9.57J | Fall 2001 | Undergraduate

Language Acquisition

Study Materials

These notes provide a comprehensive review of the course content, in order to facilitate studying for both the midterm and final exams.

Review

1. Define learnability. Would the following set of languages be learnable? If yes, describe the learning procedure. If no, explain why not.

L1 = {zot(1)}
L2 = {zot(1), zot(2)}
L3 = {zot(1), zot(2), zot(3)}
and so on through
L1000 = {zot(1), …, zot(1000)}

2. Define the following, with respect to language acquisition: positive evidence; negative evidence; Motherese. Which of these do children use in order to acquire their language? What is the evidence that some of these factors don’t play an important role in language acquisition (think Newport, Gleitman and Gleitman; Wexler and Culicover)?

3. a) What is the difference between a complement and an adjunct? What is a specifier? Keeping these terms in mind, give the tree structure for (i).
(i) A friend of my parents from Florida read a book about Chomsky’s theory of Universal Grammar in the car.

b) What thematic roles does read assign? Which phrases get these thematic roles in (i)?

4. Give the tree structures for the following sentences - include IPs and CPs.

a) Two researchers in Bill’s lab want for this gorilla to learn English.

b) The gorilla thinks that the researchers are wasting their time.

5. Give the tree structures for the following sentences, and show all movements that take place.

a) What language is little Susie acquiring?

b) Are her parents correcting her grammar?

c) Susie has not been listening to her parents.

6. a) Give the tree structure for the following German sentence, and show the movements that take place.

Hans hat einen Roman gelesen
Hans has one novel read
“Hans has read one novel”

b) Which of the following sentences is correct in (adult) German? Give the tree structure for the grammatical sentence.

i) einen Roman Hans hat gelesen
ii) einen Roman hat Hans gelesen
iii) einen Roman hat gelesen Hans

7. a) For the sentence in (i), give the form of this sentence (using English words) for each of the following languages: French (SVO / non-V2); German (SOV / V2); Swedish (SVO / V2); Irish (SVO / V1). (More than one word order may be possible for some of the languages - give all possible word orders for each language).

(i) Anne likes chocolate.

French
German
Swedish
Irish

b) Give the root infinitive version of (i) that would be produced by a young child (according to the Optional Tense theory), for each of the languages below:

French
German
Swedish
Irish

c) What would the Modal-Drop theory predict to be possible word orders for a child acquiring each of the following languages? Give all possible word orders that the Modal-Drop theory would predict.

French
German
Swedish
Irish

8. a) Give the tree structure for the (adult) French sentence in (i). (Note: I’m leaving out ne, a negative particle that precedes the verb in formal French but that can be left out in colloquial French). Assume that pas is immediately to the left of the VP. Show the movements that take place.

(i) Marie aime pas Claude.
Marie likes not Claude.
“Mary does not like Claude.”

b) If an L1-French child in the Optional Infinitive stage produces an infinitival version of (i), what would the resulting sentence be? (The infinitival form of like is aimer). Give the tree structure for this sentence.

c) If an adult L2-French learner produces an infinitival version of (i), what would the resulting sentence be? (Think Prévost and White).

9. a) What is a principle of Universal Grammar? What is a parameter?

b) Let’s say we have some L1-French speakers acquiring English as an L2. French and English differ on a particular parameter (R): French is [+R] and English is [-R]. Under each of the following models, what would we expect the learners to posit as the value for R at the start of their L2 acquisition? At the end of their L2 acquisition?

UG is fully available in L2; no transfer from L1
there is full transfer from the L1, but UG is also available
there is full transfer from the L1, and UG is not available except through the L1

10. a) Masha is a two-year-old acquiring Russian as her first language. The following lexical items are available to her: ja (I - NOM), menja (me-ACC), ona (she - NOM), ee (her - ACC), poju (sing - 1st person singular), poet (sings - 3rd person singular), pet’ (sing - infinitive). The default case in Russian is nominative. Assuming that Masha is in the Optional Infinitive stage, list all the possible sentences she might produce using the lexical items above. Why?

b) Mary is the same age as Masha, and is acquiring English as her first language. The same lexical items are available to her as to Masha, except in English (I, me, she, her, sing, sings ). The default case in English is accusative. Assuming that Mary is in the Optional Infinitive stage, list all the possible sentences she might produce using the lexical items above. Why?

c) Mary has acquired some new lexical items. She now knows play, plays, and played. According to the ATOM model, which combinations of she / her with play / plays / played will she produce? Why?

11. a) Which of the following sentences should be grammatical in (adult)Italian? The translations are as follows: lui (he-NOM), è (is - used like the English auxiliary has ), arrivato (arrived - past participle; as in has arrived ). Explain your answers. Compare your answers for Italian to the judgments given for English. What is the parameter, exemplified here, on which Italian and English differ?

ITALIAN ENGLISH
__ è arrivato “__ has arrived”
bad
lui è arrivato [neutral intonation] “he has arrived”
good
LUI è arrivato [contrastive pronoun] “HE has arrived”
good

 

b) Given only the evidence in (a), what would we expect for children learning English and Italian? Should L1-Italian children treat Italian like English with respect to presence of subjects, or should L1-English children treat English like Italian? Why?

c) What actually happens with this parameter in acquisition of English vs. Italian? How does Hyams (1987) explain this? What is the trigger that she suggests kids use to set the parameter appropriately?

12. Suppose that we analyzing data of an L1-English child, George. Early on in his acquisition of English, George sometimes uses full NP subjects, sometimes uses pronominal subjects, and sometimes uses null subjects. By around age three, George has ceased using null subjects, and has the following proportions for overt subject use: he uses full NP subjects 25% of the time, and pronominal subjects 75% of the time. Let’s assume that proportions of full NP to pronominal subjects stays constant throughout development.

a) In his early samples, George drops subjects about 40% of the time. Given this information, fill in the empty cells in the table below: what would the processing vs. grammatical models predict to be percentages of full NP and pronominal subjects in George’s early samples? Explain your answers. Which model is better supported by the data of Eve and Adam that we looked at in class?

PREDICTIONS
MODEL Full NP Subjects Pronominal Subjects Null Subjects

Output Omission Model

    40%
Grammatical model     40%

b) Suppose that, while dropping subjects 40% of the time, George drops objects only 2% of the time. On the other hand, Greta, a Dutch-speaking two-year-old, drops subjects 45% of the time and drops objects 25% of the time. How might this discrepancy be explained under the theory of Topic-Drop? (see Hyams and Wexler, section 2.3). What would this theory predict about the placement of the null subject or object in Greta’s Dutch?

Course Info

Instructor
As Taught In
Fall 2001
Learning Resource Types
Problem Sets with Solutions