The syntactic abilities of children with SLI: From Tense to Movement

37-975-01

Challenges to Language Acquisition: Bilingualism and Language Impairment

Dr. Sharon Armon-LotemBar Ilan University

Topics

Tense and Agreement Passive Binding WH-Questions Relative clauses

Tense as a clinical marker for SLI

The use of Root/Optional Infinitives

The phenomenon

Up to the age of three children use the infinitival form of the verbs in indicative matrix clauses in 50% of their verbal utterances in English (Wexler 1994), and to a lesser extent in other languages (Armon-Lotem 1996a, Hyams 1995, Rhee & Wexler 1995, Rizzi 1994a).

Finite sentences are produced at the same time Children seem to know the grammatical properties

of finiteness and non-finiteness (e.g., Deprez & Pierce 1994)

1) a. It only write on the padb. He bite mec. My finger hurts

2) M: ma at osa? what you do

'what are you doing ?L: tapuax lishtot (Lior 1;08)

apple to-drink 'I drink an apple'

Infinitival forms constitute only 5% of the Italian data.

>>> Extensive use of root infinitives correlates with non-null subject languages.

“A language goes through an OI stage if and only if the language is not an INFL-licensed null-subject language.” (Wexler 1996(

Armon-Lotem (1996) for Hebrew

There’s a gradual increase in the use of inflected verbs.

Past tense morphology is acquired prior to person morphology, but this does not correlate with a decrease in the use of root infinitives, but rather with a decrease in the use of “stem-like forms”.

The use of root infinitives reduces (from 5% to less than 1%) only when questions (and subordination) are mastered (last stage of Klima & Bellugi 1966).

Inflections in Hebrew speaking children with SLI Dromi, E. & S. Davidson. 2002. A Clinical Marker for HSLI: from

Empirical Findings to Theorizing. Paper presented at Brain and Language: Language Acquisition in Special Populations, Bar Ilan University, June.

Dromi, E., Leonard, L., Adam, G. & Zadunaisky-Ehrlich, S. 1999. Verb Agreement Morphology in Hebrew-Speaking Children with Specific Language Impairment. Journal of Speech, Language and Hearing Research, 42, 1414-1431.

Dromi, E., Leonard, L.B., & Adam, G. 1997. Evaluating the morphological abilities of Hebrew- speaking children with SLI. Amsterdam Series in Child language Development, 6, 65-78,

Dromi, E., L. B. Leonard, and M. Shteiman (1993) The grammatical morphology of Hebrew-speaking children with Specific Language Impairment: some competing hypotheses. Journal of Speech and Hearing Research 36: 760-771

The morphological richness hypothesis SLI children have a limited processing capacity.

They focus on the most salient aspects of the language they acquire. For example, in English they focus on word-order and ignore the morphology, while in German they focus on morphology and ignore the word order.

Subjects: SLI, NDA, NDL (matched by MLU)

Dromi, E., L. B. Leonard, and M. Shteiman (1993) Findings: “Hebrew speaking children with SLI

resembled their MLU controls in their use of both present and past tense inflections requiring agreement with the subject”.

In the nominal system, plural formation, adjectival agreement, and the use of the accusative case marker are all delayed, but not different from language matched controls.

>>> SLI is a delay

Dromi, E., Leonard, L., Adam, G. & Zadunaisky-Ehrlich, S. (1999)

Method: Sentence completion for 3rd person.

Enactment tasks for 1st and 2nd person. 4 conjugations: pa'al, piel, hitpael, hif'il.

The inflectional paradigm for past and present.

Findings In present tense, both SLI and NDL used past for preset In present tense, both SLI and NDL used masculine for feminine in singular

and plural. SLI found Hitpa'el more difficult – using p'iel instead. Simplifying consonants

cluster. SLI found Hif'il more difficult – using present for past and vice versa, using

infinitives. SLI found pi'el more difficult – they used also stripped forms In past tense, 3rd person singular replaced many of the inflected forms. SLI used it mostly instead of other singular forms (56/64) – mostly for 2nd

person NDL used it mostly instead of plural forms. Past tense does pose a problem for Hebrew speaking SLI children, whereas

difficulties with present tense are less pronounced. Most errors were mostly related to the use of tense (60/144)) or person

(67/144), but usually not both. Most errors were different by one feature from the target (77% in the past

tense) >>> A limited processing capacity, since more complex structures, which place

more demands on the system, seem to be more impaired.

Blass A. 2000.Method: Spontaneous speech samples of the same

children

Findings: No difference between SLI and NDL in the level of

inflections No difference between SLI and NDL in the mastery of

inflections Out of all forms in Pa’al (80% of verbs), 90% were

tensed. SLI used more bare (stripped) forms – significant, but the

numbers are small. SLI and NDL had similar errors, but SLI had more. In natural settings children do what they know and avoid

the difficult forms.>>>Delay

Davidson, S. 2002. The Language Profile of Hebrew Speaking Preschoolers with Specific Language Impairement. M.A. Thesis, TAU.

Methodology: H-IPSyn

Findings: SLI are similar to NDL but for three criteria:

Lexicon - SLI use a smaller variety of verb types than NDL

Mrpho-syntax - SLI make more errors than NDL but of the same kind

Pragmatic (??)- SLI have difficulties with reference not found in the NDL group

Passive Participle vs. Regular Past

Tense Laurence B. Leonard, Patricia Deevy, Carol A. Miller, Leila

Rauf, Monique Charest, and Robert Kurtz. 2003. Surface Forms and Grammatical Functions: Past Tense and Passive Participle Use by Children with Specific Language Impairment. Journal of Speech, Language, and Hearing Research Vol.46 43-55

The girl pushed the boy.The boy got pushed by the girl.

EOI account: different

The surface account: same

Method

Subjects 12 of the children (aged from 4,6 to 6, 10) with

SLI 12 ND-A 12 ND-MLU

Sentence completion tasks: the use of past tense verb forms the use of passive participle verb forms

Summary

The inconsistency with which children with SLI produce past –ed cannot be due to the surface property of this inflection. Its grammatical function probably plays the central role.

Children with SLI have special problems with verb morphology, even when tense is not involved. The passive participle –ed proved to be one such area of weakness.

The syntactic abilities of children with SLI: The Passive

Passive

Maryi was kissed ti by John

Passive is A-movement rather than A’-movement The subject is the patient (no necessary agent) The transitive verb has unique morphology (with or without

an auxiliary verb) which makes it intransitive The passive derives n-place predicate from n+1-place

predicate Not all languages permit an agent-phrase (by phrase), and

the same agent-phase can occur with non-passive verbs Verbal vs. adjectival passive

Issues in acquisition

Reversible vs. non-reversible Actional vs. non-actional Adjectival vs. Verbal >> Do children understand the by-phrase? Comprehension vs. Production >>

Verbal vs. adjectival

The girl is covered (by the boy)

The covered girl (*by the boy)

Ha-yalda mexusa (al yedey ha-yeled)

the-girl cover-pass (on hands the-boy)

‘The girl is covered (by the boy)’

<<

SLI Children's Delayed Acquisition of Passive

Mabel L. Rice, Kenneth Wexler, & Jennifer Francois

Paper Presented at the BU Conference on Language Development

Boston, MA, November 1-4, 2001

Subjects Study 1

19 10-year-old children 17 age-equivalent controls 16 8-year-old lexically-equivalent controls (PPVT raw

scores)

Study 2 17 5-year-old SLI children 17 age-equivalent controls 16 3-year-old lexically-equivalent controls (PPVT raw

scores)

Method

Stromswold’s 32-item task for reversible full passives, with toy animals.

Examiner: “The goal kicked the horse.”

Child: act out action with toy animals

[Verbal item set: Kiss, slap, touch, hug, kick, lick, tickle, push]

Results - Study 1Passive Comprehension

0

0.15

0.3

0.45

0.6

0.75

0.9

SLI Lexically Matched Age Matched

Per

cen

t C

orre

ct

By 10 years of age, children in the SLI group comprehended reversible full verbal passives, showing knowledge of movement (A-chains)

Results - Study 2Passive Comprehension: Identification of Agent

0

0.15

0.3

0.45

0.6

0.75

0.9

SLI Lexically Matched Age Matched

Per

cen

t C

orre

ct

At 5 years of age, children in the SLI group were below age peers in their comprehension of reversible full verbal passives, and similar to their younger lexically-equivalent peers

How do children with SLI interpret the passive?

Children with SLI consistently interpret reversible passive using SVO strategy (Bishop 1982)

Children with SLI show a mixture of correct interpretation and a reversal interpretation (Van der Lely & Harris 1990)

Children with SLI perform better on short passive than on long Passive (Van der Lely 1994)

Children with SLI adopt an adjectival interpretation (Van der Lely 1996)

Van der Lely, H. 1996. Specifically language impaired and normally developing children: Verbal passive vs. adjectival passive interpretation. Lingua, 98, 243–272.

Subjects

Method – TAPS (Picture selection task) (a) reversible active SVO (e.g.,

“the man eats the fish”); (b) reversible full passive (e.g.,

“the man is eaten by the fish”); (c) short progressive passive

(e.g., “the fish is being eaten”); and

(d) short passive with potentially adjectival passive interpretation (e.g., “the fish is eaten”).

12 items x 4 sentence types = 48 sentences

6 verbs: wash, mend, paint, eat, cut, hit

Results (p.258)

Reversal

Adjectival

Passive

p. 259

Adjectival

Passive

D. V. M. Bishop, P. Bright, C. James, S. J. Bishop, and H. K. J. Van der lely. 2000. Grammatical SLI: A distinct subtype of developmental language impairment? Applied Psycholinguistics 21, 159–181

Subjects

Sample A - LI - 46 children out of 37 same-sex twin pairs selected for the presence of language impairment in one or both twins

Sample B - LN- 32 children out of an unselected sample of 104 twin pairs from the general population

All children were 7 - 13.

Results

There was a significant difference between groups: mean correct (out of 48) for group LI = 40.4 (SD = 3.96) and for group LN = 45.3 (SD = 2.29), F(1, 76) = 39.8, p < .001.

Age was not significantly correlated with TAPS performance, r(76) = −.047

Nonverbal ability was significantly correlated with TAPS : r(76) = .420 for Raven’s Matrices and .445 for PIQ (both p < .001

Results by sentence type

*

*

The Acquisition of Passive Constructions in Russian Children with SLI

Maria Babyonyshev, Lesley Hart, & Elena Grigorenko. 2005. Paper presented at Formal Approaches to Slavic Linguistics - The Princeton Meeting

Subjects A medium-sized village (population of

approximately 900) in Arkhangelsk region where the incidence of language disorders is far greater than in the general population.

14 monolingual Russian children aged between 6;3 and 9;10 (mean age 7; 10), non-verbal IQ above 70: seven TD children (mean age 8;3 ) and seven children with SLI (mean age 7;5).

Children were grouped based on: clinical impressions, and either MLU, or syntactic complexity (the proportion of syntactically complex structures to all structures produced)

Method

A picture selection task with reversible passive sentences in the perfective form.

20 passive sentences with pairs of pictures: 10 based on actional verbs (a), 5 based on psychological predicates (b), and 5 based on perception verbs (c).

a. Petux byl oščipan gusem. ‘A rooster was plucked by a goose.’

b. Lisa byla utešena korovoj. ‘A fox was consoled by a cow.’

c. Žiraf byl obnyuxan obez’janoj. ‘A giraffe was smelled by a monkey.’

Results - percentage of success

ActionalPsychologicalPerceptionTotal

TD*77%71%80%76%

SLI71%57%40%56%

* Younger TD do not distinguish the three types of passives, performing at chance level on all of them (see Babyonyshev & Brun 2003).

Is this universal?

Leonard, L. B., Wong, A. M. Y, Deevy, P., Stokes, S. F., and P. Fletcher .2006. The production of passives by children with specific language impairment: Acquiring English or Cantonese. Applied Psycholinguistics 27, 267–299

English – movement, one-to-many often reduced morpheme, adjectival/verbal confusion,

Cantonese – movement, no morphology, bei with a contrastive tone which is unique to passive

English

Cantonese

“The findings necessitate a modification of the assumptions of the sparse morphology hypothesis, and provide only partial support for the surface account. The English get-passives and the Cantonese passives employed in this study differ in their structure but both require some type of movement. However,we found no evidence that movement was at the heart of the children’s difficulties. If optional movement is a correct characterization, then we must assume that our tasks increased the likelihood that an available but optional movement operation was selected by the children with SLI."

The syntactic abilities of children with SLI: Binding

Johni shaved himselfi

1. John likes himself2. John likes him3. He likes John4. *Himself likes John

5. John thinks that Bill likes him6. He thinks that Bill likes John7. John thinks that Bill likes himself

Binding conditions

A: anaphors must be bound in their local domain

B: pronouns must be free in their local domain

C: R-expressions are always free

The coindexation resembles A-movement, but no theta role transmission is involved

The binding local domain varies across languages

Issues in acquisition

Which words are pronouns and which are reflexives.

What the local domain is. Principle A vs. principle B. Comprehension vs. production

Solan (1987 – (Act-out task, 37 children, ages 4-7.

Sentences % correct 95%

49%

82%

1. The dog said that the horse hit himself

2. The dog said that the horse hit him

3. The dog told the horse to hit himself

4. The dog told the horse to hit him 36%

85% 5. The dog found the horse’s picture of himself

6. The dog found the horse’s picture of him 1%

86%

38%

68%

7. The dog said that the horse found the picture of himself

8. The dog said that the horse found the picture of him

9. The dog told the horse to find the picture of himself

10. The dog told the horse to find the picture of him

23%

Chien & Wexler (1990) – Pictures selection, 150 children, ages 2;6-6;6

This is Goldilocks; this is Mama bear.

Is Mama bear touching herself/her?

Children older than 5 obey principle A. Younger children allow non-local antecedent: Goldilocks = herself

Children seem to violate principle B even after 6;6

But Children obey principle B at the same age that

they obey principle A, but violate a pragmatic principle which governs the choice of reference (Reinhart 1983, 1986).

Coreference is possible without coindexing on a pragmatic basis (contrastive stress). Children who are not sensitive to contrastive stress would seem to violate principle B ( McDaniel 1992)

Grice’s principles of cooperation (maxim of manner) – use the most precise way to say what you want to say - use him only when you do not mean himself. This is hard for children (Grodzinsky & Reinhart 1993)

Binding in SLI

Franks, S. L., Connell, P. J. 1996. Knowledge of Binding in Normal and SLI Children. Journal of Child Language, 23, 431-64

Reflexives NL - pass through a long-distance binding stage LI - behave like very young NL requiring the

nearest available noun phrase to be the antecedent.

Bishop et al. 2000. Grammatical SLI: A distinct subtype of developmental language impairment? Applied Psycholinguistics 21, 159–181

Advanced Syntactic Test of Pronominal Reference (Figure 2, A)

ResultsLI 18.72 (SD=2.90)LN 21.41 (SD=2.53)(t = 5.61, p < .001).

“Baloo Bear says Mowgli is tickling him” “Baloo Bear says Mowgli is tickling himself” (X) “Mowgli says Baloo Bear is tickling him” (S)

The syntactic abilities of children with SLI: WH-Questions

Questions in English Yes/no questions are marked only by subject-auxiliary

inversion, i.e., an overt syntactic change in word order in which the auxiliary is raised into C. Do-support operates when there is no auxiliary is the declarative.

[Spec, CP] is the target for overt Wh-movement both in matrix and embedded clauses, with subject-auxiliary inversion in matrix clauses, but not in embedded clause. Do-support operates when there is no auxiliary is the declarative. a. What did the child see?

b. The teacher wondered what the child saw.

Stromswold, K. 1995. The acquisition of subject and object wh-questions.

Longitudinal study of 12 children in CHILDES.

Who and what are acquired almost simultaneously, around age 2;5. Object questions are acquired at the same age or earlier than subject questions.

All children asked at least one long distance object question (mean age 2;10), but only one child asked a long distance subject question (at 5;0).

By the age of 2;6

TD children use wh-movement properly

TD children do not show problem with wh-non-local dependency

TD children have no problem with theta-government

Wh-Errors in Leonard Corpus of Children with SLI1 Which one I can do? (C ‘Which one can I do?) 2. What Kent’s gonna play with? (C ‘What’s Kent gonna play with?)3. How you knowed? (E ‘How did you know?’) 4. What he did? (F ‘What did he do?’) 5. What you doing? (E ‘What are you doing?’) 6. What this for? (G ‘What is this for?’) 7. How much we got to do? (J ‘How much have we got to do?’) 8. How you get this out? (A ‘How d’you get this out?’) 9. What this do? (A ‘What’s this do?/What does this do’) 10. How open it up? (B ‘How d’you open it up?’) 11. What say? (B ‘What d’you say?’) 12. Where go on? (B ‘Where’s it go on/Where does it go on?’) 13. How much long gonna be? (A ‘How much longer’s it gonna be?’) 14. These do? (C ‘What do these do?’) 15. What is this is? (H ‘What is this?’)

Wh-movement in children with grammatical SLI: A test of the RDDR hypothesis

Van der Lely HKJ and Battell J (2003), Language 79: 153-181

SLI subjects fail to master the syntax of the two types of movement operation involved in wh-questions (preposing a wh-expression and preposing an auxiliary).

This is the result of difficulties they have in processing non-local dependencies.

Subjects

15 SLI subjects aged from 11;3 to 18;2 12 TD (typically developing) grammar-

matched children aged from 5;3 to 7;4 12 TD (typically developing) vocabulary-

matched children aged from 7;4 to 9;1

Method

Wh-questions containing who, what and which by getting the subjects to play a version of the board game Cluedo:

Prompt Target response Mrs Peacock saw someone in the lounge. Ask me who Who did Mrs Peacock see in the lounge? Mrs Brown placed something in the library. Ask me what What did Mrs Brown place in the library? Professor Plum wore a coat. Ask me which one Which coat did Professor Plum wear?

Findings

Types of error produced by the subjects SLI subjects Younger TD controls Older TD controls No errors 0/15 (0%) 6/12 (50%) 12/12 (100%) AUX errors only 3/15 (20%) 4/12(33%) 0/12 (0%) WH errors only 0/15 (0%) 1/12 (8%) 0/12 (0%) WH and AUX errors 12/15 (80%) 1/12 (8%) 0/12 (0%)

Wh-errors(a) Who Miss Scarlett saw somebody? (Response to ‘Miss

Scarlet saw someone in the lounge. Ask me who’ – the target response being Who did Miss Scarlet see in the lounge?)

(b) Which Reverend Green open a door? (Response to ‘Reverend Green opened a door. Ask me which one’ – the target response being Which door did Rev. Green open?).

(c) What did Colonel Mustard had something in his pocket? (Response to ‘Something was in Colonel Mustard’s pocket. Ask me what’ – the target response being What was in Colonel Mustard’s pocket?).

Summary of findings

SLI subjects have far more problems with the syntax of wh-questions than language-matched TD controls.

The pattern of errors made by the SLI subjects differs from the pattern of errors made by the TD subjects: Most SLI subjects have problems with both auxiliaries

and wh-expressions Most TD subjects have problems with neither, or only

with auxiliary inversion.

Can it account for auxiliary inversion errors?

1. What cat Mrs White stroked?

2. What did they drank?

3. Who Mrs Brown see?

The Uninterpretable Feature Deficit Model (Tsimpli and Stavrakaki 1991) SLI children have problems with movement operations,

because these are driven by uninterpretable features. Chomsky (2006) argues that wh-movement is driven by an

interpretable edge feature on C which (in an interrogative clause) attracts an interrogative wh-expression to move to the edge of CP

Pesetsky and Torrego (2001) argue that auxiliary inversion is driven by an uninterpretable tense feature on C which attracts a tensed auxiliary to move from T into C.

Can UFDM account for why the SLI children in the Leonard corpus show perfect performance on wh-movement but perform much more poorly on auxiliary inversion.

Other topics: From Singleton to Exhaustive: the Acquisition of Wh-

Roeper, T., Schulz, P., Pearson, B. Z. & Reckling, I. (2006).   From singleton to exhaustive: The acquisition of wh-. Proceedings of SULA 2005 Conference (Semantics of Understudied Languages), Buffalo NY.

Who is eating what?

Double wh-question - Paired answer

Who is wearing a hat ?

Exhaustive answer, singleton answer, plural answer

The [+variable] Feature

Necessary in order to recognize exhaustivity

Specificity: relating to pre-established elements in the discourse +Specific = - variable = singleton, -Specific = +variable = exhaustive/paired.

Child’s initial default assumption: Questions are specific in nature

Results

All children pass through a singleton stage around age 4-5.

Singleton readings in four-year-olds: English 79%, German 52%

Exhaustive responses Age 5: German 80%, English 27% Age 6: German 85%, English 75% Age 7: German 84%, English 74%

Plural responses: 6%

The syntactic abilities of children with SLI:Relative Clauses

Types of complex clauses

Complement clauses – I want to drink, I know that she is late

Coordinate clauses – I like juice and she likes water

Adverbial clauses – I went to sleep when we got home

Relative clauses – The man who Mary saw was funny

Relative clauses

The girli that John kissed ti is nice

Relative clauses involve an A'-movement which yields coindexation of an NP in the main clause with a gap in the embedded clause, through an operator.

The operator carries the theta-role of its trace/gap

subject vs. object

Some languages have resumptive pronouns in RCs

ha-yalda she dani nishek ota nexmada

the-girl that Dani kissed her nice

'The girl that Dani kissed is nice'

Types of relative clauses Subject RC

The man who _ reads the book is my friend I saw the man who _ read my book האיש ש_קרא את הספר הוא ידידיפגשתי את האיש ש_קרא את הספר

Object RC The man who David saw _ is my friend I met the man who David saw _ האיש שדויד ראה _ הוא ידידיפגשתי את האיש שדויד ראה _

The head external analysis (Chomsky 1977, Jackendoff 1977, Partee 1975) The man [CP whoi [C 0 ] [IP Mary loves ti]] is my friend The man [CP Opi [C that] [IP Mary loves ti]] is my friend The man [CP Opi [C 0 ] [IP Mary loves ti]] is my friend

The head noun is base-generated outside CP The operator undergoes A'-movement to [Spec CP] The relative clause is right adjoined to the head noun The head noun and CP are combined via predication Resumptive pronouns are either base generated (a non-

movement analysis) or traces spell out (a movement analysis).

Issues in acquisition

Production vs. comprehension Resumptive NPs Subject vs. object

Production of relative clauses by TD children Children produce preconjunctional relative clauses even before the

age of 2:

a. *ze regel koevet laxthis foot-fm hurts-fm you'This is the foot that hurts you' [Lior 1;10;08]

b. *ze shaon ose tuktuk this clock does ticktock'This is a clock that goes ticktock' [Leor 2;1]

The complementizer appears around 2-2;6

aviron she la-shamayim [Lior 2;01;27]airplane that to-the-sky'an airplane that flies to the sky'

Which dog is happy?

Reem Bshara

Resumptive pronouns and resumptive NPs

Children initially use resumptive pronouns in French, English an other languages

Children use resumptive NPs The zebra who the man sat next to the zebra.

The non-movement approach (cf. Labelle 1988, 1990, 1996, Goodluck & Stojanoviç 1996)

The movement approach (cf. Law 1992, Pérez-Leroux 1995, Guasti & Shlonsky 1995, McDaniel, Bernstein & McKee 1997, Varlokosta 1997a).

Resumptives and Wh-Movement in the Acquisition of Relative Clauses in Modern Greek and Hebrew Varlokosta & Armon-Lotem (1998)

Dependency Number +cl/pronoun -cl RNP S 55 *13 (24%) 42 (76%)

DO 44 41 (93%) 2 (5%) *1 (2%) IO 38 38 (100%) PP 40 33 (83%) *3 (7%) *4 (10%)

24 monolingual Hebrew-speaking children from 2;8 to 5;5

Comprehension and Production of relative clauses by children with SLI

Novogrodsky, R., & Friedmann, N. 2006. The production of relative clauses in SLI: A window to the nature of the impairment. Advances in Speech-Language pathology, 8(4).

Novogrodsky & Friedmann (2006)

18 Hebrew-speaking children with SLI, aged 9;3 to 14;6 years (mean =12;6).

28 TD children divided into three subgroups: 8 (7 years old), 13 (9 years old), and 7 (10 years old).

13 participated in the preference task, and 16 participated in the picture description task (11 participated in both tasks).

Picture selection

Preference (Adif) - SR

Preference (Adif) - OR

Findings

Children with S-SLI have difficulties in the production of relative clauses, especially in object relatives that were mainly related to thematic role assignment.

Age is not a factor in the production of relative clauses in that age range. Their production was either identical or virtually the same with no significant difference in both tasks.

In the preference task the S-SLI children produced significantly fewer target object relatives than the control group (60% compared to 94%), and significantly fewer subject relatives (94%compared to 99%).

In the picture description task the S-SLI children produced significantly fewer target object relatives than the control group (46% compared to 94%), and significantly fewer subject relatives (83%compared to 98%). (See figure 3).

The children used a variety of structures in order to provide a task-appropriate response without using the impaired syntactic abilities. The non target responses in both tasks included thematic errors and reduction of thematic roles, avoidance of movement from object position, relative head doubling and production of simple sentences without a relative clause.

No complementizers were omitted.