Chp 1 Review PPT.ppt

8/12/2019 Chp 1 Review PPT.ppt

1/38

Syntax

Andrew Carnie


2/38

Topic 1: Syntax: some background

What is syntax?

Syntax as a (cognitive) science

Rulesprescriptivism vs. descriptivism

Evaluating Grammars

Language as an instinct


3/38

Q. What is Syntax??

The scientific study of sentence

structure

Perspective: The psychological (or

cognitive) organization of sentence

structure in the mind.


4/38

Q. What is a sentence??

A hierarchically organized structure of

words that maps sound to meaning and

vice versa.

soundssentencesmeaning


5/38

Scientific Method

Study of synta is a science.

!ses the scientific method" #$serve some data

" %a&e some generalizations

" 'evelop a hypothesis

" Test against more data


6/38

Scientific method

1) John loves himself

2) Mary loves herself

3) John and Mary love themselves

Generalization: The form of the Xself seems to be

dependent upon the gender/number of the noun they

refer to.

Hypothesis: Anaphors (Xself) agree with the nounthey refer to in number and gender.

4) The boy loves himself/*herself/*themselves

Anaphor:A noun that

refers back to a

previously mentioned

noun: self nouns.


7/38

Rues: ! kind of hypothesis

n this class we will encode our hypotheses

a$out sentence structure using rules.

A group of rules are called a *rammar. *rammar is a scary word. +ut it doesn,t mean

what you thin& it does. A grammar in the

linguistic sense is a cognitive structure. t isthe part of the mind that generates and

understands language.


8/38

"rescripti#e #s. $escripti#e

Rues-e are always told to never split

infinitives.

-ho(m) did you give the $oo& to

/opefully we,ll never learn the rules of

grammar0


9/38

"rescripti#e #s. $escripti#e

Prescriptive rules prescri$e how we

should spea&

'escriptive rules descri$e how weactually spea&.

Which is more scientific?


10/38

"rescripti#e Rues

These are made up $y so called

language mavins

These are made up by so caed anguage ma#ins% Who are

they to te you ho& to speak?%?

"rescripti#e rues are often based on the rues of 'atin or

(ogic). Who says 'atin is so great? Why shoud anguage be

ogica?

Descriptive rules are the way to go!Descriptive rules are the way to go!


11/38

$escripti#e Rues

The rules we will use are said to

generatethe sentences of the

languages we are loo&ing at. Theyactually $uild the sentences we

produce. They are sentence $uilding

rules.The &ind of grammar we are loo&ing at

is called generative grammar(1group ofrules that generate the sentences of a language)


12/38

Sources of $ata

Corpora of Spo&en 2 -ritten 3anguage

4 Collections of recorded real world speech

4 Telephone recordings (3'C)

4 5ewspapers +oo&s %agazines

4 6ol& tales etc recorded in the field.


13/38

Sources of $ata

*Where do you wonder if he lives?

How do you know this is ungrammatical?

Have you ever heard this sentence uttered?

Will the fact that this sentence is

ungrammatical appear in any corpus?

Every day, you produce grammaticalsentences that have never been uttered

before.


14/38

Sources of $ata

Corpora are not sufficient. They dont

contain negative information (such as what

sentences are ungrammatical), and they cannever contain all the sentences of a

language.

We need to access our mentalknowledge(also called competence)

about sentences.


15/38

Sources of $ata

We use a special experimental technique for

tapping our syntactic knowledge. This

technique is called the acceptabilityjudgement. (In the psychology literature,

this is sometimes also called magnitude

estimation)


16/38

!cceptabiity *udgements

!nfortunately sometimes accepta$ility

7udgements are called intuitions.

The term 8intuition, has a negativeconnotation: ma&es us thin& of fortune

tellers and psychics.

/owever accepta$ility 7udgements are$oth experimentally validand

statistically sound.


17/38

!cceptabiity *udgements

-e will apply accepta$ility 7udgements

in this class non9statistically. 6or the

most part this will give us the rightresults. Statistical proof of 7udgements

is possi$le $ut we won,t $other.


18/38

"erformance #s. +ompetence

Performance: refers to what we actually

produce

Competence: refers to what we &nowa$out language.

-e will $e interested in bothof these

$ut will focus primarily on Competence.


19/38

,#auating -rammars #$servationally Adeuate *rammar: A

grammar that accounts for all the o$served

(corpus;performance) data.

'escriptively Adeuate: Accounts for allo$served data and all accepta$ility

7udgements (competence).


20/38

'earning #s. !cuisition

Learninginvolves conscious gaining of

&nowledge

Acquisitioninvolves su$consciousgaining of &nowledge

Chemistry is learned. Languages areacquired.


21/38

/o& do &e acuire anguages?

#$viously this uestion is too $ig to

answer here $ut =

Are we instructed $y our parents

'o we mimic our parents

NOPE!

1) Language is infinite: We produce sentences wevenever heard before

2) We know things about our language that weve never

been exposed to.


22/38

'anguage as an instinct

>ou &now things a$out your language

that you,ve never $een taught:

Who0m did you think Sha&n hit ?

Who0m did you think that Sha&n hit?

Who did you think hit 2i

*Who did you think that hit Bill


23/38


Despite what they may think, parents dont teach their

children to speak!

They correct content not form:(from Marcus et al. 1992)

Adult: Where is that big piece of paper I gave you yesterday?

Child: Remember? I writed on it.

Adult: Oh thats right, dont you have any paper down here,

buddy?


24/38


(from Pinker 1994, 281 attributed to Martin Braine)

Child: Want other one spoon, Daddy

Adult: You mean, you want the other spoon.

Child: Yes, I want other one spoon, please Daddy.Adult: Can you say the other spoon?

Child: Other one spoon

Adult: Say other

Child: other

Adult: spoon

Child: Spoon.

Adult: other spoon

Child: other spoon. Now give me other one spoon.


25/38

! shocking proposa%

5oam Choms&y

The ability of humans to use language is innate (an

instinct). We are prewired to use language!


26/38

/uh? anguages differ?%?

/ow can language $e an instinct if

languages differ

Proposal: 3anguages differ primarily interms of what words are used and in a

set num$er of ?parameters@

These things are learned $ut the rest (the$asic architecture of the grammar) is

innate.


27/38

Refining 3nnateness

A particular language is not innate (it is

acuired) $ut the $asic tools that any

given language uses are $uilt in.

-e,ll $e loo&ing at these tools. +oth

within languages and crosslinguisticallyto see what is universal (innate) and

what varies among languages.


28/38

The ogica probem of anguage

acuisitionA proof showing the following:

Premise: 3anguage is creative 2

infinite.

Premise: nfinite systems are

unlearna$le;unacuira$le

Conclusion 3anguage is unlearna$le.

2 it must $e innate.


29/38

Task of a chid acuiring ,ngish

%atch up a sentence that they hear with

a situation in the contet around them.

The cat spied the kissing fishes =

To ma&e the proof let,s turn this into analge$raic operation. -e,ll num$er sentences

and we,ll num$er situations and loo& for the

rule that matches them up.

1

1


30/38

Matching sentences to situations

Sentence

X

Situation

Y

1 1

2 2

3 3

4 4

5 5

Given the sentence 6, what situation do

you think it will match to?


31/38

Matching sentences to situations

SentenceX

SituationY

1 1

2 2

3 3

4 4

5 5

6 126

You assumed the rule was x = y

In fact: the rule is

[(x-5)*(x-4)*(x-3)*(x-2)*(x-1)]+x = y


32/38

'ogica "robem: in ayman4s

terms t is impossi$le to learn the rule(s) governing

a system until you have A33 the data.

3anguage is infinite 2 creative: you can neverhear all the relevant data. (t is impossi$le to

&now if you have 7ust coincidentally missed

hearing the crucial fact)

Therefore: The $asic $uilding $loc&s oflanguage cannot $e learned or acuired.

nstead they must $e innate (an instinct)


33/38

The content of this cass

n this class we will $e loo&ing at the

innate principles that govern sentence

structure (Called !niversal *rammar)And we will $e loo&ing at the different

ways in which languages implement

these innate principles.


34/38

5ni#ersa -rammar 05-

The $uilding $loc&s that all languages

use to construct the sentences of their

languages.All languages use the same $asic

hardwired tools. t is theparticular

implementationof these tools thatvaries $etween languages.


35/38

5ni#ersa -rammar 05-

#ther evidence for !*

4 /uman Specificity of 3anguage

4 'istinct area of the $rain

4 Crosslinguistic similarities in language

acuisition (despite cultural differences)

4 3ac& of overt instruction4 3anguage !niversals


36/38

Summary

Synta: A Science uses Scientific

method studies sentence structure

Prescriptive;'escriptive ules*enerative ules as /ypotheses

Sources of 'ata:

4 Corpora

4 Budgement tas&s


37/38

SummaryPerformance;Competence


38/38

$iscussion Topics

-hat things that we &now are learned

-hat things are acuired

3anguage is an instinct. /ow is this anargument against prescriptive rules

There are some good reasons to &eep

prescriptive rules. -hat are they

Documents

Chp 1 Review PPT.ppt