American Journal of Computational Linguistics

Microfiche 82

MODEL F O R KNOWLEDGE AND I T S APPLICATION T O

D I S C O U R S E A N A L Y S I S
BRIAN

PHILLIPS

University of Illinois at Chicago C i r c l e

60680

Copyright @ 1979 A s s o c i a t i o n f o r Computational Linguistics

TABLE OF CONTENTS

Introduction.

.......................

L . .

A model for knowledge

Nodes
Arcs

Par a d i g m a t i , ~relations Va r ie t y Instance Typical Manifestation Syn t a g m a t i c r e l a t i o n s Discursive relations The metalingual r e l a t i o n Status relations Negation Inhe r i @ a n c e E p i s o d ic and s y s t e m i c menlory Quantification Processes in the network Pa ttl-tracing Pattern-matching

Discourseanalysis. Thq st^-IIC t u r e o f c ~ h e r e t d ~ S C C U ~ S ~ n Anaphora S p a t i a l , t e m p o r a l , and c a u s a l c o h e s i o n Thematicity T h e r o l e of t h e e n c y c l o p e d i a Anaphora Spatial, temporal. and causal cohesion Thematicity
Implementation..

.......... ........ ..
a

44 44 44 45 46 52 53

5.4 56
a

..... .......... . . ...

57
58 58 58

Processes
Paradigmatic path tracing Causal connectivity conditfon Discovering g e n e r a l and apecif i c p r o p o s i t i o n s ? l e t s l i n g u a l decomposi t i s n Iletalinp,ual a b s t r a c t i o n Inference of o m i t t e d d i s c u r s i v e r e l a t i o n s The system

58 60 60 62 62

Analysis of some s t o r i e s Common p a t terns A boat capsizing A f a l l i n t o water Embedded themes

..

......... ....

63 65 67 70 72

INTKODUCTTON

An

i m p o r t a n t c s n t r i b u t i o n of n a t u r a l Language p r o c e s s i n g h a s been

to d i r e c t a ttr:n tion t o tile

structure

of

1~~1gi~aat ge

the
II

discau~.se

l e v e l , wl-rich hna l e d

r6 n

g r e a t e r awareness uE t h e r o t e o f

meaning" in

Iangtxage For "A t e x t is best: regarded a s n SEIAhTXG u n i t ; a u n i t n o t a

form

but

o f meaningt'

(Ilalliday

ct

Ilasan, 1976, p. 2 ) .
u r
tlndcrstanding

This b e i n g so,
s f
meaning

discsursc

analysis

will

deepen

and

vice-versa

In

t h i s paper 1: p r e s e n t a rnodel, of nzcaning s t r a n ~ l y inf'l~lencccl h y

Ilays (].969a,

3.969b, 1970,
of

1973) and s t l o w hob I i s a b l e t o ccciptwre t;

tf~c

otganizattnn

L ! ~ H C D U ~ S C ~ .I n

particular

T seek t o d e f i n e the or-

g a n i z a t i o n o f cahercnt d i ~ c o u r s t - and t o slaow llow knowledge is . inhex a

u'sed

to

coherent

structure when, as u s u a l l y is the case, the surface

forrr: i s e l l f ~ 3 t i c . Tt~eh y p a t h e s c s a r c a ~ s e d to b u i L d an automati-c system

tn test t h e coherence of d ibscnnrsc. c

A f hO'fJR1, FOl< IiiJCIE.JX,l? DCJ;

T h e p l ~ i l o s o p f ~ is t n r ~ c c -fs taken t h a t u t r r knowlr.dg.,c of c

the

;I

c s n c e p is

f~icarling of

c t ~ a t concept::

"5arrtcanc w t ~ o knows wi:at

tik-t.r

means

. . I s required
( F u t n a n r , 1975, p
)!any

to

know
1

that

w ~ e r c n t y p ~ t a lt i g e r s

are

striped"

249)
of

models

kr,owX.edfie 'have

bccn

d < w e l o p e b f o r use I n com1908:

p u t a t i o r a a l env-dr:,nmcnts

Samp arc. f

c nrcstr i c tt.d dl-~mrai.rrz; ( B l a c k , ~

E b b r o w , 1968; Colby, 1973; Raphael, 1968; Winograd,

1W1; etc.)

The

present model is raore'in t h e tradition of Klein, Oakley, Suurballe, and

Ziesemer
(1972),

(1 9 7 2 ) ,

Quillian

( 19 6 9 ) ,

Rumelhar t , Lindsay,
(1970),

and

Normaa
Wllks

Schank (1975ai,

Shapiro

(1971), S i m m o n s

and

(1972), where no particular context is prescrfbed*

It will be apparent

that a t many p o i n t s t h e present model draws upon these earlier syatems.

Some of
notation-

the

differences between systems are probably differences i n

However no system I s a t a s t a g e of c o ~ s t a n e yor completeness

much
effort
to

that makes it worthwhile to devote

equivalences

establishing

the

Although i t wauld be possib1.e to present only t h e paws

common
nota-

that I believe t o be novel, giving the whole system i n a

tion w i l l ease the task of the readerThe

model,

hereafter

called

the

encyclopedia, endeavors t o be

consistent w i t h available psychological and

structure
of

linguistic

views

of

the

language

and t'hought, f o r any automated language system

must c l o s e l y imitate the ttorkings of human c o g n i t i o n to

be

successful

(Collins
The

&

Q u i l l f a n , 1972). encyclopedia
encodes common knowledge sf t h e world which may
Putnam (1975)

d i f f e r from s c i e n t i f i c a l l y accurate descriptions.

calls

such knowledge "stereotypical":

f a c t that a feature l e , i n c l u d e d in the stereotype associated with a word X does not mean t h a t it is an analytic truth t h a t a l l Xs have that feature, nor that most Xtl have that f e a t u r e , nor that a l l normal Xs have that feature, nor that some Xs have t h a t feature. Dtwovering that our stereotype i s based on nonnormal or unrepresentative members of a natural. kind is n o t discovering a l o g i c a l contradiction. a [but) The fact is that

The

...

. .

could hardly communicate i f moat of our stereotypes weren't p r e t t y accurate as far a s t h e y g o e ( p p . 250-251)
we

The encyclopedia is achematized

graph;

and

ihplemented

as a

directed

in

current parlance it i s a network model.

Nodes characterize
state-

concepts a n d arcs r e l a t i o n s between conceptsv The most general

ment

to

make

about the model is t h a t r e l a t i o n s and aoneept types are

t h e necessary system p r i m i t i v e s ; some concepts may

be

primitive,

but

t h e model does not depend on the*existence o f p r i m i t i v e concepts

Discussion

wi2 1 cover

the

nodes

and

relations

of t h e model.

Attention w i l l a l s o b e given to network processes.

No p s y c h o l ~ g i c a lvalidity i s claimed for the content of any of the

structures shown; the claim extends only to the

relational

sttucture

Questions of content must be answered empirically-

Nodes

There modality.

are

four

types

o.f nodes:

event

Y entit r a t t r i b u t e , and

-.

The f i r s t three correspond t o simple verb

simple noun,

and
Its

simple modifier, r e s p e c t i v e l y .

The fourth t y p e of node is novel

r a l e in *he system w i l l become clear after a description o f arcs *

For

the

meantime

it

will, have

to

s u f f i c e t o say t h a t i t i s u s e d i n t h e

spatio-temporal causal, b e l i e f , and hiersrchic

organization

of

know-

ledge *

Its

ancestor

in

linguistic

theory

is

"modal"

in

the
has

model/proposition dichotomy of Fillmore ( 1 9 6 9 ) .

Sch~bert (1976)

predicate

nodes

that

are similar in motivation, but different in use

-6

from m o d a l i t y nodes.

Nodes of t h e encyclopedia are not

1972).

labeled

(Collins

&

Quillian,

An

arc,

termed name, p o i n t s from a node i n t o a d i c t i o n a r y of

p r i n t names.

For clarity nodes i n diagrams w i l l be annotated, b u t t h i s

implementation,

s h o u l d n o t b e taken a s r e p r e s e n t i n g the

which

is

as

shown in Figure 1.
I I
I

DICTIONARY

ENCYCLOPEDIA

1 NAME
rock

person

Peter
Adnt Sally
I

Figure 1 L a b e l i n g nodes

In
node-

all

the

following figures

is an event, e n t i t y or a t t r i b u t e

Annotations on these nodes are

enclosed

in

1 , <>,

and

[I,

respectively-

M o d a l i t y n o d e s appear as o and are never annotated.

Arcs
Five
t y p e s o f arcs are used in t h e network: paradigmatic a r c s a t e
propositions,

taxonomic, syntagmatic arcs iorm

discursive

arcs

link

propositions, t h e r n e t a l i n g u a l arc i s u s e d t o a s s o c i a t e a concept w i t h a

s t o r y i n network form that defines t h e c o n c e p t , and s t a t u s a r c s characterize b e l i e f s and d e s i r e s +

Paradigmatic R e l a t i o n s

Variety.

readily

observable

aspect of human b e h a v i o r is the

studied in

existence o f f o l k t a x o ~ ~ o m i e s .These have been

detail

by

e t t ~ n a s e m a n t l c i ss i n order t o discover t h e i r c o g n i t i ~ es i g n i f i c a n c e and t

structure:

Man

is

by

nature a classifying animal.

His c o n t i n u e d existence

d e p e n d s on his a b i l i t y t o r e c o g n i z e s i m i l a r i t i e s and d i f f e r e n c e s between o b j e c t s a n d events i n Iris p h y s i c a l u n i v e r s e and t o make known these s i m i l a r i t i e s and differences l i n g u i s t i c a l l y Ifideed , t h e very development of t h e human mind seems to have been closely related t o t h e p e r c e p t i o n of d i s c o n t i n u i t i e s i n n a t u r e . In view of t h i s , t h e s t u d y o f f o l k taxonomic systems, which have received a great d e a l o f i n t e r e s t i n recent years, h a s a h i g h s i g n i f i c a n c e i n i n t e r p r e t i n g t h e logical processes going on i n o u r minds, a s w e l l as in u n d e r s t a n d i n g t h e a p p l i c a t i o n and u t i l i t y of t h e taxoncmfc systcms themselves. (Raven, E e r l i n , 6 Breedlove, 1971, p 1210)

For example, mammal, b i r d , and r e p t i l e might b e classified as k i n d s

vertebrates*

of

Tn

the

network,

the

relation

is termed v a r i e t y (ab-

b r e v i a t e d t o VAR i n the f i g u r e s ) ,

Figure 2 diagrams t h i s knowledge.

<pecyn
(sleepwalk))

1rt)>
(talk in sleep))

' 4
((hairy William Proxm ire)

1 / CAUSE
/ n n r n r r - / * r a m \ \

William Proxmire)

(William Proxmire (makes foreign policy

statements))

Figure 2

Paradigmatic organization

Varietal

nodes

are

seen

as

representing

concepts at a categqrical

level, hence variety is the

Breedlove,

category-subcategory

relation.

Berlin,

and Raven ( 1 9 6 8 ) show the existence of covert categories in

not

folk taxonomies, L e e , nodes having s c i e n t i f i c , but

say
It

folk,

names,

vertebrate" in Figure 2.

These categories are revealed by memory,

classification,

and other experiments.

ThLs is counter t o the view o f

Conklin (1962) f o r whom concepts must have monolexemic labels.

Covert

c a t e g o r i e s e n a b l e Raven e t a l e (1971) to show a degree o f u n i f o r m i t y i n taxonomies: about

five h i e r a r c h i c l e v e l s with seldom more than f i v e

Berldn e t a l . (1968) c l a i m t h a t items i n categories+ i .e., the structure

hundred items under o n e node.

a f o l k taxonomy form n o n - i n t e r s e c t i n g

is

strictly

tree-like.

T h i s view i s n o t h e l d h e r e , f o r a t y p e w r i t e r

can b e c l a s s i f i e d b o t h a s
Cobequently , v a r i e t a l
like.

machine

and
are

as

writing

instrument.

structures

n o t r e s t r i c t e d t o being treeNor is it n e c e s s a r y

Loops, however, do n o t seem p o s s i b l e .

that

A n~de have a name.

Enstance. kogic

since A r i s t o t l e has d i s t i n g u i s h e d between c a t e member

of

gory ( o r type) and a s p e c i f k membership relation

category

(token).

This

is

termed i n s t a n c e ( I S T ) . F i g u r e 2.

For example, "William Most i n s t a n c e s are n o t

Proxmire" i s an i n s t a n c e of "person",

named, t a k h g t h e i r name from t h e i r varietal p a r e n t , but a major excep-

tion i s p e o p l e , ebg.,

"~eter", "Aunt S a l l y " , Figure 1.

which

Any p a t h through the paradigmatic o r g a n i z a t i o n of knowledge

f o l l o w s only arcs having t h e same d i x e c t i o n a l i t y (termed a p a r a d i g m a t i c

path)

contains

at

most

one

instance arc.
from

Traversing about

this

arc

represents a concepts to

cognitive thinking

transition about

thinking

categorical from thinking to the

p a r t i c u l a r c o n c e p t s , e .g

.,

about man t o t h i n k i n g about Abraham L i n c o l n , o r from

b l u e o f your car.

blueness

Rumelhart variety

st

al.

(1972)

use

an

ISA r e l a t i d n that cover8 both

and

and

gnstance,

e . 8 ~ ~ ISA(Luigi s ,tavern)

ISA(tavern,establishment).

The p r e s e n t feeling is t h a t a d i s t i n c t i o n

does e x i s t ; hence t h e two r e l a t i o n s of t h e encyclopedia Typical. Concepts

A t h i r d condition of knowledge needs t o be

represented. For example,

have both u n i v e r s a l and occasional p r o p e r t i e s .

"birds e a t wormst' i s an occasional, n o t u n i v e r s a l , f a c t about birds

as

some never do, b u t bven those t h a t do a r e sometlmeg Iound eating f r u i t , fish, false

or even

not

e a t i n g , withbut t h e p r o p o s i t i o n being n e c e s s a r i l y

However, "birds have wings'' i s expected t o b e t r u e a t all times counter-example

f o r a l l b i r d s ; i t i s a pathological s i t u a t i o n i f a
foubd.

is

To

represept

t e arguments

of occas2onal predications, t h e

D i c a l (TYP) arc is used.

in Figure 2.

Thus the "blrd" in "bfrds eat worms" is as

I i s also possible t o use the typical r e l a t i o n t o a t t a c h t

occasional proper t i e s t o members of c a t e g o r i e s , i .e., Figure

t o instances.

In

2 i s s h o w the represenation of " W i l l i a m P r o m i r e makes f o r e i g n

p o l i c y statements" where this i s a statement rather than a

of

an

occasional

habit

record

specific a c t .

N p o s i t l o n i s taken on how o
development of the

noteworthy knowledge i e recognized as such i n the encyclopedia Manifestation.

(MAN)

The

f i n a l paradigmatic r e l a t i o n is manif es-t-ation of object constancy: A n

This corresponds to- t h e phenomenon

object

may undergo change i n space and t i m e , but i t i s still perceived

a s t h e same o b j e e t.

For -ampla,

W i l l i a m Proxmire b e f o r e apd a f t e r h i s

h a i r t r a n s p l a n t i s s t i l l W i l l i a m Proxmire. ticipate
eg, ..

Also

an

object

may

par-

in

many

different

a c t i o n s but s t i l l p r e s e r v e i t s i d e n t i t y ,

A l b e r t E i n s t e i n playing a v i o l i n and A l b e r t E i n s t e i n writing on a

blackboard remains Albert s i t u a t j . 9 ~ involves

Einstefn =

TI

the

system

each

diffetefii

a d i s t i n c t node.

To a node d e f i n e d by a n i n s t a n c e

are l i n k e d , by m a n i f e s t a t i o n arcs, nodes t h a t correspond t o

in
its

an

object

different

guises

Manifestations

of " W i l l i a m Proxmirett are different

shq~lni n F i g u r e 2.
from
that

b n i f e s t a t i o n s do n o t ustPally have names

of

their

p a r e a t instance; g rare exception tp this i s t h e Venus

Evening S t a r and the Morning Star which are b o t h

at

different

times o f t h e day.
M a a l f e s t a t i a n s of v a r i e t a l and typical concepts a r e also p o s s i b l e . The

latter

are

used

f o r p r o ~ e r t i e st h a t are t r u e o f t h e concept b u t

o n l y a t spme p o i n t or period of time,

for

example,

"vertebrates

are

horn",

F i g u r e 2.

For t y p i c a l a r c s t h i s n o t i o n is redundant a s t y p i c a l inde tgrminancy

embodies s p a t i a l and temporal

does

However

manifestation

have

use

with

the

t y p i c a l a r c in representirrg ~ o r e f e r e n c e .
to

Suppose i t can happen that a person can t r i p c a u s i n g him The

Uperaon" i n

be

hurt. but i t

the

encoded event is a t y p i c a l i s e d "person", Figure

must b e t h e same person t h a t t r i p s t h a t i s h u r t .

shows

the

u s e of m a n i f e s t a t i o n relations t o i n d i c a t e t h i s i d e n t i t y . said later about the formal representation

More w i l l b e

of t h e c a u s a l r e l a t i o n

indicated in Figure 2.

If only t y p i c a l a r c s were u s e d , t h e i n t e r p r e ~ a to

tian would b,e t h a t anyone t r i p p i n g c o u l d cause literally anyone

hurt.

be

Multiple

manifestations

can

also

be

used

with

variety i f

c o r e f e r e n c e n e e d s t o b e marked.

[This n e x t paragraph is almost c e r t a i n not t o make sense u n t i l the

r e a d e r has completed

as

far

as,

and

including,

the

section

II

Tn-

h e r i t a n c e " , and so he may clloose t o leave i t a n d r e t u r n l a t e r Other

systems,

Quillian

(1969),

Rumelhart

et

a l . ( 1 9 7 2 1 , and
object constancy

Schank (1975a), d o n o t u s e m a n i f e s t a t i o n b u t c a p t u r e
by

having

one

and

the

same

node

for

a p a r t i c i p a n t in a l l of i t s

propositions.

T h i s is a v i a b l e alternative.

Nevertheless, i n f o r m a t i o n
has

on t h e r e l a t i v e s t a n d i n g of t h e a p p e a r a n c e s o f the p a r t i c i p a n t
be

to

representable-

I f a s i n g l e node were u s e d i n the e n c y c l o p e d i a , t h e

nodes

d i f f e r e n t i a t i o n c o u l d be made on the m o d a l i t y tions.

of

rile

proposi-

This r o u t e was n o t t a k e n a s i t i s more c o n v e n i e n t , f o r example,

the

to

let

nature

of

the

inherita'nce

b e determined completely i n

p a r a d i g m a t i c o r g a n i z a t i o n , r a t h e r than i n a rnix t u r e of p a r a d i g m a t i c and

discursive structures.

For even without

manifestation,

the

varietal

s t r u c t u r e w i l l require the process of i n h e r i t a d c e .

Of

the

four

arcs,

variety,

typical,

and m a n i f e s t a t i o n can b e

i t e r a t e d ; instance c a n n o t rangements

Figure

2- and 3

illustrate

iterative

ar-

of

variety

and m a n i f e s t a t i o n -

That t y p i c a l a l s o has t h i s

p r o p e r t y i s s e e n from cofisidering t h a t

"While

dreaming,

some

people

talk

o r sleep-walk".

None o t these p r o p o s i t i o n s are u n i v e r s a l l y true, Figure 2 contains t h i e paradigms

of

but o n l y o f a r b i t r a r y people.
above examples present only

situation.

The

e n t i t i e s . but e v e n t s and

a t t r i b u t e s also exhibit this kind of o r g a n i z a t i o n . I f p a r a d i g m a t i c structure i s a loopless directed graph t h e n will

be

there
Can
with

origin

nodes,

that

is,

nodes

w i t h o u t e n t e r i n g arcs.

a n y t h i n g b e s a i d about t h e number o r k i n d s of c o n c e p t s associated origin nodes?

It

i s speculated

that

e n t i t i e s can b e d i v i d e d i n t o Possible domains

doma-ins of b e i n g each of which has i t s own paradigm. are t h i n g , s o u l , role, t i m e , e t c .

Thus t o r e p r e s e n t Ford a s P r e s i d e n t
be

of t h e USA t h e s t r u c t u r e i n Figure 3 would

used.

Figure

3 also

shows

how

the

totality

of

John brown ( J B ) and h i s f r a g m e n t s , a s i~

the

" ~ o h nBrown's body 1 ies mouLdering i n

grave

but

his

soul

goes

'marching on" can be represented

Figure 3 Domains of being

To

date

scholars have only studied e n t i t y paradigms i n d e t a i l *

taken

Little investigation of attribute or event paradigms has V

place.

It

i s hard

to i n t u i t i v e l y discern the hierarchical ordering of these

concepts, i . e
etc.,

., t o know
color

which

concepts imply
of

others.

Red,

yellow,

are obviously v a r i e t i e s

c o l o r , but does having mass imply

having color?--but many gases and g l a s s have mass

but

are colorlesso

Or
etc

does having

imply having mass?--but red l i g h t , blue jokes,

Or are they q u i t e independent attributes that happen

to have a

large

intersection

in

their

domain of a p p l i c a b i l i t y ?

fiese are a l l

open question8 i n the taxonomy of attributes.

The

event paradigm

is

a l s o open t o much s p e c u l a t i o n .
Syntagmatic Relation8

(APL)

Syntagmatic r e l a t i o n s connect nodes from d i f f e r e n t paradigms ( w i t h one exception).

Relations of

participation,

similar t d Fillmore s
A

(1969) case r e l a t i o n s , connect e n t i t i e s and

mlicatioa

events.

relation

of

l i n k $ a t t r i b u t e s *to e v e n t s o r t o e n t i t i e s rn

A relaA syntag-

t i o n of part-whole

(P-W) c~nnectsa unity t o i t s

component^.

m a t i c a l l y related s t r u c t u r e is termed a prpposlltion.

agent

Four r e l a t i o n s of p a r t i c i p a t i o n are d i s t i n g u i s h e d : instrumental

(INS),

objective (ORJ), and experiencer (EXP)

(AGT),

The role

c h a r a c t e r i z e d by each i s derived from dichotomies animate/inanimate and

causal/non-carnal, a s given in Table 1 (Fillmore, 1969).

Animate
Causal
Non-causal

Inan h a t e
INSTKWE~AL

AGENT
EXPER IENC ER

OBJECTIVE

Table 1 R e l a t i o n s o f Participation

Thus

"Angry B i l l

ferociously

h i t Fred with t h e handle of an axe" Is

diagrammed a s i n Figure 4.

'
Figure 4 The

<axe>

Syn tagmatic organization

set

of case relatkbns d o e s n o t include locative and temporal

relations.

Sentence adverb i a l s (Chobsky, 1965 ) are not part of SYntagbut

matic strucrare,
represented

of

the contextual

structure, which

fs here

on modality nodes.

Bound adverbial8 are part of syntagand

matic structure, e.g.,

"ferociously" above,

are related

to

the

event node by a relation of application.

A

part-whole relation i s used in Figure 4 to show the relation of

"handle" t o "axe".
tions
in

This relation d i f f e r s from other eyntagmatic

rela-

that i t connects nodes of the same type, e m g e ,two entities. considered

a

A case can be made for t h i s relation to b e

paradigmatic

relation; for

the

present

it

has been put i n with the syntagmatic

of

mainly because i t i s n o t used by the process of inheritance,

more later.

which

Discursive RelatLons Propositions do not occur in isolation.

cognition in
They are t i e d together i n

a number o f

ways.

The spatial, temporal, and causal

connections are characterized by$iclcursive

ares.
it

Intuitively

t h ~ e

are relations between whole propositions and

is t o f a i l t o capture
nodes.

t h i s feeling i f , say, a cause relation d i r e c t l y link8 t w o event

Modality

nodes are uaed

to

represent situations in Which the whole

proposition l a involved.

Though schematically linked to an event node;

Discursive John

conceptually the modality belonse t o the whole propoeition.

arca relate the modalities o f propositions.

Thus "Mary s l a p p e d

because he chased her" is represented as i n Figure 5.

Figure 5 Discursive organization

-1 8-

The one causal relation, cswe, a d m i t s of no finer distinction.

Others

(Schank,

1975a;

Halllday &

Haaan,

1975) distinguish three kinds ~f

causation: reason, result, and purpose.

the

The single cause relation of

Purpose (or enabling)
a

encyclopedia models the f i r s t tw d i r e c r l y .

causation i s seen as separable into cauae together wf th

desire

for
The
of

the

consequent.

For example, a cup may fell causing i t t o break.

f a l l could be accidental or i t c o u l d be deliberate with the purpose

breaking

the cup

T e same causal relation e x i s t s between the actions h

in both cases, but the analysis of the putposive s i t u a t i o n w i l l in-lve "desire".

Time

arcs do permit
(SML)

subdivision.

proposition

may

be

simultaneous

with another proposition:

"Fred washed the car

of
proposi-

while John chased Mary", Figure 50 A sequenttal rn iering

tions
is

also found, characterized by a sequence (SEQ) relation.

The

suggestions made here for the organization of space are only a

working

set

for which l i t t l e justification can be offered:

location (LOC)--a

neutral statement of position, contact--in physical contact, and -* near

#

far

-9

above

9 -

below

left -*

etc.,

which are self-explanatoryo

washed

Figure 5 represents t h e location of "Fred

tt

the

car" a s being

garage".

Since t h i s work was completed Sondheimer (1977) has propaeed

an analysis of space and t h e .

The - Me t a l i n g u a l
Speech

Relation
do

acts

not

make

use

only
A

of
most

forms

having p h y s i c a l
aspect

reference,e.g.,

table,

John,

blue.

important

of

language behavior is abstraction-

Human social, scientific , and i n t e l -

lectual

development

is dependent on t h e a b i l i t y t o create and control

paragraph

a b s t r a c t concepts- A quick appraisal of t h i 9

reveals
A
system

many

such

concepts:

language,

behavior,

bocial,

etc

that

s e r i o u s l y hopes t o approach human c a p a b i l i t i e s must have a

ing a b i l i t y .

correspond-

One part of modeling abstraction 5 s representation; but what is t o

be

represen tedg

AbsOrsc t i o n involves knowing a s i t u a t i o n i n which the

the

nbst rac t term a p p l i e s and replacing the sltu~tionald e s c r i p t i o n by

abstract

term.

An
19

example

is "tragedy"a

The scene to which i t i s

that:

applfcable is, say,

death".

Someone does is

good

act

results i n

his

This

definiens

encoded

in r i g u r e 6.

"Tragedy" names a

s i n g l e node.

Figure 6

Metalingual organization

The general

propositions

of

the

definiens are
the

conjoined
by

using a

modality node linked to the modalYties of

whole

propositions

part-

relations.

In general

there nay

be any number o f levels of

the

modalities related by part-whole.

T complete the association of o

a

definiendum w i t h i t s deftniens,

metallngual

(PEL) arc l i n k s the

Figure

former t o the appropriate modality in the l a t t e r ,

6.

If

any

situation matches the d e f i n i e n s , then the abstract term is appropriate.

The process of matching will be discussed later. The definiendum

can

also be any concept, the choice is fdiosyn-

crattc;

there i s no reason why t h i s device cannot be used

with ap-

parently non-abstract concepts, for example, a dog could be "man's best iriend" for sowk, in contrast with a non-abstrabt d e f i n i t i o n of "canine

animal"

Wn-abstract

d e f i n i t i o n s have the form "genus-specificata".

In the encyclapedia, t h e r e p r e s e n t a t i o n i s made up from a node by

related

v a r i e t y t o t h e genus (animal), t o which are a t t a c h e d t h e p r o p e r t i e s

i n t h e s p e c i f i c a t a (canine)

Rumelhart and Ortony (1976) use a r e l a t i o n similar to metalingual,

ISWHEN, b u t d o n o t show how p a r t i c i p a n t s are equated f n the definiendum
and d e f i n i e n s , nor t h e p r o c e s s e s t h a t use

such d e f i n i t i o n s .
Some p r o p o s i t i o n s

The m e t a l i n g u a l a r c is used i n a n o t h e r c o n t e x t * c o n t a i n embedded p r o p o s i t i o n s . For uniformity

to it

i desirable e
nodes.

to

restrict

participation

in

propositions

entity

Thus the

matrix p r o p o s i t i o n has a n strumental

unnamed

participant

in

objective

or

in-

role

and

t h i s nsde is d e f i n e d by a m e t a l i n g u a l arc t o the For example

modality of t h e contained p r o p o s i t i o n o

(1)

Peter b e l i e v e s Fred chased t h e c a t *

is r e p r e s e n t e d a s i n F i g u r e 7.

<Fred>

Figure 7

Embedding p r o p o s i t i o n s

.Status Relations

Knowledge
person.

in

an

encyclopedia

is

a model of t h e b e l i e f s o f one

N e v e r t h e l e s s t h e knowledge i s n o t a l l of t h e same s t a t u s .

In

a d d i t i o n t o c o n t a i n i n g the p e r s o n s b e l i e f s , i t ~ n c l u d e srepresentcation

of

beliefs

a b o u t h i s own d e s i r e s and of h i s b c l i e f s a b o u t t h e beliefs

and d e s i r e s of o t h e r s . control, and

His personal
his personal

beliefs

and

desires

interpret,

direct

activities.

The knpwledge about

o t h e r s 1 s t h e b a s i s for i n t e r a c t i n g and communicating w i t h example, a

them.

For

c o n v e r s a t i o n w i t h a child about the s t r u c t u r e of m a t t e r i s different

quite d i f f e r e n t from one with a n u c l e a r p h y s i c l e t because o f

c o n c e p t i o n s a b o u t their levels o f knowledge and hence what can b e taken

for

granted.

One h a s knowledge about i n d i v i d u a l s , e-g.,

your b r o t h e r ,
sports

Nelson R o c k e f e l l e r , e t c

., and about g r o u p s ,
be
made for between example,

e.g.,

politicians,

wr l t e r s , Russlans , e t c .
A d i s t i - n c t io n can s u b c o n s c i o u s and conscious the knowledge
of

knowledge.

Tbe former i s ,

language

underlying its u s e o r (2).

(2)

The Sun C i r c l e s t h e Farth.

Conscious knowledge is l e a r n t o r communicated knowledge, e .g -, what one

h a s been t a u g h t a b o u t t h e s o l a r system.

here is no r e a s o n f o r the two

krnds

of

knowledge

t o b e i n accord r e g a r d i n g t h e same e n t i t i e s

One

h a s l e a r n e d , f o r example, t h a t t h e F a r t h c i r c l e s t h e Sun.

Subconscious b e l i e f s of self are Subconscious

unmarked

in

the

encyclopedia.

beliefs of a n o t h e r are i n d i c a t e d by a believe arc between

and

a node r e p r e s e n t i n g t h e believer
network

modality

node

aovering

the

r e p r e s e n t a i o n of the c o n t e n t o f the b e l i e f s .

The subconscious

belief of ( 2 ) by "people" is g i v e n i n F i g u r e 8.

Figure 8

Knowledge s t a t u s

Conscious beliefs are represented as propositions embedded within

an event "beLieve".
i n Figure 7.
An example i s g i v e n i n (1) w i t h i t s r e p r e s e n t a t i o n

It is not o n l y propositions t h a t
simple

have

belief

status,

but

also

concepts,

e .g

.,

ghosts.

To accomodate t h i s i n f o r m a t i o n , the
P r e v i o u s l y only proposf t i o n s

placing of modality nodes i s generalized.

were a s s o c i a t e d
modality.

with

modalities;

now

any

node

can

have

its

own

On this modality i n f o r m a t i o n about a concept's e x i s t e n c e and

belief s t a t u s can b e r e p r e s e n t e d , as i n Figure 8 f o r "Fred Smith".

I t i s u n l i k e l y t h a t e a c h node o r p r o p o s i t i o n is immediately linked

to

its

believer.

Using

part-whole

relations

and

m o d a l i t y nodes, Figure
8

domains o b b e l i e f , which may i n t e r s e c t , c a n b e c r e a t e d a s i n f o r "Hugo"

.
(1975) p a r t i t i o n s s e m a n t i c networks t o d e l i m i t domains of

Hendrix

b e l i e f ; here t h e same e f f e c t structures.

i s gained

through

the

use

of

modality

The
exist.

desires

of

people

are

situations

t h a t t h e y would l l k e t o

The c o n t e n t of t h e s e g o a l s can b e
(complexes

represented

by

modality

covering

o f ) p r o p o s i t i o n s o r single c o n c e p t s , e - g . ,

peaceto of a

I f t h e goals a r e s u b c o n s c i o u s , a d e s i r e r e l a t i o n l i n k s t h e

desirer part

the

modal t y .

For

consczous

states,
an the

the event

modality

is
In

metaliagually defined object v e of behavior, these goals provide

"desire".

modeling

situations t h a t other behaviotal

a c t i o n s a r e i n t e n d e d t o c o n t r i b u t e towards achieving

Negation Negation

is

property

t h a t i s marked on 4 m o d a l i t y . a propos:itional

The most

common s i t e for n e g a t i v e marking i s

modality-

Thus

F i g ~ e c o n t a i n s t h e p r o p o s i t i o n "I d o n o t l i k e tomatoes" 9

\like

(tomatoes)

Figure 9 Negation

When

some other constituent o f a sentence i s negated, say using strong "John did

stress, t h i s i s marked on a corresponding modality, so

not

hit Mary" i s encoded as i n Figure 9 .

It

is

not anticipated, that negation is a common feature i n knownegative

fact

ledge, f o r "A person sometimes learns a t'radicts

when

it

con-

something

that

might b e inferred by mistake or that i s true

facts

f o r a similar c o n c e p t

But, most negative

are never

learned''

(Collins & Q u i l l i a n , 1972, p

. 319).
Inheritance

node

w i l l i n h e r i t properties from nodes h i g h e r in i t s paxadig-

matic path* Quillian (1969)

purpose.

used

superset

relations

for

the

same

In Figure 2, B i n h e r i t s the properties of A, C those of B, D

those of E, and E those of D .

Inherihance i s t r a n s i t i v e ,

thus

in-

herits

properties

from A,

B,

C,

and D .

This permits parsimonious

representation of properties: ancestor

property

need

only appear

at

the

of

concepts having

t h e property.

I n h e r i t a n c e is i n h i b i t e d For
it

o n l y i f t h e i n h e r i t a b l e p r o p e r t y i s c o n t r a d i c t e d on a lower node. example,

although the p r o p e r t y "fly" may be associated with "bird",

"penguin" by

is prevented from being i n h e r i t e d by

"penguin n o t f l y " .

having

explicitly

The g e n e r a l i t y of i n h e r i t a n c e depends on t h e form of r e p r e s e n t a t i o n o f t h e p r o p e r t y a t t h e a n c e s t o r node. s a l l y t r u e a t a l l t i m e s , e.g., P r o p e r t i e s t h a t are univerdirectly

If a t

b i r d s have wings, are a t t a c h e d

t o a v a r i e t a l node and a r e o b l i g a t o r i l y t r u e o f a l l descendents. any

time

bird

without

wings were r e p o r t e d , i t would b e cause for

f u r t h e r explanation.

Some o t h e r p r o p e r t i e s are always t r u e b u t o n l y a t

people eat, whose r e p r e s e n t a t i o n i n v o l v e s t h e
It i s n o t odd that a person can

i n t e r m i t t e n t times, e.g., manifestation r e l a t i o n . eating,

be

seen n o t

but

i f you watched long enough, i t would b e f u l l y expected t o Finally there a r e occasional proper-

obeerve t h i s b e h a v i o r sometime

ties

that

make use of t h e t y p i c a l arc i n their r e p r e s e n t a t i o n .

being merely noteworthy

These

proper t i e s are not u n i v e r s a l ,

recollections

about

a concept, e.g.,

The French a r e rude.

It would w e l l be p o s s i b l e

t o have a complete h i s t o r y of an example of the concept and n o t w i t n e s s t h e p r o p e r t y w i t h o u t being d i s t u r b e d by its absence.

E p i s o d i c andL Sys ternic Memory Tulving

(I

972 )

d i s t i n g u i s l l e s e p i s o d i c from s e m a n t i c memory.

The

former " r e c e i v e s and s t o r e s i n f o r m a t i o n aboht t e m p o r a l l y d a t e d episodes o r e v e n t s , and t e m p o r a l - s p a t i a l r e l a t i o n s between

events"

(p.

385).

T H '~l a t t e r

is

knowledge a p e r s o n posse s s e s a b o u t words and o t h e r v e r b a l symbols, t l ~ e a rmeaning and r e f e r e n t s , about r e l a t i o n s among them, and about r u l e s , f o r m u l a s , and a l g o r i t h m s f o r t h e m a n i p u l a t i o n of these symbols, c o n c e p t s , and r e l a t i o n s . Sqhan t i c memory d o e s not r e g i s t e r p e r c e p t i b l e p r o p e r t i e s of ~ n p u t s ,but r a t h e r cogni t l v e r e f e r e f i t s of i n p b t s i g n a l s . ( p . 386) Abelson (1975 ) d i s t i n g u i s h e s e p i s o d i c from Woods propositional memory,
and

( 1 9 7 5 ) c o n t r a s t s i n t e n s i o n s w i t h e x t e n s i o n s a l o n g similar l m e s . Hays
(1978),

The term I p r e f e r , f o l l o w i n g semantic, propositional

The

is

systemic

rather

than

,or

i ntensional

l o c a l i z a t i o n i n space and t i m e o f knowledge i s r e p r e s e n t e d . i n of proper propositions s u b p a r t of

Nanifesta-

t h e e n c y c l o p e d i a by s p a t i a l and temporal o r g a n i z a t i o n

using

the

appropriate

d&scursive relations.

e p i s o d i c memory i s c o n t a i n e d i n p a r a d i g m a t i c o r g a n i z a t i o n . tions

of i n s t a n c e s (remember t h e r e a r e a l s o m a n i f e s t a t i o n s of varietal
so
it

and t y p i c a l n o d e s ,

must

be

thus

stated)

represent

spatioConsetheir

Qemporally l o c a l i z e d

i n f o n n a t i o n about members of c a t e g o r i e s

quently' knowledge r e p r e s e n t e d on m a n i f e s t a t i o n s of i n s t a n c e s , o r manifestations,

is

i n e p i s o d i c memory.

This i s o n l y p a r t of episodic
For example, in

memory a s c a t e g o r i c a l knowledge can a l s o be p r e s e n t .

"Jung

changed our view of dreams", t h e reference is t o t h e c a t e g o r i c a l

notion of dreams, n o t to any s p e c i f i c ones. proposition to have

Nor i s i t s u f f i c i e n t for a

a
the

non-categorical p a r t i c i p a n t t o b e m episodic

memory f o r "Prior t o
serfs"
cnotdins

Revolution.

Russf an

peasants

were

feudal

catpgorical g a r t i c i pants, y e t i s e p i s o ~ i c s The total

decided through

extent of episodic memory i s y l t i m a t e l y

temporal

spatial

and

relatiion

of

discursive

o r g a n i z a t i ~ n , not

b y paradigmatic

structure.

Ouan tif i c a t i o n

para dig ma ti.^

arcs have the c a p a b i l i t y o f capturing the essence of

To i l l u s t r a t e the facility, cons:ider Figure

q u a n t i f i c a t i o n , .including scope

(3)

and ( 4 )

whlch are equivalettt t o the Iormulae ( 5 ) and ( 6 )

10 encodes ( 3 ) and ( 4 ) .

(3) (4) (5) (6)

There Every a x Vy Vydx

i s a book chorlgter [ (book(x) [ (song(x)

t h a t :js read by every scholar, knows a song. & s c h o l a r t y ) ) 3 read (y,x)]

& ohorister(y)) 3 know(y,x)

Figure 10

Quantification

Xf for

a given c . o r i s t e r in ( 4 ) it i s necessary t o determine the song

b e . ,

he knows, i
present
as

t o evaluate the Skolem

function,

the

information

is

predication

of that individual and should be retrieved

using his name, say "George" and r rink t o m only w i t h t h i n e eyes" - i n e

Figure 10.

It

is

also

possible

t o g i v e d i s t i n o t representation t o unquanin

t l f i e d statements, such as ( 7 ) , a s

Figure 10.

(7) A person l i k e s candy.

Paradigmatic arcs are here achieving reoresentational power equiva l e n t t o the p a r t i t i o n i n g of networks by Hendrix (1975)-

-30-

The above i s a s y s t e m i c r e n d i t i o n o f
can

"all".

The

quantification

also

be

characterized

episodically

by every m a n i f e s t a t < o n of a
1975),

concept h a v i n g t h e p r o p e r t y .

I n t e r p r e t i n g "all" (Woods,

could

call

upon

e i t h e r systemic o r episodic f a c t s * either

A q u e s t i o n containing a

u n i v e r s a l q u a n t i f i e r may b e answered by
node

examining

varietal

(Are

all

moil-boxes blue?), or b = examining e v e r y m h i f e s t a t i o n y

(Do a11 m a i l - b o ~ e s stand a t street corners?).

It should be n o t e d t h a t "all" r e q u i r e s
that

the

predicacron

be

true only

at

some

time,

e.g.,

A l l p e o p l e d i e ; i t d o e s n o t require
Thus

c o n t i n u i t y i n time, e .g.,

A l l b i r d s have wings.

untiversal

quan-

t i f i c a t i o n i s a l s o true i f t h e p r e d i c a t i o n i s found for a m a n f f e s t a t i o n

o f t h e v a r i e t a l node,

os i s found for every i n s t a n c e of t h e concept.

P r o c e s s e s i n t h e Network

--

The

model

for knowledge described above is only p a r t o f a system

Thought is simulated by processing

t o model c o g n i t i v e behavior.
ledge.

know-

Different

a s p e c t s o f behavior c o r r e s p o n d t o d i f f e r e n t proces-

ses, but with o n e and the same e n c y c l o p e d i a common t o

all.

system

for d i s c o u r s e analysis r e q u i r e s processes t h a t u s e t h e e n c y c l o p e d i a t o

f i n d patterns of o r g a n i z a t i o n i n a d i s c o u r s e . describe

I t would b e p o s s i b l e

to

solely

the

requirements

of d i s c o u r s e analysis, but g r e a t e r

o v e r a l l insight i s g a r n e d through a p r e l i m i n a r y g e n e r a l e x a m i n a t i o n and

classification o f c o g n i t i v e

processes.

Once

this

is

accomplished,

discourse

analy4s

i s seen not t o be a unique process but a s composed

S i m u l a t i o n of many

of more b a s i c general ones.

asp'ec ts

of

cognitive

behavior

can

he

porfomed

by

complexes o f t h e s e g e n e r a l processes:

d i s c o u r s e a n a l y s i s i s just an6 s u c h complex

Processes can b e c l a s s i f i e d i n

various

ways:

functionally,

by

c o m p l e x i t y , o r b y the c l a s s of relation involved.

The

ftlnction

of some p r o c e s s e s i s e x t e r n a l ; t h e y d e a l w i t h i n p u t

and o u t p u t .

Some internal processes f i n d r e l a t i o n s between new

infor-

mation

and

knowledge

already i n t h e e n c y c l o p p d i a , o t h e r s i n v e s t i g a t e

t h e v a 1 i d i . t ~ new knowledge, e t c . of P r o c e s s e s a r e of two type oE complexity, either path-tracini

or

ptn. aetr*
tasks

The dichdtomy i s j u s t i f i e d by showing t h a t there a r e

that
can o n l y b e done b y p a t t e r n - m a t c h i n g *
This t o p i c i.s con-

sidered in d e t a i l l a t e r .
Of the infinite number of possible ordered s e t s of a r c s , o n l y some

d e f i n e s i g n i f i c a n t p a t h s in the network. of

An example of a r e l e v a n t

set

arcs

i s

the

arcs

of

a paradigmatic p a t h ; t h i s defines possible

inheritances. represented
by

Other s i g n i f i c a n t s e t s

are

causal

chains,

whi-ch

are

a s t r i n g of c a u s e arcs between modalities.

Th~s suggidentical

e s t s t h a t p r o c e s s e s t h a t use tile same

kind of

r e l a t i , o n s or

relations a r e significant.
A

functronal

classification

o f processes d o e s n q t g i v e a deeper However, classification

by

understanding

of3 c o g n i t i v e

processes.

complexity

and by kind o f arc i s r e v e a l i n g .

Path-tracing and pattern-

matching d i f f e r i n power. the kind of arc be found described

For the tormer, s u b p a t h s can b e

i n t h e subpath. according
as

defined

by

Henceforth processes i n the they are path-tracers

or

network w i l t

pattern-matchers.
Path-tracing -

Path-tracing

processes t r y t o establish paths between nodes along Q u i l l i a n ( 1 9 6 9 ) e s t a b l i s h e d t h i s methodology

arcs of t h e network.

for

semantic
path.

nets.

A p a r t i c u l a r l y common type of path i s t h e paradigmatic

In rigure 3 there i s

a
but

paradigmatic

path

betweqn

"Ford

(as

President)"

and

"thing",

not

between

"rock"

and "soul". events,

The
and

d e f i n i t i o n of a paradigmatic p a t h i s v a l i d f o r attributes.

entities,

Any paradigmatic path i n the n e t w o r k w i l l conform to the structure

sl~own in Flgure 11 where

i n d i c a t e s any number o f occurrences includ-

i n g zero of t h e marked r e l a t i o n .

F i g u r e 11

Paradigmatic paths

The

structure

follows

directly

from

the

iterativity

of

variety,

manifestation, and typical arcs and t h e i r

tions. Strings

posbible

relative

orien t a ~

o f arc l a b e l s r e p r e s e n t i n g p a t h s t h r o w h t h e tree a r e
are
sentences

o b v i o u s l y regular expressions, L e e , the strings

of

type

language* P a r a d i g m a t i c p a t h - t r a c i n g can t h u s b e c h a r a c t e r i z e d

by a f i n i t e s V a t e a u t o m a t o n ( H o p c r o f t & Ullman, 1969).

Any p r o c e s s t h a t c a n h e c h a r a c t e r i z e d by a f i n i t e s t a t e
i s f o r m a l l y termed

automaton

a p a t h - t r a c i n g p r w e s s i n the s y s t e m *

One

such

p r o c e s s i s testing t h e a p p l i c a b i l i t y of a n attribute t o

an e n t i t y , e .go, w h e t h e r " f r e s h fish" o r "round when

smoke"

is

acceptable

the r e l a t i o n s h i p i s not e x p l i c i t l y i n the encyclopedia

Assuming instance

t h e named e n t r y p o i n t s t o t h e e n c y c l o p e d i a a r e a t v a r i - e t y

or

n o d e s , a n e n t i t y F1 ( e . g . ,

h o r s e ) can i n h e r i t p r o p e r t i e s from an e n t i t y
i s a p a t h between F 1 a n d T2 of t h e form

C2 (I)
)

(e.g.,

animal)

if

there

iiihR*, where

i n d i c a t e s a r e l a t i o n that i s t h e c o n v e r s e o f X

and

i n d ~ c a t e an optional arc.

Properties may b e a t t a c h e d t o E2 either

d i r e c t l y ar with typical a n d / o r manifestation arcs, i m e . , t h e p a t h from

C p t o t h e node F3 i n the
TYP*
MAN*.

representation of t h e property

has

the

form

MAN*.

Thus

t h e p a t h from E l t o Eg h a s t h e form

(m)WR* TYP*
is

A n a l o g o u s l y , an a t t r i b u t e A l can a p p l y t o an e n t l t y i f t h e r e
path

similar

to

an

attribute

t h a t :is encoded as a p p l y i n g t o the

entity.

Thus i f t h e r e i s a p a t h

(8) '<El > (ET) KW TYP* MAN*

G IIAN* TYP* VAK* E

( I S T ) [Al

t h e n A, can f e a s i b l y a p p l y t o E l .
ing

T h a t i s t o say, t h e e x p l i c i t
make it

encod-

of

"emotional

animal"

would

r e a s o n a b l e t o i n f e r "sad
patjis

horse".

The p a t h ( 8 ) is composed o f p a r a d i g m a t i c

linked

by

single a p p l i c a t i o n a r c . 3
languages

Each segment is a regular expression.

As t y p e

are

closed under concatenation (Hopcroft & Ullman, 1969,

theorem 3 . 8 ) , i t EollowS t h a t (8) is also a regular e x p r e s s i o n and that

a t t r i b u t e g p p l i c a b i l i t y t e s t i n 8 is a p a t h - t r a c i n g * process. P r o p o s i t i o n s i n a d i s c o u r s e should b e c o n s i s t e n t wit11 encyclopedic

knowledge.
encyclopedia
e.g.,

Consistency i s established by f i n d i n g a p r o p o s i t i o n i n
that

the

is a

generalization o f the discourse p r o p o s i t i o n ,

g i v e n t h e discourse p r o p o s i t i o n

(9)

Harv g o b b l e d t h e c a v i a r .

and f i n d i n g t h e g e n e r a l i z a t i o n

(10)
A

P e o p l e e a t food.
statement,
e .g., "fiarry munched

novel

the s p i d e r " , which is

variety
of

not

consistent w i t h (10) (assuming " s p i d e r "

would

is not a

"food"),

evoke a demand for f u r t h e r e x p l a n a t i o n , o r similar.

Consistency
In

judgment can b e f o r m u l a t e d a s a complex of path-tracing processes.

the network form of ( 9 ) , "Marv"
t i v e of "gobble1'.

i s the agent and

t8

caviar" is t h e objec-

Flgute 12 encodes (10).

Figure 1 2

C o n s i s t e n c y j udgment

The

words

in

t h e d i s c o u r s e p r o p o s i t i o n provide entry p o i n t s i n t o t h e
and

network of F i g u r e 12 t h r o u g h t h e d i c t i o n a r y
tions*

converse

name

rela-

From

"gobble",

node

1,

paths

along

paradigmatic arcs are

traversed tio locate nodes from which "gobble" c o u l d i n h e r i t properties,

e e g . , n o d e 2.

Kext from t h e e n t r i e s f o r "tlarv"

(A),

and

It

caviar", ( B ) ,
(among

analogous p a t h s a r e followed, reaching C other nodes).

From
C

and

D,

respectively

and TI arcs c o r r e s p o n d i n g t o the p a r t i c i p a t o r y

r e l a t i o n s of "Marv" a n d o f " c a v i a r " t o "gobble", i.e., t i v e , respectively, are followed.

agent and objec-

If all p a t h s i n t e r s e c t a t

single

node, e . g o , node 2, then the p r o p o s i t i o n c o n t a i n i n g t h e i n t e r s e c t i o n i s

the

general

proposition

sought.

Each p a t h from an entry p o i n t t o an

i n t e r s e c t i o n can b e c h a r a c t e r i z e d by a r e g u l a r e x p r e s s i o n .

There a r e

o n l y f o u r case relations, which s e t s a f i n i t e upper bound t o t h e number of paths to be followed.

Hence t h i s p r o c e s s is also a p a t h - t r a c i n e

proceqs.
L b c a t l n g e x i s t i n g knbwledge, p r o p o s i t i o n s

that

are

already

ex-

plicitly

i n t h e e n c y c l o p e d i a , i s e f f e c t i v e l y i d e n t i c a l to t h e c o n s i s t -

ency t e s t i n g process above, b u t w i t h downward p a r a d i g m a t i c p a t h s followed

instead

being

of

upward

ones.

Thus

given "Oswald assassinated

Kennedy" and t h e network of F i g u r e 13,

< Oswald >

ate

Oswald

'I

Kennedy

F i g u r e 13

F i n d i n g known p r o p o s i t i o n s

paths

can

be

traced

from node 1 t o node 2, from node A t o node B t o The common intersection

node 2 , and from node C t o node D t o node 2.

IS

i n the known p r o p o s i t i o n .

Pat tern-Ma t c h l n g
Pattern-atching is used i n processes where two c o n f i g u r a t i o n s of
me

n o d e s and a r c s must match. d e f i n e d terms

If

such

process

is

with

metalingually

discourse

conf l g u r a t i o n matches a m e t a l i n g u a l

d e f i n i t i o n , then t h e p a r t o f t h e d i s c o u r s e so matched may

be

replaced

by

t h e term.

Figure 14 c o n t a i n s representations of ( a ) "Fred a t e some

11

cake t h a t made him s i c k " and ( b ) t h e d e f i n i t i o n o f

poisonf':

11

Someone

ingests somethin8 t h a t makes him ill".

(a)

(b)

F i g u r e 14

Pattern-matching

I f t h e l a t t e r m a t c h e s t h e former, t h e n
situation.

II

p o i s o n " d e s c r i b e s the d i s c o u r s e

Earlies

a path-tracing

p r o c e s s was u s e d to e s t a b l i s h conThe same proc-

s i s t e n c y between a general and a s p e c i f i c p r o p o s i t i o n .

ess
tion.

can

b e used t o p a i r propositions or the d i s c o u r s e and t h e d e f i n i -

However, t h e r e i s an aspect of complexes path-tracing

of

propositions

that

prevents

from b e i n g a complete solution.

this

I f t h e complex
coreferentiality

c o n t a i n s c o r e f e r e n t i a l items, a s "poison" does,

must

be

examined;

if

i t were n o t e r r o r s c o u l d r e s u l t .

For example, Fred

c o n s i d e r a d i s c o u r s e c o n t a i n i n g "John's sick". Each proposition

eating

the

worms made

matches p h r t o f t h e d e f i n i t i o n of "poison",

but i t s h o u l d n o t be t a k e n a s an
tiality condition plevents

act

of

poisoning.

The

coreferen-

a match.

As, i n general, t h e r e c a n h e any

number o f c o r e f e r e n t i a l p a r t i c i p a n t s i n a complex o f
is

propositions,

it

not

possible

to

define

a r e g u l a r e x p r e s s i o n ta c h a r a c t e r i z e t h e

c o r e f e r e n t i a l i t y test. of an

This c a n b e shown by c o n s i d e r i n g

definition T h e r e is number
of

a b s t r a c t term that c o n t a i n s n c o r e f e r e n t i a l c o n c e p t s . any

i n g e n e r a l no bound on d s s t h e d e f i n i t i o n c a n contain
propositions.

If

a complex of d i s c o u r s e p r o p o s i t i o n s i s t o match t h e proposition

d e f i n i t t o n then there must f i r s t b e a unique c o r r e s p o n d tng

in

the

definition

for

each d i s c o u r s e p r o p o s i t i o n . But

This c a n b e done
over and beyond

using t h e p a t h - t r a c i n g p r o c e s s d e s c r i b e d above.
t h i s , the c o r e f e r e n c e c o n d i t i o n must be s a t i s t t e d

F o r each m a n i f e s t a -

tion

of

the

coreferential

c o n c e p t i n t h e d e f i n i t i o n t h e r e must b e a in

c o r r e s p o n d i n g m a n i f e s t a t i m o f o n e and always t h e same c o n c e p t discourse

the

Also

the syntagmatic r o l e of corresponding manifestations

be

i n their respective propositions m w t c o n d i t i o n i n v o l v e s pair-wise strings counting.

the

same*

The

acceptance

This i s e q u i v a l e n t t o a c c e p t i n g

o f t h e form anbn , which a r e n o t s e n t e n c e s i n a t y p e 3 l a n g u a g e

( H o p c r o f t & Ullman, compare complexes

1969).
of

This

demonstrates

that

processes

that

p r o p o s i t i o n s c o n t a i n i n g c o r e f eren t i a l i t e m s a r e

n o t , i n general, p a t h - t r a c ~ h g p r o c e s s e s .

-3 9-

A p r o c e s s charac t e r i r e d by a device more powerful

than

finite

?utomrrton i s f o r m a l l y d e s i g n h t e d a p a t t e r n - h a t c h i n g process.

Paraphrasing
another

discourse

usinp

metalingunlly
lfatalinaual

d e f ineri

terms
can

is be

sattern-matching
embedded.

process*

definitions

recursively

For

example, "buy1' may b e d e f i n e d i n terms of Recursion

3

"pive", which i n turd may b e defined i n t e n s of "have". not

is

a p r o p e r t y of regular languages, h e n c e t h i s p r o c e s s i s not

path-

tracing plocesse Matching discourse con i g u r a t i o n s

of
the

against
process

definitions,

called

abstraction,

is

an

extension by

that

substantiates

discourse p r o p o s i t i o n s
encyclopedia,

seeking
earlier*

g e n e r a l i zed

proposi t i o n s

in

the

discussed

The components o f a d e f i n i t i o n a r e

g e n e r a l i z e d p r o p o s i t i o n s and hence t h e s u b s t a n t i a t i o n process w i l l f i n d them i f t h e y c o r r e s p o n d t o p a r t o f t h e d i s c o u r s e . discourse Schematically, two

p r o p o s i t i o n s DP, and DP1, may match g e n e r a l i z e d p r o p o s i t i o n s

rP1 and CP2 an$ GP3, G$,

and CP5, r e s p e c t i v e l y , a s shown i n Figure 15.

MLD

DEFINITION

GENERALIZED PROPOSITIONS DISCOURSE PROPOSITIONS

DP1

DP2

Figure 15 A b s t r a c t i - o n

This

i s t h e normal o u t p u t when j u d g i n g c o n s i s t e n c y .

Propositions of a

d e f i n i t i o n are u n d e r a c o n j o i n i n g m o d a l i t y , t o

which the m e t a l i n g u a l

arc

points.

I f i t i s found t h a t Some a f t h e general p r o p o s i t i o n s a r e

these d e f i n i t i o n s

p a r t of d e f i n i t i o n s , L e a , CP2 and GP3 i n MLD, t h e n

are
*i.e.,

examined t o see i f a l l t h e c o n d i t i o n s f o r their use a r e s a t i s f i e d , c o r e f e r e n c e and c o n t e x t u a l (e.g., cause

arcs)

conditions.

For

example,

in llpoison", Figure 14, the c o r e f e r e n c e of t h e agent o f "eat"

is
satisfied, tKen
the

and t h e a p p l i c a n d of "ill". If a d e f i n i t i o n

p a r t o f t h e a i s c o u r s e matching t h e d e f i n i e n s c a n be p a r a p h r a s e d .

The

definitional

nets

so

far

presented

are

n o t adequate f o r

p a r a p h r a s i n g , b u t must b e augmented 40 i n c l u d e t h e r o l e s of e n t i t i e s of the definiens with respect t o manifestation the definiendum.

This i s d o n e

with

arcs.

network definition o f "buy"

16.

( i n "A buys t h i n g is
"A

from B f o r money1') is given i n F i g u r e

The

verbalization

g i v e s money t o B and B g i v e s t h i n g t o A".

Figure 16 R o l e correspondence

The

m a n i f e s t a t i o n arcs indicate the role correspondences between "buy"

and the d e f i n i n g s i t u a t i o n a s well

latter.

as

coreferentialities within

the

The case correspondences are essential information f o r the prot ess

of abstraction

and

for

i t s inverse, decomposition, which produces a

l e s s abstract description from a network containing a term t h a t

metalingual d e f i n i t i o n .

has

For example, given t h e sentence "John bought a

bicycle

from Jane", the definition o f Figure 16 enables the paraphrase

"Jane gave a b i c y c l e t o John generated.

and

John

gave

money

to

Jane"

to

be

"Money" was unexpressed i n t h e o r i g i n a l , b u t i s p r e s e n t i n

The process fills empty

the d e f i n i t i o n , and a p p e a r s i n t h e p a r a p h r a s e .

s l o t s by t h e a p p r o p r i a t e concept from
1)

the

definition,

in

this

case

money".

The a g e n t , e x p e r i e n c e r , and o b j e c t i v e s l o t s are f i l l e d i n t h e

s o u r c e s t a t e m e n t and are t r a n s f e r r e d t o t h e p a r a p h r a s e . Another a b s t r a c t term can p o i n t t o t h e same d e f i n i t i o n a l network,

say "sell" i n t h e case o f F i g u r e 16.

The n e t t h e n h a s all t h e informaterms as

<ion f o r p a r a p h r a s e s between t h e two a b s t r a c t

decomposition and a b s t r a c t i o n *

well

as

for

There

i s no p r o d u c t i v e relationship between t h e r o l e s o f t h e same

Case r e l a t i o n s

p a r t i c i p a n t a t d i f f e r e n t Levels o f a b s t r a c t i o n . sent

repre-

only the causal/animate perception o f p a r t i c i p a t i o n i n a n event.

participants

More detailed d e s c r i p t i o n s of t h e roles o f

can

only

be

given i n c o n t e x t .
"buytt,

For example,

If

money" is p e r c e i v e d as i n s t r u m e n t a l i n

but

at

t h e next l e v e l of decomposition, i t i s i n an o b j e c t i u e

r o l e i n "give".
The o u t p u t s o f both a b s t r a c t t o n and d e c o m p o s i t i o n are s t r u c t u r a l l y i n d i s t i n g u i s h a b l e from any o t h e r p r o p o s i t i o n i n
the

encyclopedia

and

t h e r e f o r e can again be subject t o e i t h e r o f t h e p r o c e s s e s . matching

As p a t t e r n -

is a r e c u r s i v e p r o c e s s t h i s a b i l i t y f o r o u t p u t o f t h e p r o c e s s

t o b e accepted a s input i s e s s e n t i a l .

T h e d i s t i n c t i o n between path-tracing and pattern-matching

ses

proces-

may

be

psychologically

significant use

Inhelder and Piaget (1 964)

logical equations such
as

f i n d t h a t prepuberty c h i l d r e n cannot
-(A A B ) 3 -A
V

-Be

The
a

e q u a t i o n s i n v o l v e c g r e f e r e n c e and hence t h e i r pattern-matching process

application speculated

requires

It

could

be

t h a t t h i s more powerful p r o c e s s o n l y a p p e a r & a t m a t u r a t i o n .

DISCOURSE ANALYS IS

The of - S t r u c t u r e - Coherent D i s c o u r s e

In

this

section t h e hypothesrs c o n c e r n i n g t h e k i n d s o f organizaA

t i o n present i n c o h e r e n t d i s c o u r s e i s o u t l i n e d . can b e found elsewhere ( P h i l l i p s , 1975).

fuller

description

The r o l e of t h e e n c y c l o p e d i a

i n d i s c o u r s e i s t h e n exemplified.
A d i s c o u r s e i s judged c o h e r e n t if i t s c o n s t i t u e n t p r o p o s i t i o n s a r e

connected

Various t y p e s of c o h e s i v e l i n k s a r e observed i n temporal, c a u s a l , and t h e m a t i c .

discourse :

anaphoric , s p a t i a l ,

I w i l l formally terms of these

d e s c r i b e t h e s t r u c t u r e of a well-formed d i s c o u r s e connect v e s L Anaphora

A

in

discourse of the

has

reference

to

objects. kinds of

Coherence i s g i v e n by anaphora can

be

repetition

reference.

Two

distinguished.

The first i s marked by t h e p r e s e n c e o f a proform ( o r b y

the

repetition

of

fonn):

[It i s u s u a l f o r c o h e r e n t d.lscourse t o ex-

h i b i t s e v e r a l kinds of cohe'sive l i n k s

Thus

the

examples

invariably

c o n t a i n more than t h e one s p e c i f i c a l l y being i l l u s t r a t e d . ] (10) Antecedents

Henry t r a v e l s t o o much.

H is g e t t i n g a f o r e i g n a c c e n t . e

may

b m

nominal,

v e r b a l , o r clausal

The second k i n d of

anaphore has a dependent t h a t i s a n a b s t r a c t term f o r

the

antecedent.

for example

(11) John put the c a r into I1 r e v e r s e " i n s t e a d of "drive". The mistake c o s t him $300 t o r e p a i r .
"Plistake"
i n ( 1 1 ) i s an a b s t r a c t c h a r a c t e r i z a t i o n of t h e gear s e l e c t i o n

expressed
issue.

i n t h e f i r s t sentence.

Nagao anti T s u j i i (1976) address this

A conventi-onal way t o l a b e l t h e r e c u r r i n g c h a r a c t e r s i n

discourse

is

as

tf

dramatis personaetf.

Ifowever, c o h e s i o n c a n r e s u l t n o t only from

m u l t i p l e a p p e a r a n c e s of p e o p l e ( l o ) , b u t o f a n y c o n c e p t , as i n ( 1 1 ) . S p a t i a l , Temporal, and C a u s a l Cohesion Space, time, and cause give c o h e r e n c e sentence(12)

to

set

of

clauses

or

The King was i n t h e c o u n t i n g house, c o u n t i n g o u t h i s money. The Queen was i n the p a r l o r , e a t i n g bread and honey. in (12). are
set

The
11

actions

in

different

rooms, b u t o f .the same

palace".

(13)

A f t e r Richard t a l k e d t o t h e r e p o r t e r , h e went t o l u n c h .

The temporal sequence o f e v e n t s i n (13) i s e x p r e s s e d by " a f t e r " (14)

John e a t s g a r l i c .

Martha a v o i d s him*

To

nun-aficionados

garlic

is

known anly f.or its aroma, d e t e c t i o n of

which causes evasive ' ~ cion. t

Cat~sc, i l l u s t r a t e d i n ( 1 4 ) , is an important

discourse

connective

(Schank,

1975b).

Ttle

importance

is
to

perhaps e t h n o c e n t r i c ; i n o t h e r
be

c u l t u r e s d i f f e r e n t p o s i t i o n s may have
t e l e o g i c a l world v i e w ( k h i t e , 1975).
The

taken,

for

mcain~le. n

causal

chaYn of p r o p o s i t i o n s i n d i s c o u r s e i s termed i t s p l o t

structure. Thematicity Coherent discourse i s expected t o have a theme, t o have

a

topic*

tor example

(15) DF drowned today i n IlB resevoir a f t e r restuing his son

who had f a l l e n i n t o t h e water w h i l e on a f i s h i n g t r i p .
is
tt

news s t o r y 6rom t h e New York Times with a theme t h a t I w i l l c a l l

tragedy".

I n t h i s s e c t i o n I wish t o j u s t i f y t h e claim t h a t a thematic

different
exarnrles
and

s t r u c t u r e c o n d i t i o n is u n i v e r s a l by examining

analyses o f general discourse f o r evidertce.

The notion of
clar'ity.
discourse
It

theme

is

much
stated

used to

but n o t o f t e n d e f i n e d with
be "The subject
of

is

variously

. . a topic1' (Oxford
view
towards admiration, betrayal,

English Dictionary); "the playwright's

point

of

his m a t e r i a l 1 ' (Elabley, 1972, p . 1 4 ) , e t c .

In

Abelson (1973) t h e r e i s a list of themes (admitted t o b e n e i t h e r

nor love, exhaustive) : alienation, devotion, victory, appreciation,

fixed

cooperation, mutual

dominance,

rebellion,

antagonism,

opposition,

and

conflict

Occasionally one f i n d s overt about

comment p n t h e l a c k of a t h e g e :

"The t h i n g t h a t p u z z l e d m most e

The of Beau -Last Remake -- Geste was

Took, "cinema"
p -

i t s lack o f a p o i n t of view" (Barry

Punch, December 7, 1977)

E q u a l l y Infrequently one c a n
a

f i n d a s u c c i n c t a m p l i f i c a t i o n of t h e s t r u c t u r e of
other hand,

thehe:

"On

the

t h e s u s p e n s i o n of d i s b e l i e f i s what t h r i l l e r s a r e a b o u t . I t

( S h e r i d a n Horlqy,

heatre re" 8 - Punch, November 19, 1975).

explicitly quite usual stated to in discourse. a complete as

A theme may b e writing

it

In

technical definition,

is

express

d e f iniendum-def i n i e n s : Kuhn (1962) d e f i n e s "paradigm" ment"

an

"achieve-

t h a t i s " s u f f i c i e n t l y u n p r e c e d e n t e d t a a t t r a c t an e n d u r i n g group [and] suf-

o f adherents away from competing s c i e n t i f i c a c t i v i t y ficiently

open

ended t o l e a v e a l l s o r t s o f problems f o r t h e r e d e f i n e d
Much o f t h e rest o f t h e book

group o f p r a c t i o n e r s t o r e s o l v e " ( p . 1 0 ) .

t h e n d i s c u s s e s paradigms a s models f o r sclen tific r e v o l u t & o n s . I f a d i s c o u r s e h a s an i m p l i c i t theme, i t h a s t o b e i n f e r r e d by t h e reader. reader

An a u t h o r , t h e r e f o r e , s h o u l d u s e themes t h a t a r e k n o w t o
One

the

.
the

possibilty

is

that

there

is

o n l y a f i n i t e number o f
I

themes.
that

But l a c k i n g e v i d e n c e for this

positfon,

will

hypothesise

number

of

themes may be u n l i m i t e d in the same way t h a t t h e

A reader may n o t know a word t h a t i s

v o c a b u l a r y o f a l a n g u a g e is open.

used b y an a u t h o r ; i n a similar f a s h i o n h e may n o t r e c o g n i z e a theme.

There are studies that indicate the existence o f

i n language. In folk-tales,
the

abstract

themes

Propp (1968) a n a l y s e s a r e n d e r ' s expectanPropp starts b y comparing the

cies

about

s t r u c t u r e o f t h e tale.

f o l l o w i n 8 e v e n t s from d i f f e r e n t tales:

1.
a

A t s a r g i v e s an e a g l e t o a h e r o . The e a g l e c a r r i e s t h e h e r o away t o a n o t h e r kingdom*

2.

A princess g i v e s Ivan a r i n g Young men appear'tilg from otrt o f the ring c a r r y Ivan i n t o a n o t h e r kingdom.

Propp i n f e r s t h a t "a t a l e o f t e n a t t r i b u t e s i d e n t i c a l a c t i o n s t o v a r h u s personages.

This

makes p o s s i b l e t h e s t u d y o f the f o l k t a l e according

20).
Falk
tales

t d t h e f u n c t i o n s o f the d r a m a t i s personae" ( p a

are

analysed

in

terms

of Eunct'ions.

The above examples a r e d e s c r i b e d as

It

con'taining two f u n c t i o n s :

"Aquisi t i o n of a m a g i c a l agent" and

Trans-

f e r e n c e t o a d e s i g n a t e d place".
(22)

An example o f Propp s a n a l y s i s i s

ACTION A t s a r , three daughters. The d a u g h t e r s go walking, o v e r s t a y i n t h e garden. A dragon kidnaps them A call f o r a i d . Quest o f t h e t h r e e h e r o e s .

FUNCTION I N I T I A L SITUATION A SENTATION B

VIOLATION VILLAINY
PlEDIATION CONSENT' T O COUNTERACTION

DEPARTURF
Thxee b a t t l e s w i t h t h e

dragon.
Rescue o f t h e mazdens. Return. WedcFing.

STRUGGLE
VICTORY IN IT IAL MISPORTUNE LIQUIDATED

RETURN WEDDING
( p * 128)

Functions

correspond

to

m e t a l i , n g u a l l y d e f i n e d c o n c e p t s o f t h e ency-

clopedia.

Propp s h o w that t h i . s g e n r e 01 d i s c o u r s e can b e a n a l y s e d

as

an o r d e r e d string of a b s t r a c t c o n c e p t s .

-4 8-

Linde

(1974) finds t h a t t h e r e i s a p r e s c r i b e d pattern i n v e r b a l

Only two d i s c o u r s e s t r a t e g i e s are used
by

d e s c r i p t i o n s of a p a r t m e n t s . her

subjects

to

e x p r e s s the s p a t i a l s t r u c t u r e s , and o f t h e s e , o n e is

c o n s i d e r a b l y more f r e q u e n t t h a n t h e o t h e r :

There are a t l e a s t two l o g i c a l p o s s i b i l i t i e s f o r [the t h e s p e a k e r may o v e r a l l d e s c r i p t i o n of a p a r t m e n t l a y o u t s ] d e s c r i b e a map o f t h e a p a r t m e n t . o r h e may describe a tour of it. F x m p l e s of each a r e the following:
I'd say i t ' s laid o u t i n a huge square p a t t e r n , broken down into f o u r u n i t s . If you were l o o k i n g down a t t h i s a p a r t m e n t f r o m a h e i g h t , i t would be like like I s a i d before, a huge square w i t h two l i n e s drawn through t h e c e n t e r t o make f o u r smaller s q u a r e s . Now on the e n d s uh i n the two boxes f a c i n g o u t on the s t r e e t you have t h e l i v i n g room and a bedroom. I n between thC?se two b o x e s y o u h a v e a Now between the n e x t two b o x e s , facing t h e bathroom. c o u r t y a r d you have a small foyer and then two boxes, o n e o f which i s a bedroom and t h e o t h e r o f which i s a k i t c h e n and a small foyer a a l i t t l e beyond t h a t .

... . .

...

.. .. .

...

Well you walk i n t h e door and t h e r e ' s a k i t c h e n and then o f f t h e k i t c h e n i s one bedroom. As you g o s t r a i e h t i n from t h e doorway t h r o u g h t t h e k i t c h e n you go i n t o t h e l i v i n g room. And t h e n to t h e l e f t of the l i v i n g room aye two bedrooms. The two bedrooms are o n the same s i d e of t h e b u i l d i n g and t h e l i v i n g room and t h e k i t c h e n a r e o n the same s i d e of t h e building.

Both o f t h e s e d e s c r i p t i o n s a r e r e a s o n a b l e agsyers t o t h e q u e s t i o n "would you describe t h e l g y o u t o f your a p a r t m e n t ? " Our i n t u i t i o n c e r t a i n l y i n f o r m s us t h a t b o t h speakers h a v e f u l f i l l e d cne task t h a t was proposed them. What o u r i n t u i t i o n s d o not tell us i s t h a t d e s c r i p t i o n s like [the first] a r e extremely rare, while d e s c r i p t i o n s l i k e [the second] are e x t r e m e l y common. O f 72 a p a r t * while 6 9 ment d e s c r i p t i o n s , o n l y 3 are o f t h e form of a map are the form of a t o u r ( p p , 8-9)

..

The t o u r may b e a c o m p o s i t i o n o f s e p a r a t e between rooms

of t h e a p a r t m e n t

episodic

events

of

moving

The p l a n i s more o b v i o u s l y s y s t e m i c ,

involving s p a t i a l ( l e f t , r i g h t ,
organiea tion. Longacre

e tc * )

and

comnonential

(part-whole)

(1968)

n o t e s t h a t i n a given language t h e r e i s a f i n i t e
types

number o f d i s c o u r s e Discourse from

which

can

never

be

mixed

or

confused.

various

P h i l i p p i n e l a n g u a g e s s u g g e s t four contrasting

d i s c o u r s e prose g e n r e s : N a r r a t i v e : r e c o u n t s some s o r t of s t o f y P r o c e d u r a l : t e l l s how t o do something. E x p o s i t o r y : a n y sort o f e x p l a n a t o r y e s s a y . H o r t a t o r y : a t t e m p t s t o i n f l u e n c e o r t o change c o n d u c t * N a r r a t i v e d i s c o u r s e i s composed o f t h e f o l l o w i n g tagmemes :

APERTURE p r o v i d e s t e m p o r a l and s p a t i a l s e t t i n g and i n t r o d u c e s some o f t h e p r i n c i p a l dramatis p e r s o n a e . CLOSURE gives f i n a l commentary on t h e main p a r t i c i p a n t s , " t h e y lived h a p p i l y e v e r after". Nuclear tagmemes EPISODE, DENOUEMENT, anti ANTIDENOUEEIENT show a g r e a t v a r i e t y o f exponence * typically
any paragraph t y p e may b e an e x p o n e n t plus d r s c o u r s e o f t h e PROCEDURAL o r EXPOSITORY g e n r e .
A

embedded

correspondence

can b e

informally tagmemes.

r e c o g n i z e d between some of F o r example, between

It

Propp's f u n c t i o n s and Longacre's

tial

Ini-

S i t u a t i o n 1 ' and " ~ ~t u r er" , and "Reward" and losu sure". e "Initial

For Propp
Lie

t h e peak o f t h e discourse i s i n t h e f u n c t i o n

PIisfortune

q u i d a t e d " , and f o r Longacre i t i s i n t h e tagmeme "Anti-Denouement"

The

idea of

h i e r a r k h i c o r k a n i z a t i o n o f tagmemes mentioned b y paralleled in Lakof f ' s

Longacre, above, i s generative nodel

(1992)

transf:ormational
A phrase s t r u c -

t h a t u s e s Propp's set o f f u n c t i o n s .
11

t u r e component g e n e r a t e s a

deep s t r u c t u r e " .

For example, t h e t a l e

of

( 2 2 ) may b e represented b y the tree stxuc t u r e o f F i g u r e 17.

-50-

SITUATION
REUARD

N
QUEST RESCUE

CALL FOR AID

RELEASE

MEDIATION

C W ~ E R - DEPARTURE S R G L T U GE ACTION

VICTORY

ilaimm

M ISFORTUNE

RETURN

WEDDING

Figure 17 Textual d e e p structure

The conclusion i s that there are prescibed patterns in all genres

of discourse; I term these patterns "themes".
plete

I do not offer

com-

inventory of

themes;

t h e i r discovery i s a matter of empirical

investigation
Any extended discourse i s unlikely t o b e organized according to
a

single theme* I hypothesise that a coherent aiscourse is characterized

by

a single rooted tree of themes, as schematized in Figure 17.

A

All

themes must be proper subthemes of the matrix theme.

text

with an

overlapping thematic structure is incoherent :

(23) Eating f i s h made John s i c k .

He caught measles l a s t May*

-51-

shown s c h e m a t i c a l l y i n F i g u r e 18.

Jdhn's poisoning

n illnesses John's

4 '
John eat a

John alck

John he8
meaales

Figure 18 I n c o h e r e n t t h e m a t i c s t r u c t u r e

An i m p o r t a n t p o i n t t o c o n c l u d e t h i s s e c t i o n .
tions
may

The inferred connecT h i ~s i

not

correspond w i t h t h o s e i n t e n d e d by t h e a u t h o r .
Here I o n l y a d d r e s s t h e a n a l y s i s

a n o t h e r problem. reader. If
he

of

story

by

c o n n e c t s i t i n t h e manner d e s c r i b e d above, t h e n i t i s

c o h e r e n t for him.

The R q l e of ---Not

t h e Encyclopedia

all

of

discourse

s t r u c t u r e is o v e r t l y s t a t e d ; d i s c o u r s e is

highly e l l i p t i c

I n (13) t h a discourse c o n n e c t i v e "sf ter"

is

present

t o mark a temporal sequence, b u t i n ( 1 4 ) t h e r e is no r e a l i z a t i ~ n f t h e o

causal

relation

between

the

two p r a p o s i t i o n s .
is

Normally o n e assumes acceptable if the

t h a t a d i s c o u r s e i s c o h e r e n t ; hence ( 1 2 ) rooms are taken

most

w i t h i n t h e same h a b i t a t i o n .

E v i d e n t l y a r e a d e r must from his cognitive

i n f e r omitted structure. s t o r e o f world knowledge.

The i n f e r e n c e s a r e made

There

is much

discussion

a t p r e s e n t a b o u t i n f e r e n c e as p a r t o f
the

understanding. right ones.

To make i n f e r e n c e s i s e a s y ; t h e problem is t o make

It h e l p s t o have a g o a l .

It i s s u g g e s t e d t h a t d i s c o u r s e judged coherent,

can b e s a i d to have been u n d e t s t o o d when i t has b e e n a s d e f i n e d above.

I n t h e n e x t s e c t i o n s a r e p r e s e n t e d t h e r o l e of t h e e n c y c l o p e d i a i n determining above. Anaphora

I f the dependent i s a proform then part

and

representing

t h e d i m e n s i o n s of c o h e r e n c e spelled out

of

understanding

is

to

determine

the correct

antecedent

There

are syntactic c o n s t r a i n t s
choices for antecedents

(Langacker , 1969) which s e r v e t o narrow down

and

to

give an o r d e r of p r e f e r e n c e .

Winograd (1971) a l s o e s t a b l i s h e d Nash-Webber antecedents (1976) used

an ordering f o r the choice of

antecedents. possible

lambda

abstraction

to

establish

The chosen

a n t e c e d e n t , when s u b s t i t u t e d f o r t h e proform, must produce a meaningful proposition t h a t i s coherent i n context.

1

A meaningful

proposition

is

one t h a t h a s a c o u n t e r p a r t i n t h e e n c y c l o p e d i a .

Wilks (1975) d i s c u s s e s

method

of

f i n d i n g t h e most s e m a n t i c a l l y a c c e p t a b l e a n t e c e d e n t . self-same

In

e n c y c l o p e d ~ c terms, t h e c o u n t e r p a r t may b e t h e o?, more l i k e l y , a s y s t e m i c p r o p o s i t i o n . p r o p o s i t i o n has been d e s c r i b e d earlier.

proposition,

The p r o c e s s o f f i n d i n g s u c h a

If no g e n e r a l i z a t i o n is found

t h e i n p u t p r o p o s i t i o n i s n o t c o n s i s t e n t with e n c y c l o p e d i c knowledge.

Abstract

terms can be

d e f i n e d by complexes of general proposi-

t i o n s , each having s u f f i c i e n t c o n c e p t u a l c o n t e n t t o in

define

situations

which

they a p p l y .

For example, a d e f i n i t i o n of "mistake" must b e

The

such thar i t a p p l i e s t o p a r t of t h e f i r s t sentence i n (11 1.

e s s o f a b s t r a c t i o n needed h e r e was p r e s e n t e d above. S p a t i a l , Temporal, - Causal Cohesion and To i n er

proc-

omitted spa t i o - temporal and causal r e l a t i o n s , i .e

., the
to

d i s c u r s i v e r e l a t i o n s of t h e

encyclopedia, Systemic

it

is

also

necessary

locate

general

propositions.

memory,

of course, includes discourse proposi-

these r e l a t i o n s .
tion

S c h e m a t i c a l l y , F i g u r e 19, from a

PI w e c a n l o c a t e P2, b y t h e means a l r e a d y d e s c r i b e d .

P2 may have
A proposi-

a d i s c u r s i v e r e l a t l o n R t o a n o t h e r s y s t e m i c p r o p o s i t i o n P3.
tion and

Pq , a

p a r t i c u l a r i z e d v e r s i o n of P3, and t h e r e l a t i o n R, 'between

P2

P3, c a n b e added t o t h e d i s c o u r s e .
then

Of t e n
merely

Pq

w i l l be a proposition
relation need be

already s t a t e d i n the discourse,

the

i n f e r r e d co augment t h e p l o t s t r u c t u r e .

R
P2

P4

ENCYCLOPEDIA

P1

DISCOURSE

F i g u r e 1 9 inference of d i s c u r s i v e s t r u c t u r e

It

may

may, however, b e n e c e s s a r y t o infer a chain of p r o p o s i t i o n n t o

original discourse Intuitively there

l i n k t h e p r o p o s i t i o n s of t h e

must be a l i m i t on t h e number of p r o p o s i t i o n s t h a t can be i n f e r r e d i n a

s e n s i b l e p a t h , b u t a t p r e s e n t no % m i g h t can be offered.

To

exemplify

t h e p r o c e s s i n g r e a t e r d e t a i l , l e t us c o n s i d e r some
"In

o f the knowledge t h a t i s used i n the a n a l y s i s of ( 1 5 ) :

water

and

not

ah1 e

to

act

c a u s e s drowning".

In F i g u r e 20 t h e network form o f

t h i s knowledge i s p r e s e n t e d .

F i g u r e 20

Example of causal i n f e r e n c e

Prom

the

discourse

propositions

"DF

i n watert' and "DF cannot act",

be

paradigmatic s t r u c t u r e e n a b l e s t h e systemic p r o p o s i t i o n s A and B t o found.

mere

is e c o r e f e r e n t i a l i t y c o n d i t i o n t h a t must b e tested in

the manner d e s c r i b e d earlier

The d i s c o u r s e

propositions

pass

the

test,

so

the

complex

represented

by

the

m o d a l i t y C e x i s t s i n the

discourse.

Ttle d i s c u r s i v e r e l a t i o n cause c a n be followed from C t o

D.

The l a t t e r i s a p l a ~ i s i b l ei n f e r e n c e , and i n f a c t , o s p e c i f i c equivolerlt

of

is

o n e o f the o r i g i n a l p r o p o s i t i o n s o f the d i s c o u r s e , i . e m , "DF

drom".

The c o n c e p t s o f t h e sys.temic p r o p o s i t i o n s .ire

the

linked

to

thcr

rest

of

encyclopedia

by

t y p i c a l o r L s m It i s s o because o f t h e

n a t u r e o f the hnowledge. is s u c h that i t i s only happen i n t h e given c i r c u n ~ s t a n es c

The

son~ething that

could

.
testing reality of Thorndyke ( 1 9 7 6 ) a r e that. in understanding natural

indications a

from

the

~ n i e r e n c e s are language t e x t s . Thematicity

psychological

In, t h e p r e s e n t s y s t e m , a t h e m a t i c c o n c e p t i s d e f i n e d s t r u c t u r a l l y ,

i t i s a n y t h i n g having a metalingual d e f i n i t i o n *

1\

theme i s t h e r e t o r e a

complex

of

generalized

propositions.

The

p r o c e s s o f detecting t h e themes,

a p g l l c a b i l i t y o f a b s t r a c t terms, and hence o f f i n d i n g s t r a ct i o n , d e s c r i b e d above.

is

ab-

A b s t r a c t i o n i s a r e c u r s i v e p r o c e s s and IS to exist

t h u s one way t o c a p t u r e t h e embedding of themes h y p o t h e s i s e d i n discourse. The

p a r a p h r a s e s e l l c i t e d by Pfandler , Johnson, and D e F o r e s t (1'976

and Rumelhart (1976) show t h a t s u b j e c t s d o c r e a t e d e s c r i p t i o n s of t e x t s t h a t vary i n abstractness proposed here. in accord with the
hierarchy

of

themes

IMPLEMENTATION

I n t h i q s e c t i o n t h e i m p l e ~ n e n t a t i o nof t h e structures and processes presented above

is

described*

The

o r i g i n a l program was w r i t t e n i n

SNOBOL f o r a CDC6400.
I n a c o m p l e t e s y s t e m there s h o u l d , o f c o u r s e , b e

parser.

For

t h e present t h i s does n o t e x i s t ; t h e system o n l y embodies the c o g n i t i v e component.

be
T h i s means t h a t t h e o v e r a l l o r g a n i z a t i o n i s n o t a s i t would

in

c o m p l e t e t e x t a n a l y s i s system, where i n t e r a c t i o n b e t w e e n the

(Woods,

syptac t i c and s e m a n t i c components is essential

1971;

Schank,

1975a;

Winograd,

1971;

Erdman,

1975)

The

justification for this establish only the

ommission i s t h a t for t h e p r e s e n t I a s e e k i n g to m

n a t u r e of t h e semantic organization of a coherent discourse. structure

systern-

Once t h i s

has

b e e n i d e n t i f i e d i t w i l l p r o v i d e the g o a l f o r a complete

Input t o t h e spcern

is

accordingly

in

cognipive

form

that

r e t a i n s t h e l o g i t a l e l l i p s i s o f t h e s u r f a c e form. Most of
t h e p r o c e s s i n 8 i s p e r f o r m e d by " N o n n a l i z e r " w h i c h i n f e r s

o m i t t e d l o g i c a l and t h e m a ~ i cs t r u c t u r e . then

A judgement

of

coherence

is

simple

task:

i f t h e d i s c o u r s e is not l o g i c a l l y c o n n e c t e d o r
it

d o e s n o t have a s i n g l e theme, t h e n

is

incoherent;

otherwise

the

m a t r i x theme i n d E c a t e s t h e t o p i c o f t h e d i s c o u r s e .

Processes
A

component

of

all

processes

i s a breadth f i r s t path-tracing
a

r o u t i n e , c a l l e d "Ripple".

A s e a r c h p a t h i s d e f i n e d by

sequence of

arc

types

path

does

n o t explicitly s t a t e whether an a r c can be

and

repeated. this

The network i s assumed t o b e s y n t a c t i c a l l y well-formed

controls repetitions.

An a r c can b e marked as o b l i g a t o r y ; o t h e r -

wise i t i s o p t i o n a l .

A g o a l of the search can b e d e f i n e d .

This may b e
tag",

a p a r t i c u l a r node, o r a node marked with a s p e c i f i e d " a c t i v a t i o n

i.e.,

node

reached

by a n o t h e r p a t h , when seeking an i n t e r s e c t i o n .

P a r a d i p i a t i c path,-tracinq Paradigmatic p a t h - t r a c i n g is implemented by

sequence

Ripple

with

path

is

'EfAN TYP I S T VAR.

----

VAR IST TYP MAN

(see Figure 11).

A converse paradigmatic path

The p r o p e r t i e s a s s o c i a t e d w i t h a v a r i e t a l concept the concept.

may b e found by R i p p l e with a path TYP MAN s t a r t i n g from

Causal c o n n e c t i v i t y c o n d i t i o n This process
uses

R i p p l e w i t h cause a s t h e p a t h d e f i n i t i o n *

It

a l s o has t o i n c l u d e P-W arid P-W

conj unc t s .

to

be

able

to

reach

from

and

to

Discovery of gener a1 and spec i f i c propo s i t i ons

A l l p r o p o s i t i o n s of a d i s c o u r s e must match g e n e r a l p r o p o s i t i o n s i n

The

the

encyclopedia.

procedure

is t o make c y c l i c c a l l s of Ripple.

The f i r s t i s from the modality node of t h e d i s c o u r s e proposition.

node reached, o t h e r than t h e modality, i n i t i a t e s a n o t h e r s e a r c h i n

Each
the

encyclopedia.

For

example,

given

t h e d i s c o u r s e and e n c y c l o p e d i a o f

F i g u r e 12, t h e p r o c e s s i s as f o l l o w s : from "gobblef' node 1, a c o n v e r s e p a r a d i g m a t i c p a t h p l u s a t y p i c a l arc p l u s m a n i f e s t a t i o n i a followed t o , f o r example, node 2 i n t h e e n c y c l o p e d i a . discourse gives nodes R i p p l i p g from "gobble" i n the

"Marv" and

If

caviar".

The

syntagmatic

arcs

traversed are n o t e d .

From "Marv", node A, a c o n v e r s e p a r a d i g m a t i c p a t h

p l u s t y p i c a l p l u s m a n i f e s t a t i o n plus c o n v e r s e a g e n t i s f o l l o w e d , w i t h a g o a l of a node a c t i v a t e d from t h e p r i o r d i s c o u r s e node, i ~ e . ,"gobble".

[Not a l l of t h e s e arcs h a v e t o b e p r e s e n t , t h e y a r e o p t i o n a l except 0,

the syntagmatic

arc.] Node 2 s a t i s f i e s t h i s goal.

plus

From

tt

caviar" node plus

B, a c o n v e r s e p a r a d i g m a t i c p a t h p l u s t y p i c a l

manifestation

converse o b j e c t i v e is followed w i t h a g ~ a of a node a c t i v a t e d from t h e l

prior d i s c o u r s e node "gobbleff* Again node 2 s a t i s f i e s t h e g o a l . Thus

t h e p r o p o s i t i o n a t 2 i s a g e n e r a l i z a t i o n o f tihe d i s c o u r s e
The

proposition.

condition

on an a c c e p t a b l e g e n e r a l i z e d match i s t h a t i t must discourse proposition,;

c o n t a i n a l l t h e s y n t a g m a t i c informirtion o f t h e the less generalizatdon

may c o n t a i n more i n f o r m a t i o n b u t i t c a n n o t c o n t a i n syntagmetic structure and for

S e p a r a t e searches a r e made f o r

s p a tio-temporal

i n f o r m a t i o n on t h e m o d a l i t y o f a p r o p o s i t i o n .

It i s o n l y n e c e s s a r y t o change t h e p a t h d e s c r i p t i o n from t h a t used

in

Figure

12 from converse p a r a d i g m a t i c p a t h t o p a r a d i g m a t i c p a t h f o r

the r o u t i n e t o l o c a l e more s p e c i f i c p r o p o s i t i o n s .

k i e t a l i n g u a l decomposition

The s e a r c h f o r g e n e r a l p r o p o s i t i o n s a l s o
metalingual definitions.

flags

nodes

that

have

New propositions having the structure of the

names
drawn

d e f i n i e n s are made by copying the d e f i n i e n s but w i t h node

from t h e p r o p o s i t i o n t h a t i s being p a r a p h r a s e d

These new p r o p o s i t i o n s To make t h e copy, t h e the definiens.

are

then

considered

a s p a r t of t h e d i s c o u r s e .

b r e s d t h f i r s t s e a r c h r o u t i n e i s used t o For

pass

through

each

ode and

arc i n the d e f i n i t i o n , an e q u i v a l e n t s t r u c t u r e i s

created
by

The end of s c a n n i n g a p r o p o s i t i o n of the d e f i n i e n s i s marked

a typical a r c , i.e.,

reaching

a t t h e p o i n t a t which t h e d e f i n i t i o n
If

i s linked i n t o paradigmatic organization.

participant

in

the

g e n e r a l i z e d p r o p o s i t i o n matches a p a r t i c i p a n t i n t h e d i s c o u r s e proposit i o n t h e n t h i s p a r t i c i p a n t f i t s i n t o t h e c o r r e s p o n d i n g role s l o t i n t h e definiendum, otherwise the concept

i n the definiendm i s used.

definition

For
of

example, given "Peter buys a b i c y c l e from Jane" and the

"buy"

as

in

rigure

16.

In

l o c a t i n g the systemic definiendm, t h e

c o r r e s p o n d e n c e s o f " ~ e t e r "t o "A", e t c reaches

.,

a r e a l s o found.

When

copying

the

node t h a t matched "peter", t h i s name is inserted into the There i s no c o r r e s p o n d e n c e f o r t h e i n s t r u m e n t , s o

t1

paraphrase.

money"

1s t h e i n s e r t e d from t h e d e f i n i i e n d u m ,
M e t a L i n ~ u a la b s t r a c t i o n In s e a r c h i n g f o r g e n e r a l p r o p o s i t i o n s , some may b e found t h a t a r e

components o f m e t a l i n g u a l d e f i n i t i o n s . p d i n t e d t o b y a part-whole arc.

These have m o d a l i t i e s t h a t

are

-60-

The

proqess

that

tests

c o r e f e r e n c e and c o n t e x t u a l requirements

u s e s Ripple to t r a v e r s e i n p a r a l l e l t h e tions and those o f t h e d e f i n i e n s .

candidate

discourse proposi-

Typical arcs i n the d e f i n i e n s l i m i t

t h e search. nodes in

Each node of the d e f i n i e n s

is

compared

with

equivalent

the

discourse

p r o p o s i t i o n s a t each s t e p . not

A p r o p o s i t i o n is

r e j e c t e d i f a node has no e q u i v a l e n t o r i t does properties

possess

all

the

of

the

nodes

of

the
See.,

definiens, including arcs t o nodes

if

matched i n the previous (discourse) were

step,

nodes

(systemic)

and

taken as corresponding nodes a t one step, then i f on definition

the n e x t step a node o f the

has

an

arc

to

then

the

discourse

must

have

t h e same a r c t o Ya

Only t h o s e p r o p o s i t i o n s t h a t

match the d e f i n i e n s w i l l n o t have been rejected and

can

be

rewritten

using

the

abstrwt

term.

The

p r o c e s s can -be i l l u s t r a t e d using the the

d e f i n i t i o n o f "poison" given i n F i g u r e 14b, and i t s a p p l i c a t i o n t o discourse

in

Figure 14a.

The c r u x of t h e t e s t i s a t t h a t node o f t h e from

it*

d e f i n i t i o n having t h e two m a n i f e s t a t i o n s arcs emanating

If

the

discourse

p r o p o s i t i o n d i d n o t have two m a n i f e s t a t i o n s i t would be

rejected

This i s how "John's

eating t h e

worms

made

Fred

sick"

is

eliminated*

O r i f i t does have two m a n i f e s t a t i o n s , t h e y must p o i n t t o

Thus

nodes t h a t were s a t i s f i e d Qn the p r e v i o u s step of the comparison.

if o n e o f t h e m a n i f e s t a t i o n s p o i n t e d into a n o t h e r p r o p o s i t i o n t h e
would f a i l .

test

I n f e r e n c e of o m i t t e d d i s c u r s i v e r e l a t i o n s
I

search

along

d i s c u r s i v e a r c s i n s y s t e m i c memory from c o u n t e r lend to a proposition

p a r t s o f d i s c o u r s e p r o p o s i t i o n s may flagged

that

is

a s a g e n e r a l i z ~ t i o nof a n o t h e r d i s c o u r s e p r o p o s l t i o n . discusive
If

If t h i s

is s o then the

arc

may

be

added

between

the

discourse

propositions.

the

p r o p o s i t i o n r e a c h e d i s n o t flagged then i t and copying

t h e d i s c u r s i v e a r c a r e c o p i e d , and added t o t h e d i s c o u r s e . Ttle r o u t i n e was g i v e n i n t h e d i s c u s s i o n o f d e c o m p o s i t i o n .

The - System
The

flow

c h a r t of the a n a l y s t s f s shown in Figure 21.

The mean-

i n g s of t h e a n n o t a t i o n s are:
OLDINFO h a s a d i s c o u r s e p r o p o s i t i o n as i t s argument. It f i n d s It c a l l s a r o u t i n e SPACETIME t o cornsystemic equivalents. p a r e spatio-temporal contexts. SPACETLME is a l s o c a l l e d

d u r i n g t h e s e a r c h f o r g e n e r a l p r o p o s i t i o n s when a non-event node is found w i t h a n a t t a c h e d m o d a l i t y . I f OLDINFO i s p r e s e n t e d w i t h a m o d a l i t y that has o n l y part-whole r e l a t i o n s t a o t h e r nodes, i t d o e s n o t h i n g . It s u c c e e d s i f LOGCON h a s a s y s t e m i c p r o p o s i t i o n a s i t s argumfnt. i t f i n d s d link t o a g e n e r a l p r o p o s i t i o n c o r r e s p o n d i n g t o a p r o p o s i t i o n o f t h e d i s c o u r s e ( i n c l u d i n g p r o p o s i t i o n s added by inference). It a l s o g e n e r a t e s INTERLIST, a l i s t o f c a u s a l i n f e r e n c e s from p r o p o s i t i o n s o f the d i s c o u r s e . IKST i s a l i s t o f n o d e s found t o h a v e m e t a l i n g u a l d e f i n i t i o n s . CONJLIST i s a l i s t o f c o n j o i n e d p r o p o s i t i o n s . When a d i s c o u r s e p r o p o s i t i o n i s matched a g a i n s t t h e e n c y c l o p e d i a , i t s e e s if t h e encycl o p e d i c p r o p o s i t i o n is a c o n s t i t u e n t of a n o t h e r m o d a l i t y . A CONJUNCTION TEST r o u t i n e u s e s CONJLIST t o l o c a t e d i s c o u r s e p r o p o s i t i o n s t h a t c a n be grouped. TRANSFORM has two modes I n o n e i t i s used t o decompose p r o p o s i t i o n s t h a t c o n t a i n a m e t a l i n g u a l l y d e f i r i e s c o n c e p t * A second mode c e a t e s c a u s a l l y i n f e r r e d p r o p o s i t i o n s

/\

/ \

Y
'C /

/\

CONJUNCTION

Figure 21

F l o w chart of the system

nNALYSIS OF SOME STORIES I

want to show that abstract patterns are quite general, that all

l i n g u i s t i c behavior i s based on such patterns.

Obviously such a

claim

must

be

s u b s t a n t i a t e d by t h e d i s c o v e r y of s u c h patterns.
test

A number of

s t o r i e s o f drowning were used t o

this

hypothesis.

The

second

claim

of

p r o p e r embedding,of themes was a l s o t e s t e d by a more complex

drowning s t o r y . I n t h e examples a r e f i n e d hypo t h e s i s of d i s c o u r s e c o n n e c t e d n e s s i s used.

One h a b i t i n d i s c o u r s e i s t o s e t
~ o n ~ a c r e '"Aperture"). s

the

stage

(kropp's

"Initial

Situation",

I n terms of t h e model t h i n a s p e c t

s h o u l d b e r e c o g n i z a b l e by t h e o c c u r r e n c e of s p a c e and

time

relations.

W e

find

?I

todayf',

"in

MB

resevoir",

"On

October 11, 1974". "DF of

Quekns", e t c . A g r e a t e r s t r u ~ t u r a lc o m p l e x i t y o f expression

is

to

be

expected

elsewhere

i n t h e s t o r i e s ( s e e Longacre's comments on n u c l e a r Longacre (1972) i n c l u d e s i n t h e n m u r a l o u t l i n e of a V a r i o u s surface long sentences,

tagmemes, a b o v e ) .

d i s c o u r s e , r e c o g n i t i o n o f a peak within t h e discourse. m a r k i n g s f o r t h e peak a r e given: rhetorical ( c f . White's t e n s e change, e x t r a

u n d e r l i n i n g s , etc. Taking an e t h n o c e n t r i c v i e w o f t h e world t e l e o g i c a l commune), i t i s s u g g e s t e d t h a t I n t h e underlyIt

ing

form,

t h e peak w i l l l i e w i t h i n c a u s a l l y r e l a t e d p r o p o s i t i o n s .

i s t h u s e x p e c t e d to f i n d t h e theme w i t h i n t h e c a u s a l s t r u c t u r e and s o I

f o c u s on t h i s o r g a n i z a t i o n .

T h i s would be i n a p p r o p r i a t e i f t h e s t o r i e s

were a e s c r i p t i o n s o f t h e kind e l i c i t e d by L i n d e , above.

Common p a t t e r n s S h a r t f a c t u a l accounts o f drowning8 were e l i c i t e d from freshmen fn Linguistics

and English.

The i n s t r u c t i o n s given sought o n l y t o define

"Write a drowning s t o r y
that, for

a topic and an approleimate length:

example,
Times."

you

might

expect t o f i n d as a column filler i n t h e New York

A sample is

(Story 1) The body o f Horatio Smith was found last night i n the Niagar a River H was drowned when his b o a t overturned on the e river

The hypothesis formed i s t h a t an a c c e p t a b l e drowning

g i v e the following information:

story

must

(a) (b)

W y t h e victim was i n t h e water. h Why the v i c t i m was not able t o save himself

The r a t i o n a l e s for t h e s e r e q u i s i t e s are:

(c)

(d)

A p e r s o n i s n o t u s u a l l y found i n water, and therefore some e x p l a n a t i o n of t h i s l o c a t i o n is expected. By an i n s t i n c t o f s e l f - p r e s e r v a t i o n , one would expect t h e victim t o t r y t o e x t r i c a t e himself from his predicament. The s t o r y should say why he couldn't.

F i g u r e 22 shows t h e c o g n i t i v e form of t h i s requirement.

MTL

E
CAUSE

@<drowning

1
CAUSE

Idrow

CAUSE

act 1

<person>

F i g u r e 22

The drowning theme

The empty m o d a l i t i e s i n d i c a t e t h a t a matching s t o r y must have something

that

stand

in

a c a u s a l r e l a t i o n s h i p t o the o t h e r p r o p o s i t i o n s , L e e ,

e x p l a i n why t h e y happened (what caused them) plicit, this information must

If n o t o r i g i n a l l y

ex-

be

recoverable

through encyclopedic

knowledge.

One way i n which the c o n t e n t of systemic memory

tiated
is by

may

be

substan-

examination

of

negative

propositions i n the s t o r i e s ;

writers presumably o n l y need to negate normal expectancies, and this i s

what g e n e r a l i z a t i o n s are.
stories. Unfortunately t h e r e are few n e g a t i v e s i n the

S t o r i e s 31 and 5 indicate t h e assumption of swimming ability. success

Stories 12, 32, and 42 show t h a t t h e

anticipated.

of

rescue

attempts

is

-66-

The

stories

elicited

fell

into

several

categories.

The two

a n a l y s e d here are a b o a t c a p s i z i n g and

a fall i n t o water.
at
an abstract

I n s p i t e of
level the

surface d i f f e r e n c e i t will be shown

stories conforn to the

that:

thematic p a t t e r n .

The analysis o f one s t o r y

from each c a t e g o r y i s p r e s e n t e d
S t o r i e s n o t a n a l y s e d i n c l u d e such happenings a s too much

person

eating

t h e n g o i n g f o r a swim; Jesus f r e a k s t r a n g t o walk on water; occupants

and w a t e r - s k i e r s h a v i n g a c c i d e n t s w h i l e watching b i k i n i - c l a d
of passing boats
A b -- o a t

capsizing

( S t o r y 1) The body of H o r a t i o Smith was found l a s t n i g h t i n the Niagara R i v e r H e was drowned when h i s b o a t o v e r t u r n e d i n t h e

river.
( S t o r y 2) Eggbert W i l l i s , 56, of Bayside, drowned t h i s morning a f t e r t h e b o a t h e was rowing o v e r t u r n e d n e a r D e v i l ' s Cove. ( S t o r y 3) The body o f John Smith, 58, was d i s c o v e r e d today a t t h e foot of West Ferry S t r e e t * H was r e p o r t e d m i s s i n g f o u r d a y s ago e by h i s wife a f t e r he f a i l e d t o r e t u r n from a b o a t i n g t r i p . HIS b o a t had capsized. Death was d u e t o drowning. ( S t o r y 31) A small s a i l b o a t was a f l o a t on a calm p e a c e f u l lake when suddenly t h e mast of t h e v e s s e l s t r u c k some cables overhead and the boat c a p s i z e d . The two Inen aboard drowned, ane b e c a u s e he was h i t by t h e b o a t and r e n d e r e d unconscious, t h e other didn't know how t o s w i m * Story 1 is analysed*

In

it,

some o f t h e c a u s a l and t h e m a t i c

s t r u c t u r e i s a b s e n t and is r e c o n s t r u c t e d u s i n g the f o l l o w i n g knowledge:

(i)
(ii) (iii)

I f a p e r s o n i s i n a b o a t and t h e b o a t o v e r t u r n s , t h i s may c a u s e him t o be i n j u r e d and t o b e i n t h e w a t e r . I f injured a p e r s o n may n o t b e a b l e t o 11act". I f a p e r s o n i s i n water and cannot "act" t h e n he may drown. F i g u r e 18 shows t h e e n c y c l o p e d i c form of t h i s p i e c e o f knowledge. -6 7-

The nodes appearing i n t h e e n c y c l o p e d i c e n t r i e s for all of t h e s e Eacts

e x e m p l i f i e d i n t h i s a e c t i o n a r e linked t o

varietal

nodes

by

typical

arcs.

T h i s is so because the Eacts qre n o t obligatory on some concept

examples of this category

b u t are something t h a t may happen t o some some o f t h e t i m e .

In

the

analysis

o f the c o l l e c t i o n o f drowning s t o r i e s , o n l y t h e dtowning

parts of t h e s t o r y t h a t a r e relevant t o t h e For example, in Story

are

conside'ted.

1, o n l y t h e second s e n t e n c e is p r o c e s s e d ; t h e

f i r s t d e a l s w i t h an event a f t e r t h e d e a t h and c ~ s e q u e n t l y excluded. is

The result o f t h e a n a l y s i s i s shown i n Fip;ure

p r o p o s i t i o n s of t h e d i s c o u r s e a r e : Boat c o n t a i n s H o r a t i o Smith Boat o v e r t u r n s "on Niagara R i v e r . H o r a t i o Smith drowns.

23.

The

original

The

a n t e c e d e n t c o n d i t i o n s e x p r e s s e d i n a g e n e r a l form i n ( i ) match t h e Thus i t i s i n f e r r e d t h a t Moratio Smith

s p e c i f i c s i t u a t i o n i n S t o r y 1.
is in

t h e w a t e r and t h a t he is i n j u r e d .

From (if) i t f o l l o w s t h a t h e

is not: a b l e t o a c t .

By b e i n g i n t h e water and n o t a b l e t o a c t , he

can

drown, a f a c t s t a t e d e x p l i c i t l y i n t h e s t o r y , a s shown i n Figure 23-

(Horatlo Smith)

Figure 23 A boat capsizing

hrther

we have explanations for Horatio Smith being i n the water: the

boat he was in ovexturned; and f o r him not being able t o

i nj ur ed

act:

he

was

The

theme

fits.

W e

have
is

a connected
coherent.

discourse

with

(trivially) a single-rooted theme; i t

that
a

The p o s s i b i l i t y

boat over turning only puts a person i n t o the water i s added to

i s not

t h e encyclopedia to account for part o f Story 31, where i t

an

injury but
himself.

the i n a b i l i t y to swim t h a t prevents one victim from saving

These two facts match the

This has consequences f o r Story 1.

same propositions but a r e an exclusive conjunction.

When an complex is
already
cow

found that has the same constituents, as an

conjunction

s t r u c t e d , the l a t e r episode i s stacked and used i f the current analysis

path

fails*

In Story 1 the use of t h e a l t e r n a t i v e does n o t lead t o a

connected structure.

The subsequent backup

then

takes

the

correct

I -

A f a l l into water

The second c a t e g o r y of dromina s t o r y requires t h e a d d i t i o n o f t h e

knowledge t h a t " I f

a person falls, he may i n j u r e himselftta Ten s t o r i e s

i n t h i s c a t e g o r y a r e l i s t e d below.
(StoPy 5) Early titis morning, James R. S m i t h , age 7 , was for~nd In a swimming pool near liis hame. I n v e s t i g a t o r s say the boy stumbled into t h e pool i n t h e d a r k n e s s e s r l y t h i s morning whilst looking for his p e t k i t t e n , Unable to swim, the boy drowned. ( S t o r y 7) A t the home o f blrs. ~ o h nSmith on Elmwood Avenue, a b o y , . Mark, age 15, drowned in h i s pool. The boy was w i t h t w o o t h e r f r i e n d s . They were performing w a t e r s t u n t s when Mark f e l l and smashed h i s head on t h e bottom of t h e pool ( S t o r y 12) Y e s t e r d a y a f t e r n o o n , the life o f a B u f f a l o youth was t a k e n when he s l i p p e d on rocks-at a l o c a l q u a r r y . The failure of attempted rescues resulted i n the drowning of llichael Smith, age 7, of 29 Oak Street, B u f f a l o .

(Story 19) A 1 2 year-old boy was found drowned i n E l l i c o t t Creek. Sources say t h e boy r a n away from home and f e l l a c c i d e n t a l l y i n t o t h e water.
(Story 26) - A l0-yea1 d d boy d i e d l a s t n i g h t when he fell into a s m k l pond. H i s f r i e n d s say he w a s c h a s i n g his p a r a k e e t which had escaped from its cage, when the i n c i d e n t o c c u r r e d .

( S t o r y 32)- S t e v e Smith, o f Hickstown, drowned t o d a y while sailing on G l a s s l y k e Eake. Er I Smith, who was knocked- overboard when struck on t h e head by 8 s e a g u l l , perished , b e f o r e h e l p could r e a c h * him. H i s son Edgar's attempts t o save his l i f e proved f u t i l e .

( S t o r y 37) hn u n i d e n t i f i e d man was seen by s e v e r a l p e r s o n s f a ; l i n g s i n t o the Niagara R i v e r a t the foot oE F e r r y S t r e e t . He was l a t e r pulled from t h e w a t e r and pronounced dead a t t h e scene. The cause of d e a t h was drowning. ( S t o r y 38) Today, t h e world's greatest s w i m m e r died. John Whale was preparing t o take a bath when he t r i p p e d and fell i n t o the bath Cause o f d e a t h was# drowning. -70-

( S t o r y 40) On October l l t h , 1974, an u n i d e n t i f i e d man drowned i n his b a t h t u b a t t h e Hotel Sheraton* The drowning was due t o t h e f a c t t h a t h6 f e l l into t h e tub i n t r y i n g t o make himself sobet. ( S t o r y 42) An 11 year-old boy drowned today after f a l l i n g i n t o t h e c a n a l where he and his f r i e n d s were playing. The two o t h e r boys, both eleven, t r i e d t o save t h e i r companion but were unable t o do s o * ( S t o r y 4) A body was found early y e s t e r d a y a t t h e f o o t o f t h e Mango River, near C l u b s p o r t * The body i s b e l i e v e d t o be t h a t o f Jose Gepasto. It seems a s i f M Gepaeto's car made a wrong t u r n r o n t h e highway and plunged i n t o t h e water S t o r y 4 i s a n a l y s e d * Note t h a t i t does n o t e x p l i c i t l y mention t h e motion of the person, o n l y t h a t o f the car. Understanding t h e s t o r y

r e q u i r e s t h a t i t be known t h a t :

(h) If a person is "contained" by something t h a t f a l l s ,

then he a l s o f a l l s .

(b) I f a person is "contained" i n c o n t a c t with something p e r s o n is i n c o n t a c t with (But n o t i f t h e something

Further

by something t h a t i s (emgm, water), t h e n t h e t h e something ( w a t e r ) too. is a submariner)

it

i s n o t given t h a t Jose Gepasto drowns.

ended.

This can b e i n f e r analysis continued

r e d , b u t the i n f e r e n c e chain i s open

The

making

causal

i n f e r e f i c e s and c o n j u n c t i v e groupings, some o f which l e d Only when t h e system ran o u t of l o g i c a l

it

t o t h e d i s c o v e r y of t h e theme*

and c o n j u n c t i v e p o s s i b i l i t i e s , did

make

the

ccannec tedness

test.

F i g u r e 24 shows t h e network developed i n the analysis-

<car>

Figure 24 A fall i n t o water

Embedded themes
Story

22,

in

fact

taken
shall

from the New York T i m e s , looks like a

drowning story, and

as

be

shown,

does contain

this theme.

However, i t contains more.

T e claim i"s made that i t is a "tragedyt'. h

(Story 22) DF, 43 years o l d , of Queens, drowned today i n MB resevoir a f t e r rescuing his son D, who had fal,lem into the water w h i l e on a f i s h i n g t r i p a t TF, near here, the p o l i c e said.

This theme is defined i n Figure 6 as a situation i n which "Someane does

a good a c t

[ e . g . , rescue) and dies ( e . g . , drowns)".

the

It will be seen
theme.

that the tragedy i s a proper sub theme of

drowning

Thus,

though

the

story may be edid t o have two themes, one is part of t h e At each

other, and by our hypothesis, the discourse is still coherent.

s t e p the encyclopedic knowledge used in the inference and an outline of

the inferred nodes are indicated

Figure 25 shows an o u t l i n e of

the

evolved

atruc t u r e , where the original discourse propositions are shown

by

and inferfed proposi tiona by 0.

Father not
able to act

Sc I fall

Son injured

Son not able

to act

Figure 25

Embedbd themes

Step Step Step Step

0.

1 .
2.

3.

I n t i a l s t a t e . (8odes 1, 2, 3, 4 ) . Fall causes injury. (Node 5 ) . Injury causes i n a b i l i t y t o act (Node 6 ) In water and not able to acr: causes fescue. (Node 7 and a l i n k t o node 3 ) .

Step 4 .

To rescue someone who is i n the water, get i n t o

S t e p 5. Step 6 . S t e p 7.

(Nodes 8, 9 ) Acting c a u s e s weariness. (Node 10). Weariness causes i n a b i l i t y t o a c t . (Node 11). In water and n o t a b l e t o a c t causes drowning. (Node 12 and a link to node 4 ) .

t h e water.

Note t h a t t h e antcedent c o n d i t i o n i n S t e p 3 is t h e same as i n Step 7. Both r e s u l t a n t s i t u a t i o n s are p o s s i b l e and are noted

T e system h
a con-

can select e i t h e r .

However, t h e wrong choice does not lead t o

nected s t r u c t u r e and backup t o the a l t e r n a t i v e h a s t o b e made. After

Step

7 t h e d i s c o u r s e has an i n f e r r e d caugal s t r u c t u r e con-

nec t i n g all the o r i g i n a l p r o p o s i t i o n s . The theme "tragedy" f i t s , the rescue i s a ( p a r t i a l ) cause demisem of the

Rescue is a variety of a c t and good can apply t o i t and Brown

The drowning theme is a l s o present.
the tragThe

i s a v a r i e t y of d i e .

Although t h e drowning theme i s n o t defined i n terms of edy,

it

can

be seen t h a t one is properly embedded i n the o t h e r . present incomplete

Process t h a t performed the a n a l y s i s i s a t the notion of

because

embedding i s nQt well understood f o r the h i g h l y s t r u c -

tured network. conj oining

The process used the

transitivity

of

cause

and

the

of p r o p o s i t i o n s

Thus Che tragedy encompasses p r o p o s i t i o n s 10,

3, 10, 11, and 4 and the drowning 8, 9,

11,

and

4.

The

tran-

sitivity 10 and 11.

of

cause l e t s t h e chain 3, 10, 11 be e q u i v a l e n t t o the chain

A postmorrem on t h i s example r e v e a l s a s e r i o u s flaw.

As

can

be

seen

the

rescue

is t h e c a u s e of t h e f a t h e r being i n the water

The

analysis has f a i l e d t o d i e t i n g u i s h desire for an action, a goal,

the
execution of the action.

from

A more s a t i s f y i n g analysis would include

some of the mechanisms t o be found i n the

C4975).

robot

planner of

Furugori

Step 2 should be seen as s e t t i n g up the conditions for the son This provides a

to

drown, which i s an event that should be prevented.

goal for the subsequent a c t i v i t i e s

drowning

One way t o

prevent

someone

from

is t o save him, t h i s is a subgoal that would directly achieve

the goal.

If you want t o rescue someone who is in the water,

t o g e t into the water.

then i t

may b e necessary

With t h i s subgoal included,

5.

the goal can b e achieved, and the analysis resumes a t Step

Figure
This

26 shows

this preferred a n a l y s i s of t h i s fragment of the story.

would not c'hange the r e l a t i v e status of the themes.

Son In water

Father res

son

Son not drown

\
b father
Figure 26

+$*#

Father save son

Improved story a n a l y s i s

DISCUSSION

Much

of t h e r e p r e s e n t a t i v e power of the encyclopedia is unused in

the system for discourse analysis and t h e r e f o r e remains t o t o b e tested and evaluated
f i e r e is dlso n o t a t p r e s e n t a parsing system t o effect form.

t r a n s d u c t i o n from s u r f a c e t o encyclopedic first to

The methodology

is

t r y t o e s t a b l i s h a n adequate c o n c e p t u a l r e p r e s e n t a t i o n which Although one

provides a g o a l for a complete system.

example

demon-

strated
tion.

the embedding of themes, i t d i d n o t exhibit recursive abstrac-

Further examination of themes i n d i s c o u r s e should overcome this.

There a r e two aspects of that the
set

encyclopedia, this

paradigmatic

and

metalingual

organization,

model a p a r t from any o t h e r

c u r r e n t system.

Discussion w i l l be d i r e c t e d to comparative comments on

these a s p e c t s
It i s evident that t h e present sytem makes much

use

of

parsdig-

matic

organization.

Pet Skhank (1975a) seeks t o minimize the need f o r

H i s conlrlusion

t h i s kind of knowledge. that people

arises

from

the

observation

do not make responses based on paradigmatic a s s o c i a t i o n s ,

b u t r a t h e r on e p l s o d i c a s s o c i a t i o n s . against the

This i s n o t

telling

evldence

existence

of such s t r u c t u r e , r a t h e r i t may s a y something

about the c o g n i t i v e process o f f r e e a s s o c i a t i o n .

In

Schank's

system

there

is

no need f o r paraddgms a t t h e l e v e l of conceptual representa-

t i o n as words are transformed into conceptual p r i m i t i v e s by t h e concep-

t u a l p a r s e r * The p a r s e r thus c o n t a i n s knowledge t h a t

e q u i v a l e n t t o paradigmatic s t r u c t u r e * -75-

is

functionally

The

question

then

to

ask i s whether having a s i n g l e l e v e l a f

r e p r e s e n t a t i o n o f c o n c e p t s , the primitives, is t h e most beneficial conceptual processing. Firstly Metalingual thematically

for

I would claim n o t f o r two r e a s o n s *

b t h e r e is t presence o f thematic s t r u c t u r e i n d i s c o u r s e
organization deacribed enables the content

of

text

to

be

a many l e v e l s of p r e c i s e n e s s e t

It i s p o s s i b l e

t o b e quite s u p e r f i c i a l , o r by decomposition t o become detailed, or vice-versa using a b s t r a c t i o n .

more

and

more

The d e p t h o f analysis can a

given links

b e determined by

the

requirements or

of

understanding until

text:
are

essentially

abstracting

decomposing

causal

e s t a b l i s h e d over the t e x t .

It 3 s not apparent

that

definitions

of

themes

can

be

c o n t r o l l e d i n Conceptual Dependency Theory, i n t h a t if abstract

term

s t a t e d i n terms o f primitives, each

tremelv l a r g e .
bf l e s s e r

could

become

ex-

In c o n t r a s t , t h e encylopedia can d e f i n e themes i n terms

themes.

Secondly paradigmatic s t r u c t u r e e n a b l e s comparatively small chunks

of
knowledge, say involving a s i n g l e causal r e l a t i o n , t o b e r e t r i e v e d complete the underlying form
of

and pieced t o g e t h e r t o

discourse.

Rather

then

a t t e m p t i n g t o p a t c h g e n e r a l knowledge, Schank and Abelson

(1975 ) have introduced "Scripts",

Wose

large preformed knowledge

structures

function

is

t o l i m i t the p o s s i b l e i n f e r e n c e s i n understanding*

This has a danger o f e s s e n t i a l l y

consequent difficulty in

idiomatizing understanding,
deviant s i t u a t i o n s

with

handling

And a s Wilks

( 1 9 7 6 ) p o i n t s o u t , one problem w i t h Scripte i s that they are invoked $n

t h e i r e n t i a e t y by word association.

Thus i t

is

suggested

that,

for

example,

"I bought

some beer

from the supermarket, drove home, and

would
evoke

drank i t w h i l e watching a football game on t e l e v i s i o n "

multitude "drive", of

Scripts by the presence of s u c h ~ r d as "supermarket", s

etc

"football",

Heace

the

desired

reduction

of

possible

inerencea i s n o t achieved.
tion

Paradigmatic o r g a n i z a t i o n e n a b l e s recogni-

of

higher level s t r u c t u r e s , i n c l u d i n g p r o p o s i t t a n s t h a t a r e p a r t
P a r t i a l l y r e c o g n ~ z e da b s t r a c t terms

of m e t a l i n g u a l l y d e f i n e d concepts.

cgn b e used t~ predict t h e i r completion.

The

encyclopedia

thus h a s

general, p r o d u c t i v e bottom-up and top-down c a p a b i l i t i e s .

Even

though

an

abstract

definition

should b e a c t i v a t e d by t h e
will
not

appearance of an appropriate word in the t e x t , t h e structure

in

general

be

large,

and

so

n o t produce an overhelming number of

e x t r a n e o u s a c t i v e nodes

.
Searches can b e I n i t i a t e d from

O the o t h e r hand i t i s c e r t a i n l y advantageous to have a m u l t i t u d e n

of o v e r t themes i n a n a l y s i n g d i s c o u r s e

these terms as w e l l as from t h e more s p e c i f i c

discourse

propositions.

To i l l u s t r a t e t h i s , c o n s i d e r an e x h a u s t i v e undirected search f o r a g o a l

states d i s t a h t

in

space

where each s t a t e is linked t o m o t h e r m**n.

If

states.

The number o f nodes a c t i v a t e d w i l l b e o f t h e o r d e r

the

goal is known then a b i d i r e c t i o n a l search w i l l reduce t h e exponent

But a laore s i g n i f i c a n t r e d u c t i o n

t o n/2.

takes place

if

there are

stated

intermediate

s u b g o a l s , ie, ..

themes i n t h e hierarchy, s a y g af

them, w i t h a n average s e p a r a t i o n of n / g , t h e exponent becomes n/(2g)

ACKNOWLCDGEMENTS

would

like

to

thank David Hays f o r h i s c o n t i n u a l s u p p o r t and

Ray Bennett and Randy Walser helped

guidance throughout t h i s work. trying mine. to make

in

this

discourse

c o h e r e n t , t h e f a u l t s t h a t remain a r e

REFERENCES

Abelson,

R .

P .

The

structure of b e l i e f systems.
Models - Thought of a n

In R C. Schank & .

K M Colby (Eds-) . .

, Computer

Language.

San

Francisco: W * H Freeman, 19/3. . Abelson,

R Po Concepts f o r r e p r e s e n t i n g mundane r e a l i t y i.n plans. .

Representati.on and
h -

In

D O G. Bobrow & A. C o l l i n s (Eds.),

S t u d i o s i n C o g n l u v e Science.

Understanding.

New York:

Academic P r e s s , 1975. Covert c a t e g o r i e s i n f o l k

B e r l i n , Ba, Breedlove, DO E m , & Raven, P W. . taxonomies. Black,

American A n t h r ~ p o l o g i s, 1968, 70, 290-299. t

F1 A d e d u c t i v e question-answering system.

I n Me Minsky (Ed .) ,

Semantic I n f o r m a t i o n Processinq. Bubrow, D G O .

Cambridge: M I T P r e s s , 1968.
a
computer problem-solving Processin&.

N a t u r a l language i n p u t f o r

system.

In

M .

Minsky

(Ed .), Semantic

Information

Cambridge: MIT P r e s s , 1968.

Ch~msky, N.
1965.

Aspec tg of a Theory of

--

Syntax.

Cambridge:

MIT

Press,

Colby,

K O M.

S i m u l a t i o n of b e l i e f systems. Computer Models of Thou-

In R.

C. Schank &

K. M.
Ran-

Colby (Eds.),

- Language. and
-

San

cicrcd: W e H Freeman, 1973. .

Collins,
A*,
&

Quillian,

MI R e

How t o make a language user.

Memory.

In E .
York:

Tulving & W. Donaldson ( E d s . ) , Organization of

Academic Press, 1972. Conklin,

New

H .

C .

Lexigraphical treatment
0)

06 folk

taxonomies.

I n F. W. (Publica-

Householder & S. Saporta ( E d s

t i o n No. 21)

, Problems - Lexicography in

Blooming ton: I n d i a n a Research Center i n Anthropology,

F o l k l o r e , and L i n g u i s t i c s , 1962.

Erdman,

L.

D.

Overview of the Hearsay s p e e c h understanding research

--

(Computer S c i e n c e s Review 1974-1 975) U n i v e r s i t y , 1975. Fillmore,

Pittsburgh :

Carnegie-Mellon

C.

J o

Toward
(Eds.),

a modern t h e o r y of case.
Nodern

I n D A. Reibel .
Readings

&

S. A.

Schane

Studies

in
of

English:

in
,

Transf o n n a t i o n a l

Grammar.

Englewood

Cliffs

NJ:

Prentice-Ha11

1969.
Furugari, T .

- memory A

model and simulation

memory processes

for

d r i v i n g a ear (Technical Report No. 77).

-M.

Buffalo:

State University

of New York, Department of Computer S c i e n c e , 1975.

Hallfday,

A. KO, & Hasan, R.

Cohesion i n E n g l i s h .

Lopdon:

Long-

man, 1975.

-80-

Hays.
In

D o

G.

L i n g u i s t i c foundations f o r a theory of

content

analysis.

GO Gertner,
J o

B .

0 .

Ro

H o l s t i , K k i p p e n d o f f , W. 30 P a i a e l y , & .

P.

Stone (Eds.),

The
(a)

Analysis

of

Communication

Content.

New

York: Wiley, 1969.

Nays,

Do

Go

Applied

Computational

Linguistics. Linguistics.

In GI E Perren & .

J L Trim (Eds .) A p p l i c a t i o n s of . . ,
bridge U n i v e r s i t y Press, 1969.
Haya,
Do

(b)

Cambridge:

Cam-

GO

Linguistic

problems

of d e n o t a t i o n .

In M. BForwischC
The Hague:
Mouton,

K Ec Heidolph (Eds.), Progreps - Linguistics. . in

19700 Hays, D. C .

V p e s of processes on ~ o g n i t i v e networks

Paper presented
Pisa

a t the I n t e r n a t i o n a l Conference on Computational Linguistics,

I t a l y , 1973.
Ilays, D. G O Cognitive S t r u c t u r e s .

Book i n p r e p a r a t i o n , 1978.
u t i l i t y of semantic networks throwh

Hendrix,

C .

C.

Expanding

the

partitioning. Hopcroft, J. E, .
to - Automata.
Inheldet,
Be,

Menlo Park, CA: Stanford Research Institute, 1975.

&

Ullman, J. DO Fofmal Ladguages and

the - -i r

Relation

Reading, MA: Addlson-wesley, 1YbY.

&
We

Piaget,

J .

The of i n the - Earlv Growth - Logic -- c h i l d .

New York: W. Uein,

S o ,

Norton, 1964.

Oakley,
for

J.

Da,

Suurballe,
reports
on

Do

J, .
the

&

Ziesemer,

R e

A.

A program

generating

status

and h i s t o r y of

s t o c h a s t i c a l l y m o d i f i a b l e semantic models of

arbitrary

universes

Statistical Methods in Linguistics, 1972, 8 , 64-93. -81-

Kuhn, T S. .

- Structure - S c i e n t i f id R e v o l u t i o n s . The of
Structural complexity i n f a i r y tales.

Chicago: Univer-

sity of Chicago Press, 1962.

Lakoff , G.

- S t u d y - -sMsn The of
In D. A,

1972, 1, 128-150.

Langacker, R .

On p r o n o m i n a l i z a t i o n and t h e cha'ln o f command.

Reibel & S. A. Schane (Eds.),

Transformational

Elolern S t u d i e s i n E n g l i s h : Readin=
Englewood Clifts,

in

Grammar.

N3:

Prentice-Hall,

1969.
Linde, C .
- . The

l i n g u i s t i c e n c o d i n g of s p a t i a l i n f o r m a t i o n .

Unpublished

doctoral b i s s e r t a t i o n , Columbia U n i v e r s i t y , 1974.

Longacre,

R .

D i s c o u r s e , Paragraph, and S e n t e n c e S t r u c t u r e _in S e l e c t e d Languages San ta

Philippine

Ana,

CA:

Summer

Institute

of

L i n g u i s t i c s , 1968. Longacre,

Re

Hierarchy and U n i v e r s a l i t y of D i s c o u r s e C o n s t i t u e n t s i n

New - Guinea

Languages.

Washington,

D. C. :

Georgetown

University

Press, 1972.
Eiabley, E .

Drarilatic C o n s t r u c t i o n . Johnson, N. S.,

P h i l a d e l p h i a : C h i l t o n , 1972.
& DeForest

Flandler , J q M. .

, M. - s t r u c t u r a l A
2
time"

analysis
"Happily

From of and - s t o r i e s - their recall: - "Once upon

to -

ever -

after" ( T e c h n i c a l Report No. 57).

La Jolla:

C e n t e r f o r Human

I n f o r m a t i o n P r o c e s s i n g , U n i v e r s i t y of C a l i f o r n i a a t San Diego, 1976. Nagao, Ma, & T s u j i i , J.

Analysis

of

Japanese

sentences.

American

Journal of Computational Linguistics, 1976, M i c r o f i c h e 41.

Nash-Webber

, IS: L.

NO. 3335)
P h i l l i p s , B.

Semantic interpretation r e v t s i t e d (Technical Report

Cambridge: Bolt, Beranek, & Neman, 1976.

Topic analysis.

Unpublished doc tore1 d i s s e r t a t i o n , S t a t e

Universit,y of Nw Yqrk a t Buffalo, 1975. e

Propp, V,

Morphology of t h e F o l k t a l e .

--

Austin: The University of Texas

Press, 1968.
Putnam,
N o

The meaning of "meaningw*

Mind

Language,

Philosophica< Papers

(Vol

- Reality< and
1969,

2).

Cambridge:

Cambridge University

P r e s s , 1975.

Q u i l l i a n , M. R.

The teachable language comprehender.

-Cornme ACM,
M o

12, 459-475.

Raphael, B .

SIR: Semantic information r e t r i e v a l .

In

Minsky

(Ed .) ,

Semantic Information Processinq.

Cambridge: MIT P r e s s , 1968.

Raven,

P o H e , Berlin,

Bo,

& Breedlwe,

Dm E .

The origins of taxonomy.

Science, 1971, 174, 1210-1213.

Rumelhart, D Em, Lindsay, P. H e , & Norman, .

Do

A.

A process model

for

long term memory.

and - Memory. Rumelhart,

In E Tulviog 8 Yo Donaldson ( E d s . ) .

, Organization

New York: Academic Press, 1972.

E, .
8

D .

Ortony,

A.

- r e p r e s e n t a t i o n - knowledge i n The of
stories
(Tech-

memory (Technical Report No. 55).

La 3olla:'Center f o r Human Infor-

mation Processing, University of C a l i f o r n i a a t San Diego, 1976.

Rumelhart,

D o

E .

Understandinq and summarizing brief

f l i c a l Report

No.

58).

La Jolla:

Center

f o r H m n Information u a

Processink, University of California a t San Diego, 1976. -83-

Schank,

R.

C.

Conceptual I n f o r m a t i o n P r o c e s s i n q .
(a)

Amsterdam:

North-

Holland, 1975.

Schank, R C. .
A.

The structure of episodes in memory.

(Eds.), Repreeentation
New

In

I. )

G. Bobmw &

Collins

and

Understandinq: Studies (b)

&

Cognitive Science.

Y ~ r k : Academic P r e s s , 1975.

Schank, R e C, .

h Abelson, R P . .

Scrlpts, plan?, and k n c w l - d ~ . e

Paper

p r e s e n t e d a t t h e F o u r t h I n t e r n a t i o n a l . J o i n t Conference on 4r t i f ic ial I n t e l l i g e n c e , T b i l i s i , USSR, 1975. Schubert,

1. ;

K.

Extending the e x p r e s s i v e power o f s e m a n t i c networks.

A r t i f i c i a l I n t e l l i g e n c e , 1976 Shapiro, S .

7,

163-198.
A data --

C .

The MIND -processing

system:

structure

for

semantic

information

(Report R-83 7-PR )

.
for

San t a ilonica, CA:

RAND

C o r p o r a t i o n , 1971. Simmons, R e F .

-Some Spatial

semantic

structures

meanings

( T e c h n i c a l Report NL-1)

representing

En~lish
Instruc-

Austin : Computer-Aided

t i o n Ltrboratory, The U n i v e r s i t y o f Texas, 1970.

Sondheimer , N. K .

reference

and

semantic

nets.

Americlan

~ o u r n a io f C o ~ n p u t a t i o n a lL i n g u i s t i c s , 1977, Microfiche 71. Thomnclyke,

P .

W.

The r o l e of i n f e r e n c e s i n d i s c o u r s e comprehension.

Journal of V e r b a l Learninq and Verbal Behavior, 1976, 15, 4 3 7 4 4 6 .

of

Tulving, E . son
1972.

E p i s o d i c and semantic memory. Memory.

I n E Tulving & W . .

Donald-

(Eds .) O r g a n i z a t i o n ,

New York:

Academic Press,

White,

M .

Abstract

definition

,& the

cognitive

network:

The

metaphysical

terminology

- .a.. contemporary m i l l e n a r i a n community. of

Yark a t

Unpublished d o c t o r a l d i s s e r t a t i o n , State U n i v e r s i t y of New

Buffalo, 1975.
Wilks,

Y.

Grammar,

Meaning 9.- and

- Machine Analysis - Language the of

semantics

London: Routledge, 1972. Wilks

Ye

prefential,

pattern-seeking,

for

Natural

Language inference. Wilks

the

Artificial I n t e l l i g e n c e , 1975, - 53-74. 6,

Yb

Frames, s c r i p t s , s t o r i e s , and f a n t a s i e s .

Paper presented a t

International

Canada, 1976
Winogfad, T .
proaram

Conference on Computational L i n g u i s t i c s , Ottawa,

Procedures as a r e p r e s e n t a t i o n for
under standdnq

--

---data

in

computer:

for

natural

language

(Repart

MAC-TR-84)

.
In

Cambridge: A r t i f i c i a l I n t e l l i g e n c e Laboratory, MIT, 1971.

Woods, W A. Y

What s in a link: Foundations f o r semantic networks.

D m G Bobrow 6 A. C o l l i n s (Eds.) .
Studdes i n Cognitive Science.

, Representation

and - Understanding:

New York: Academic P r e s s , 1975.