregiver speech ( motherese ) - General cognitive development 24
5.1 Acquisition is not just imitation Some people assume that children simply learn language by imitating adult speakers - Sometimes a child can imitate more accurately than they generally speak However, imitation is far from the only process involved in language acquisition - What characteristics of language are incompatible with the imitation hypothesis? - What often happens when a child is explicitly encouraged to imitate? 25
5.1 Acquisition is not just imitation What characteristics of language are incompatible with the imitation hypothesis? - Creativity/productivity of language - Morphological overgeneralization and other child-specific linguistic rules goed, bringed What would the child be imitating? What often happens when child asked to imitate? Parent: Why can t kitty stand up? Child: Why kitty can t stand up? (CL, p 378) 26
5.2 Acquisition is not just via feedback Some people assume that parents teach language to children by correcting their errors - Do children learn from direct correction by caregivers? - Do children learn from recasts (repeating child s utterance with readjustment)? 27
5.2 Acquisition is not just via feedback Children often ignore (or, their mental grammars can t interpret) direct correction Child: Want other one spoon, Daddy. (CL, pp 378-379) Father: Child: Father: Child: Father: Child: Father: Child: Father: Child: 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. other...spoon. Now give me other one spoon? 28
5.2 Acquisition is not just via feedback When are recasts used? Do they help identify non-adult-like utterances? i. Child: Mama isn t boy; he s a girl! Parent: That s right. ii. Child: Boy chasing dog. Parent: Yes, the boy is chasing the dog. iii.child: The dog is barking. Parent: Yes, the dog is barking at the kitty. 29
5.2 Acquisition is not just via feedback When parents correct their children, it s more likely to be about the truth of their utterances than phonology/morphology/syntax Recasts do not always serve as a cue for nonadult-like utterances 30
5.2 Acquisition is not just via feedback BUT! Is there a link between recasts and acquisition? Two experiments: - No? Recasts had no effect in a study of the acquisition of determiners a, the - Yes? A single recast was often enough to permit learning the irregular past tense of a novel (made-up) verb Is there a principled difference between these two examples of recasts that might explain the different results? 31
5.2 Acquisition is not just via feedback Is there a principled difference between these two examples of recasts that might explain the different results? - Learning when/where to use determiners is mostly a matter for the... - Learning the irregular past tense of a new verb is mostly a matter for the... There is some additional evidence that parent recast rates and child language development may be linked (see the Busting Language Myths course!) 32