speech disorder in schizophrenia: review of the literature and

Speech Disorder in Schizophrenia: Review ofthe Literature and Exploration of Its Relation to

the Uniquely Human Capacity for Language

by Lynn E. DeLisi

Abstract

The language capacity of modern humans is thoughtby some to be clearly distinct from that of nonhumanprimates (Bickerton 1990). Crow (1997, 1998a) hasproposed that a disturbance in the uniquely humanaspects of language is central to the genetic etiology ofschizophrenia. A review of the literature on languagedisorder in schizophrenia provides evidence for wide-spread deficits in comprehension, production, atten-tion, and cerebral lateralization of language. Wefocused here on those anomalies that are uniquelyhuman aspects of language. Bickerton's five distinctlyhuman language devices were examined in patientswith schizophrenia and their families by using a struc-tured scoring format on oral soliloquies. The chronicpatients showed reduced use of clausal embedding andused fewer words than first episode patients or wellfamily members. The amount of sentence complexitywas found to be familial and to cosegregate with schiz-ophrenia within families. These data are consistentwith previous literature and additionally show a famil-ial component to these measures, thus suggesting thatdeficits in specifically human aspects of language maybe related to the genetics of schizophrenia.

Keywords: Language, speech, evolution, genetics,families, relatives, schizophrenia.

Schizophrenia Bulletin, 27(3):481-496,2001.

There has been much debate in linguistic literature aboutwhat constitutes specifically human aspects of language(reviewed in Hauser 1997) and whether there was anabrupt or gradual evolutionary transition in language abil-ities from nonhuman primates (early hominids) to Homosapiens (Gardner and Gardner 1978; Bickerton 1990,1995). Bickerton (1990) proposed five unique features ofhuman language: (1) differences in the superficial order ofconstituents; (2) the use of null elements (e.g., "it"); (3)the use of a subcategorized argument structure for verbs;

(4) mechanisms for the expansion of utterances; and (5)the use of grammatical items. These aspects of languagemust result from newly formed neuronal organization inthe transition from early hominids to modern Homo sapi-ens. Genetic control of the development of these neuronalcircuits responsible for language is thus likely. However,the genes responsible and the pathways that are specifi-cally human are unknown.

The core phenomenologic characteristics of schizo-phrenia can be considered to stem from anomalous orga-nization, retrieval, and communication of complex"thought" and language. Thus, it follows that schizophre-nia is likely to be a uniquely human condition and thatsome deficits in the neural organization of language mustexist in schizophrenia. Numerous aspects of both speechand its perception have been found to be deviant in stud-ies of patients with schizophrenia (table 1). Crow has sug-gested that language deviance is the basis for the develop-ment of the nuclear symptoms of schizophrenia (Crow1998a) and that "Schizophrenia is the price Homo sapienspays for language" (Crow 1997). He further states that thedevelopment of cerebral structural asymmetries duringhominid evolution is responsible for the uniquely humancomponents of language (as discussed by Geschwind andGalaburda 1987 and Corballis 1991) and that these asym-metries are anomalous in schizophrenia (Crow 1998b). Insupport of this hypothesis, we previously have shown thatreduced cerebral asymmetries are present early in thecourse of schizophrenia and that some components ofasymmetry are inherited.

Other investigators ascribe a less central role to lan-guage and have focused on alternative explanations forlanguage deviance in schizophrenia, such as dysfunctionalexecutive control (frontal lobe; Chaika 1990; Morice1995) or a deficit in either working or semantic memory,

Send reprint requests to Dr. L.E. DeLisi, Professor of Psychiatry, NewYork University School of Medicine, Millhauser Laboratories, 550 FirstAvenue, New York, NY 10016; e-mail: [email protected].

481

Dow

nloaded from https://academ

ic.oup.com/schizophreniabulletin/article/27/3/481/1835137 by guest on 27 D

ecember 2021

Tabl

e 1

. S

elec

tive

rev

iew

of

liter

atu

re o

n th

e la

ng

uag

e of

pat

ien

ts w

ith

sch

izo

ph

ren

ia1

Stu

dy

Su

bje

cts

Fu

nct

ion

tes

ted

Fin

din

gs

i O

I. S

pee

ch c

om

pre

hen

sio

n

Con

dray

at a

l.19

96

Con

dray

et a

l.19

92,1

995

Gro

ve a

ndA

ndre

asen

1985

Fab

er e

t al.

1983

Fab

er a

ndR

eich

stei

n19

81

Mill

er a

ndP

hela

n 19

80

Che

adle

and

Mor

gan

1972

25 C

Sz

mal

es11

con

trol

mal

es

36 C

Sz

mal

es41

bro

ther

s18

con

trol

mal

es

100

Sz

subj

ects

50 c

ontr

ols

14 S

z su

bjec

tsw

ith T

D (

13 m

ale)

13 a

phas

ia s

ubje

cts

24 S

z su

bjec

ts(5

man

ia, 5

dep

r)28

con

trol

s

25 C

Sz

subj

ects

25 c

ontr

ols

65 S

z su

bjec

ts25

oth

er p

ts57

chi

ldre

n 8-

11 y

rs

Com

preh

ensi

on o

f co

mpl

ex v

s. s

impl

e se

nten

ces

Com

preh

ensi

on;

Luria

-Neb

rask

a R

elat

iona

lC

once

pts

Sca

le

Syn

tax

proc

essi

ng a

nd p

erce

ptio

n; "

embe

dded

clic

k" a

nd "

mem

ory

for

gist

task

s" te

sts

Tra

nscr

ibed

inte

rvie

ws

subj

ectiv

ely

rate

d

Aph

asia

exa

m; n

amin

g te

sts;

wor

d re

petit

ion;

audi

tory

com

preh

ensi

on

Com

preh

ensi

on o

f sen

tenc

e ac

cept

abili

ty

Und

erst

andi

ng o

f com

mon

Eng

lish

expr

essi

ons

Dec

reas

ed c

ompr

ehen

sion

ass

ocia

ted

with

dec

reas

edw

orki

ng m

emor

y lo

ad;

no d

iffer

ence

in

abili

ty t

o pe

rcei

vein

divi

dual

wor

ds

Mor

e er

rors

in S

z an

d S

z sp

ectr

um s

ubje

cts

than

inno

rmal

bro

ther

s or

con

trol

s; n

ot r

elat

ed to

med

icat

ion

Sz

subj

ects

did

wor

se o

n m

emor

y th

an o

n se

nten

cepr

oces

sing

; con

clud

es t

hat

they

hav

e a

mem

ory

defic

it,no

t a

perc

eptio

n de

ficit.

Sz

pts

show

ed fe

wer

wor

d-fin

ding

pro

blem

s, b

ette

rco

mpr

ehen

sion

, and

less

pov

erty

of

spee

ch, b

ut m

ore

wor

d ap

prox

imat

ions

; som

e ch

arac

teris

tics

over

lapp

ed.

Sz

subj

ects

with

TD

per

form

ed w

orse

on

all t

ests

;la

ngua

ge c

ompr

ehen

sion

and

rep

etiti

on d

ysfu

nctio

n in

Sz

subj

ects

com

pare

d to

oth

ers.

No

diffe

renc

e be

twee

n gr

oups

in a

bilit

y to

det

erm

ine

acce

ptab

ility

of

sent

ence

rul

es

Sz

subj

ects

faile

d to

und

erst

and

appr

oxim

atel

y 1/

3 of

all

ques

tions

and

wer

e cl

ose

to th

e le

vel o

f 8-

9-yr

-old

child

ren.

II.

Att

enti

on

/inte

rfer

ence

Land

re a

ndT

aylo

r 19

95

Spi

tzer

et a

l.19

94

Kw

apil

et a

l.19

90

Gra

nd e

t al.

1975

21 S

z su

bjec

ts (

TD

)16

Sz

subj

ects

(no

TD

)

70 S

z su

bjec

ts44

con

trol

s

21 S

z su

bjec

ts18

BP

sub

ject

s21

con

trol

s

10 C

Sz

subj

ects

,al

l mal

e

Por

ch in

dex

of c

omm

unic

ativ

e ab

ility

and

oth

erfa

ctor

s

Sem

antic

and

pho

nolo

gica

l pr

imin

g

Wor

d re

cogn

ition

mea

sure

of

sem

antic

prim

ing

Str

oop

colo

r w

ord

inte

rfer

ence

tas

k fo

llow

ed b

yre

cord

ed u

nstr

uctu

red

inte

rvie

w; 4

dim

ensi

ons

ofla

ngua

ge s

core

d; #

of c

laus

es a

nd g

ram

mat

ical

lyin

com

plet

e ve

rbal

uni

ts s

core

d

Sz

subj

ects

with

TD

wer

e m

ore

impa

ired

than

tho

sew

ithou

t; re

duce

d at

tent

ion

was

bes

t pr

edic

tor

ofla

ngua

ge d

ysfu

nctio

n.

Sz

subj

ects

with

TD

exh

ibite

d m

ore

intr

usio

nfr

om th

e pr

imin

g.

Sz

subj

ects

exc

eede

d bo

th n

orm

al c

ontr

ols

and

BP

sub

ject

son

faci

litat

ion

(e.g

., se

man

tic p

rimin

g he

ight

ened

);B

P s

ubje

cts

wer

e no

t di

ffere

nt f

rom

con

trol

s.

Spe

cific

nar

rativ

e la

ngua

ge d

efic

its w

ere

asso

ciat

ed w

ith in

ter-

fere

nce

but

not w

ith w

ord-

nam

ing

or c

olor

-enc

odin

g sk

ills.

r tn a a

Dow



ecember 2021

Stu

dy

Su

bje

cts

Fu

nct

ion

tes

ted

Fin

din

gs

on n 3.

III.

Sem

anti

cs

Kar

eken

et

al19

96

Gol

dfar

b et

al.

1994

Cut

ting

and

Mur

phy

1990

Hof

fman

et a

l.19

85

Man

schr

eck

eta

l. 19

80

Will

iam

s et

al.

1976

Pav

y et

al.

1969

29 S

z su

bjec

ts20

con

trol

s

29 le

ft he

mis

pher

e st

roke

and

tum

or p

ts26

CS

z su

bjec

ts

20 A

c S

z su

bjec

ts20

man

ia s

ubje

cts

20 d

epr

subj

ects

24 S

z su

bjec

ts24

non

-Sz

psyc

hiat

ric p

ts

10 S

z su

bjec

ts10

con

trol

s

40 S

z su

bjec

ts20

con

trol

s

6 C

Sz

subj

ects

9 A

c S

z su

bjec

ts

Cal

iforn

ia V

erba

l Lea

rnin

g T

est t

o ex

amin

e pr

o-ac

tive

inhi

bitio

n an

d se

man

tic p

roce

ssin

g

Sem

antic

ass

ocia

tions

Sem

antic

s; d

enot

ativ

e vs

. con

nota

tive

asso

ciat

ions

of

wor

ds

Hun

t Tes

t re

quiri

ng u

se o

f sy

ntac

tic s

trat

egie

s to

synt

hesi

ze a

n ex

tend

ed t

ext w

ith a

con

trol

led

set

of i

nput

s

Lang

uage

sam

ples

in r

espo

nse

to 2

pic

ture

s,sp

oken

and

writ

ten;

clo

ze p

roce

dure

use

d(d

elet

ing

ever

y 4t

h or

5th

wor

d);

norm

alvo

lunt

eers

wer

e as

ked

to fi

ll in

mis

sing

wor

d.C

loze

sco

re =

% o

f w

ords

gue

ssed

rig

ht.

Sem

antic

res

pons

e te

sts

100

wor

ds fr

om w

ritte

n di

arie

s an

alyz

ed fo

r th

ety

pe-t

oken

rat

io

Sz

subj

ects

hav

e si

gnifi

cant

ly l

ess

proa

ctiv

e in

hibi

tion

and

did

not

orga

nize

rec

all a

ccor

ding

to

sem

antic

cate

gory

; pt

s ha

d m

ore

phon

emic

err

ors

than

con

trol

s;re

sults

sug

gest

tha

t re

duce

d se

man

tic p

roce

ssin

gpr

even

ted

proa

ctiv

e in

hibi

tion.

Err

ors

show

sim

ilarit

ies

betw

een

Sz

subj

ects

and

apha

sia

subj

ects

.

Sz

subj

ects

had

bia

s to

war

d de

nota

tive

asso

ciat

ions

.

Mor

e m

eani

ng m

isre

pres

enta

tions

of

inpu

ts a

mon

gS

z su

bjec

ts

Writ

ten

sam

ples

wer

e no

t di

ffere

nt b

etw

een

grou

ps; o

ral

sam

ples

dis

tingu

ishe

d S

z su

bjec

ts fr

om c

ontr

ols

for

the

5th

dele

tion

proc

edur

e an

d co

rrel

ated

with

clin

ical

for

mal

TD

.

Sz

subj

ects

diff

ered

; chr

onic

sub

ject

s pe

rfor

med

wor

seth

an A

c su

bjec

ts.

Chr

onic

ity a

ssoc

iate

d w

ith r

educ

tion

in th

e to

tal #

of

diffe

rent

wor

ds u

sed

IV.

Flu

ency

a- i I I to o o o

Alo

ia e

t al.

1996

Joyc

e et

al.

1996

Alle

n et

al.

1993

Cha

ika

1990

;C

haik

a an

dLa

mbe

198

9

28 C

Sz

subj

ects

32 c

ontr

ols

50 S

z su

bjec

ts25

con

trol

s

20 C

Sz

subj

ects

10 c

ontr

ols

14 S

z su

bjec

ts8

man

ia s

ubje

cts

Spo

ntan

eous

ly c

lust

er e

xem

plar

s fr

om a

spe

cific

cate

gory

dur

ing

a flu

ency

tas

k

Ver

bal f

luen

cy,

lette

r, a

nd c

ateg

ory

Ver

bal f

luen

cy

Spe

ech

in r

espo

nse

to v

iew

ing

a vi

deo

stor

y;co

hesi

ve t

ies

coun

ted

Sz

subj

ects

sho

wed

less

sta

ble

2-di

men

sion

alor

gani

zatio

n of

exe

mpl

ars

into

sub

ordi

nate

clu

ster

s th

anco

ntro

ls, s

ugge

stin

g di

sorg

aniz

ed s

eman

tic n

etw

orks

.

Cat

egor

y w

as s

uper

ior

to le

tter

fluen

cy i

n S

z su

bjec

tsan

d ca

tego

ry i

mpr

oved

with

cue

s; r

esul

ts s

ugge

stin

effic

ient

acc

ess

to s

eman

tic s

tore

s.

Few

er w

ords

; re

duce

d ab

ility

to

retr

ieve

wor

ds,

lexi

con

inta

ct

No

diffe

renc

es b

etw

een

grou

ps; S

z su

bjec

ts m

ore

likel

yto

use

nov

el r

efer

ence

s an

d di

gres

sion

Dow



ecember 2021

Tabl

e 1

. S

elec

tive

rev

iew

of

liter

atu

re o

n t

he

lan

gu

age

of p

atie

nts

wit

h s

chiz

op

hre

nia

1—C

on

tin

ued

Stu

dy

Su

bje

cts

Fu

nct

ion

tes

ted

Fin

din

gs

I a- i I- CO I p

Rag

in a

ndO

ltman

ns19

86

Alle

n an

d A

llen

1985

Wyk

es a

ndLe

ff 19

82

Rau

sch

et a

l.19

80

Roc

hest

er a

ndM

artin

197

9;R

oche

ster

1978

Rut

ter1

979

10

Sz

subj

ects

11 m

ania

sub

ject

s11

sch

izoa

ffect

ive

subj

ects

10 c

ontr

ols

10 S

z su

bjec

ts w

ith p

ossy

mpt

oms

9 S

z su

bjec

ts w

ith n

egsy

mpt

oms

9 co

ntro

ls

8 S

z su

bjec

ts4

man

ia s

ubje

cts

10 S

z "p

roce

ss"

subj

ects

10 S

z "r

eact

ive"

sub

ject

s10

aph

asia

sub

ject

s10

con

trol

s

40

Sz

subj

ects

20 c

ontr

ols

10 S

z su

bjec

ts10

con

trol

s

Inte

rvie

w s

peec

h sa

mpl

es; l

exic

al t

ies

per

15cl

ause

s w

ere

coun

ted

betw

een

and

with

incl

ause

s.

Spe

ech

prod

uced

in r

espo

nse

to p

ictu

res

Sie

gel e

t al.

1976

15

CS

z su

bjec

ts

Mea

sure

men

t of

the

# of

str

uctu

ral l

ink

Wor

d-or

derin

g ta

sks

Spe

ech

sam

ples

; bet

wee

n-cl

ause

lex

ical

tie

s

Uns

truc

ture

d sp

eech

sam

ples

; ps

ycho

logy

stud

ents

ask

ed to

firs

t pl

ace

logi

cal p

unct

uatio

nin

eac

h tr

ansc

ript

and

then

rec

onst

ruct

the

tran

scrip

tions

fro

m c

ards

of

cuto

ut s

ente

nces

from

eac

h pa

ssag

e.

Fre

e sp

eech

No

diffe

renc

es i

n be

twee

n-cl

ause

lex

ical

coh

esio

n; S

zsu

bjec

ts s

how

ed fe

wer

with

in-c

laus

e co

hesi

ons

than

the

othe

r pt

s or

con

trol

s.

Org

aniz

atio

n w

ithin

and

bet

wee

n id

eas

was

not

diff

eren

t;ho

wev

er,

pos

spee

ch-d

isor

dere

d S

z su

bjec

ts o

mitt

edre

fere

nts

for

noun

phr

ases

.

Man

ia s

ubje

cts

had

mor

e lin

ks t

han

Sz

subj

ects

.

Aph

asia

sub

ject

s ha

d si

gnifi

cant

ly l

onge

r tim

e an

d er

rors

in ta

sk t

han

othe

rs;

Sz

subj

ects

diff

ered

fro

m n

orm

als

only

in #

of

erro

rs r

earr

angi

ng s

ente

nces

inv

olvi

ngob

ject

s.

Red

uced

in S

z pt

s

Pas

sage

s fr

om S

z su

bjec

ts w

ere

sign

ifica

ntly

les

sco

nnec

ted

than

con

trol

s, b

ut th

e se

nten

ces

them

selv

esw

ere

logi

cally

ord

ered

.

Pts

with

LT

hos

pita

lizat

ion

had

sign

ifica

ntly

gre

ater

pau

city

of

spee

ch, p

erse

vera

tion,

repe

titio

n, a

nd d

evia

nt v

erba

lizat

ions

.

V.

Co

mp

lexi

ty

DeL

isi e

t al.

1997

Tho

mas

et a

l.19

96

Tho

mas

et a

l.19

93

41 C

Sz

subj

ects

26 c

ontr

ols

Firs

t-on

set

pts,

45 w

ith S

z, 1

4 w

ithm

ania

18 c

ontr

ols

Firs

t-on

set

pts,

38 w

ith S

z, 1

1 w

ithm

ania

16 c

ontr

ols

Sen

tenc

e co

mpl

exity

fro

m s

olilo

quie

s

Writ

ten

lang

uage

; Hun

t Tes

t m

easu

ring

com

plex

ity,

erro

rs in

syn

tax,

and

sem

antic

s

Ora

l int

ervi

ew t

rans

crip

ts; s

ynta

ctic

ana

lysi

s

Mor

e m

orph

olog

ical

err

ors,

less

deg

ree

of e

mbe

ddin

g

than

con

trol

s

No

diffe

renc

e in

writ

ten

com

plex

ity;

both

pat

ient

gro

ups

mad

e m

ore

erro

rs, w

ith in

crea

sed

com

plex

ity r

elat

ed to

atte

ntio

n an

d w

orki

ng m

emor

y.

Less

sen

tenc

e co

mpl

exity

am

ong

Sz

subj

ects

r m

Dow



ecember 2021

Stu

dy

Su

bje

cts

Fu

nct

ion

tes

ted

Fin

din

gs

00 •o

•fc.

00

Tho

mas

et a

l.19

90

Fra

ser

et a

l.19

86

Hof

fman

et a

l.19

86

Mor

ice

and

McN

icho

l19

85

Mor

ice

and

Ingr

am 1

982

50 A

c S

z su

bjec

ts27

CS

z su

bjec

ts50

con

trol

s

50 S

z su

bjec

ts51

man

ia s

ubje

cts

50 c

ontr

ols

29 S

z su

bjec

ts24

man

ia s

ubje

cts

40 c

ontr

ols

17 S

z su

bjec

ts17

man

ia s

ubje

cts

19 c

ontr

ols

34 S

z su

bjec

ts11

man

ia s

ubje

cts

18 c

ontr

ols

Ora

l int

ervi

ew t

rans

crip

ts

Fre

e sp

eech

sam

ples

; syn

tact

ic c

ompl

exity

(em

bedd

ing)

and

oth

er s

truc

ture

ana

lyze

d on

1,00

0 w

ords

eac

h.

Fre

e sp

eech

sam

ples

;sc

ored

com

plex

ity a

nd le

ngth

of

utte

ranc

es

Tok

en te

st (

mea

sure

s co

mpr

ehen

sion

of

synt

ax)

and

free

spe

ech

sam

ples

of

1,00

0 w

ords

Fre

e sp

eech

sam

ples

of

1,00

0 w

ords

; gr

amm

atic

alan

d sy

ntax

ana

lyse

s; 9

8 va

riabl

es,

refle

ctin

gco

mpl

exity

, in

tegr

ity,

and

fluen

cy o

f spe

ech;

disc

rimin

ant

func

tion

anal

ysis

per

form

ed

Chr

onic

pts

had

less

inte

grity

, co

mpl

exity

, an

d flu

ency

than

Ac

pts,

but

Ac

pts

perf

orm

ed w

orse

tha

n co

ntro

ls.

Dec

reas

ed c

ompl

exity

was

rel

ated

to n

eg s

ympt

oms

and

poor

out

com

e.

Con

tinuu

m o

f lin

guis

tic d

egen

erat

ion

acro

ss t

heps

ycho

tic s

pect

rum

fro

m c

ontr

ols;

incl

udes

red

uced

com

plex

ity

Man

ia s

ubje

cts:

shi

fts f

rom

one

coh

eren

t st

ruct

ure

toan

othe

r; S

z su

bjec

ts: d

efic

ient

abi

lity

to c

onst

ruct

disc

ours

e st

ruct

ure

Tok

en te

st n

ot im

paire

d in

Sz

or m

ania

sub

ject

s; fo

r S

zsu

bjec

ts o

nly,

ther

e w

as a

pos

cor

rela

tion

betw

een

scor

eon

toke

n te

st a

nd d

epth

of c

laus

al e

mbe

ddin

g.

Gro

ups

wer

e di

stin

guis

hed

by t

his

anal

ysis

.

VI.

Ref

eren

ce f

ailu

res

Doc

hert

y an

dH

eb

ert

19

97

Bar

ch a

ndB

eren

baum

1996

Doc

hert

y et

al.

1994

a

Doc

hert

y et

al.

1994

b

Har

vey

and

Ser

per

1990

Har

vey

and

Bra

ult1

98

6

Kan

toro

witz

and

Coh

en 1

977

29 S

z su

bjec

ts

39 C

Sz

subj

ects

No

cont

rols

30 A

c S

z su

bjec

ts

10 s

tabl

e S

z su

bjec

ts18

una

ffect

ed p

aren

ts10

con

trol

s

38 S

z su

bjec

ts30

man

ia s

ubje

cts

25 c

ontr

ols

22 S

z su

bjec

ts21

man

ia s

ubje

cts

30 C

Sz

subj

ects

15 c

ontr

ols

Fre

e sp

eech

sam

ples

ass

esse

d fo

r co

mm

unic

atio

nfa

ilure

s: o

verin

clus

ive

refe

renc

es,

ambi

guou

sw

ord

mea

ning

s, a

mbi

guou

s re

fere

nces

, an

daf

fect

ive

reac

tivity

Ref

eren

ce e

rror

s; d

isco

urse

pla

nnin

g; n

eolo

gism

s;gr

amm

atic

al a

nd p

hono

logi

cal

enco

ding

Fre

e sp

eech

sam

ples

; dev

ianc

e in

lang

uage

scor

ed in

pos

and

neg

affe

ctiv

e co

nditi

ons

Fre

e sp

eech

sam

ples

; fre

quen

cy o

f un

clea

rlin

guis

tic r

efer

ence

s in

pos

and

neg

affe

ctiv

eco

nditi

ons

Mea

sure

d se

rial r

ecal

l and

enc

odin

g, d

istr

actib

ility

,po

s an

d ne

g T

D, a

nd r

efer

ence

fai

lure

s

Sco

ring

of u

ncle

ar o

r am

bigu

ous

refe

renc

es

Str

uctu

red

task

des

crib

ing

colo

red

chip

s so

that

anot

her

pers

on c

ould

iden

tify

them

.

Affe

ctiv

e re

activ

ity a

ssoc

iate

d w

ith th

e po

s sc

hizo

phre

nic

proc

ess.

Ref

eren

ce e

rror

s an

d di

scou

rse

plan

ning

cor

rela

ted;

neol

ogis

ms

and

gram

mat

ical

and

pho

nolo

gica

l enc

odin

gco

rrel

ated

Lang

uage

on

neg

topi

cs c

onta

ined

mor

e di

sord

er t

han

on p

os to

pics

.

Sz

subj

ects

and

par

ents

sco

red

wor

se th

an c

ontr

ols

onpo

s si

tuat

ion;

pts

det

erio

rate

d on

neg

situ

atio

n, p

aren

tsan

d co

ntro

ls d

id n

ot.

Pos

TD

and

ref

eren

ce f

ailu

res

amon

g S

z su

bjec

ts w

ere

best

pre

dict

ed b

y m

easu

res

of d

istr

actib

ility

.

Pre

ssur

e of

spe

ech

acco

unte

d fo

r re

fere

nce

failu

res

inm

ania

, whi

le s

ever

ity o

f de

railm

ent

and

pove

rty

ofsp

eech

wer

e re

late

d to

ref

eren

ce fa

ilure

s in

Sz.

Poo

r co

mm

unic

atio

n ac

cura

cy i

n S

z su

bjec

ts; f

ailu

re t

opr

ovid

e pr

oper

ref

eren

ts

a- D I GO o ! u' S I 3 5' to I a' O

Dow



ecember 2021

Tabl

e 1

. S

elec

tive

rev

iew

of

liter

atu

re o

n t

he

lan

gu

age

of p

atie

nts

wit

h s

chiz

op

hre

nia

1—C

on

tin

ued

00

Stu

dyS

ub

ject

sF

un

ctio

n t

este

dF

ind

ing

s

Sm

ith 1

970

Coh

en a

ndC

amhi

196

7

12 S

z m

ales

with

TD

10 c

ontr

ols

18 s

peak

er-li

sten

er p

airs

(Sz-

cont

rol;

Sz-

Sz;

Con

trol

-con

trol

)

Ref

eren

t ta

sk:

spea

ker

mus

t pr

ovid

e cl

ues

tohe

arer

abo

ut r

efer

ent

wor

d

Ref

eren

t ta

sk:

spea

ker

prov

ides

clu

es t

o lis

tene

rto

dis

tingu

ish

refe

rent

fro

m n

onre

fere

nt i

tem

s

Sz

subj

ects

per

form

ed s

igni

fican

tly w

orse

tha

n co

ntro

ls.

Sz

spea

kers

per

form

ed s

igni

fican

tly w

orse

tha

n co

ntro

ls,

whi

le S

z lis

tene

rs d

id n

ot.

VII.

G

ram

mar

Sie

gel e

t al.

1976

VIII

. L

ang

uag

e

15 C

Sz

subj

ects

late

ral i

zati

on

Fre

e sp

eech

Pts

with

LT

hos

pita

lizat

ion

had

sign

ifica

ntly

gre

ater

pauc

ity o

f sp

eech

, per

seve

ratio

n, r

epet

ition

, and

devi

ant

verb

aliz

atio

n.

Wei

sbro

d et

al.

1998

Man

oach

199

4

24 S

z su

bjec

ts (

no T

D)

16 S

z su

bjec

ts (

with

TD

)38

con

trol

s

29 S

z su

bjec

ts42

BP

sub

ject

s25

con

trol

s

Sem

antic

ass

ocia

tions

and

prim

ing

both

hem

isph

eres

sep

arat

ely

Deg

ree

of m

anua

l sup

erio

rity

asse

ssed

Pts

with

TD

sho

wed

indi

rect

sem

antic

prim

ing

in le

fthe

mis

pher

e, e

nhan

ced

spre

ad o

f se

man

tic a

ssoc

iatio

nsin

TD

, and

dis

orga

niza

tion

of t

he f

unct

iona

l as

ymm

etry

of s

eman

tic p

roce

ssin

g.

Aty

pica

l han

dedn

ess

was

ass

ocia

ted

with

lan

guag

edy

sfun

ctio

n in

Sz.

IX.

Mis

cella

neo

us

Bar

r et

al.

1989

Kin

ney

et a

l.19

85

Hal

pern

and

McC

artin

-C

lark

19

84

Sie

gel e

t al.

1976

Got

tsch

alk

etal

. 19

61

15 C

Sz

subj

ects

No

cont

rols

Ado

ptee

s20

Sz

subj

ects

26 c

ontr

ol

61 C

Sz

subj

ects

61 a

phas

ia s

ubje

cts

15 C

Sz

subj

ects

152

CS

z su

bjec

ts28

mal

es12

4 fe

mal

es

Ver

bal

pers

ever

atio

ns

Sol

iloqu

ies;

sco

res

on s

ocia

l alie

natio

n an

dpe

rson

al d

isor

gani

zatio

n sc

ales

Tes

ted

for

impa

irmen

t in

10

lang

uage

cat

egor

ies

Fre

e sp

eech

Ver

bal s

ampl

es r

ecor

ded

and

tran

scrib

ed;

diso

rgan

izat

ion

and

soci

al a

liena

tion

scor

ed.

Hig

her

verb

al p

erse

vera

tive

erro

rs t

han

publ

ishe

d no

rms

Hig

h sc

ores

Aph

asia

sub

ject

s m

ore

impa

ired

in w

ritin

g di

ctat

edw

ords

, na

min

g, s

ynta

x, a

nd a

udito

ry r

eten

tion;

Sz

subj

ects

mor

e im

paire

d in

rel

evan

ce

Pts

with

LT

hos

pita

lizat

ion

had

mor

e pa

ucity

of s

peec

h,re

petit

ion,

per

seve

ratio

n, a

nd d

evia

nt v

erba

lizat

ion

Pau

city

of

spee

ch w

as a

n in

dica

tor

of s

ever

ity o

f ill

ness

.

Not

e.—

# =

num

ber;

Ac

= a

cute

; BP

= b

ipol

ar;

CS

z =

chr

onic

sch

izop

hren

ia; d

epr

= d

epre

ssio

n; L

T =

long

-ter

m; n

eg =

neg

ativ

e; p

os =

pos

itive

; pts

= p

atie

nts;

Sz

= s

chiz

ophr

enia

; TD

=th

ough

t di

sord

er.

1 Stu

dies

usi

ng o

nly

stan

dard

neu

rops

ycho

logi

cal

test

s w

ere

omitt

ed, a

s w

ere

stud

ies

only

clin

ical

ly a

sses

sing

"fo

rmal

tho

ught

dis

orde

r,"

stud

ies

of t

he s

peec

h of

chi

ldre

n w

ith s

chiz

o-ph

reni

a, a

nd s

tudi

es in

lang

uage

s ot

her

than

Eng

lish.

Not

all

publ

icat

ions

fro

m th

e sa

me

rese

arch

er a

nd to

pic

are

pres

ente

d. In

thes

e ca

ses

the

publ

icat

ion

with

the

larg

est

sam

ple

size

was

sel

ecte

d. A

ll st

udie

s w

ere

mix

ed-s

ex p

artic

ipan

ts a

nd c

ontr

ols

mat

ched

to p

atie

nts

unle

ss o

ther

wis

e no

ted.

In g

ener

al, m

ore

mal

es p

artic

ipat

ed t

han

fem

ales

.

to I; i a S'

Co s S" z p r m o

Dow



ecember 2021

Speech Disorder in Schizophrenia Schizophrenia Bulletin, Vol. 27, No. 3, 2001

or both (Grove and Andreasen 1985; Mortimer et al.1995).

The present report consists of two parts: (1) a reviewof the extensive past literature on language in schizophre-nia to determine whether consistent patterns emerge; and(2) an extension of our previous results (DeLisi et al.1997; Shedlack et al. 1997) to a new group of first episodepatients with schizophrenia and families with a high den-sity of schizophrenia. In this new study, a structuredmethod for soliciting verbal production from each subjectis used and scored. The data are then used to test thehypotheses (1) that uniquely human aspects of languageare deviant in individuals with schizophrenia, (2) that thedeviance is familial, (3) that it is associated with theinheritance of schizophrenia within families, and (4) thatreduced cerebral dominance is associated with these lan-guage anomalies.

Methods

Literature Review. A comprehensive computer search ofthe literature was performed using Paperchase, a Webvariant of Medline. The search items were "language,""schizophrenia," "speech," "syntax," and "semantics."Studies of children with schizophrenia, studies using onlystandard neuropsychological testing, studies only clini-cally assessing "formal thought disorder," and studiespublished in a language other than English were excluded.When multiple publications from the same researcherusing overlapping samples were found, only the publica-tion with the largest sample size was selected for inclu-sion. A total of 57 journal articles or books dating back to1959 were reviewed and are included in table 1. The fol-lowing topic areas were covered: speech comprehension,attention and interference, semantics, fluency, complexity,reference failures (i.e., when the speaker incorrectly infersthat what he or she is referring to has been clearly stated),grammar, lateralization of language, and miscellaneousother aspects (e.g., disorganization and social alienation).

Schizophrenia Study

Subjects. Twenty-nine patients with DSM-IV chronicschizophrenia (22 males, 7 females; mean age 33.8 ± 8years), 9 first episode schizophreniform or subchroniccases (6 males and 3 females; mean age 23.4 ± 5 years),and 12 controls (7 males and 5 females; mean age 32.6 ±7 years) consented to be audiotaped for the followingstudy. The patients with chronic schizophrenia were ill amean of 12.4 ± 8 years and the first episode patients, amean of 1.2 ± 2.0 years (time from first appearance ofpsychotic symptoms). The controls were solicited from

the community as previously selected for other studies(DeLisi et al. 1991) on the basis that they were the samesex, age, and social class of consecutively obtained firstepisode patients. Thus the 12 controls were matched to 12patients, 9 of whom satisfied DSM-IV criteria for schizo-phreniform or subchronic schizophrenic illness and 3chronic schizophrenia at the time of a first hospital admis-sion. All controls were interviewed using a structured for-mat (Schedule for Affective Disorders andSchizophrenia-Lifetime [SADS-L]; modified fromSpitzer and Endicott 1978) and anyone with an Axis I orII psychiatric disorder was eliminated prior to the study.

Eleven families with two or more ill siblings withschizophrenia also participated in the present study. Thesewere local New York families drawn from a larger cohortof national U.S. families recruited for molecular geneticlinkage studies (see Garner et al. 1996 and Shaw et al.1998 for cohort details). In two of these families, one ofthe ill siblings refused to participate. Thus the familieswere composed of 20 subjects with chronic schizophreniaor schizoaffective disorder and 13 nonpsychotic members(two sisters, seven mothers, and four fathers; mean age58.9 ±16 years). Of the 13, one mother had a past episodeof an acute psychosis (not otherwise specified), twofathers had paranoid personality disorder, two sisters hadrecurrent major depression, and one mother had a pasthistory of panic disorder; the remainder had no majordiagnosis.

All individuals with schizophrenia were medicated atthe time of this study. The subjects with schizophreniaand those members of the family study with schizophreniawere all on conventional antipsychotic medications, withthe exception of three chronic patients and the nine firstepisode patients, all of whom were participating in a dou-ble-blind haloperidol versus risperidone research protocol.

Family members were interviewed using theDiagnostic Interview for Genetic Studies (Nurnberger et al.1994). The interviews, information from family members,and medical records were used to determine diagnoses.Diagnoses for all individuals studied were made usingDSM-IV (APA 1994) criteria by two independent diagnosti-cians and any disputes settled by consensus in the presenceof a third trained clinical professional. All individuals per-forming diagnoses previously underwent diagnostic reliabil-ity studies to maintain a kappa statistic above 0.90. None ofthe nonpsychotic family members had an Axis I diagnosis.

Study Design. Oral soliloquies were recorded and latertranscribed as described below. An examiner asked sub-jects to use their imagination and describe what is occur-ring in each of seven pictures taken from the ThematicApperception Test (Murray 1971). The seven pictures areas follows:

487

Dow



ecember 2021

Schizophrenia Bulletin, Vol. 27, No. 3, 2001 L.E. DeLisi

Picture 1 is a person of ambiguous gender lying on abed or couch, with a man bending over the figure, onearm on the figure's left leg, the other arm raised over thereclining person's head.

Picture 2 is a woman lying on a bed face up, rightarm hanging to the floor. She appears naked from thewaist up. A man is standing facing away from her with hisright arm covering his face. He is dressed and wearing atie.

Picture 3 is a woman opening a door into a room witha wall bookcase, flowers, and a lamp on a table. Shadowsand light mark this picture.

Picture 4 is a boy sitting in a dark doorway of awooden building.

Picture 5 is a woman holding a person by his or herhead at the bottom of a staircase.

Picture 6 is a boy sitting with his hands holding uphis head and looking down at a violin on a table.

Picture 7 is scenery consisting of a tree with a row-boat underneath. A river is nearby.

The same individual recorded all subjects in thisstudy. Pictures were presented to subjects in the sameorder. Subjects were given as much time as they needed tointerpret each picture. The following are examples of aportion of a soliloquy from each of two patients withschizophrenia (dots indicate pauses):

(One subject responding to picture 1) "I'm afraidto say something lewd but it could connotate thatbe a grandfather coming in to touch his preciousgrand pre-post pubescent granddaughter...andyet he has sorta of a halo around him...so itcould be an angel...or maybe I am saying thatbecause there was a chance I was sexuallyabused which I have been exploring in the lastyear or two...the innocent part of me says its anangel visiting the virgin Mary and the other partof me says its incest the beginning of an inces-tual relationship that has been going on for awhile and he's not nervous...so I'm a little upsetlooking at that one"

(Another subject responding to picture 2) "Itlooks like he is tired.. .he wants to go tobed.. .he's tired.. .1 guess he has his hand over hisface like he is really tired and he is aggravated hecould be aggravated.. .he is just getting up out ofbed putting his clothes on...getting out of bedgetting something to drink or something...hecould be getting up to go to work in the morningyou know...his wife or his girlfriend is sleepingor maybe the wife died or something...he proba-bly is going to go to work"

In addition, subjects were asked to describe what theydo in the morning to get ready for an appointment, from

waking up to arrival at the destination (a sample of verbalsequential ordering of events). This ability is used as ameasure suggested by E. Chaika (personal communica-tion, August 1997) to test intactness of executive controlover language production. Cerebral dominance as deter-mined by handedness was assessed by a hand use ques-tionnaire (Annett 1967) and by a relative hand-skill test(Tapley and Bryden 1985). The latter is performed byeach hand separately placing dots in circles as fast as pos-sible in 20 seconds. The test is performed first by the righthand and then by the left. The number completed with theright hand is compared with those completed by the left asa ratio of right over left. These tests were incorporated inorder to examine the relationship of cerebral dominanceto language disorder.

Data Analysis. The tapes were transcribed by an experi-enced transcriber without punctuation but with pausesindicated. These were then rechecked for accuracy by theoriginal recorder. All soliloquies were coded by theresearcher, who recorded them with subject identifiersremoved. The author, who did not have knowledge ofpatient, family, or control status, then analyzed transcribedrecordings. A structured score sheet (figure 1) was used toanalyze each soliloquy recording and included a total wordcount. An assessment of syntax was based on Bickerton'sfive unique characteristics of human language: (1) order(number of awkward sentences and number with incorrectword order; the examiner judged correctness of word orderfor all sentences); (2) number of null elements (use ofdummy words, such as "it," that do not refer to objects);(3) number of sentences missing arguments for verbs; (4)mechanisms for expansion (number of conjoined and num-ber of embedded clauses in the entire soliloquy); and (5)number of grammatical mistakes. Semantic deviance wasscored by counting the number of misused words and thenumber of neologisms. All items were scored per the totalnumber of words with the exception of order and missingarguments, which were adjusted for number of assumedsentences. Miscellaneous items, such as appropriateness ofoverall content, logical order to sequential task, andamount of detail in this task, were also graded as eitheryes/no or appropriate/inappropriate (1 or 2).

All analyses were performed using the StatisticalPackage for Social Sciences, version 9.1, for Windows 95.For quantitative variables, two separate analyses of vari-ance (ANOVAs) were performed: one for the first episodecases compared to chronic patients and controls andcovarying age, with sex and diagnosis as independentvariables; and the second for the 11 families only, withsex, diagnosis, and family membership as independentvariables, covarying for age. Family membership wascoded 1-11 corresponding to each family (on averagethree individuals per family). In the variables in which

488

Dow



ecember 2021


Figure 1. Contents of structured score sheet forevaluation of soliloquies

Subject Code #

Date:

SOLILOQUY ANALYSIS:

Total # Words:Total # Sentences:

SYNTAX: TOTAL #:I. Order:Overall awkwardness of sentence:

Incorrect order:

II. # of Null Elements:(e.g., use of dummy subject "if)

III. Arguments of Verbs:(# sentences missing arguments)

IV. Mechanisms for Expansion:Noun Phrases:Conjoined clauses:Embedded clauses:

V. Grammatical Items:Ratio of grammatical to lexical items:Grammatical mistakes:

SEMANTICS:I. Misuse of words:II. Neologisms:

OTHER:(e.g., adequacy of content and order of sequential task;appropriateness of content of verbal responses)

types of words were counted, the total number of wordswas covaried. For categorical variables, both chi-squareanalyses and ANOVAs were used. Because the number offamilies was small and thus the numbers of individuals ineach cell of the three-way analysis small, two further setsof analyses were performed to supplement these data: (1)in order to examine the familial effect, without regard fordiagnosis, a two-way ANOVA was performed with sexand family group as the factors and age as a covariate;and (2) without regard for family membership, a two-wayANOVA was performed with sex and diagnosis as factors,covarying for age (and where appropriate for total numberof words and presumed sentences). A third analysis wasperformed comparing all family members with schizo-phrenia to their healthy relatives and controls.

Results

Literature Review. As can be seen in table 1, severalaspects of language comprehension and production havebeen found to be abnormal in patients with chronic schiz-

ophrenia compared to patients with bipolar psychosis,their nonpsychotic siblings, or controls: comprehension,attention, semantic organization, fluency, complexity, ref-erence failures, paucity of speech, and language lateral-ization. Almost all studies focused on only one aspect oflanguage and neglected others in the same individuals.Thus, whether patients with language disorder, or a sub-group of patients, have all the anomalies listed in table 1,or whether some patients have one abnormality while oth-ers have another, cannot be ascertained from the literatureto date. A literature bias must also be recognized in thatstudies failing to find similar differences between patientsand controls tend not to be published, whereas positiveones are more often reported.

Nevertheless, some patterns do emerge. The studiesof speech comprehension seem to attribute reduced com-prehension to deficient working memory (e.g., Grove andAndreasen 1985; Condray et al. 1996). On the other hand,deficits found in semantic processing (e.g., Kareken et al.1996), fluency (e.g., Rochester and Martin 1979), andcomplexity (e.g., Thomas et al. 1996) are interestingbecause they may have a relationship to the cerebral orga-nization of language. Reference failures, while shown tobe frequent in patients with schizophrenia, may havemore to do with distractibility than with primarily verbalorganization. Very little was found about grammatical dis-turbance in schizophrenia and it could not be separatedfrom general paucity of speech as characteristic of poorprognosis patients. At least one study has shown that thedegree of language dysfunction was associated with atypi-cal handedness (Manoach 1994). However, none of thestudies reviewed addressed whether any of these traits arefamilial or related to the heritability of schizophrenia.

The literature suggests a lack of concentration on theunique features of human language with the exception ofmechanisms for expansion, which are clearly associatedwith the degree of fluency and complexity of speech.Regardless of the aspect of speech studied, taken as awhole these studies point to the value of free speech sam-ples for examining several components of language inpatients with schizophrenia. When specifically devisedtests are used, their relevance to overall language produc-tion is uncertain. Thus the following study of schizophre-nia uses a standardized but free speech sample and con-centrates on specific characteristics of human language inthe analysis.

Schizophrenia Study

Unrelated Subjects With Schizophrenia versusControls. As presented in table 2, there was a significantdifference between all patients and controls for the num-ber of words (F = 3.2, p < 0.05) and for the number of

489

Dow



ecember 2021


Table 2. Patients versus controls (tests of between-subject effects)1'2

Variable

Age (yrs)# words# awkward sentences# incorrect order# null elements# grammatical mistakes# missing arguments# conjoined clauses

Covarying # words# embedded clauses

Covarying # wordsMisuse of wordsNeologismsPoor overall contentVaguenessSentences lacking

connectivityIncomplete sentencesNumerous false startsSequential task

Poor contentPoor order

Handedness (% R)Hand-skill (R - L)/0.5(R

1 s t episode patients(n = 9)

23.4 ± 5594 ±406

2.7 ±30.3 ±0.76.8 ±51.8±22.0 ±2

29.9 ± 24

6.9 ±7

0.44 ± 0.50

11%33%22%

44%22%

20%20%80%

+ L) 1.3 ±0.3

Chronic patients(n = 29)

33.8 ±8530 ± 267

3.3 ±40.3 ± 0.96.2 ±41.3±21.1 ±2

20.2 ±16

5.6 ±5

0.38 ± 0.60.10 ±0.6

4 1 %28%14%

38%35%

46%50%84%

1.3 ±0.4

Controls(n = 12)

32.6 ± 7862 ±514

1.8 ±10

11.0±80.9 ±10.9 ±1

45.8 ± 27

12.2 ±12

0.17 ±0.40000

8%25%

9%18%

100%1.3 ±0.3

F(Dx){df= 2,49)

3.210.441.060.461.500.315.922.602.680.070.540.303.322.110.95

1.090.80

1.441.430.670.33

P <

0.050.650.360.640.240.730.0050.0850.080.930.590.750.050.130.39

0.350.46

0.250.250.510.73

Note.—# = number of; Dx = diagnosis; L = left-handed; R = right-handed; SD = standard deviation.1 Mean ± SD or percent of whole.2 There were significant sex and sex-by-diagnosis effects. F(sex): # embedded clauses/* words, F = 4.77, p < 0.034, males > females.F(sex x Dx): # embedded clauses/* words, F = 3.94, p< 0.027, where female chronic patients had more embedded clauses than malepatients, but male controls had more than female controls.

conjoined clauses (F = 5.9, p < 0.005), the latter becominga trend when controlled for number of words (F = 2.7, p <0.085). The number of embedded clauses only had a trendfor significance without controlling for the number ofwords (F = 2.7, p < 0.08), and with the number of wordscontrolled there were no differences (F = 0.07, p < 0.93).However, in post hoc analyses (Tukey-HSD), it was foundthat these variables were significantly reduced in thechronic patients when compared with controls, but not infirst episode patients. Chronic patients have significantlyfewer conjoined clauses (p < 0.002), fewer embeddedclauses (p < 0.04), and fewer words (p < 0.03). Firstepisode patients in general scored between chronicpatients and controls, not reaching significance in compar-ison to either. There were no sex-by-diagnosis effects.

Using Spearman's nonparametric correlations, poorcontent (limited number of items mentioned) in thesequential task was inversely correlated with both thenumber of conjoined clauses (r = -0.47, p < 0.000) andthe number of embedded clauses (r = -0.52, p < 0.000);however, neither were significantly correlated with theappropriateness of the sequential order of items.

Chi-square analyses for the independent indicatorvariables overall (presence or absence of vagueness, logi-cal ordering, incomplete sentences, false starts, as well asorder of and poor content of the sequential task) were per-formed using the expected amounts for the total group.Both first episode and chronic patients had significantlymore vagueness to their speech (x2 = 4.7, p = 0.03 and x2

= 4.1, p = 0.04, respectively) and incomplete sentences(X2 = 3.7, p = 0.055 and x2 = 3.6, p = 0.058) than con-trols. However, only the chronic patients showed a higherfrequency of overall inappropriate content (x2 = 7.0, p =0.008) and poor content to the sequential task (x2 = 4.7, p= 0.03).

Neither hand skill nor handedness were correlatedwith any language variables, except that the degree ofasymmetric hand skill (right > left) was positively corre-lated with the ability to order events appropriately in thesequential task (r = 0.55, p < 0.000). Age of onset wascorrelated with the number of embedded clauses (r =0.35, p < 03; r = 0.32, p < 0.05 when controlled for num-ber of words); however, it was not correlated with numberof conjoined clauses.

490

Dow



ecember 2021

Tab

le 3

. A

nal

ysis

of f

amili

es d

ense

wit

h s

chiz

op

hre

nia

1

Dep

end

ent

vari

able

s

# w

ords

# aw

kwar

d se

nten

ces

# in

corr

ect

orde

r#

null

elem

ents

# gr

amm

atic

al m

ista

kes

# m

issi

ng a

rgum

ents

# co

njoi

ned

clau

ses

Per

# w

ords

# em

bedd

ed c

laus

esP

er #

wor

dsM

isus

e of

wor

dsP

oor

over

all c

onte

ntV

ague

ness

III r

elat

ives

527

± 29

94.

7 ±

50.

5 ±

15.

5 ±

41.

2 ±

11.

4 ±

319

.7 ±

19

0.03

± 0

.02

5.24

± 7

0.00

8 ±

0.0

060.

53 ±

0.7

47%

35%

Sen

tenc

es la

ckin

g co

nnec

tivity

18

%In

com

plet

e se

nten

ces

Num

erou

s fa

lse

star

tsS

eque

ntia

l tas

kP

oor

cont

ent

Logi

cal

Han

dedn

ess(

% R

)H

ands

kill

(R -

L)/0

.5(R

+ L

]

41

%4

1%

60%

60

%80

%I

1.3

±0

.5

Mal

es

Wel

l re

lativ

es

588

± 20

11

.3±

20.

3 ±

0.5

4.0

±5

0.8

±0

.50.

023

.8 ±

18

0.04

± 0

.01

7.3

±2

0.01

3 ±

0.0

04

0.25

± 0

.525

%0% 0% 25

% 0

25%

50

%10

0%1.

3 ±

0.2

III

rela

tives

499

± 22

61.

0 ±

10.

07.

0 ±

51.

3 ±

0.6

1.0

±2

22.0

±1

60.

04 ±

0.0

15.

0 ±

50.

009

± 0

.01

0.33

± 0

.60% 0% 33

%33

%33

%

33% 0

66%

1.3

±0

.2

Fem

ales

Wel

l re

lativ

es

785

± 55

83.

6 ±

60.

010

.3 ±

12

0.9

±1

0.9

±2

38.6

± 3

60.

05 ±

0.0

28.

7 ±

80.

010

±0

.00

70.

11 ±

0.3

0% 22%

22% 0

11%

25%

50

%10

0%1.

3 ±

0.2

(df=

F

1.59

1.18

0.70

8.62

1.13

0.18

3.76

0.04

5.71

5.10

0.87

0.97

0.75

1.53

5.54

0.83

5.92

0.34

1.43

0.44

Fam

ily

10,3

2) P<

0.26

0.42

0.71

0.00

50.

450.

991.

660.

260.

010.

020.

590.

530.

670.

280.

010.

61

0.01

0.93

0.33

0.89

Dia

gn

osi

s

(df=

F

0.74

00.

150

0.31

00.

040

0.22

00.

060

0.96

01.

030

0.00

21.

280

3.17

00.

200

1.22

00.

230

0.13

00.

530

0.12

00.

620

0.00

30.

200

1,32

) P<

0.42

0.71

0.59

0.84

2

0.66

0.82

0.03

0.34

0.97

0.30

2

0.12

0.66

0.30

0.65

0.73

0.49

0.74

0.46

0.96

0.67

a o a- n>isor<ierii n N' i |

Not

e.—

# =

num

ber o

f; L

= le

ft-ha

nded

; Ft

= r

ight

-han

ded.

1 Ana

lysi

s of

11

fam

ilies

with

two

sibl

ings

dia

gnos

ed w

ith s

chiz

ophr

enia

. Thr

ee-w

ay A

NO

VA

was

per

form

ed w

ith s

ex, d

iagn

osis

(sch

izop

hren

ia o

r sc

hizo

affe

ctiv

e di

sord

er v

s. n

o ps

y-ch

osis

), a

nd

fam

ily g

roup

as

fact

ors

with

age

as

a co

varia

te.

Mea

n ±

SD

or

perc

ent o

f w

ho

le.

2 A s

igni

fican

t fam

ily-b

y-di

agno

sis

effe

ct c

ovar

ying

for

tota

l num

ber o

f wo

rds

was

pre

sent

for

embe

dded

cla

uses

(F =

5.8

7, p

= 0

.02)

an

d fo

r nu

mbe

r of

null

elem

ents

(F =

8.7

4, p

<0.

007)

.f I o

Dow



ecember 2021


Family Analyses. In the three-way ANOVA examiningfamily membership, diagnosis, and sex (table 3), therewas a significant familial effect for the number of embed-ded clauses when controlled for the number of words (F =5.1, p < 0.02), the number of incomplete sentences (F =5.5, p < 0.01), and poor content of the sequential task (F =5.9, p < 0.01). A family-by-diagnosis interaction per num-ber of words was present for embedded clauses (F = 5.9, p< 0.02) and per number of null elements (F = 8.7, p <0.007).

In the two-way analysis with family membership andsex as factors, significant familial effects were observedfor the number of null elements (F = 2.7, p < 0.05), num-ber of conjoined clauses (F = 7.1, p < 0.001), number ofembedded clauses (F = 3.5, p < 0.02), poor content of thesequential task (F = 3.4, p < 0.02), and number of incom-plete sentences (F = 8.0, p < 0.000). There was also a sexby family interaction for number of null elements (F =4.8, p < 0.009), number of conjoined clauses (F = 6.3, p <0.003), number of embedded clauses (F = 3.3, p < 0.03),and number of incomplete sentences (F = 6.5, p < 0.002).In general, within these families with a high density ofschizophrenia, females had more null elements in theirspeech than males and more conjoined and embeddedclauses than males; males had more incomplete sentencesthan females.

The analysis comparing family members with schizo-phrenia, their healthy relatives, and controls withoutregard to family membership revealed differences in thenumber of conjoined clauses (only when not controlledfor number of words: F = 3.8, p < 0.03, with the relativeshaving schizophrenia showing significant difference fromcontrols [p < 0.02] and the well relatives not differingfrom controls) and in poor sequential task content (F =3.5, p < 0.04, relatives with schizophrenia again showingsignificant difference from controls [p < 0.01], but notnonpsychotic relatives). There was a trend for a differencein only the number of embedded clauses (F = 2.0, p <0.15, present only for the ill [p < 0.06] and not for thenonpsychotic relatives).

DiscussionThe extensive literature analyzing the speech of patientswith schizophrenia was searched for any evidence ofunique characteristics that appear consistently across stud-ies (table 1). No one aspect of speech appeared deviant;however, several observations emerged, the majority ofwhich could be attributed to an underlying disturbance ineither working memory or attention, or both, rather thanin the ability to use the uniquely human tools of languagecorrectly. For example, Thomas et al. (1996) have shownthat when given a writing task, patients with schizophre-nia wrote with as much complexity as controls, yet their

speech, which depends on working memory and attentionfor focused fluent production, was less complex than thatof controls (Thomas et al. 1990). Thus, Thomas and col-leagues suggest that linguistic abilities are normal inschizophrenia (as seen in writing), but that quick retrieval(as in speech) is problematic. Similarly, Manshreck et al.(1980) showed that the correct use of words (theirretrieval from long-term memory) was deficient in thespeech of patients with schizophrenia but not in their writ-ing. In addition, the connection between words in long-term memory storage, as suggested by the counting ofcohesive ties (Rochester and Martin 1979), may be anom-alous. Thus, quick retrieval of words from long-termmemory when speaking may also be an underlying diffi-culty in schizophrenia.

The studies on increased reference failures in schizo-phrenia could indicate working memory problemswhereby the speaker does not realize that the "hearer"does not recognize the referent. Failure to comprehend thesentence structure of speakers, particularly as it becomesmore complex, is another indication that working memorycapacity may be reduced.

However, neither working memory nor its ties tolong-term memory storage and attention are clearlyuniquely human, except that they may be suggested tooperate at a quantitatively higher capacity in Homo sapi-ens than in nonhuman primates. Whether the deficits inlanguage seen in the previous studies and our present oneare due to primary cognitive defects cannot be discernedbecause the overall IQ of each subject was not measuredand controlled for in the analyses. Future studies will needto examine the relationship of specific language measuresto other measures of cognition.

Despite the controversial nature of the theoreticalbasis for the present study, we attempted to examinewhether those "uniquely" human tools of language asdescribed by Bickerton (1990) could be determined by theanalysis of structured oral soliloquies to be specificallydeviant in schizophrenia. While the number of subjectswas small, a characteristic reduction in sentence complex-ity (reduced conjoined and embedded clauses) wasobserved in patients with chronic schizophrenia. Theyused fewer words overall, more inappropriate content, andless connectivity between sentences than controls did.These data are consistent with previous reports of reducedsentence complexity in chronic schizophrenia and werealso correlated with age at onset—the earlier the onset,the less complex the sentence structure (Morice andIngram 1983; Thomas and Leudar 1995).

In addition, familial basis for clausal expansion issuggested from the family analyses in our present study;clausal embedding in particular appeared to be familialand also to segregate with schizophrenia within families.Overall, the nonpsychotic relatives of individuals with

492

Dow



ecember 2021


schizophrenia did not appear to have language productionthat differs from controls; the familial deviance wasoccurring only in those family members with schizophre-nia. It is of interest that of Bickerton's five unique aspectsof language, mechanisms for expansion (clausal embed-ding) were reduced, but the use of null elements, orderingof sentence constituents, grammatical deficiencies, andappropriate use of arguments for verbs were not differentfor patients and controls.

An interesting sex effect also emerged and is not sur-prising, given the several sex differences in the clinicaland biological expression of schizophrenia (reviewed inDeLisi et al. 1989). In the present study, whereas malecontrols had significantly more sentence complexity thanfemales, the reverse was true for schizophrenia (i.e.,chronic male patients had significantly fewer embeddedclauses than females). These sex differences could be seenas consistent with Crow's hypothesis that the underlyinggenetic basis for psychosis is located on the sex chromo-somes (Crow 1997, 19986; DeLisi and Crow 1989), butnevertheless the results should be interpreted with cautiongiven the small number of males and females studied.

The additional aspect of Crow's hypothesis (1997),that anomalous cerebral lateralization underlies languagedeficits, was also tested. Although the degree of handed-ness or relative hand skill was unrelated to clausal embed-ding, asymmetric hand-skill was significantly correlatedwith the ability to order events appropriately in a sequen-tial task (r = 0.55, p < 0.000).

In attempts to associate the defects in language thatwere observed with an alternative underlying mechanism,we examined their relationship to executive functioningas measured by being able to construct a sequential taskorally. The ability to perform this task adequately was notcorrelated with the degree of clausal embedding (i.e.,complexity). However, this was only a superficial look atcomplex executive control over language production.

The lack in the present study of detectable languageabnormalities at the time of the first episode suggests pro-gressive development after the onset of illness or hetero-geneity in outcome after only one episode of psychosis(DeLisi 1997, 1999). The reduction of significance whencontrolled for total number of words used may mean thatchronic patients with schizophrenia simply use fewerwords than controls but that their mechanisms for expan-sion are intact.

In summary, patients with chronic schizophrenia haveless verbal production than controls, and they appear tohave reduced use of expansive mechanisms, which wasshown in the present study to be familial and related to ill-ness within families. Overall, this and other studies pro-vide some evidence that some uniquely human qualitiesof language underlie the core deficits of schizophrenia.

However, further exploration in larger groups of individu-als is warranted. It is presently difficult to say what thesignificance of reduced numbers of words implies. Dosubjects with schizophrenia speak less yet have an intactcapacity for uniquely human expansive mechanisms, ordo they produce fewer words because their capacity forexpansion is less?

This was an extensive review of the literature and apreliminary study with a relatively small number of fami-lies. It was an initial attempt to explore the hypothesis thatsome uniquely human deviations of language are relatedto anomalous cerebral dominance and the inheritance ofschizophrenia. Further extension of this work to the eval-uation of a greater number of subjects and their familieswill hopefully clarify the significance of these results.

References

Allen, H.A., and Allen, D.S. Positive symptoms and theorganization within and between ideas in schizophrenicspeech. Psychological Medicine, 15:71-80, 1985.

Allen, H.A.; Liddle, P.F.; and Frith, CD. Negative features,retrieval processes and verbal fluency in schizophrenia.British Journal of Psychiatry, 163:769-775, 1993.

Aloia, M.S.; Gourovitch, M.L.; Weinberger, D.R.; andGoldberg, T.E. An investigation of semantic space inpatients with schizophrenia. Journal of the InternationalNeuropsychological Society, 2:267-273, 1996.

American Psychiatric Association. DSM-IV: Diagnosticand Statistical Manual of Mental Disorders. 4th ed.Washington, DC: APA, 1994.

Annett, M. The binomial distribution of right, mixed, andleft handedness. Quarterly Journal of ExperimentalPsychology, 19:327-333, 1967.

Barch, D.M., and Berenbaum, H. Language productionand thought disorder in schizophrenia. Journal ofAbnormal Psychology, 105:81-88, 1996.

Barr, W.B.; Bilder, R.M.; Goldberg, E.; Kaplan, E.; andMukherjee, S. The neuropsychology of schizophrenicspeech. Journal of Communication Disorders,22:327-349, 1989.

Bickerton, D. Language and Species. Chicago, IL:University of Chicago Press, 1990.

Bickerton, D. Language and Human Behavior. Seattle,WA: University of Washington Press, 1995.

Chaika, E., and Lambe, R.A. Cohesion in schizophrenicnarratives, revisited. Journal of CommunicationDisorders, 22:407^121, 1989.

Chaika, E.O. Understanding Psychotic Speech.Springfield, IL: Charles C. Thomas, 1990.

493

Dow



ecember 2021


Cheadle, J., and Morgan, R. Does the chronic psychiatricpatient understand plain English? British Journal ofPsychiatry, 120:553-560, 1972.

Cohen, B.D., and Camhi, J. Schizophrenia performance ina word-communication task. Journal of AbnormalPsychology, 72:240-246, 1967.

Condray, R.; Steinhauer, S.R.; and Goldstein, G.Language comprehension in schizophrenics and theirbrothers. Biological Psychiatry, 32:790-802, 1992.

Condray, R.; Steinhauer, S.R.; van Kammen, D.P.; andKasparek, A. Working memory capacity predicts languagecomprehension in schizophrenic patients. SchizophreniaResearch, 20:1-13, 1996.

Condray, R.; van Kammen, D.P.; Steinhauer, S.R.;Kasparek, A.; and Yao, J.K. Language comprehension inschizophrenia: Trait or state indicator? BiologicalPsychiatry, 38:287-296, 1995.

Corballis, M.C. The Lopsided Ape: Evolution of theGenerative Mind. Oxford: Oxford University Press, 1991.

Crow, T.J. Is schizophrenia the price that Homo sapiens paysfor language? Schizophrenia Research, 28:127-141,1997.

Crow, T.J. Nuclear schizophrenia symptoms as a windowon the relationship between thought and speech. BritishJournal of Psychiatry, 173:303-309, 1998a.

Crow, T.J. Sexual selection, timing and the descent ofman: A theory of the genetic origins of language. CurrentPsychology of Cognition, 17:1079-1114, 1998b.

Cutting, J., and Murphy, D. Preference for denotative asopposed to connotative meanings in schizophrenics. Brainand Language, 39:459-468, 1990.

Davis, K.L.; Buchsbaum, M.S.; Shihabuddin, L.; Spiegel-Cohen, J.; Metzger, M.; Frecska, E.; Keefe, R.S.; andPowchik, P. Ventricular enlargement in poor-outcomeschizophrenia. Biological Psychiatry, 43:783-793, 1998.

DeLisi, L.E. Is schizophrenia a lifetime disorder of brainplasticity, growth and aging? Schizophrenia Research,23:119-129, 1997.

DeLisi, L.E. Defining the course of brain structuralchange and plasticity in schizophrenia. PsychiatryResearch Neuroimaging, 92:1-9, 1999.

DeLisi, L.E.; Boccio, A.M.; Riordan, H.; Hoff, A.L.;Dorfman, A.; McClelland, J.; Kushner, M.; Van Eyl, O.;and Oden, N. Familial thyroid disease and delayed lan-guage development in first admission patients with schiz-ophrenia. Psychiatry Research, 38:39-50, 1991.

DeLisi, L.E., and Crow, T.J. Evidence for a sex chromo-some locus for schizophrenia. Schizophrenia Bulletin,15(3):431^M0, 1989.

DeLisi, L.E.; Dauphinais, D.; and Hauser, P. Gender dif-ferences in the brain: Are they relevant to the pathogene-

sis of schizophrenia? Comprehensive Psychiatry,30:197-207,1989.

DeLisi, L.E.; Sakuma, M.; Kushner, M.; Finer, D.L.;Hoff, A.L.; and Crow, T.J. Anomalous cerebral asymme-try and language processing in schizophrenia.Schizophrenia Bulletin, 23(2): 255-271, 1997.

Docherty, N.M.; Evans, I.M.; Sledge, W.H.; Seibyl, J.P.;and Krystal, J.H. Affective reactivity of language in schiz-ophrenia. Journal of Nervous and Mental Disease,182:98-102, 1994a.

Docherty, N.M., and Hebert, A.S. Comparative affectivereactivity of different types of communication distur-bances in schizophrenia. Journal of AbnormalPsychology, 106:325-330, 1997.

Docherty, N.M.; Sledge, W.H.; and Wexler, B.E. Affectivereactivity of language in stable schizophrenic outpatientsand their parents. Journal of Nervous and Mental Disease,182:313-318, 19946.

Faber, R.; Abrams, R.; Taylor, M.A.; Kasprison, A.;Morris, C; and Weisz, R. Comparison of schizophrenicpatients with formal thought disorder and neurologicallyimpaired patients with aphasia. American Journal ofPsychiatry, 140:1348-1351, 1983.

Faber, R., and Reichstein, M.B. Language dysfunction inschizophrenia. British Journal of Psychiatry,139:519-522,1981.

Fraser, W.I.; King, K.M.; Thomas, P.; and Kendell, R.E.The diagnosis of schizophrenia by language analysis.British Journal of Psychiatry, 148:275-278, 1986.

Gardner, R.A., and Gardner, B.T. Comparative psychol-ogy and language acquisition. Annals of the New YorkAcademy of Science, 309:37-76, 1978.

Garner, C; Kelly, M.; Cardon, L.; Joslyn, G.; Carey, A.;LeDuc, C; Lichter, J.; Harris, T; Loftus, J.; Shields, G.;Comazzi, M.; Vita, A.; Smith, A.M.; Dann, J.; Crow, T.J.;and DeLisi, L.E. Linkage analyses of schizophrenia tochromosome 6p24-p22: An attempt to replicate.Neuropsychiatric Genetics, 67:595-610, 1996.

Geschwind, N., and Galaburda, A.M. CerebralLateralization. Cambridge, MA: MIT Press, 1987.

Goldfarb, R.; Eisenson, J.; Stocker, B.; and DeSanti, S.Communicative responsibility and semantic task in apha-sia and "schizophasia." Perceptual and Motor Skills,79:1027-1039,1994.

Gottschalk, L.A.; Glesner, G.C.; Magliocco, B.; andD'Zmura, T.L. Further studies on the speech patterns ofschizophrenic patients. Journal of Nervous and MentalDisease, 132:101-113, 1961.

Grand, S.; Steingart, I.; Freedman, N.; and Buchwald, C.Organization of language behavior and cognitive perform-

494

Dow



ecember 2021


ance in chronic schizophrenia. Journal of AbnormalPsychology, 84:621-628, 1975.

Grove, W.M., and Andreasen, N.C. Language and think-ing in psychosis: Is there an input abnormality? Archivesof General Psychiatry, 42:26-32, 1985.

Halpern, H., and McCartin-Clark, M. Differential lan-guage characteristics in adult aphasic and schizophrenicsubjects. Journal of Communication Disorders,17:289-307, 1984.

Harvey, P.D., and Brault, J. Speech performance in maniaand schizophrenia: The association of positive and nega-tive thought disorders and reference failures. Journal ofCommunication Disorders, 19:161-173, 1986.

Harvey, P.D., and Serper, M.R. Linguistic and cognitivefailures in schizophrenia: A multivariate analysis. Journalof Nervous and Mental Disease, 178:487-493, 1990.

Hauser, M.D. Language evolution: Linguists take a look.In: Hauser, M.D., ed. The Evolution of Communication.Cambridge, MA: MIT Press, 1997. pp. 30-50.

Hoffman, R.E.; Hoggben, G.L.; Smith, H.; and Calhoun,W.F. Message disruptions during syntactic processing inschizophrenia. Journal of Communication Disorders,18:183-202, 1985.

Hoffman, R.E.; Stopek, S.; and Andreasen, N.C. A com-parative study of manic vs. schizophrenic speech disorga-nization. Archives of General Psychiatry, 43:831-838,1986.

Joyce, E.M.; Collinson, S.L.; and Crichton, P. Verbal flu-ency in schizophrenia: Relationship with executive func-tion, semantic memory and clinical alogia. PsychologicalMedicine, 26:39-49, 1996.

Kantorowitz, D.A., and Cohen, B.D. Referent communi-cation in chronic schizophrenia. Journal of AbnormalPsychology, 86:1-9, 1977.

Kareken, D.A.; Moberg, P.J.; and Gur, R.C. Proactiveinhibition and semantic organization: Relationship withmemory in patients with schizophrenia. Journal of theInternational Neuropsychological Society, 2:486—493,1996.

Kinney, D.K.; Jacobsen, B.; Bechgaard, B.; Jansson, L.;Faber, B.; Kasell, E.; and Uliana, R.L. Content analysis ofspeech of schizophrenic and control adoptees and theirrelatives: Preliminary results. Social Science andMedicine, 21:589-593, 1985.

Kwapil, T.R.; Hegley, D.C.; Chapman, L.J.; andChapman, J.P. Facilitation of word recognition by seman-tic priming in schizophrenia. Journal of AbnormalPsychology, 99:215-221, 1990.

Landre, N.A., and Taylor, M.A. Formal thought disorderin schizophrenia: Linguistic, attentional, and intellectual

correlates. Journal of Nervous and Mental Disease,183:673-680, 1995.

Manoach, D.S. Handedness is related to formal thoughtdisorder and language dysfunction in schizophrenia.Journal of Clinical and Experimental Neuropsychology,16:2-14, 1994.

Manschreck, T.C.; Maher, B.A.; and Rucklos, M.E. Clozeprocedure and written language in schizophrenia.Language and Speech, 23:323-328, 1980.

Miller, W.K., and Phelan, J.G. Comparison of adult schiz-ophrenics with matched normal native speakers ofEnglish as to "acceptability" of English sentences.Journal of Psycholinguistic Research, 9:579-593, 1980.

Morice, R. Language impairments and executive dysfunc-tion in schizophrenia. In: Sims, A., ed. Speech andLanguage Disorders in Psychiatry. London, U.K.:Gaskell Press, 1995. pp. 57-69.

Morice, R., and McNicol, D. The comprehension and pro-duction of complex syntax in schizophrenia. Cortex,21:567-580, 1985.

Morice, R.D., and Ingram, J.C. Language complexity andage of onset of schizophrenia. Psychiatry Research,9:233-242, 1983.

Morice, R.D., and Ingram, J.C.L. Language analysis inschizophrenia: Diagnostic implications. Australian andNew Zealand Journal of Psychiatry, 16:11-21, 1982.

Mortimer, A.; Corridan, B.; Rudge, S.; Kho, K.; Kelly, F.;Bristow, M.; and Hodges, J. Thought, speech and lan-guage disorder and semantic memory in schizophrenia.In: Sims, A., ed. Speech and Language Disorders inPsychiatry. London, U.K.: Gaskell Press, 1995. pp.70-80.

Murray, H.A. Thematic Apperception Test. Cambridge,MA: Harvard University Press, 1971.

Nurnberger, J.I.; Blehar, M.C.; Kaufman, C.A.; York-Cooler, C ; Simpson, S.G.; Harkavy-Friedman, J.;Severe, J.B.; Malaspina, D.; and Reich, T. Diagnosticinterview for genetic studies: Rationale, unique featuresand training. Archives of General Psychiatry,51:849-862, 1994.

Pavy, D.; Grinspoon, L.; and Shader, R.I. Word frequencymeasures of verbal disorders in schizophrenia. Disordersof the Nervous System, 30:553-556, 1969.

Ragin, A.B., and Oltmanns, T.F. Lexical cohesion and for-mal thought disorder during and after psychotic episodes.Journal of Abnormal Psychology, 95:181-183, 1986.

Rausch, M.A.; Prescott, T.E.; and DeWolfe, A.S.Schizophrenia and aphasic language: Discriminable ornot? Journal of Consulting and Clinical Psychology,48:63-70, 1980.

495

Dow



ecember 2021


Rochester, S.R. Are language disorders in acute schizo-phrenia actually information processing problems?Journal of Psychiatric Research, 14:275-283,1978.

Rochester, S.R., and Martin, J.R. Crazy Talk: A Study ofthe Discourse of Schizophrenic Speakers. New York, NY:Plenum Press, 1979.

Rutter, D.R. The reconstruction of schizophrenic speech.British Journal of Psychiatry, 134:356-359, 1979.

Shaw, S.H.; Kelly, M.; Smith, A.B.; Shields, G.; Hopkins,P.J.; Loftus, J.; Laval, S.H.; Vita, A.; De Hert, M.; Cardon,L.R.; Crow, T.J.; Sherrington, R.; and DeLisi, L.E. Agenome-wide search for schizophrenia susceptibilitygenes. American Journal of Medical Genetics(Neuropsychiatric Genetics), 81:364-376, 1998.

Shedlack, K.; Lee, G.; Sakuma, M.; Xie, S.-H.; Kushner,M.; Pepple, J.; Finer, D.L.; Hoff, A.L.; and DeLisi, L.E.Language processing and memory in ill and well siblingsfrom multiplex families affected with schizophrenia.Schizophrenia Research, 25:43-52, 1997.

Siegel, A.; Harrow, M.; Reilly, F.E.; and Tucker, G.J.Loose associations and disordered speech patterns inchronic schizophrenia. Journal of Nervous and MentalDisease, 162:105-112, 1976.

Smith, E.E. Associative and editing processes in schizo-phrenic communication. Journal of Abnormal Psychology,75:182-186,1970.

Spitzer, M.; Weisker, I.; Winter, M.; Maier, S.; Hermle,L.; and Maher, B.A. Semantic and phonological primingin schizophrenia. Journal of Abnormal Psychology,103:485^94, 1994.

Spitzer, R.L., and Endicott, J. Schedule for AffectiveDisorders and Schizophrenia—Lifetime Version. 3rd ed.New York, NY: New York Psychiatric Institute, 1977.

Tapley, S., and Bryden, M. A group test for the assessmentof performance between the hands. Neuropsychologia,23:215-221, 1985.

Thomas, P.; Kearney, G.; Napier, E.; Ellis, E.; Leudar, I.;and Johnston, M. Speech and language in first onset psy-

chosis: Differences between people with schizophrenia,mania, and controls. British Journal of Psychiatry,168(3):337-343, 1996.

Thomas, P.; King, K.; Fraser, W.I.; and Kendell, R.E.Linguistic performance in schizophrenia: A comparisonof acute and chronic patients. British Journal ofPsychiatry, 156:204-210, 1990.

Thomas, P., and Leudar, I. Syntactic processing and com-munication disorder in first-onset schizophrenia. In: Sims,A., ed. Speech and Language Disorders in Psychiatry.London, U.K.: Gaskell Press, 1995. pp. 96-112.

Thomas, P.; Leudar, I.; Newby, D.; and Johnston, M.Syntactic processing and written language output in firstonset psychosis. Journal of Communication Disorders,26:209-230, 1993.

Weisbrod, M.; Maier, S.; Harig, S.; Himmelsbach, U.; andSpitzer, M. Lateralized semantic and indirect semanticpriming effects in people with schizophrenia. BritishJournal of Psychiatry, 172:142-146, 1998.

Williams, R.M.; Hemsley, D.R.; and Denning-Duke, C.Language behaviour in acute and chronic schizophrenia.British Journal of Social and Clinical Psychology,15:73-83,1976.

Wykes, T, and Leff, J. Disordered speech: Differencesbetween manics and schizophrenics. Brain andLanguage, 15:117-124, 1982.

Acknowledgments

The author wishes to acknowledge the assistance ofMaureen Kushner, who interviewed and audiotapedall individuals; Patti Heine, who transcribed all tapes;and Kimberly P. Greene, who performed the dataanalysis.

The Author

Lynn E. DeLisi, M.D., is Professor of Psychiatry, NewYork University, New York, NY.

496

Dow



ecember 2021

speech disorder in schizophrenia: review of the literature and

Documents