.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Syntax
Pavel Caha
20 Sept 2020
1 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Studying syntax
▶ Language – human ability to communicate
2 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Studying syntax
▶ Language – human ability to communicate
▶ Syntax – pu ng words together
2 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Studying syntax
▶ Language – human ability to communicate
▶ Syntax – pu ng words together
▶ Using nite means to create in nite range of sentences
2 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Studying syntax
▶ Language – human ability to communicate
▶ Syntax – pu ng words together
▶ Using nite means to create in nite range of sentences
▶ How can we characterize/describe this ability?
2 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Studying syntax
▶ Language – human ability to communicate
▶ Syntax – pu ng words together
▶ Using nite means to create in nite range of sentences
▶ How can we characterize/describe this ability?
▶ What is it that we have and animals don’t?
2 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Humans
(1) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
3 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
Varia on
Beyond humans
Language as a linear string
Conclusions
4 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
(2) black cab drivers
5 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
(2) black cab drivers
a. drivers of black cabs
5 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
(2) black cab drivers
a. drivers of black cabs
b. cab drivers who are black
5 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
(2) black cab drivers
a. drivers of black cabs
b. cab drivers who are black
black cab
drivers
5 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
(2) black cab drivers
a. drivers of black cabs
b. cab drivers who are black
black cab
drivers black
cab drivers
5 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(3) 5 + 3 x 2
6 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(3) 5 + 3 x 2
a. ( 5 + 3 ) x 2 = 16
6 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(3) 5 + 3 x 2
a. ( 5 + 3 ) x 2 = 16
b. 5 + ( 3 x 2 ) = 11
6 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(3) 5 + 3 x 2
a. ( 5 + 3 ) x 2 = 16
b. 5 + ( 3 x 2 ) = 11
+16
+8
5 +3
x 2
6 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(3) 5 + 3 x 2
a. ( 5 + 3 ) x 2 = 16
b. 5 + ( 3 x 2 ) = 11
+16
+8
5 +3
x 2
+11
5 +6
+3 x 2
6 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(4) 5 + 3 x 2
a. ( 5 + 3 ) x 2 = 16
b. 5 + ( 3 x 2 ) = 11
black cab
drivers black
cab drivers
7 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(5) hit the dog with a hat
8 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(5) hit the dog with a hat
a. the dog has a hat on
8 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(5) hit the dog with a hat
a. the dog has a hat on
b. the hat is used to hit the dog
8 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(5) hit the dog with a hat
a. the dog has a hat on
b. the hat is used to hit the dog
event
hit the dog
with a hat
8 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(5) hit the dog with a hat
a. the dog has a hat on
b. the hat is used to hit the dog
event
hit the dog
with a hat hit thing
the dog with a hat
8 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
(6) hit the dog with a hat
a. the dog has a hat on
b. you use the hat
event
hit thing
the dog
with a hat
event
hit thing
the dog with a hat
9 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
Varia on
Beyond humans
Language as a linear string
Conclusions
10 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(7) a. 1 + ( 2 x 3 )
11 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(7) a. 1 + ( 2 x 3 )
b. 1 + ( 3 x 2 )
11 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(7) a. 1 + ( 2 x 3 )
b. 1 + ( 3 x 2 )
c. ( 3 x 2 ) + 1
11 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(7) a. 1 + ( 2 x 3 )
b. 1 + ( 3 x 2 )
c. ( 3 x 2 ) + 1
d. ( 2 x 3 ) + 1
11 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(1) a. 1 + ( 2 x 3 ) [123]
b. 1 + ( 3 x 2 )
c. ( 3 x 2 ) + 1
d. ( 2 x 3 ) + 1
12 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(1) a. 1 + ( 2 x 3 ) [123]
b. 1 + ( 3 x 2 ) [132]
c. ( 3 x 2 ) + 1
d. ( 2 x 3 ) + 1
13 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(1) a. 1 + ( 2 x 3 ) [123]
b. 1 + ( 3 x 2 ) [132]
c. ( 3 x 2 ) + 1 [321]
d. ( 2 x 3 ) + 1
14 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(1) a. 1 + ( 2 x 3 ) [123]
b. 1 + ( 3 x 2 ) [132]
c. ( 3 x 2 ) + 1 [321]
d. ( 2 x 3 ) + 1 [231]
15 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(8) SVO vs. SOV
a. (Hans says that) the dog eats the bone
16 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(8) SVO vs. SOV
a. (Hans says that) the dog eats the bone
b. (Hans
Hans
sagt
said
dass)
that
der
the
Hund
dog
den
the.ACC
Knochen
bone.ACC
frisst.
eats
‘(Hans says) that the dog eats the bone.’
16 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(8) SVO vs. SOV
a. (Hans says that) the dog eats the bone
b. (Hans
Hans
sagt
said
dass)
that
der
the
Hund
dog
den
the.ACC
Knochen
bone.ACC
frisst.
eats
‘(Hans says) that the dog eats the bone.’
sentence
subject
the dog
event
verb
eats
object
the bone
16 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Meaning and ordering
(8) SVO vs. SOV
a. (Hans says that) the dog eats the bone
b. (Hans
Hans
sagt
said
dass)
that
der
the
Hund
dog
den
the.ACC
Knochen
bone.ACC
frisst.
eats
‘(Hans says) that the dog eats the bone.’
sentence
subject
the dog
event
verb
eats
object
the bone
sentence
subject
der Hund
the dog
event
object
den Knochen
the bone
verb
frisst
eats
16 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
Varia on
Beyond humans
Language as a linear string
Conclusions
17 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
How to study cogni ve abili es
(9) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
▶ How can you gure out whether, e.g., monkeys have such
structures?
18 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
How to study cogni ve abili es
(9) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
▶ How can you gure out whether, e.g., monkeys have such
structures?
▶ How can you gure out whether children have such structures?
18 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
How to study cogni ve abili es
(9) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
▶ How can you gure out whether, e.g., monkeys have such
structures?
▶ How can you gure out whether children have such structures?
▶ Preferen al looking paradigm
18 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
coun ng
19 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
coun ng
20 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
coun ng
21 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
Varia on
Beyond humans
Language as a linear string
Conclusions
22 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(10) a. Mary invited Sue
b. Sue invited Mary
c. AGENT > PATIENT
23 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(10) a. Mary invited Sue
b. Sue invited Mary
c. AGENT > PATIENT
(11) a. John pushed Bill
b. Bill pushed John
c. AGENT > PATIENT
23 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(10) a. Mary invited Sue
b. Sue invited Mary
c. AGENT > PATIENT
(11) a. John pushed Bill
b. Bill pushed John
c. AGENT > PATIENT
23 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
WALS
Figure: This is an image from WALS
24 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(12) a. Mary turned Sue (around)
25 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(12) a. Mary turned Sue (around)
b. Mary and Sue turned (around)
25 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(12) a. Mary turned Sue (around)
b. Mary and Sue turned (around)
c. ¬AGENT > PATIENT
25 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(12) a. Mary turned Sue (around)
b. Mary and Sue turned (around)
c. ¬AGENT > PATIENT
Mary
turned Sue
25 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Strings
(12) a. Mary turned Sue (around)
b. Mary and Sue turned (around)
c. ¬AGENT > PATIENT
Mary
turned Sue Mary and Sue
turned
25 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Structure vs. Linearity
26 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
The setup
27 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
The experiment
28 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
The results
29 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Kanzi
30 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Kanzi
31 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Kanzi
32 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Kanzi
33 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Kanzi
34 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Ambiguity
Varia on
Beyond humans
Language as a linear string
Conclusions
35 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
conclusions
(13) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
36 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
conclusions
(13) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
a. ambiguity
36 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
conclusions
(13) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
a. ambiguity
b. language varia on
36 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
conclusions
(13) Humans have a mul -domain capacity and proclivity to infer
tree structures from strings, to a degree that is di cult or
impossible for most non-human animal species.
a. ambiguity
b. language varia on
c. li le reliance on ordering cues
36 / 37
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
References
37 / 37