Ling 240: Language and Mind
Structure Dependence in
Grammar Formation
TWO HYPOTHESES ABOUT
YES/NO QUESTIONS



How do we form yes/no Qs?
a. The man is tall. Is the man tall?
b. The book is on the table. Is the book
on the table?

c. I can go.
Can I go?
Move the first Aux



Move the first Aux
The ordinal order of X is determined by
position in a sequence => a structure
independent
A structure independent rule ignores
higher-order structural units, relying only
on linear order
Does ‘Move the first Aux’ rule
work?





‘Move the first Aux’ rule doesn’t work when
there is more than one Aux
(1) The man who is tall is in the other room.
which becomes:
(2) *Is the man who ___ tall is in the other
room?
(3) Is the man who is tall ___ in the other
room?
Structure Dependent
the correct hypothesis must therefore
recognize the internal structure of sentences
 Move the highest Aux
 The height of X is determined by the
number of nodes that dominate X =>
structure dependent

UG narrows down the
hypotheses space



Input is compatible with infinite number of
hypotheses
Claim: Innate principles (and parameters)
guide learning (Universal
Grammar)
Prediction: children will not make types of
errors that violate UG principles
Summary: Adult Mental Grammar
of English

Rule: Move the highest Aux.

Mental Grammar only allows the structuredependent version of the rule.

What do kids do? What types of mistakes
do they make? (Crain & Nakayama 1987)
Do children ever consider “move
the first” rule?


Null hypothesis: children do not have
innate grammatical knowledge that makes
them ignore structure-independent rules.
Prediction: children should consider the
“move the first” rule, because the rule is
simple, concrete and perfectly compatible
with their experience.
Crain and Nakayama 1987




“Elicited production” experiment
Participants: English-speaking children
(N=30, Age: 3;2 – 5;11)
Group 1: 3;2 – 4;7 (Mean 4;3)
Group 2: 4;7 – 5;11 (Mean 5;11)
Hey Tommy, look at this! Look at
this picture!
Tommy
Experimenter
Jabba the Hutt
Tommy, do you think the girl is tall?
Experimenter
Tommy
Jabba the Hutt
Experimenter
Noooo! She is not tall!
Tommy
Jabba the Hutt
Yeah I agree…but I wonder what
Jabba would say…
Tommy
Experimenter
Jabba the Hutt
Hey Tommy, why don’t you ask
Jabba if the girl is tall? Ask Jabba
if the girl is tall!
Tommy
Experimenter
Jabba the Hutt
Experimenter
Is the girl tall?
Tommy
Jabba the Hutt
Experimenter
No…
Tommy
Jabba the Hutt
He was right! Tommy, give him
a strawberry!
Tommy
Experimenter
Jabba the Hutt
Materials

Pretest sentences: to ensure that children
understood the task and could form
simple yes/no questions.
a. The girl is tall
b. The man is tired
c. The pig next to the tree is red
Materials


Test sentences
a.
b.
c.
d.
e.
f.
[The dog that is sleeping] is on the blue bench.
[The ball that the girl is sitting on] is big.
[The boy who is watching Mickey Mouse] is
happy.
[The boy who is unhappy] is watching Mickey
Mouse.
[The boy who is being kissed by his mother] is
happy.
[The boy who was holding the plate] is crying.
“Ungrammatical” questions


We are interested in whether children
consider the “move the first” rule…
*Was the boy who __ watching TV is
crying?  structure-independent, or
“TYPE III” error
“Ungrammatical” questions


“Good” results would be
If children never made any ungrammatical
questions at all.
Even stronger results would be:
If children made various ungrammatical
questions, but they never made TYPE III errors.
Results I: did they make
ungrammatical questions?

YES.
Group I
Group II
Total
81
87
168
Grammatical
31 (38%)
70 (80%)
101 (60%)
Ungrammatical
50 (62%)
17 (20%)
67 (40%)
Are there any Type III (structureindependent) errors?

Was the boy who __ watching TV is
crying?
Group I
Group II
Type I
30(60%)
9 (53%)
Type II
10 (20%)
5 (29%)
Type III
0
0
Type I = extra Aux. Is the boy who is watching TV is happy?
Type II Restart. *Is the boy who is watching TV, is he happy?
*Was the boy who __ watching TV is crying?  structure-independent, or
“TYPE III” error
Children never made Type III
errors.

suggests that they never consider
structure-independent movement rules:
“move the first…”.
Do children ever consider “move
the first” rule?


Null hypothesis: children do not have
innate grammatical knowledge that makes
them ignore structure-independent rules.
Prediction: children should consider the
“move the first” rule, because the rule is
simple and perfectly compatible with their
experience.
Remaining question

Could they get the right pattern from
input directed to them? (Legate & Yang
2002)
The CHILDES database

A huge database of spontaneous speech by young children (age 2 –
6)

Transcriptions of parent-child linguistic interactions

What kinds of questions do children hear from adults?
Legate and Yang (2002)

Child: Adam (2;7 - )

Total parental utterances:
Number of questions:

The crucial sentences:

46,499
20,651
0
Child-directed speech does not
involve the crucial data


Is John happy?
is = the first, and the highest
Is the boy who was watching TV crying?
is = NOT the first, but the highest
So…




Input directed to children does not involve the crucial
data to distinguish “move the first Aux” from “move the
highest Aux”
Therefore, the input does not tell children that “move
the first” is wrong.
Innate linguistic knowledge /UG (Universal Grammar)
UG restricts the range of possible rules that children
consider in the course of language acquisition
Descargar

Ling 240: Language and Mind