Title: Material specific hemispheric activation
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00097467/00001
 Material Information
Title: Material specific hemispheric activation
Physical Description: xi, 128 leaves : ill. ; 28 cm.
Language: English
Creator: Bowers, Dawn, 1950-
Copyright Date: 1978
 Subjects
Subject: Laterality -- Testing   ( lcsh )
Attention   ( lcsh )
Cerebral hemispheres   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 112-127.
Additional Physical Form: Also available on World Wide Web
General Note: Typescript.
General Note: Vita.
Statement of Responsibility: by Dawn Bowers.
 Record Information
Bibliographic ID: UF00097467
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000084467
oclc - 05267309
notis - AAJ9811

Downloads

This item has the following downloads:

PDF ( 5 MBs ) ( PDF )


Full Text









rIATERIAL SPECIFIC HEMISPHERIC ACTIVATIO:I


By

DAWN BOWERS





















A DISSERTATION PRESENTED TO THE GRjDI'UTE COUNI:CIL OF
THE I1III VERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF TIE PEQUIREIHENTS FR0 THE
DEGREE OF DOCTOR OF PHILOSOPHY









UNIVERSITY OF FLORIDA


1 978 :,


























TO DIANE












ACKNOWLEDGEMENTS

I heartily thank the members of my doctoral committee (Ken Heilman,

Paul Satz, Steve Zornetzer, Mike Levy, and Nate Perry) for their enthu-

siastic, although sometimes painful supervision and direction. I es-

pecially thank Ken Heilman and Paul Satz for their participation in

the training and development of my career as a psychologist. Not only

have these two delightful, creative people had significant impact on

my professional interests in neuropsychology, they have been people

whose friendship I value highly. Appreciation is extended to Mike Levy

for his "early" direction and guidance, from my undergraduate honors

thesis to the present study. Steve Zornetzer and Nate Perry are also

thanked for their contribution to my development as a "scientist" through-

out my graduate education.

There are numerous other significant people who have been actively

supportive in various "specific" and "nonspecific" ways during the past

six years. I particularly thank Molly, Harry, Maureen, my early bird

racquetball partners (C.S. and J.G.), Bill, Margie, Eileen, Tom, Edith,

Andrea, Carol, Larry, BTM, Alice, Mrs. Lombardi, Rick, Jan, and my parents.













TABLE OF CONTENTS

ACKNOWLEDGEMENTS..................................................... iii

LIST OF TABLES....................................................... v

LIST OF FIGURES.....................................................viii

ABSTRACT............................................................. ix

INTRODUCTION......................................................... 1

Overview of the Problem... ................................... 1
Laterality of Cerebral Function................................ 2
Attention ................................................... 32
Notes .......................................................... 60

STATEMENT OF THE PROBLEM............................................. 62

METHOD ................................ ............................... 67

Subjects .................................. ..................... 67
Apparatus ..... ................................................. 67
Warning Stimuli ............................................... 68
Go-No Go Condition............................................ 68
Simple RT Condition........................................ .. 70
Analyses .................................................... 71

RESULTS.............................................................. 73

Analyses of the Go-No Go Condition..................... ......... 74
Analyses of the Simple RT Condition.......... ................... 87

DISCUSSION............................ ............................... 96

BIBLIOGRAPHY.........................................................112

BIOGRAPHICAL SKETCH..................................................128













LIST OF TABLES


TABLE 1. SYNOPSIS OF DICHOTIC LISTENING STUDIES WITH
RIGHT-HANDED ADULTS ACCORDING TO THE TYPES OF
STIMULI THAT PRODUCE RIGHT EAR (RE), LEFT EAR (LE),
OR NO BETWEEN-EAR DIFFERENCES (ND) FOR THE RECALL
RECOGNITION OF DICHOTIC STIMULI.............................. 7

TABLE 2. SYNOPSIS OF DICHOTIC LISTENING STUDIES WITH
RIGHT-HANDED ADULTS ACCORDING TO THE TYPES OF
STIMULI THAT PRODUCE RIGHT EAR (RE), LEFT EAR (LE),
OR NO BETWEEN-EAR DIFFERENCES (ND) IN REACTION
TIME TO DICHOTIC STIMULI................................... 10

TABLE 3. SYNOPSIS OF VISUAL HALF FIELD STUDIES (VHF)
WITH RIGHT-HANDED ADULTS ACCORDING TO THE TYPES
OF STIMULI THAT PRODUCE RIGHT VHF, LEFT VHF, OR
NO VHF DIFFERENCES (ND) FOR THE RECALL-RECOGNITION
OF VISUAL STIMULI ............................................ 11

TABLE 4. SYNOPSIS OF VISUAL HALF FIELD STUDIES (VHF)
WITH RIGHT-HANDED ADULTS ACCORDING TO THE TYPES
OF STIMULI THAT PRODUCE RIGHT VHF, LEFT VHF, OR
NO VHF DIFFERENCES (ND) IN REACTION TIME TO
VISUAL STIMULI.......................................... .. 13

TABLE 5. SYNOPSIS OF DICHOTOMOUS STUDIES WITH RIGHT-HANDED
ADULTS ACCORDING TO THE TYPES OF STIMULI THAT
PRODUCE RIGHT HAND (RH). LEFT HAND (LH), OR NO
BETWEENI-HAND DIFFERENCES FOR THE RECOGNITION
OF TACTILE STIMULI .......................................... 14


TABLE 6. DICHOTOMIES OF LATERALIZATION............................. 17

TABLE 7. SUMMARY' OF STUDIES USING "PRIMING" TASKS TO
TEST THE COGNITIVE SET HYPOTHESIS .......................... 51

TABLE 8. SUMARPY OF STUDIES USING INTERilIING PARADIGMS TO
TEST THE COGNITIVE SET HYPOTHESIS.......................... 55

TABLE 9. MEAN REACTION TIMES IN THE GO-NO GO AND SIMPLE
RT TASKS ................... ............................. .. -74


SUMMARY OF ANALYSIS OF COVARIANICE OF THE IlEAIN
"GO" REACTION TIMES IN1 THE GO-NO GO CONDITION......


......... 75


TABLE 10.







TABLE 11.



TABLE 12.



TABLE 13.



TABLE 14.


TABLE 15.



TABLE 16.


TABLE 17.


TABLE 18.



TABLE 19.



TABLE 20.


MEAN COVARIATE VALUES OF THE RIGHT AND LEFT
HANDS FOR VERBAL AND NONVERBAL SESSIONS IN THE
GO-NO GO CONDITION........................................... 75


MEAN REACTION TIMES OF RIGHT AND LEFT HANDS
WITH WORD OR FACE WARNING STIMULI IN THE
GO-NO GO CONDITION ...................................


DUNCAN'S POST HOC COMPARISONS OF MEAN REACTION
TIMES OF RIGHT AND LEFT HANDS WITH WORD OR
FACE WARNING STIMULI.........................................78


SUMMARY OF ANALYSIS OF COVARIANCE OF THE MEAN
"TONIC" REACTION TIMES IN THE GO-NO GO CONDITION......

MEAN REACTION TIMES OF RIGHT AND LEFT HANDS TO
AN UNWARNED TONE EMBEDDED AMONG VERBAL OR NON-
VERBAL WARNED TRIALS IN THE GO-NO GO CONDITION........

SUMMARY OF ANALYSIS OF VARIANCE OF THE MEAN
"PRE-POST" REACTION TIMES IN THE GO-NO GO CONDITION...


MEAN "PRE-POST" REACTION TIMES PRIOR TO AND
FOLLOWING THE MAIN TASK IN THE GO-NO GO CONDITION....


.... 79


......81


........82


MEAN "PRE-POST" REACTION TIMES WITH EITHER VERBAL
OR NONVERBAL INTERVENING TASK IN THE GO-NO GO
CONDITION................................................... 82

MEAN REACTION TIMES PRIOR TO AND FOLLOWING EITHER
A VERBAL OR NONVERBAL INTERVENING TASK IN THE GO-
NO GO CONDITION............................................85

DUNCAN'S POST HOC COMPARISONS OF MEAN REACTION TIMES
PRIOR TO AND FOLLOWING A VERBAL OR NONVERBAL INTER-
VENING TASK IN THE GO-NO GO CONDITION....................... 85


TABLE 21. FREQUENCY OF ANTICIPATIONS BY RIGHT AND LEFT HANDS
WITH VERBAL AND NONVERBAL WARNING STIMULI IN THE
GO-NO GO CONDITION......... ...........................


TABLE 22.



TABLE 23.


......86


FREQUENCY OF OMISSIONS BY RIGHT AND LEFT HANDS
WITH VERBAL AND NONVERBAL WARNING STIMULI IN THE
GO-NO GO CONDITION......................................... 86

FREQUENCY OF COMMISSIONS BY RIGHT AND LEFT HANDS
WITH VERBAL AND NONVERBAL WARNING STIMULI IN THE
GO-NO GO CONDITION.......................................... 86


SUMMARY OF ANALYSIS OF COVARIANCE OF THE MEAN "GO"
REACTION TIMES IN THE GO-NO GO CONDITION..........


..........88


TABLE 24.


..... 78







MEAN "GO" REACTION TIMES WITH WORD OR FACE
WARNING STIMULI IN THE SIMPLE RT CONDITION.................... 88

SUMMARY OF ANALYSIS OF COVARIANCE OF THE MEAN
"TONIC" REACTION TIMES IN THE SIMPLE RT CONDITION............ 90

MEAN "TONIC" REACTION TIMES EMBEDDED WITHIN VERBAL
OR NONVERBAL TRIALS IN THE SIMPLE RT CONDITION............... 90


TABLE 25.


TABLE 26.


TABLE 27.


TABLE 28.


TABLE 29.


.......... 91



.......... 93


DUNCAN'S POST HOC COMPARISONS OF MEAN REACTION TIMES
BY RIGHT AND LEFT HANDS PRIOR TO AND FOLLOWING AN
INTERVENING TASK IN THE SIMPLE RT CONDITION................


FREQUENCY OF ANTICIPATIONS BY RIGHT AND LEFT HANDS
WITH VERBAL AND NONVERBAL WARNING STIMULI IN THE
SIMPLE RT CONDITION................................


........ 94


FREQUENCY OF OMISSIONS BY RIGHT AND LEFT HANDS
WITH VERBAL AND NONVERBAL WARNING STIMULI IN THE
SIMPLE RT CONDITION........................................


SUMMARY OF ANALYSIS OF VARIANCE OF THE MEAN "PRE-
POST" REACTION TIMES IN THE SIMPLE RT CONDITION....

MEAN "PRE-POST" REACTION TIMES BY RIGHT AND LEFT
HANDS PRIOR TO AND FOLLOWING AN INTERVENING TASK
IN THE STIMPLE RT CONDITION ........................


TABLE 30.


TABLE 31.


TABLE 32.












LIST OF FIGURES


FIGURE 1. MEAN "GO" REACTION TIMES OF RIGHT AND
LEFT HANDS WITH EITHER WORD OR FACE
WARNING STIMULI IN THE GO-NO GO CONDITION................. 76


FIGURE 2.




FIGURE 3.



FIGURE 4.


MEAN "TONIC" REACTION TIMES OF RIGHT AND
LEFT HANDS TO AN UNWARNED TONE EMBEDDED
AMONG VERBAL AND NONVERBAL WARNED TRIALS
IN THE GO-NO GO CONDITION ................................ 80

MEAN "PRE-POST" REACTION TIMES PRIOR TO
AND FOLLOWING A VERBAL OR NONVERBAL
INTERVENING TASK IN THE GO-NO GO CONDITION ............... 84

MEAN "PRE-POST" REACTION TIMES PRIOR TO
AND FOLLOWING AN INTERVENING TASK IN THE
SIMPLE REACTION TIME CONDITION............................ 92


viii













Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



MATERIAL SPECIFIC HEMISPHERIC ACTIVATION

By

Dawn Bowers

December 1978

Chairman: Kenneth M. Heilman
CoChairman: Paul Satz
Major Department: Psychology

It has been previously demonstrated that binaurally presented verbal

warning stimuli (WS) asymmetrically reduce reaction times (RT) of the

right hand to a midline neutral stimulus. However, nonverbal auditory

WS reduce RTs of both hands equally. These findings have been inter-

preted as providing partial support for the hypothesis that verbal and

nonverbal materials differentially activate and prepare the hemispheres

for action.

The purpose of the present study was: (a) to determine whether ver-

bal and nonverbal WS in the visual modality differentially reduce RTs of

the hands; (b) to determine the extent to which this asymmetric RT re-

duction is temporally related to the WS; and (c) to determine whether a

response-linked decisional process is necessary for inducing these asym-

metries.

Thirty-two dextral college students performed a RT task in which

manual RTs were obtained to a binaurally presented tone (500 ms) that was






preceded at random intervals (500-1500 ms) by either a word or face WS.

The WS were tachistoscopically presented at central fixation for 250 ms.

During any one session, only one type of WS (verbal or nonverbal) was

presented, and right and left hand responses were measured.

If verbal and nonverbal materials differentially activate and pre-

pare the hemispheres for action, then verbal WS should selectively re-

duce right hand RTs and nonverbal WS should reduce left hand RTs. To

determine whether the asymmetric RT reduction, induced by the verbally

or nonverbally warned trials, persists on the order of 15-20 sec, 48

unwarned trials in which no WS preceded the RT stimulus were presented.

These unwarned trials were randomly embedded among the 192 warned trials

of each session. Thus, if asymmetric activation is persistent during

the session, the RT asymmetries should also be obtained for these un-

warned embedded trials.

Prior to and following each verbal and nonverbal session, simple un-

warned RTs were also obtained for each hand. This was done to determine

whether asymmetric hemispheric activation might extend beyond the verbal/

nonverbal session itself and affect post-session trials by inducing RT

asymmetries.

Sixteen subjects (Ss) were assigned to a "go-no go" RT task in which

half the WS during any one session signalled that a response should be made

to tone onset and the remaining WS signalled that no response should be made

Four sessions of 192 trials each were administered in the "go-no go" task.

In two sessions, words were used as WS and in two sessions, faces were used.

The other 16 Ss were assigned to a simple RT task in which responses were

made on every trial. Two sessions (one verbal and one nonverbal) of 192

trials each were administered in the simple RT task. If the predicted RT








asymmetries are found only the the "go-no go" task, in which a decision

to respond is based on the information conveyed by the WS, this would

suggest that the decisional process is important for inducing asymmetric

hemispheric activation (i.e., preparation for action or readiness to

respond). If the predicted RT asymmetries are also found in the simple

RT task, this would suggest that asymmetric activation is not decisionally

dependent.

The results showed that: (a) In the "go-no go" task, verbal WS asym-

metrically reduced RTs of the right hand more than left hand RTs. How-

ever, nonverbal WS reduced RTs of both hands equally; (b) For the embedded

unwarned trials, there was only a trend for faster RTs by the right hand

when verbal WS were used, suggesting that asymmetric activation is tempo-

rally related to the WS. For the post-session unwarned trials, the pre-

dicted between-hand asymmetries were not found. The RTs following the

verbal sessions were slower than those following the nonverbal sessions,

suggesting that verbal/nonverbal WS induced differential fatigue effects;

and (c) In contrast to the "go-no go" task, no RT asymmetries were found

in the simple RT task, suggesting that asymmetric hemispheric activation

is linked to a decisional process.

These findings are discussed in terms of the implications for atten-

tional versus structural models of perceptual asymmetries on laterality

tasks. The role of right hemisphere attentional mechanisms is also ad-

dressed.













INTRODUCTION

Overview of the Problem

An extensive body of data collected over the past century has been

generally interpreted as supporting the notion that the two cerebral hemi-

spheres of man's brain are differentially involved in linguistic and cog-

nitive functions (Gazzaniga, 1970; Geschwind, 1970). Based on current

ways of interpreting structure-function relationships, it appears that

the left hemisphere processes most, if not all, linguistic stimuli and

the right hemisphere processes many, if not most, visuospatial stimuli.

This hypothesis concerning cerebral laterality function was ini-

tially drawn from clinical studies of patients following unilateral neu-

rologic insult to the left or the right hemisphere. The development of

various lateral perceptual stimulation techniques approximately 20 years

ago enabled researchers to address for the first time cerebral laterality

of function in normal intact individuals, thereby circumventing some of

the difficulties inherent in the classical "lesion" approach to this prob-

lem. Research using these laterality tasks with normal adults has demon-

strated that language stimuli are more readily perceived when presented

contralateral to the left hemisphere and that nonlanguage stimuli are

better perceived when presented contralateral to the right hemisphere.

While it has been widely accepted that the perceptual asymmetries

obtained on such laterality tasks reflect the divergent functions of the

two hemispheres, the precise neurophysiological mechanisms underlying

such asymmetries has been controversial. At least two such mechanisms




2


have been proposed to explain perceptual asymmetries. One is based on

a structural anatomical model (Kimura, 1961, 1967) and the other is

based on a more functional attentional-activational model (Kinsbourne,

1970a, 1975).

The present study represents an attempt to experimentally differen-

tiate between these two models of perceptual asymmetries and specifically

addresses the hypothesis that verbal and nonverbal materials asymmetri-

cally activate and prepare the hemispheres for action. Since the evalua-

tion of this "material specific hemispheric activation" hypothesis

bridges two main content areas, that dealing with cerebral laterality

of function and that dealing with attention and its underlying neurophys-

iological mechanisms, each of these topics will be addressed in the fol-

lowing sections.

Laterality of Cerebral Function

Historical Perspectives

The view that the dual function of the human brain relies on a func-

tional verbal-nonverbal dichotomy is historically rooted in the early lo-

calizationist school of classical neurology. During the early half of

this century, ideas regarding cortical functioning were dominated by two

"streams of thought," that initiated by Broca (1861) and Wernicke (1874)

which emphasized localization, and that including Head (1926), Goldstein

(1927, 1948), and Lashley (1929) which was anti-localizationistic (Gesch-

wind, 1963, 1967).

The rapid historical movement towards the idea that different cog-

nitive functions are localized in different parts of the brain was

prompted by Gall in the early 1800's with his theory of phrenology (Blake-

more, 1977). It was not until Broca's (1861) discovery, that anterior







lesions of the left hemisphere produced expressive language disturbances,

that the localizationist school gained its first scientific respectabil-

ity. Subsequently, Wernicke (1874) identified another form of aphasia

that was produced by more posterior lesions in the left hemisphere. Not

only did Wernicke provide new evidence for the localization of aphasia,

he also advanced a theory which tied these phenomena to existing know-

ledge of anatomical fiber tracts in the brain (Geschwind, 1963, 1967).

On the basis of Wernicke's theory, it was possible to predict the exist-

ence of clinical syndromes not previously seen. It was the theoretical

aspect of Wernicke's approach that gave rise to the "golden age of neu-

rology," during which a myriad of neurological syndromes were identified

over the next 40 years and related to lesions of particular cortical

areas.

The anti-localizationists, on the other hand, argued that one could

not pinpoint one-to-one relationships between language and cortical func-

tions. One of the foremost proponents of this view was Goldstein (1927,

1948), whose psychological approach to the study of brain injuries coin-

cided with the academic gestalt movement in this country and Germany.

Goldstein stressed the "equipotentiality" of the brain and argued that

differences in clinical syndromes might reflect the extent of brain dam-

age rather than the precise site of the lesion. In particular, he empha-

sized that abstract categorical behavior, i.e., the "abstract attitude,"

was more dependent on the actual mass of brain tissue involved than on

the participation of discrete cortical areas.

According to Geschwind (1964), the seemingly contradictory perspec-

tives of the localizationists and the holists are somewhat artificial

and misleading, in that the similarities between these two points of







view far outweigh their differences. The holists, for example, supported

the main anatomical tenets of the localization school, and like the lat-

ter, even postulated the existence of "conceptual centers" which corre-

sponded to the "centers" and schemass" to which they objected (Geschwind,

1964). What remains indisputable is that both views acknowledged that

verbal functions were more easily disrupted by lesions of the left than

the right hemisphere. This observation consequently led to the hypothe-

sis that language was lateralized to the left hemisphere and that non-

language functions were lateralized to the right hemisphere (Jackson,

1874).

Until approximately 20 years ago, research in cerebral asymmetry

of function was limited to the study of individuals following neurologic

insult. This clinical approach to the neural localizations of function

is based on the association of cognitive deficits with known cortical

lesions and is subject to logical limitations. In studies of patients

with discrete cortical lesions, one is never assessing directly the

function of the lesioned area, but rather inferring its function from

the manner in which the remainder of the intact brain has compensated

for its loss.

This inferential problem is further compounded when models of "nor-

mal" brain functioning are then based on findings from neurologically

deviant populations, i.e., split-brain patients whose neocortical coIn-

missures have been surgically sectioned as a "last resort" attempt to

control seizures. A voluminoLu almoulnt of research over the past decade

using such "split-brain" patients has given rise to models which attempt

to explain the divergent function; of the left and right hemispheres

(Nebes, 1974). The problem which emerges is that the majority of these







individuals have had long-standing seizure disorders, which greatly mag-

nify the probability of major cortical reorganization. Data directly

bearing on this caveat has been recently provided by Whitaker (1978)

from a series of cortical mapping studies carried out on epileptics

prior to the removal of seizure foci. In Whitaker's series, the vast

majority of these epileptics had grossly deviant topographical arrange-

ments of motor, sensory, and language areas within the left hemisphere.

Such findings suggest that one should proceed cautiously when generating

models of brain function based on pathological populations.

The development of various lateral sensory stimulation tasks, i.e.,

dichotic listening, visual half field procedures, dichotomous stimulation,

has enabled researchers to assess cerebral laterality of function in nor-

mals. Use of these laterality tasks has involved the presentation of

sensory stimuli (i.e., visual, auditory, tactile) to one or both lateral

sides of the body (i.e., ear, visual half field, hand), to which a sub-

ject is required to make some discriminative judgement about its occur-

rence (i.e., recall, recognition, reaction time). In dichotic listen-

ing (DL), initially developed by Broadbent (1958) for the study of atten-

tion and later adopted by Kimura (1961), competing auditory stimuli are

simultaneously presented to each ear via stereo headphones. Similarly,

visual half field (VHF) tasks involve the tachistoscopic presentation

of visual stimuli to either or both left and right VHFs. Finally, di-

chotomous tactile stimulation procedures, recently developed by Witel-

son (1974), entail the manual palpation of different three-dimensional

forms by the left and right hands.

Most investigators have accepted Kimura's (1961, 1967) initial

claim that the perceptual asymmetries obtained on such tasks reflect







the underlying asymmetric functions of the two hemispheres for proces-

sing verbal and nonverbal materials. There is, in fact, so much evi-

dence in favor of this hypothesis that it would be difficult to refute.

It has been repeatedly found that language stimuli are better detected

when presented to the right than left channel (ear, VHF, hand), where-

as nonlanguage stimuli are better detected when presented to the left

channel. The right-sided advantage for language stimuli and the left-

sided advantage for nonlanguage stimuli have been thought by some in-

vestigators to reflect the most direct access of stimuli, via contra-

lateral sensory pathways, to the hemisphere most specialized for pro-

cessing them (Kimura, 1961, 1967; Moscovitch, 1973). Other investiga-

tors have attributed these perceptual asymmetries to attentional mecha-

nisms (Kinsbourne, 1970a, 1975).

Over the past decade, the focus of laterality research has taken

several different directions. At one end of the spectrum are a myriad

of analytic studies which have attempted to identify various aspects of

stimulus parameters which give rise to perceptual asymmetries. Some of

these parameters have included, among others, various acoustic, phonetic,

semantic, syntactic, and configurational aspects of stimuli. A synopsis

of the major findings of some of these studies using right-handed adults

as subjects (Ss) is provided in Tables 1-5. At the other end of the

spectrum are studies concerned with individual differences in perceptual

asymmetries and with their relationships to other behaviors, i.e., learn-

ing disabilities (Satz, 1975), sex (McGlone & Davidson, 1973; Waber. 1976;

Witelson, 1976), handedness (Fennell, Satz, Van Den Abell, Bowers. &

Thomas, 1978; Levy, 1974), and personality characteristics (Zoccolatti &

Ottman, 1978).














C
C 6
CC

ULL---



1- r' (



N (/A I


01 -a a
-0 0


0 r ~ 4c
S -- ro _0 >



CL U



SO 0000CD W NC L -.
---.-- O --'-.--.-. -,- -u ( -


CcrC(Cnicn ro o
c *,1.- -.- r .r -- "- *r- CU r
C Wr- o -s Io Q ~o
k- k- 10 C oa 3 S-
M- -05 c- a -- '- fL U


S- S- a a) a ) cW WW C -o Q x< N C Ul
s- LS- s- c c C C C E S- C 0 4-' ra c
3 a C ) *- *- *- *- *- 0 *'- C ra C ra
OU(J M r LrrrrLi- ^c ,) YZV V I=


LUJ
LUJ
I




LLUJ








I--
CO


CZ








S---
S- '













F-LU



C) C)



















O-
O C)
- LJ I--




0










C 1








LU C).)
L)-F-


C) 0
1-4 "--







t- U





.-J 0i LlJ
U-c-










C) L







_F- cU
LUi
F-





a)
zw











00







CCvv



r rar
CC
-( -






CCC 'hh
a) a)0- 0
) .L O -0
I I --
L 4J LLM


55 -C*L
o c -- 4 .a)

4-' 4-' ) M 0
S > O L-) (D


4-J
Cra


O P
C *4-
0
C:

0 *
0--- 4-) -

4-j r-


C- 5- c .

---- 0 a)-
4- U .
.- (A ra
(A C:4-) 4- --
4-' *r- C n V r
C ro 4-- ra (A (A W
ro CI C (A C -' >: (A
C-- 0 ra 0 (A C-) ,--

( c) C 4-' C C) a- 0 C
C c- O C OC '--1 >a)


U L I U (Ar-- -0 '
a- - 0 >v
0Cl- W 4-- ra J > 3 c
o4- 3 C C : C 0- c3
4-) r- 0 o *r- 0 3-
(- > C 4- > -
1 0 1 1 4-
~ -' (- 1- C +- -
= a) c C c4- C ra c 3
U-J r 0r- r0 r
/l (/ (n/ (n in a -0 C U 3 C -O C (/l
4-)4- 4-4-i)J rO 0 W 0'-
. -r 4-* (n (n (n U (A
--,-, CI- c L-) C C 3O
*.- *I- *.- -.- *- >) > 0)0 0 0
M oooo LV iL ><-)>


(A (A
0 0
C C

COCO

O O
C 4-' c W
U 4

r 4J-' QJ U
S.- n S- C (n
(Al 4 -Q U 00

J 0 0
L O O0
Or.- C U-
3 ra0 r

C *5- *r- *r- *r-

-0 -0 *r- a *r- r U
S- S- 4 E E c

O 0-C O C ra C O
33: Q =U- L-=


-L


4- O
3
ca


*i Q(
C C

(aC
WO 0
: Z




























SCO

r-C C

0-) rl_ c 0
-O -C -
-.- LL.

S.- - 0.

r- L- CO-

S *r- ,. c"
U0 C C
.. 4-U .-i 0

E S- -) CS -
*r-"3 "- O Q.0


























CL
C )
e

JC 0

=0 .0



d v--
U 04
0 *O -
( A







On C0 .C
C di C 4-

SU 4-

0 u C

0 T- *-



S .. 4-n C C -
0.. U i r/- C
O ,- -r
C 0 )S-


Q- c *o- *
Li C C 'r-

diC o *.- s e

S. *- 3 r- a
(- 0i S.- r (A



C3 S.- C4- >,
0c 0L LI 2


_0 Zf
i en


- C


0Oh CD
O C r- O
* .- o r- c
O V)
U C" -0 0-' o C



CI >-- S C ,-- r---
*r- m C r- :---- o .
-4 I C "-- c


E Mr 0 S X-0-0 C-
3 Li S- S- 0 S.- S- E di ,--
r- A4- 3 C O O C di 0 C
SO WJ_ O (D CO : 3-D U


4-'
S-

0
C-
L.

0 0 (
OV 0
s-.. C C -


4-- ) e-
ac-n a- e2
O 5 C C- S- -

C-' C C -C -C
OC U0 JU
*r- o EE (A
C C r--

C C ..C S.- S- U U-
O U U *Ir- *- *r- *r- 0O

C *r- *r- C C 3 3 di
Scl. C1.LlJLU 2 SC


(U
V)
*-c
O






C
r-










(n

014--'
r-

0










0 1 4-'
v (A






* S*- I
= n0 L

0 *r-







-i-"
4- E C 4~.
- *r- e_
S- *r



oE


hO



r- or-


C d-





4-' r~
I Cl

S.C-


= OJ







r--
+- 4-' C
Si S.- d
-0L
-0 oa Li
0C 3

-I VC)
-0) 0o
C/) +-/)!


(n
*-- )
0- U
tC



me i
S-

> *r-






0L



4- -'
C
lm



4-Q


O

d4-'
U S
0a I
-- *i

0 0 E-

> > Le)






9
















0




-- -
C~



r -, 0 r-.




(t >- -a) -- O
Sr-
on c






4- C C


















U







4-'
*r- C -
t) C i- -1J O )















L CC S-






















E


4--
(nW










W 4-














r J ]
c









S r

S4J
-r



E
Ln
v)






















S- 4- 0 -
O C/ O -0'
0- 3 CL










(1 C U-
4- 0 *o C m
0 ) 0 i nr-
=3 v
CT U *r- Cn
0) *r- 3
Q E cn) -r-



C C O (0- C
I UT C



r-- O S- 4 *r- C

E i >U-coE
*r- 0 1 S
S 0 O *r-O (U n
<) 1-O Z Q S-- 3I






10













LU
LU

LU


z -


0 d-
o



-- LLO T 0 So- a -

Qj -j 0- o r-


OLU 4-C)) :nW 0C
cl) 0








H- "0C)
M -j 4J on Ln OZ5 c





F- -C

C LU Cl


(_)


= --l L)









---
2 0 OU

L) I




LUJ 0 LJU
fC\ > 0



- 1- I 7: )





< 2 '- CD .-



c,- c

SH- LU 0







H-- C 0)












CD)

I>
4-' 0<-4-
U U 9 CI -










LO
-- r-


S-
E-
c-_ s-





U OU
(-)



S- r- r-




U- 1) a)

0 --- n "-

4- 4-" r s rr- O o o

4- 4- Mo L- -a -- M

( ( ( ( ( (
U) U) C) U) cr U) r) -- ( 0 r fD o.


*r- .r 0 L .r-r- U) L U) LS-
=.---r-:Cc, :3_. _jJCO


V) -~--00 -
s- r _




r D u-
" 0 -- r2 U-.-- s (--




c- Cr- cr- D -- -- L -C -



C c cS- S- -0 S
03 0 0 ra T- C >C .-

0003:ca3:c3 c3 Co3n3:S


zz


00


UIJ
LLJ




MI,
I-u--




LLc

-I C


30










O0
_CD LL.
-LL-
Q0


0











(-) C3
bibCD






D
I- --
0 "- SZ














(DQ
Q- C> I-
1-r



OL
L. j"
3 I-
LL. -I
11LLJ











00


JOJ
LLJ -" >_












LWO


11
OO
000




C0 > U


0>-
LLLL





I >-






O-
CD
00


---


0
I -
U-=: 3
QJ3 -


0L S-
.'- 4-'
S- c
S0
(U
CD-

4-.


in -- X
E -r- *l
4-J



S I
V/) V) 4-3 4-J
_0 E *- C -:3
S- M S- A- 0 V
0 S- 4J S- -9 a)
3 n 04 4- (Z

(D E 4- r r-

W S- 0 (A ) W W =
L c'n l) (A 0o CL C 0
C -r-. 3 c s- n3 M E
O 0 O 0 -=
Z- )> (-) a0
ZH>U~C CI)L


-


0- I-
S- > no
S- 4--




0 S-) 0 *
O L r- *- )


0 o 0 O .. *"- S
4-' 3 0 C-

> 0 a (A
ou >, ,aa=


0) w cr 0- = W o W
*--- O a ia 0
CA 0 S- S- )0 > O
4- 4-W 4- 4- -0 4 M ()

*- *r- 4 +- -0 )- 0 0 -!= r E C =r V)
C0)r C7H4-)+-> 4- S. L 0 0r 1 3: 3 Q -Q C
r- *r- *- 1) o 0 c'u 0 0 E 0
QC~3C3_13333c3:0























co

Sr-

r- r--a C J a) t-






r- L (1) a 4 -r -0
= 1 -- o~ r- r -C
r-- Zc E 4 O oZ 51


r-. 00c0IE C C v,--
-- I ( r --
S 4--- V 0 E v6- m
"-(,C) C
SU 0 O c 1 -- o Z3
-L 0 o 4 o O o LL 4- E
- d.- C I O C- i-


-- --- U ) > V: ) ?
'O 1 0) 0 Cd M Cd
*r- r- -- O V) Qm m1- L
"r".0.L JL-DOQcLno


C


u




4L ) .4-

Lv 3 0 *- C)


'- Li a) I -,
4-)
Lo 14- U '4-- L.--
c ) *- C -- 0
U 'LA v 0 '-
CO d3 '3 *- U
0 ,- -) ,'- 0

4-o L a4- a c- -
1 0 .-13
4-) 4-) a a 4-)
S.I 'LA 'A.. C: Q -O iLn C
a) a IDa> ww = -'1a )
U U *' 0 C IU C0 C-
. ro1L c U 0 0 -
U. U. .LL- C :> M a:


C)
O"1


C')


(3)- C
CD








0 a
-- '--
CCO








E E -- -0 W tC N a) () -0 C N
CO (a >) > aC U a a C:

- -- Sr- S- r-- a) -r ro

'- I L --

EE E-oo D ra cu a cro
( n )C- U C C-.r- C L
t 5o 5.s -r- -i *r- "T- __ -r--3
c

a)

Cd
0-

4-O
--J






OO

0*
L4- 4-- -

C 4-0
O f

c
L..
4-- .r-- r


L




o 0 O
-a



S- C
0
S)









U O C T-
0- 2









a) C -0










U C .
0 c



- O3 ,-
C Cd



IZ l) (-0 .
> 0 0) en
0 U C


L -CL OC' LU
-) UC 0 0
C 0 U 0

U 'I- 0. rO 4-
0 C -C
C Q 1 3 i-- <1)
0 E 0 E O l
4._ Cn .,.. C Cn O
Cd '4- do t-CO
X U a) 3 CT

14- A 3 l- 0 **-

-CE .C to I C0 d-
S4-) G O 0
.- '4- -*- U C *-
S O 3 C- 3-

'. L'A L' LA 0 0 CT.
2 C o) a *= '.. a)


ULL U ~3 0.. L L


















O u
U-0 =
r~CO S o

UC'- --c0 r- i a)
a) 3 r .- r.--
U-Oh coo' LO

i- a)- n-a)- N n
r _- C *r- .- -
n3 C >) r3-
2 cz a)
E r3 3 -. S-3-

4-) -C c.O 4-) r" Ln
3 3- - M u *-- N -0
cv .-' U -)C (a)






54- Nl- UC)- '4.-' ) 4 C
+3 r- 4- I- u n3 m
-o ra > ^ U r N c
n3 4-) 0 N (r 4-) oV




Ea 0 E +
L D L DU:-,-Z :r- +D


C3G;3S:^3:Q=> 0


0



LU
w

UJ
Cr
LL
- LL
:--LL

D Q

r-

0




Ir>





F--
SLL




I-








C3 I-




ZD
SLLJ










r1 U i-<




M 0
1 -



LL-









OF-
I-L



-i-
LO
-10 -


L)I
SLU



Fi-i
C-D






L3:LL




D0
_I-_









en
C)


0 0 aU
4-) 4-) -U
4-) 4-) S-
0W )0
.. 3
000C~


r-






OJ
a)



4-)


-0 4-
C


O n3
S- 0-

C 0 (a
(U .-
1) -C 4-'
mUL
a) 4-) 1)

r- E


-c n" "
C --


n3 C n3
:0 C

) *r- O
*- N
r- a) *,-

UZ


-' .+4-)



r- o


*- 06
N
N

.) a )









>, >
CI o CO 0



UO h- 4) C
o- r- c )











7- U L
4-. --. u r-









4- a) s 0 0- c
rs -. ,-..-^ o {r O3 r--.-
oh i- -- c .n o r .~
S- S) - ro l ogJ
3-''-'- :3 n3 > O"-- .
-- -- ( S- N 4 U


*r- -3-C U WL
N-4-4-- U C 0- .r- 4--
4-' a)j, ooa.C C

L-r- C > S-',- 3 (U
U C a) C o a) s- (') oC+-O C
3 N I '-.i- U 4-' a)r- 4-'- 0-
- NI a) 4-) -o > S- r- C
*ir a) O 'U 3 =r aW E O
=D-2ro-o CDco3 -


UUUUU
r1 (v rL r( (a
LL- LL- LL- LL- LL-


a)




U,
a)
4-


0 =
0O 4-)









0 4



C0 0




4-0L- -

*^(>0>
C n3 4-)





*-- u
S4-) a) 0 W
C 0 -
m -C W o



C U 0-' U: C

1) 4-) U, C *r- a)
U C >*- i- C
Ca E L C a) .*-.
LL- LUJ 0- en0 -)


4-> 4-)
4-) -)
_0_
Mc
-1-



























00- r-






4- -

r:- 4 -'

Ocr a)


'.,-- U

4J 00

-. -I-, U

-- ,) 0
CDoon









(/)




s-
,J







Ll-



0
2:































C
4r_











































0
-4-
S-
Sr-

















































a) 0 'n
m LL
,'D ,

._ -7 LI


4-,


4-'








ro
S-O
ra)










-C E
"o .-




OO





.-i m


r
Z
2-










cm
LI







Z2




C C)


-i J
_W.




:-) -r
u3






-JO-







cL)













SLUL



CD CK.





d- LL L
C-
I b ^l





























Cor)U 3
C3D--





























2-
I--
C3D













2: (^
0 -L









I-


LU


-J
':1


CC CV

Wa





0 0
;a)
cyc
C/C C/C

ID )


E
4-'






4-'


4-'







a(
4-3









4-'
7)









S-
0
4-








0)
0-















0
















a)
-0






4-,











s-






(3

4-







Current Models of Cerebral Laterality of Function

Two laterality models have been proposed which attempt to describe

the functional specialization of the two cerebral hemispheres. According

to one, each hemisphere is specifically organized to mediate different

types of stimuli. This "stimulus-specific" model maintains that verbal

stimuli are processed by the left hemisphere and that nonverbal stimuli

are processed by the right hemisphere (Kimura, 1967). The second model

postulates that both hemispheres can mediate all types of stimuli, al-

though each employs qualitatively different processing strategies (Gold-

berg, Vaugh, & Gerstman, 1978). This "strategy-specific" model main-

tains that each hemisphere has its own characteristic processing systems

for analyzing component parts of a stimulus.

According to the "stimulus-specific" model, there is a pronounced

separation of linguistic and nonlinguistic skills into the left and right

hemispheres, respectively (Kimura, 1967). Underlying this model is the

implicit assumption that the two halves of the brain are similar in the

degree of their neural specialization. Namely, different functions are

localizable to the same extent within each hemisphere. According to

this assumption, one would expect laterality effects to be equally con-

sistent in terms of reliability and magnitude for the left and the

right hemispheres. Effects of this sort, however, have not been obtained.

Laterality effects for nonverbal materials with neurologically intact

adults using either DL or VHF procedures are generally smaller than ef-

fects with verbal materials and less consistently obtained (White, 1973).

In addition, the effects of brain damage to the right hemisphere are much

less clearcut than effects of damage to the left hemisphere (DeRenzi &

Faglioni, 1965).








Furthermore, the "stimulus-specific" model has produced few detailed

explanations of the mechanisms underlying such specialization. Semmes

(1968), for example, has stated that:

Unfortunately, the concept of cerebral dominance is
not helpful for it proposes nothing about mechanism;
in fact, it is little more than a label, a restate-
ment of findings that lesions of one hemisphere pro-
duce deficits that lesions of the other hemisphere
do not. (p. 11)

Just as the label "cerebral dominance" fails to explain, so the la-

bels "linguistic skill" and "visuospatial skill" offer little help in

understanding the formal nature of the abilities referred to (Marshall,

1973). Vygotsky (1965) claimed that classical investigations ". .failed

to arrive at an adequate solution of the problem because of the lack of

structural-psychological analysis of the functions they try to localize"

(p. 381). He further argued that ". . the problem of what can be local-

ized is not at all irrelevant to the problem of 'how' it can be local-

ized in the brain" (p. 382).

Several hypotheses have been proposed which attempt to explain hemi-

spheric specialization in the spirit of Vygotsky's remarks. Recent can-

didates for the mode of operation of the left and right hemispheres, re-

spectively, include: Analytic versus Gestalt processing (Leiy & Sperry,

1968); Serial versus Parallel processing (Cohen, i972); Categorical ver-

sus Noncategorical judgements (Goldberg, Vaughn, S Gerstman. 1973); Simi-

larity versus Dissimilarity judgements (Egeth & Epstein. 1972);, lame

versus Physical Identity judgements (Cohen, 1973; Klatzky, 1972); and

Categorical versus Place judgements (Veroff, 1978). Historically, these

modes of information processing, which have been attributed to the two

hemispheres, have been labelled in many different ways, some of which are







TABLE 6

DICHOTOMIES OF LATERALIZATION

Goldberg, Vaughn, & Gerstman (1978) Categorical-Noncategorical

Veroff (1978) Categorical-Placement

Cohen (1973) Serial-Parallel

Egeth & Epstein (1972) Similarity-Dissimilarity

Klatzky (1972); Cohen (1972) Name-Physical Identity

Bogen (1969) Propositional-Appositional

Bogen & Gazzaniga (1969)* Verbal-Visuospatial

Hall & Hall (1968) Analytic-Gestalt

Levy & Sperry (1968) Analytic-Gestalt

Hecaen, Ajuriaguerra, &
Angelergues (1963)* Linguistic-Preverbal

Zangwill (1961)* Symbolic-Visuospatial

Semmes, Weinstein, Ghent, &
Teuber (1961)* Discrete-Diffuse

Milner (1958)* Verbal-Perceptual

McFie & Piercy (1952)* Education or relations-


Humphrey & Zangwill (1951)*


Anderson (1951)*

Weisenberg & McBride (1935)*

Jackson (1876)*

Jackson (1874)*


Jackson (1864)*


*Taken from Bogen (1969)


Education of correlates

Symbolic-Visual
Propositional-Imaginative

Storage-Execution

Linguistic-Visual

Propositional-Visual imagery

Audio articulator-Visual
imagery

Expressive-Receptive








found in Table 6. Whether these presumed modes represent identical,

related, or independent hypotheses regarding a formal mechanism is not

known.

Thus, a "strategy-specific" model of cerebral laterality of func-

tion maintains that each hemisphere has its own "descriptive systems"

for processing and encoding stimuli (Goldberg, Vaughn, & Gerstman, 1978).

These putative "descriptive systems" do not necessarily coincide with

the simple language-nonlanguage dichotomy which the "stimulus-specific"

model of laterality maintains. Rather, the emphasis of the "strategy"

approach involves a delineation of the processing characteristics of the

two hemispheres, regardless of the verbal-nonverbal or modality classi-

fications of stimuli. From this perspective, laterality effects would

depend on the degree of relevance that classes of stimuli have to exist-

ing "descriptive systems," thus forming gradients of relative left-right

hemispheric involvement in processing these materials.

Underlying the "strategy-specific" model are the implicit assump-

tions that the two hemispheres are organized differently and that both

can participate concurrently and independently in the processing of dis-

parate components of the same stimulus. Evidence of anatomical differ-

ences between the two hemispheres in adult (Geschwind & Levitsky, 1968),

neonate (litelson S Paille, 1973), and pre-gestational brains (Wada,

Clarke, & Hamm, 1975) has been taken as support for the assumption of

differences in functional organization. These post-mortem studies ha,'e

found that the language mediating areas of the superior surface of the

temporal lobe are larger in the majority of cases on the left than on

the right.

Although there is precedence for more complex functions to be repre-

sented by larger neural areas, structural asymmetries, in and of








themselves, do not necessarily imply that there are intrinsic differ-

ences in functional and physiological organization between the hemi-

spheres. Furthermore, one must be cautious when hypothesizing func-

tional differences based on gross anatomy, since Broca's area (a speech

mediating region in the left frontal lobe) has been found to be smaller

in the left hemisphere than in the right hemisphere (Lemay, 1976; Wada

et al., 1975).

The second assumption that the two hemispheres process information

in parallel with each other has been most extensively addressed by Dimond

(1972). Based on a series of studies, he has concluded that:

The picture of brain function which emerges is that
of a two-brain individual in which each half brain
analyzes separately the information it receives.
The total capacity of the brain is different from
that which assumes that it carries out only one
function at a time, or that it consists of only
a single channel of limited capacity. It is an
arrangement in which there is both individuation
as well as integration between each half of the
system and the other. (p. 287)

Support for the assumption of parallel processing has been drawn

from studies of split-brain patients (Nebes, 1974) as well as from find-

ings with normals (Dimond, 1972, 1977). Goodglass and Calderon (1977)

studied the concurrent processing of verbal and tonal materials in trained

musicians. In this study, a DL task was used in which competing digits

sung in competing tonal patterns were presented simultaneously to each

ear. In another condition, spoken numbers were superimposed on piano

noted and presented to each ear. In both conditions, Ss were asked to

report both the digits and the tonal patterns they had heard. A right

ear superiority was found for the verbal component of the complex stimu-

lus, and a left ear superiority for the tonal component. These data were








interpreted as suggesting that each hemisphere can encode the compo-

nents of a complex stimulus for which it is particularly equipped to

handle.

One advantage of starting from the position that the two hemi-

spheres are characterized by particular "descriptive systems" is that

it enables one to reinterpret behavioral syndromes associated with neu-

rologic insult. Constructional apraxia, for example, is operationally

defined as extreme difficulty in the reproduction of drawings, geomet-

ric figures, or constructions with sticks and blocks, which cannot be

attributed to motor or sensory problems (Benton, 1967). According to

the "stimulus-specific" laterality model, constructional deficits should

occur following lesions to the right hemisphere, because of the strong

visuospatial task demands. The "strategy-specific" model, however,

would maintain that this disturbance might occur with lesions to either

hemisphere, with the types of errors differing on these constructional

tasks, depending on the side of the lesion.

Warrington (1966), who investigated constructional deficits in

brain injured patients, found impaired performance in individuals with

lesions to the left hemisphere and likewise in those with lesions of the

right hemisphere. Close scrutiny of the quality of the deficits and

the types of errors that were made showed qualitative differences with

respect to lesion side. Warrington (1966), as did McFie and Zangwill

(1960), observed that left hemisphere damaged patients made errors with

respect to details and that patients with right hemisphere lesions made

errors with respect to configuration and spatial organization.

This type of approach is reflected in the thinking of the Russian

neuropsychologist Alexander Luria. According to Luria (1973), no psycho-

logical-cognitive function is located in a specific, single neural tissue







as a strict localization model might hold. Rather, he argues for a

theory of dynamic localization and conceptualizes the brain as consist-

ing of hierarchically organized neural systems. Luria proposes that

during any single psychological-cognitive activity, such as reading,

various neural zones participate as component parts of a complex "func-

tional system," and each neural zone that is involved contributes a

different aspect to the total psychological structure.

To illustrate this concept, Luria (1973) discusses the syndrome

of apraxia, which is an inability to make purposeful movements due to

brain injury. The ability to make purposeful movements, for example,

relies on the conjoint participation of several neural areas which com-

prise the functional system. Depending on which of these neural areas

is injured, the structure of the movement is affected differently. In-

jury to the post-central cortex may result in a disturbance of finely

differentiated movements, whereas injury to other parietal-occipital

areas may produce an inability to place the hand in its necessary posi-

tion in space. Similarly, damage to the basal ganglia or premotor

cortex may result in a lack of smooth consecutive organization of hand

movements. Damage to other areas affects the structure of the activity

in different ways. Consequently, Luria demonstrates that a "syndrome"

might arise from diverse lesions to areas involved in a "functional

system" and stresses the importance of obtaining careful phenomenologi-

cal descriptions of the quality of the resulting deficit.

Despite the advantages offered by the approach outlined by Luria

(1973), more systematic research must be carried out before making de-

finitive statements about the processing strategies of the hemispheres.







An additional caveat bears on the assumption that there are isomorphic

relationships between types of strategies and brain localization. De-

spite gotsky's (1965) and Luria's (1973) stand on this issue, evidence

about strategies tends to be inferential and it cannot always be af-

firmed that differential successes or failures are the result of the ap-

plication of different strategies.

Current Research Directives in Laterality

One popular takeoff on the idea of cerebral asymmetry of function

has been the recent trend to identify individuals as having a "left"

or "right" hemisphere cognitive mode, based on the extent that they ex-

hibit characteristics presumably associated with the left and right hemi-

spheres (Arndt & Berger, 1978). It has been assumed that people who

tend to be verbal and analytic use their left hemispheres more exten-

sively and efficiently, while the reverse has been assumed for individ-

uals who favor visuospatial or gestalt approaches. In part, this in-

ferential leap has been based on reports that there is less alpha activ-

ity in the left hemisphere during performance on verbal tasks and less

alpha in the right hemisphere during performance on visuospatial tasks

(Ornstein & Galin, 1976).

Investigators have shown that various paper-and-pencil tests of pre-

ferred cognitive mode (i.e., verbal-analytic versus spatial-gestalt)

can discriminate between individuals who favor one approach over another,

as defined by their occupational choice, i.e., lawyers versus artists

(Galin & Ornstein, 1974). However, there is little direct evidence to

support the claim that an individual's preferred cognitive mode is actu-

ally related to asymmetry of cerebral functioning. In fact, recent re-

search addressing this problem has found no correlation between







behavioral measures of laterality (i.e., DL or VHF tasks) and referred

cognitive mode (Arndt & Berger, 1978). Similarly, there has been lit-

tle evidence supporting a relationship between cerebral laterality and

occupational choice.

Even more radical have been proposals (Galin, 1974) that the em-

phasis on verbal-analytic cognitive approaches, characteristic of West-

ern culture and thought, mirrors an over reliance on left hemisphere

functions. This has been contrasted with the more intuitive, esoteric

approaches of Eastern philosophies, which presumably reflect a dispro-

portionate emphasis on right hemisphere functions. This line of rea-

soning has been further extended to the realm of psychotherapy. Galin

(1974) has argued that the goal of psychotherapy should involve the bet-

ter integration of left and right hemisphere modes, with therapists ser-

ving the role of the corpus callosum. While it is not argued that such

syntheses of approaches prove humanistically valuable, there is, again,

no evidence for the speculative proposal that different cultural and

philosophical orientations actually reflect differences in cerebral lat-

erality of function.

Clearly, what is strongly needed are new directives in research.

Much recent work by Dimond and associates (Dimond, 1972, 1975; Dimond &

Beaumont, 1971, 1972) has stressed the importance of interhemispheric

sharing of cognitive load and emphasized that the nature of information

flow within and between the hemispheres should be carefully scrutinized.

There appears to be some major disagreement between those who think of

information flow as primarily competitive (Levy, 1969, 1974) and those

who see the double brain as primarily a cooperative arrangement (Dimond,

1972, 1975).







Equally salient and controversial have been recent proposals of

systematic variations in topological brain organization among particu-

lar groups of individuals, i.e., males versus females, right-handers

versus left-handers. There is current disagreement as to the existence

of possible sex differences in the cerebral representation of language

and nonlanguage functions. Some researchers postulate that females have

a more pronounced separation of linguistic and visuospatial functions

into the left and right hemispheres, respectively (Buffery & Gray, 1972),

whereas others view bilateral language representation as being more char-

acteristic of females than males (Lake & Bryden, 1976; McGlone & David-

son, 1974; McGlone, 1976, 1978). Still others have argued that such pur-

ported sex differences are artifactual and actually related to differences

in maturation rates for acquiring various developmental milestones, re-

gardless of sex (Waber, 1976).

Similarly, the cortical representation of language functions among

left-handers is far from being resolved. Although the vast majority of

right-handers (795%) are thought to have speech and language functions

lateralized primarily to the left hemisphere, the nature of such organi-

zation among left-handers remains enigmatic. Some theories postulate

bilateral representation of language in left-handers (Hecaen, 1976),

other theories postulate a more variable, though unilateral representa-

tion (Penfield & Roberts, 1959; Warrington & Pratt, 1973), whereas

others view left-handers as being a heterogeneous group consisting of

at least three different subtypes: those with bilateral language, those

with left hemisphere language, and those with right hemisphere language

(Levy, 1974; Satz, 1978a, 1978b). Recent research has attempted to

specify the actual proportions of left-handers who might fall into each

of these three subtypes (Satz, 1978a, 1978b).








To a large extent, investigators of sex and handedness differences

among normal individuals have used findings from various lateral sensory

tasks (i.e., DL, VHF, and dichotomous stimulation) to buttress their

arguments for a relationship between such variables and topological brain

organization. Individuals have been classified into left hemisphere,

right hemisphere, or bilateral speech groups on the basis of the direction

and magnitude of perceptual asymmetries obtained on these laterality

tasks. Thus, right channel superiorities for verbal stimuli have been

interpreted as reflecting left hemisphere dominance for language and

speech. Left channel, or even reduced right channel, superiorities for

nonverbal stimuli have been thought to indicate right hemisphere domi-

nance or less complete lateralization of language and speech functions

to the left hemisphere.

The problems with this approach are twofold. The first problem

relates to the actual validity of using perceptual asymmetries to

classify individuals into one hemisphere speech group or another and

has recently been addressed by Satz (1977). According to Satz (1977),

the probability of left brain speech in a right-hander, given a right

ear advantage (REA) on dichotic listening is approximately 97%. Thus,

researchers can, with a remarkably high degree of confidence, assign

right-handers with a REA to a predicted left brain speech group. How-

ever, the probability of right brain speech, given a left ear advantage

(LEA) on dichotic listening, is abysmally low (i.e., 10%). Consequently,

when right-handers are predicted to have right hemisphere speech based

on a LEA, they are incorrectly classified 90% of the time. Satz (1977)

warns that ". if the validity of these laterality measures is ques-

tionable, then we shall begin to construct theories which have no refer-

ents in the state of nature" (p. 211).







The second problem involves several unresolved, critical questions

as to the underlying neurophysiological mechanisms which account for per-

ceptual asymmetries obtained on laterality tasks. At least two such

mechanisms have been proposed to explain the occurrence of perceptual

asymmetries in normal right-handed individuals. Intrinsic to both is

the assumption that perceptual asymmetries reflect underlying asymmetric

functions of the two hemispheres for processing language and nonlanguage

stimuli.

One, however, is a structural account based on a pathway-transmis-

sion model, which states that stimuli are better processed if they have

direct access to the hemisphere most specialized for processing them.

The two foremost proponents of this model have been Kimura (1961, 1967)

and Moscovitch (1973). The form of their argument is as follows: when

verbal inputs are presented to the right channel, they are projected

directly, via contralateral pathways, to the left hemisphere which is

most adept at processing verbal stimuli. Verbal inputs presented to the

left channel, however, must traverse an indirect pathway from the left

ear or VHF to the right hemisphere and then across the corpus callosum

(or other interhemispheric commissures) to the left "language" hemisphere.

Consequently, left channel inputs are less readily perceived because of

the greater amount of time taken by the indirect route and because the

input is degraded in terms of strength or clarity as it crosses the cal-

losum. The converse is argued for the input of nbnlanguage stimuli.

Until approximately eight years ago, the pathway-transmission model

was widely accepted by neuropsychologists as providing the most cogent,

parsimonious explanation of perceptual asymrnetries. For the most part,

it made good sense to do so, based on the large number of studies which







seemed to support this model. Kinsbourne (1970a), however, challenged

this position on the grounds that it could not account for the magni-

tude of perceptual asymmetries obtained on laterality tasks. He argued

that callosal transfer of information involved only one extra synapse

taking approximately four milliseconds. This interval (four milliseconds)

did not seem to be sufficiently long enough to result in such large be-

tween-channel discrepancies as obtained on laterality tasks. Similarly,

he further argued that the pathway-transmission model could not account

for intra-individual variations in the magnitude of perceptual asymmetries

during performance across a particular laterality task.

Taking a more functional approach, Kinsbourne (1970a) postulated

that perceptual asymmetries were induced by differential changes in acti-

vation or attention between the hemispheres. Asymmetric hemispheric

activation not only made the hemisphere more receptive to incoming stim-

uli, but also produced changes in physical orientation contralateral to

the activated hemisphere. Thus, when the left hemisphere was activated,

it directed attention to the right side of body space such that stimuli

presented there were better perceived than those occurring in the left

side of space. The converse occurred with activation of the right hemi-

sphere.

Central to Kinsbourne's attentional model was the hypothetical con-

struct of "cognitive set." An individual's verbal or nonverbal cognitive

set was relegated the role of actually "priming" or activating the left

and right hemispheres, respectively. A cognitive set was itself induced

by various pre-task and task instructions which presumably set up "ex-

pectations" as to the type of processing that would be required, i.e.,

verbal or nonverbal.







Thus, according to Kinsbourne's attentional model, an individual's

cognitive set not only induces perceptual asymmetries, but also modulates

the magnitude of these asymmetries by differentially activating the two

hemispheres. This contrasts with the pathway-transmission model which

maintains that perceptual asymmetries arise because of the direct access

of stimulus input to the target hemisphere. While both perspectives

might be correct, it is necessary to more carefully scrutinize the assump-

tions underlying each as well as experimental evidence which might dif-

ferentiate between these two models.

The Pathway-Transmission Model of Perceptual Asymmetries: The Direct
Access Hypothesis

The initial purpose of Kimura's (1961) pathway-transmission model

was to account for the REA obtained on dichotic listening tasks. It was

subsequently extended to explain perceptual asymmetries with nonverbal

stimuli as well as with other laterality tasks. Central to the initial

formulation of this model was the assumption of total asymmetry of per-

ceptual functions between the hemispheres. This assumption was later

modified in light of evidence suggesting that some residual low level

language skills were mediated by the right hemisphere (Gazzaniga & Hill-

yard, 1971; Zaidel, 1978).

The modified version of the pathway-transmission model is conse-

quently based on a four-part argument which postulates that: (a) there

are hemispheric differences in the processing of verbal and nonverbal

materials; (b) contralateral sensory-hemisphere pathways are stronger

or more "prepotent" than ipsilateral connections; and (c) information

is relayed between the hemispheres via the corpus callosum and other neo-

cortical commissures. From this, it is deduced that stimuli are better







perceived when presented to the ear, VHF, or hand which has the most

direct access to the hemisphere specialized for processing these mater-

ials.

The critical test of any model is how well it can predict experi-

mental outcomes in research designed to test hypotheses generated from

the model. The question, therefore, for the pathway-transmission model

is what experimental evidence exists for the specific hypothesis that

stimuli are more readily processed when presented to the ear, VHF, etc.,

having the most direct access to the target hemisphere. Support for

this hypothesis has been drawn from a massive body of literature over

the past 15 years finding that verbal and nonverbal stimuli are recalled,

recognized, and responded to more quickly when presented contralateral

to the target hemisphere (See Tables 1-5).

A major problem which arises, however, is one of logistics. Later-

ality findings which are consistent with a pathway-transmission model are

the same ones which Kinsbourne's attentional model would predict. Conse-

quently, one cannot differentiate between these two models of perceptual

asymmetries based on findings from traditional laterality paradigms, since

both are compatible with these data.

This line of argument, of course, does not speak to the integrity

of the direct access hypothesis per se. Two types of recent studies,

however, report results that are incompatible with a simple wiring ac-

count of perceptual asymmetries in terms of sensory input-hemisphere

output connections. The first type is concerned with characterizing

laterality effects as a "side of stimulus entry" or spatial position

phenomenon and the second deals with the extent that laterality effects

can be modified and altered by contextual variables.







Goldstein and Lackner (1974) have demonstrated that laterality

effects on dichotic listening may be influenced by Ss' perceived spa-

tial orientation. In this study, the usual REA was obtained for con-

sonant-vowel syllables (CVs) in a dichotic recall task. If, however,

their Ss wore prisms which displaced their visual environments to the

left, then the REA was substantially reduced. Prisms that displaced

visual environments to the right significantly increased the magnitude

of the REA.

Similarly, Morais and colleagues (Hublet, Morais, & Bertelson,

1976; Morais, 1975; Morais & Bertelson, 1975) have found a right-sided

advantage for divergent verbal messages that are presented through loud-

speakers situated to the left and right of the S's median plane. This

right-sided advantage can be totally abolished by misleading Ss about

the position of the loudspeakers. In one study, Morais (1975) placed

"dummy" loudspeakers that were visible to the Ss in left and right hemi-

space. The loudspeakers from which the divergent messages actually orig-

inated were hidden from the Ss. When both the real and dummy loud-

speakers were placed 900 to the left and to the right of the median

plane, a marked right-sided advantage was found for the recall of CVs.

However, when the dummy loudspeakers were moved to 450 left and right

of midline, no advantages occurred for the CVs which, in actuality,

were still presented from the concealed loudspeakers located 900 to the

left and to the right of midline.

Furthermore, Anzola and coworkers (Anzola, Bertoloni, Buchtel, &

Rizzolatti, 1977) have demonstrated that choice reaction time (RT) asym-

metries to lateralized visual stimuli are dependent upon the spatial

stimulus-response compatibility of the stimulus and the responding hand.







When lateralized stimuli were presented to Ss with hands crossed and un-

crossed, faster RTs were obtained by the hand positioned in the same vis-

ual space as the stimulus. For example, if the hands were uncrossed such

that the left hand was on the left side of the body and the right hand

on the right side of the body, then stimuli presented to the LVF were

responded to faster by the left hand and stimuli in the RVF were responded

to faster by the right hand. If, however, the hands were crossed, such

that the left hand was on the right side of the body and the right hand

on the left side, then stimuli presented in the LVF were responded to

faster by the right hand (in left body space) and stimuli in the RVF

were responded to faster by the left hand (in right body space).

These findings are incompatible with a simple anatomical wiring

account of perceptual asymmetries. Instead, they implicate the impor-

tance of spatial localization mechanisms in producing laterality effects

and suggest that the routing of signals to the hemispheres depends, at

least in part, on some decision as to the spatial origins of signals.

While it is not the present purpose to speculate on the nature of such

localization mechanisms, one can readily see that in all traditional

laterality paradigms, the ear or VHF of stimulus input is perfectly con-

founded with the direction of stimulation (Kinsbourne, 1975; Morais,

1975). In some respects, these findings are more consistent with an

account of lateral asymmetries in terms of hemispheric specialization

and attention.

The effects of contextual variables on perceptual asymmetries have

been investigated by a number of researchers. Several early studies

have shown that dichotically presented vowels typically exhibit no ear

advantages, in contrast to the right-sided advantage obtained with







consonants. When, however, these vowels are embedded within either lin-

guistic or nonlinguistic contexts, these stimuli then exhibit right or

left ear asymmetries, depending on the context in which they occur (Hag-

gard, 1971; Spellacy & Blumstein, 1970). More recent studies have also

examined the effects of various task and contextual factors on percep-

tual asymmetries, although discussion of these findings will be deferred

until the section on attentional asymmetries.

The import of the preceding findings is that a structural anatomi-

cal pathway model cannot easily accommodate the effects of changing con-

text or spatial orientation. While evidence exists which is consistent

with this model, there is a body of data which cannot be accounted for

by a simple input-output wiring explanation. Logically, confirmation

of the direct access hypothesis leads one to accept the postulates upon

which it is based. However, failure to confirm this hypothesis across

a wide spectrum of data leaves one uncertain as to the source of the

deductive error, i.e., either the postulates themselves are faulty, or

the stated postulates are correct, although other relevant assumptions

have been overlooked and excluded from the model.

For the most part, there is considerable empirical evidence in fa-

vor of the three postulates of the pathway-transmission model. Despite

some controversy over the suppression of ipsilateral auditory sensory

pathways during dichotic listening', most investigators have accepted

the assumption that contralateral sensory-hemisphere pathways are

stronger or more "prepotent" than ipsilateral connections (Postulate

b). This has been based, to a large extent, on findings that evoked

responses to lateralized stimuli are faster or larger in amplitude

over the contralateral than ipsilateral hemisphere (Majkowski, Eochenek,








Bochenek, Knapiefijalowska, & Kopec, 1971; Matsumiya, Tagliasco, Lom-

brosco, & Goodglass, 1972; Mononen & Seitz, 1977). Given the rather

robust support for the postulates underlying the pathway-transmission

model leads one to believe that perhaps some critical information and

assumptions have not been considered and incorporated into this model

of perceptual asymmetries.

Before turning to a discussion of Kinsbourne's attentional-cogni-

tive set model, a brief overview of various neurophysiological and neu-

ropsychological models of attention will be provided.

Attention

Definition and Measurement of Attention

The functional significance of the construct "attention" in academ-

ic psychology has been to provide a label for some of the internal mecha-

nisms that determine the significance of stimuli, thereby making it more

possible to predict behavior. In this context, behavior represents the

end result of the organizational and information processing aspects of

brain function consequent to external stimulation. Attention is a hypo-

thetical construct which is thought to modulate the manner in which in-

formation is processed and organized in the brain.

Despite the sporadic emphasis on attention by psychological investi-

gators over the past century, a voluminous amount of research has ad-

dressed this topic during the last ten years. This literature can be

roughly divided into studies which have dealt with this problem by one of

two main approaches (Posner, 1976). First are studies of information

processing in simple tasks which have given rise to psychological-cogni-

tive theories of attention. This approach has fallen primarily within

the domain of cognitive psychology and has been of little interest to








neuroscientists because it has not addressed neural mechanisms underly-

ing attention. Second are physiological studies involving lesions of

particular brain areas, recordings of autonomic activity, measures de-

rived from the electroencephalogram (EEG), and responses of one or a

small group of cells at some specific level of the nervous system.

This approach has given rise to neurophysiological and neuropsycholog-

ical models of attention and arousal, with particular emphasis on de-

lineating brain systems which are important for mediating these pro-

cesses.

A third approach, which will not be addressed in the present over-

view, has involved the incorporation of the concept "attention" into

more molar accounts of behavior and motivation. This approach has been

characterized by behavioral research which has discussed such issues

as adaptation level (Helson, 1964), intrinsic motivation (Hunt, 1965),

and expectancy (Bruner, 1957).

While it is beyond the scope of the present discussion to address

in depth the literature on attention, an overview of the major .naurophys-

iological and neuropsychological models will be provided in the following

section. This will be followed by discussion of Kinsbourne's attentional

model of perceptual asymmetries. To insure a common frame of reference,
r
attention will be conceptualized as consisting of three components, ac-

cording to the taxonomy of Posner and Boies (1971): an intensive com-

ponent (arousal, activation, alertness); a selective component (selec-

tive attention); and a limited central processing component.

The intensive component, most frequently referred to as arousal

or alertness, is the general nonspecific receptivity of the organism

to incoming stimuli. While being "awake" implies that an organism is








in some basal state of arousal, the term "arousal reaction" denotes a

specific intensification of this basal level and has been operationally

defined in terms of neurophysiological concomitants that result from

the input of strong painful stimuli (Lynn, 1966). The types of stimu-

li that produce arousal have been studied extensively and labelled by

Berlyne (1960) as collative variables. These stimulus properties in-

clude, among others, those of novelty, incongruity, and complexity.

The selective component, i.e., selective attention, refers to the

selective guiding effect that is superimposed upon the arousal reaction.

Any time one incorporates into their model of arousal an analyzing mecha-

nism via which aspects of stimuli or responses of one kind are selected,

they are, by definition, referring to attention. Sokolov's (1963) model

of the orienting response, which is based on an arousal mechanism and a

mechanism which determines whether stimuli are significant, represents

perhaps the most primitive form of attention. An organism can be aroused

without being attentive, although the converse is never true. The rela-

tionship between arousal and attention has been described as an inverted

U-shaped curve, in which performance deteriorates at both low and high

levels of arousal (Tecce, 1972).

The selectivity component of attention has been divided into two

further components: involuntary selective attention and voluntary selec-

tive attention. Involuntary selective attention refers to the "involun-

tary" surge of activation that follows presentation of novel or signifi-

cant stimuli, including those that have acquired significance through

learning. In this case, the organism has little volitional control over

this biological response. With voluntary selective attention, however,

the organism consciously chooses to attend to stimuli because they are







relevant to a task he has chosen to perform. Studies of voluntary se-

lective attention have been classified according to what Ss are required

to select (Treisman, 1969): stimuli from a particular source, targets

of a particular type, specific stimulus attributes, or responses of a

particular type.

Finally, a third sense of attention as used in current psychologi-

cal research relates to the idea of a limited central information pro-

cessing capacity. Since various stimulus properties also define the con-

cept of "information" as used in the study of communication systems, it

has become customary to treat organisms subject to arousal and atten-

tion as "information processing systems." Various models have been pro-

posed which attempt to account for the manner in which information is

processed and the role of attention has been incorporated into these

models (Kahnemann, 1973; Posner, 1976).

One of the most critical issues facing investigators of attentional

components of behavior concerns how these phenomena are measured. Re-

searchers have variously used measures of brain electrical activity in-

cluding the EEG, contingent negative variation (CNV), evoked potential
-
(EP), recordings of autonomic activity such as the galvanic skin re-

sponse (GSR), and heart rate (HR). and more overt behav iora l indices

such as reaction time (RT). Whiile the distinction is not entirely clear-

cut, electroplhysiological measurements are generally used to inde.,.

arousal, whereaS more behavioral performance measures (i.e., accuracy,'

and/or speed of task performance) are used to inde;., selective attention.

Although recent research suggests that subtle measures derived from the

EEG such as the P300 wave of the EP and the CNV, are correlated with

attention (Picton & Hillyard, 197-a, 1974b; Tecce, 1972), attention is








generally inferred from the manner in which experimental task performance

is affected. In this context, attention is a hypothetical construct

which can only be defined in operational terms such as task performance.

In most cases, electrophysiological measures have been shown to be

reasonably accurate and sensitive to variations in the level of arousal

or alertness. However, one major problem concerns the implicit assump-

tion that arousal is unidimensional and that all indices of it should

move in the same direction as arousal fluctuates. This is based, in part,

on the work of Lindsley (1960), who demonstrated that as an individual

moves from a state of least to most arousal, there is a corresponding

increase in the frequency of EEG rhythms from high amplitude slow waves

to low amplitude fast waves (i.e., beta).

Difficulty with the unidimensional arousal assumption is reflected

in studies which have examined the correlations among various electro-

physiological measures of arousal. Graham and Clifton (1966), for ex-

ample, found that none of the correlations among a variety of these

measures exceeded a value of .5, with the highest correlation between

EEG and HR.

Furthermore, there is sometimes found a dissociation between EEG

activation and behavioral arousal, in that the EEG may reflect activa-

tion while the organism's behavior clearly does not, or vice versa.

This is noted during paradoxical sleep (Dement & Kleitnan, 1957) and

in the "sleeping" EEG records of behaviorally awake animals that have

been given various drugs. Bradley and Key (1958) demonstrated that

animals administered atropine showed slow brain wave activity on the

EEG, although they were behaviorally awake. Conversely, Feldman and

Waller (1962), who lesioned hypothalamic nuclei in cats, found that these

behaviorally comatose animals produced "active" EEG patterns.







Neuropsychological Models of Attention

Neuropsychological models of attention have been closely linked to

the neurophysiological work on the brainstem reticular activating system

(RAS). This system is a multisynaptic neuronal network extending the

reticular core of the brainstem, receiving inputs from major sensory

systems, and having ascending and descending connections to major brain

areas (Thompson, 1967). In an early landmark study, Moruzzi and Magoun

(1949) hypothesized that the RAS was critical for mediating arousal via

its activating influence on the cerebral cortex. Support for this hypoth-

esis was initially drawn from studies finding that electrical stimulation

of the RAS produced cortical EEG activation, as well as autonomic and

behavioral components of the arousal reaction (Moruzzi & Magoun, 1949).

Stimulation of the RAS has also been shown to exert enhancing ef-

fects on task performance. Fuster (1958), who trained monkeys on a

stimulus discrimination task, found that electrical stimulation of the

RAS resulted in faster discriminative RTs and more correct responses on

this task. In a human analogue study, Lansing, Schwarz, and Lindsley

(1959) measured RTs to light flashes that were sometimes preceded by

an auditory warning stimulus (WS). They found that reduced RTs in the

WS condition were correlated with EEG activation.

In contrast to these stimulation studies, ablative paradigms have

found that lesions to the ascending components of the RAS attentuate

EEG activation (Lindsley, Schreiner, Knowles, & Magoun, 1950). The be-

havior of these lesioned animals is characterized by hypokinesis ind coma.

While the ascending "reticular-cortical" connections were initially

sufficient for understanding arousal and sleep-wakefulness cycles, they

could not account for more complex components of attention. Since higher







cognitive functions involved in selective attention seemed to have

emerged with the development of the neocortex, it was logical to hypoth-

esize that the neocortex had functional connections and influence on the

RAS. This inference led French and coworkers (French, Hernandez-Peon,

& Livingston, 1955) to investigate "cortical-reticular" connections. As

predicted, stimulation of cortical areas projecting to the RAS not only

produced evoked responses of the RAS, but also induced behavioral and au-

tonomic components of the arousal reaction.

The foremost investigator to tie together this knowledge of cortical-

reticular influences into a model of attention was Sokolov (1963). He

proposed an attentional model based on the orientation reaction. This

refers to a set of behaviors, both neurophysiological and more overt,

which is prompted by the presentation of a novel of biologically signifi-

cant stimulus. The most conspicuous aspect of the orienting response (OR)

is that the organism turns (i.e., orients) toward the source of stimula-

tion. In addition to behavioral turning, various neurophysiological con-

comitants of increased arousal occur, including the lowering of sensory

thresholds, EEG desynchronization, HR deceleration, increased GSR, a

pupillary response, and vasodilation of the head. The functional signi-

ficance of all these concomitants of the OR is to make the organism more

receptive to incoming stimuli ("attentive") as well as to prepare the

organism for action ("intention").

Sokolov's attentional model consisted of two formal components, an

analyzing mechanism which determined whether stimuli necessitated an OR

and another mechanism via which the OR could be initiated. The analyzing

mechanism was located in the cortex and the "activating" mechanism in

the brainstem RAS. According to Sokolov, the neurophysiological basis







of the OR was a cortico-reticular feedback loop involving the RAS and

neuronal templates in the cortex, which were representations of stimulus

parameters. When sensory input did not match these neuronal templates,

i.e., were significant or novel, the cortex sent excitatory discharge

to the RAS via cortico-reticular pathways. This, in turn, activated the

RAS which further activated the cortex via reticular-cortical pathways.

To the extent, however, that there was a match between incoming stimuli

and cortical templates, the OR was inhibited or habituated.

Sokolov further delineated two components of the OR, the localized

and generalized components, which were mediated by the thalamic and mesen-

cephalic RAS, respectively. These two components paralleled the distinc-

tion made by Sharpless and Jasper (1956) between the phasic and tonic

arousal reactions. The tonic/generalized component involved diffuse EEG

activation, slow recovery time, and rapid habituation. The phasic/local-

ized component was characterized by activation of a discrete cortical

area, fast recovery time, and slow habituation. Thus, any stimulus that

elicited an OR could be described in terms of these three parameters:

generalization-specificity; rate of recovery time, and rate of habitua-

tion. The generalization-specificity parameter referred to the extent that

stimulus presentation induced diffuse versus discrete cerebral activa-

tion. The recovery time parameter referred to the temporal duration of

activation following a single stimulus presentation, and habituation re-

ferred to the diminished activation which occurred with repeated presen-

tation of a stimulus.

The major import of Sokolov's (1963) model was that it integrated

knowledge of cortical and subcortical mechanisms into a uniform theory

of attention. It further generated an abundance of research which greatly








enhanced the understanding of simple attentional phenomena. The lim-

itations of this model, however, was that it attempted to explain only

two components of attention: arousal and involuntary selective at-

tention. It is easy to see how one might extend the general trappings

of this model to explain more cognitively complex aspects of selective

attention.

Other neurophysiological models of attention-arousal have included

those of Hernandez-Peon (1966), Lacey (1956), Routtenberg (1968), and

Pribram and McGuinness (1975). Hernandez-Peon (1966) essentially pro-

posed a model of attention in which selective peripheral gating occurred

prior to sensory analysis. According to this model, selectivity was

mediated by a reticulofugal inhibitory mechanism which acted on the

first sensory synapse. However, no provisions were made for the selec-

tion of significant stimuli.

Lacey (1956) proposed a two-dimensional arousal model which distin-

guished between the "taking in" of stimuli from the environment and the

"rejection" of stimuli. He argued that the "taking in" of stimuli cor-

responded to increased arousal and the "rejection" of stimuli corresponded

to decreased arousal. He further postulated an inverse relationship be-

tween heart rate (HR) and EEG activation: HR deceleration reflected

increased arousal and HR acceleration reflected decreased arousal. The

neurophysiological underpinnings of Lacey's reasoning were based on find-

ings that stimulation of the carotid sinus inhibited the nucleus soli-

tarious in the brainstem. This, in turn, inhibited cortical activation,

which then improved the probability that stimuli would be rejected.

The major problem with Lacey's formulation is that no objective cri-

teria were provided for determining what was considered "rejectionable"








and what was considered "acceptable." In a convincing article which

attempted to synthesize various models of arousal, Graham and Clifton

(1966) proposed that Lacey's "taking in" of stimuli was equivalent to

Sokolov's OR, and that the "rejection" of stimuli was equivalent to

Sokolov's defensive reaction (OR). These authors further argued that

HR deceleration represented the cardiac component of the OR. Thus,

Lacey's model can best be subsumed under Sokolov's model of the OR.

Routtenberg (1968) proposed an arousal model which attempted to

explain those findings which have suggested a dissociation between EEG

and behavioral indices of arousal. According to Routtenberg, arousal

is a two component process which is mediated by two distinct, though

interacting neural systems. One component (Arousal System I) is pri-

marily concerned with producing neocortical desynchronization subse-

quest to external stimulation and is mediated by the RAS. The second

component (Arousal System II) is critical for the maintenance of basic

vegetative functions and ". . provides control of behavior through

incentive related stimuli" (p. 51). This system is mediated by the

limbic-midbrain regions, particularly the medial forebrain bundle.

According to Routtenberg, the most profound arousal deficits re-

sult from extensive lesions involving both neural systems. Lesions to

the RAS alone, however, eliminate or reduce EEG activation, although

wakefulness of the animal is maintained via Arousal System II. Con-

versly, lesions to the limbic-midbrain region induce coma or disruption

of primary vegetative functions, although EEG activation, mediated by

the RAS, may persist.

Based on this line of reasoning, Routtenberg then argues that the

EEG-behavioral dissociation; observed with atropine administration or







hypothalamic lesions are based on the differential involvement of the

two neural systems responsible for mediating arousal. He proposes that

atropine primarily suppresses the RAS (i.e., decreased EEG activation),

whereas the limbic-midbrain regions are relatively unaffected. These

predictions parallel the findings of a "sleeping" EEG in behaviorally

awake animals consequent to atropine administration (Bradley & Key, 1958).

Routtenberg similarly explains Feldman and Waller's (1962) findings of

an "alert" EEG in comatose cats consequent to hypothalamic lesions. In

this case, the limbic-midbrain region important for the wakefulness of

the animal has been damaged, although the RAS remains intact.

Routtenberg's (1968) delineation of two aspects of arousal and their

underlying neural mechanisms represents a critical contribution to the

understanding of arousal phenomena. The limitation of this model is

that it does not address the mechanisms underlying selective attention.

Pribram and McGuinness (1975) have proposed an extremely compli-

cated model of attention. Like Sokolov (1963), they claim that localized

in the brain are "central representations," or memory traces of stim-

ulus configurations. Their model consists of three separate, though

interacting attentional control systems which are centered in the

amygdala, basal ganglia, and hippocampus, respectively. The neural

system centered in the amygdala controls arousal, which they define as

a phasic physiological response to stimulus input. These arousal cir-

cuits function by modulating serotonergic neurons in the brainstem.

The neural system centered in the basal ganglia controls activation,

which is defined in terms of a tonic physiological "readiness to re-

spond" or execute perceptual or motor acts. The third neural system,

centered in the hippocampus, is concerned with coordinating the







relationship between arousal and activation. This coordinating activ-

ity is defined as requiring effort and involves ". .. uncoupling stimuli

from responses so that appropriate changes in central representations

can occur" (p. 111). According to Pribram and McGuinness, such changes

in central representations can be conceived as changes in state, set,

or attitude.

After defining the three basic parameters of this model, Pribram

and McGuinness (1975) assert that arousal, activation, and effort are

differentially involved in various activities. They maintain that the

OR involves arousal, but not activation, vigilant readiness to respond

involves activation, but not arousal; the DR involves both arousal and

activation; when neither activation or arousal are present, behavior is

automatic in that stimulus-response contingencies are direct and without

intervention from any of the attentional control systems. The latter

is called automatizedd" behavior.

Pribram and McGuinness (1975) further attempt to explain more

cognitively complex aspects of voluntary selective attention that are

involved in the performance of problem solving tasks. They draw a dis-

tinction between "categorizing tasks" and "reasoning tasks." They

define "categorizing tasks" as involving a response to some invariant

combination of stimuli (i.e., a discrimination judgement). In these

tasks, stimuli must be tected, categorized, and responded to, although

the critical aspect is stimulus categorization. Based on this line

of reasoning, they propose that arousal precedes activation in stimulus

categorization tasks.

In "reasoning tasks," stimulus events are variable, but compu-

table (i.e., arithmetic problems). According to Pribram and IlcGuinness,







"reasoning problems" involve an "'ncoupling" of attention from the im-

mediate stimulus variables. In this case, attention is voluntary and

initiated by the organism. They consequently propose that activation

precedes arousal in "reasoning tasks".

While many of Pribram and McGuinness' arguments seem rather vague,

perhaps the most important aspect of their model is the explicit dis-

tinction drawn between "attention" (called arousal in their model) and

"intention" (called activation) which are presumably mediated by dif-

ferent brain regions. This parallels the distinction made by Sokolov

(1963) in his description of the OR as making the organism more recep-

tive to incoming stimuli (attentiveness) and preparing the organism

for action (intention). For Pribram and McGuinness, attention refers

to the visceral-autonomic components of the arousal reaction that facil-

itate stimulus receptivity and categorization. Intention encompasses

not only somato-motor responses to stimuli, but also includes behaviors

involved in an organism consciously directing his cognition and behavior.

Even within the realm of information processing models of atten-

tion, the distinction between attentiveness and intention is implicitly

drawn. Two general types of cognitive theories have dominated informa-

tion processing models of attention: those that view information proces-

sing as being serial and limited by some filtering stage or bottleneck

(Broadbent, 1958; Deutsch & Deutsch, 1963) and those proposing that in-

formation can be variably allocated among diverse, parallel channels

(Kahnemann, 1973). In Broadbent's (1958) filter model, attention plays

the role of "setting the filter" to select a certain class of stimuli

and to reject others. This contrasts with the filter model of Deutsch








and Deutsch (1963) who propose that attention determines response selec-

tion, i.e., it prevents the initiation of more than one response at a

time and "selects" the response that best fits the requirements of the

task. Kahnemann's (1973) variable allocation model, on the other hand,

incorporates both the notion of a stimulus selection stage and that of

a response selection stage.

The distinction between attention and intention has interesting im-

plications on both neurophysiological and behavioral levels. Based on

the work of Sokolov (1963), Routtenberg (1968), and Pribram and McGuin-

ness (1975), it appears that arousal is mediated by a complex system

involving the RAS and limbic-midbrain regions. Attentiveness to stimuli

probably involves cortical systems, whereas intention (i.e., readiness

to respond) involves basal ganglia-frontal systems. These three systems

can be viewed as functionally interactive and interdependent in the nor-

mal individual, such that he is appropriately alert, attentive, and in-

tentive. Discrete lesions to any one of these systems should result in

qualitatively different types of attentional deficits.

On a psychological level, it seems reasonable that an individual

can be attentive to stimulus characteristics, without necessarily being

intentive, i.e., organizing one's behavior so that information conveyed

by the stimulus is appropriately acted on. Since, however, attentive-

ness is generally inferred from task performance, experimental paradigms

for behaviorally differentiating between attention and intention in nor-

mal adults are not immediately apparent.

Recent findings with neurologically impaired individuals suggests

that such a distinction can be made, however. Parkinson patients, who

have chemical lesions to the dopaminergic systems in the basal ganglia,







are clinically described as being hypokinetic and having difficulty

initiating action (intention). Their RTs to simple stimuli are signi-

ficantly slower than those of matched patient controls (Heilman, Bowers,

Watson, & Greer, 1976). However, presentation of warning stimuli signi-

ficantly reduces the RTs of the Parkinson patients. This RT reduction

with WS is proportionate to that obtained with matched controls, sug-

gesting that Parkinson patients are attentive, without being equally

intentive.

What is not entirely clear is whether an individual can be inten-

tive without being attentive. Equally unclear is whether the attentional-

intentional systems are differentially mediated by the two cerebral hemi-

spheres, i.e., whether each hemisphere has relatively independent atten-

tional-intentional neural systems. Since cortical systems involved in

the processing of linguistic and nonlinguistic stimuli are asymmetrically

represented in the left and right hemispheres, one might hypothesize

that verbal and nonverbal materials asymmetrically activate and prepare

the hemispheres for action. This is essentially the basis of Kinsbourne's

attentional model of perceptual asymmetries.

An Attentional Model of Perceptual Asymmetries: The Cognitive Set
Hypothesis

Kinsbourne's attentional-cognitive set model (1970a, 1975) repre-

sents an attempt to account for intra-individual variability in the magni-

tude of perceptual asymmetries obtained on various lateral sensory tasks.

It further attempts to explain the difficulty in replicating a number of

laterality effects, especially those purporting to underlie right hemi-

sphere advantages. Kinsbourne argues that the difficulty in obtaining

"right hemisphere" effects is largely related to the fact that individuals







frequently adopt implicit verbal "cognitive sets" which tend to mask or

diminish clear cut left-sided perceptual asymmetries.

Kinsbourne's model of perceptual asymmetries is based on a number

of provocative assumptions which, for the most part, have not been ad-

ressed empirically. Because of the complexity of his model, it can be

best conceptualized as consisting of two levels, which, when combined,

generate an hypothesis concerning the basis of perceptual asymmetries.

The first level attempts to describe "what happens" when one hemisphere

or the other is asymmetrically activated and is based on two assumptions

stating that: (a) there is an equal, interdependent balance of activa-

tion-attention between the hemispheres, such that when one hemisphere

is activated, the other is hypoaroused; and (b) each hemisphere controls

or directs attention to the contralateral side of body space. From this,

it is deduced that activation of one hemisphere induces increased recep-

tivity of that hemisphere for incoming stimuli, regardless of their ver-

bal-nonverbal dimensions. Furthermore, hemispheric activation produces

shifts in physical orientation contralateral to the primed hemisphere,

so that stimuli presented in the contralateral side of space are better

perceived.

The second level of Kinsbourne's model attempts to explain the man-

ner in which or "how" one hemisphere or the other can be asymmetrically

activated and is based on the following assumptions: (a) there are hemi-

spheric differences in processing verbal and nonverbal materials; (b)

stimuli or tasks that call for the specialized processing of a particu-

lar hemisphere asymmetrically activate the target hemisphere; and (c) a

cognitive set, which refers to the probabilistic expectations that stim-

uli will be processed in a particular manner or according to a particular








strategy, can be induced by pre-task instructions, task performance it-

self, or by implicit expectations that the individual brings into the

experimental setting. From this, Kinsbourne hypothesizes that verbal

and nonverbal cognitive sets can asymmetrically and selectively activate

the hemispheres.

Kinsbourne then proposes that perceptual asymmetries arise from the

adoption of verbal and nonverbal cognitive sets. According to his rea-

soning, a verbal cognitive set activates the left hemisphere, so that

physical orientation and attention are then directed to the right side

of space. Stimuli occurring there are more readily perceived than stim-

uli presented in the left ipsilateral side of space. The converse oc-

curs with nonverbal cognitive sets.

In this \:w;y, findings from traditional laterality tasks, i.e., bet-

ter perception of verbal stimuli presented to the right channel and bet-

ter perception of nonverbal stimuli presented to the left channel, can

be accounted for by Kinsbourne's model. Since, however, the latter find-

ings are also compatible with the direct access hypothesis, they offer

little help for distinguishing between these two models of perceptual

asymmetries. Those studies, however, which suggest that laterality ef-

fects are related to spatial side of stimulus entry are more consistent

with an attentional model.

Experimental support for the cognitive set hypothesis in studies

specifically designed to test its predictions have been controversial at

best. Not only have the findings themselves been contradictory, but the

predictions have been vague and often times loose. One type of predic-

tion generated from this model is that performance on a task which in-

duces a verbal or nonverbal "cognitive set" will affect performance on








concurrent or subsequent laterality tasks by altering either the direc-

tion or magnitude of perceptual asymmetries that are obtained on the lat-

ter.

This prediction was initially addressed by Kinsbourne and coworkers

in a series of experiments (Earl & Kinsbourne, 1975; Kinsbourne, 1970a;

Kinsbourne & Bruce, 1975). These investigators found that the silent

rehearsal of word lists during a tachistoscopically presented gap de-

tection task resulted in better gap detection in the RVF than LVF. With-

out the concurrent verbal task, gaps were detected equally well in both

VHFs. Hellige (1978) also found that rehearsal of lists during the

tachistoscopic recognition of nonverbal figures shifted perceptual asym-

metries from the left to the right. In these studies, the performance

of the verbal rehearsal task was interpreted as inducing a verbal cog-

nitive set which "primed" the left hemisphere. A summary of the findings

of studies which have used "priming" tasks to test the cognitive set

hypothesis is provided in Table 7.

Since the direct access hypothesis does not in any straight forward

manner predict such effects, findings of altered perceptual asymmetries

argue against a pathway model as the sole mechanism responsible for pro-

ducing perceptual asymmetries. However, some researchers (Gardner &

Branski, 1976) have been unable to replicate any of Kinsbourne's origi-

nal findings with gap detection.

Using a different type of paradigm, Klein, Moscovitch, and Vigna

(1976) presented "priming" tasks that were then followed by "target" lat-

erality tasks. Without the "priming" tasks, a LVF superiority was ob-

tained for face recognition and a RVF superiority for word recall.

With a verbal "priming" task, however, the LVF superiority for faces





51


O D -Wi
rr L

L-2 L L L E L





.LL -4-. Ln fl

.0) 4-) a) 0) 4-' -- Q a) U Z 4-1 0
=1 -) 4-) 4-) .- 4-' 3- D 14-3 1 L E 4-) U C
0 )- 0 0 0 ) 0 D r) C. -
=C O .- -C .C Q LL- = OQ 4-)- (, -0 L4-- _
4I- 4 -3I + M > -C U E] C *
*r- *r- Ll- -- : 3 C 0 D S-
3 v: 5 3 3 L* c D
ie Cx Y C _M LL. 0 4- LL-
) V* r- i lvn (v in *r- i i > *- 3 1 > 0 .. 4-)
a)O m 0 Dm m O C 3. m D31 Q +-) 0 4n- 1 3 11 =3
U 4-) C- -) U *-r U 4-"- C 4) u c 4- U U O
C O C C O C CM (Ua C.C
C:0 0 c0 or c 0 Q)*r v 4 -*- 4O 0 a
S= 4-) C LS C O GC S- C: 4 C a D a o S1 3r-











I.-

I--

I--"
- - - U 0 0 0 0 03
SE E E E E EI I I I
*+ )r *r- *4- *C-- *r-- + 01 .*- *r- 0) QC C: 3 C: 4--
*r- O r- C *- C *i(c CS-. 4- c m na A- () *.- o
AL 0 0 0 - 0 0 *r0- 1 *- ) -0
0, o E DE r-c- 3 -C
-T LL- D L- a) U ) LL- ..- LL- W 4-) S 4- 0 0 -- LL- 4--
C- r- O-C -C -C =-r ftL Ir SC L -- r -- M 1 ,-i-

) 0O 0 0 0 O(3) 0









0n I
U)


















I-
LLU 0 0 _0 _0 -0 _0














I -
0 0 0 0 0 0















S. W u w 1.
-" '- t-- t4-- t-- 4 4-
0 0 0 0 0 0 0 0



















C) _a
j *i c --- V *ri- *-- -

5. 0C OC n C 0C L0





I-
CD E O0) vi O } O O








LU 4-
LQ 0LJ
) n 0 0 0




m- CU LW > L
*- S. C S- r- < S-





t--




CA C C
L2 3D 3 3 C C
a ) O O










0 0 1n 0 0 0 0
0 C4 *r- *-- .n 4 4-)o












O L O S- 4a +-- +- -
- 4e D4- D D

L o 00L CL oL
m r o
CD a r i a r a











0) r- c 3

c Ln Ln co co

su r -i r- r~

o 0 S- oZ$ cZ$ (3 Qi



O) LU LU CD CD5










C



4-)
E








u 0
r-



4-'







LL. 4-)
.r









0
5->














4 --
,- 0
-0

4--










0










2:





zr
U






4-)
0C










-J






















71
ci"
S-









C


C




I--'


















17J
0-

























,,




C
c.


4 -



4-


r-
E

=-r

4-'
.r- a



El 4-)
UO 3
L 0

Sc-


4-
4- S-
*r- 0
0



-C-









0









0



u
c-"
rO

0
5-

Ccd










IC

5-


0-



0
.r_
a)




















0-
L'










c-
3
























tO




1 *,--

C :


0 E


S-oE
rC --
4-) ,--
4-) U S


L a,


0 C.
-C -
L- 4-
4 .r---


* C -
1,1 *r- U

0 *r- Q

on Q- (D
4-


4-) O "L

O *r- "I


-

Ir
4-4-)

C10





I0
rE






"r-
4--




-.r-
cnU

- 0-


CS -
4- .
0
0-4-


I 0)


> 4-
-0 C
ro ro



LL O"
Ll- 0

i= LIC C 4>
0 4- .r--



3 *-r- o o
4--

mt,-0 0
X -o *r- (I U
-- 4-P 4-
4- S- n3 *-- -
o x n o0
-4- -4- -r- S-
r- 4-0 4-'

s- cJn- s- 0
W o3 n 0 U
> 4 -)aM 4-












4-) C-
C 0

2: > -









O




4- ro
W r


X


S3








S-
> -








0 U





S- Ll






-r- T3


71 0
0



















--.

: -
a-
5--






, t >


0C 4-
(3)
al
L- -
I- Q

I-
Scd--~
L
I- W- 0
r 3 -r-
a- EC- -


mCkC *-- -P -0





E 4-3 -- ra 0
.*- 4- CS



C S. 0-i 04-
4- O


S *











u
4-) (U
-- 4- -
o N 4-'
















0 0
r-- S- 0
U 0 *r-- -> .0 4->







Ln 0









o3
U






0 C




CD
?-




rdr






.cr-
c, c








C)

.-'
0- )


a)














-I-














C .',
r-s














1.0
-r*-

H-
+-J












I-


0

I3
NS


c
4-
0


tL 4-'
-r



()










3 0 0
0- C
.- -
)U 4-'
-O C IC
"OOL,

> 0- 3

C\I





0-
0 '- O-



a,
O2 3- 12






















a.'
C.--
-O0v'
C Q







was abolished. With a nonverbal "priming" task, the RVF superiority

for words was significantly reduced. While these findings appear, at

first glance, to be consistent with Kinsbourne's model, close scrutiny

of this data raises questions regarding the manner in which the percep-

tual asymmetries were altered.

Kinsbourne's model, which predicts that verbal priming activates

the left hemisphere, also predicts vis a vis the assumption of reciprocal

hemispheric balance of attention, that the right hemisphere should be

"hypoactivated." Consequently, a verbal priming task should alter per-

ceptual asymmetries on a nonverbal laterality task by increasing recog-

nition accuracy in the RVF and decreasing recognition accuracy in the

LVF. This was only partially confirmed by Klein et al.'s data (1976).

Their findings indicated that while the verbal priming task did, in fact,

increase accuracy of facial recognition in the RVF, no decrease in ac-

curacy was found in the LVF. A similar pattern occurred with the non-

verbal priming task.

Another type of prediction generated from the cognitive set hypoth-

esis deals with the effects of randomly intermixing verbal and nonverbal

stimulus trials in a laterality task. According to Kinsbourne (1970),

it should be difficult for a S to develop a clearcut cognitive set for

the appropriate type of stimulus on each trial of such a randomly mixed

list. Consequently, he has predicted that both verbal and nonverbal

laterality effects will be diminished when the two types of stimuli

are randomly mixed. Hellige (1978), on the other hand, has argued that

in such paradigms, Ss might be "biased" to emphasize the processing of

one type stimulus more than the other. In this case, Hellige predicts

that the same laterality pattern should occur for all types of stimuli







in a randomly mixed paradigm, i.e., all right-sided, all left-sided,

or no asymmetries. One problem with both Kinsbourne's and Hellige's

predictions are that no objective criteria are provided for estimating

what type of laterality patterns might emerge.

Findings from studies using the intermixing paradigm (See Table 8)

have been contradictory and subject to methodologic criticism. Some

researchers have not included the necessary "pure-list" versus "mixed-

list" comparisons, and conflicting interpretations have resulted. These

investigators (Berlucchi, Brizzolara, Marzi, Rizzolatti, & Umilta, 1974;

Dee & Hannay, 1973; Geffen, Bradshaw, & Nettleton, 1972; Kallman, 1978)

have found that perceptual asymmetries for verbal and nonverbal stimuli

do occur in a mixed-list paradigm, although these studies have not in-

cluded conditions in which only verbal or nonverbal stimuli were presented.

Other investigators (Donnefeld, Rosen, MacKacey, & Curcio, 1976; Hellige,

1978) have reported alterations in perceptual asymmetries when pure-list

versus mixed-list comparisons were made. In these studies, the left-

sided advantages for nonverbal stimuli obtained in the pure-list condi-

tions were either attenuated or shifted to the right, whereas the right-

sided asymmetries for verbal stimuli remained intact and unaffected.

Other laterality experiments have tested the cognitive set hypothe-

sis by cueing Ss prior to stimulus presentation as to either the side

of stimulation or the type of required processing, i.e., Name versus

Physical Identity (Geffen et al., 1972; Hellige, 1978). According to

Kinsbourne's model, pre-stimulus cueing should increase the magnitude

of perceptual asymmetries. This. however, has not been demonstrated.

The culmination of these contradictory findings regarding cognitive

set is that they cast doubt on the validity of K'insbourne's attentional










I LLW

0 > C

LL- CDO -
> rZ w0
Qf 4-J 4--


1I > (

L .
0 0-

0 -=
3Z V)


>



r-)
0






LL
>






4-)







Qi U

-0 -






U,
4C
c0




















I--





































Ur-
r3r





































r"-
L

























0)
4O)




4-





s-










co
c,


























*r t


I-


C:




-r
--)




O0







I-
Sf--0
Cd-)






LLJ
1--00
0













X>-
I0

















u-


L)O

U

C.


> r-
u C.


4--) 4-
c c



rr


>>






0-

o -r
s-: Vo


'-)
ro
E

C <




0i- C-
Sr-


4-)



c- 4J
o -












CE






I-




4-)
E r-.






























4-
4-
CD
OEJ




















aro


C: r3
Cro

> LL.
(0 in>

LL






E

zu
roI (--


L
0
4- 4J
0
O


lr- +"

C,

oU


c u
*- C
4-- U
O 4-/1













I-)
0



4O
c










3
c --'~



0-0
U








4--













0 Qi
r-- 0
4-)





O
- r-
4J--








model of perceptual asymmetries. Failure to confirm the cognitive set

hypothesis across a wide spectrum of data again leads one to question

whether or not the postulates upon which it is based are faulty. Ex-

cept for the assumption of hemispheric processing differences, the postu-

lates underlying this model either have not been extensively addressed

or the evidence in support of them is ambiguous.

Kinsbourne's speculative assertion that there is a reciprocal bal-

ance of attention-activation between the hemispheres was initially ad-

vanced as the mechanism underlying unilateral spatial neglect. This

syndrome, most frequently observed following lesions to the right hemi-

sphere, is characterized, in its most extreme form, by the total dis-

regarding (i.e., neglecting) of the left side of hemispace, in the

absence of primary sensory or motor disturbances. According to Kins-

bourne (197Cb), unilateral neglect is caused by a hemispheric imbal-

ance in attention: the right hemisphere, because of its lesion, is

hypoaroused, thereby resulting in the left hemisphere being h/peraroused.

With hyperarousal of the left hemisphere, attention is focally directed

to the right side of hemispace, resulting in disregard and inattention

to left hemispace.

Although an interesting formulation, there has been no experimental

support for this account of the unilateral neglect syndrome. According

to Kinsbourne's reasoning, patients with unilateral left-sided neglect

should exhibit depressed responses on electrophysiological measures of

arousal obtained from the right hemisphere and elevated responsivity

from the left hemisphere. This has not been found. Studies of patients

with this syndrome have demonstrated a bilateral slowing of the EEG

(Watson, Andriola, & Heilman, 1977) and bilaterally depressed 'SR's








(Heilman, Schwartz, & Watson, 1978). These findings contradict Kins-

bourne's predictions and argue against his hemispheric balance assump-

tion. Similarly, there has been no support for this assumption in indi-

viduals who are neurologically intact (Klein et al., 1976). Clearly,

this is one assumption which could unequivocally be eliminated from

Kinsbourne's model.

The assumptions that hemispheric activation induces a hemispheric

state of "perceptual readiness" for stimulus input, as well as produces

shifts in physical orientation contralateral to the activated hemisphere

is analogous in form to behaviors accompanying the OR described by Soko-

lov (1963). Instead of the "whole brain" or "whole organism" phenome-

non implied by the OR, Kinsbourne postulates that this activational phe-

nomenon can be restricted to one hemisphere or the other.

Support for increased stimulus receptivity by the activated hemi-

sphere has been drawn from reports of lowered sensory thresholds for

words exposed to the RVF than LVF (Bryden, 1966; Heron, 1957; Mishkin &

Forgays, 1952). The problem with Kinsbourne's formulation, however, is

his assertion that with unilateral hemispheric activation, any type of

stimulus (verbal or nonverbal) in contralateral hemispace is more easily

perceived. At first glance, it is difficult to see how a hemisphere

(i.e., the left hemisphere) can process a stimulus (i.e., visuospatial)

for which it is not specifically equipped to handle. This, of course,

is predicated on the laterality model stating that the hemispheres are

differentially organized to mediate specific types of stimuli (i.e., the

"stimulus-specific" model). No problem arises if one holds the view

that each hemisphere can simultaneously process component parts of any

complex stimulus according to its characteristic descriptive systems

(i.e., the "strategy-specific" model).








Support for the assumption that hemispheric activation produces

shifts in physical orientation (i.e., eye, head, and body turning) con-

tralateral to the primed hemisphere has been drawn from electrical stim-

ulation and ablation studies with monkeys (Crosby, 1953; Kennard & Ec-

tors, 1938) and man (Penfield & Roberts, 1959). Less direct support for

this postulate is based on findings from studies of lateral eye movements

during reflective thought.

In this paradigm, Ss are presented with either verbal or spatial

reasoning problems and their eye movements are then monitored. Kinsb-

bourne (1972) has argued that verbal questions, which presumably pro-

duce left hemisphere activation, should induce lateral eye movements

to the right because of the functional proximity of left-frontal right-

ward turning centers. The converse is argued for visuospatial problems.

While a number of investigators have been successful at eliciting the

predicted gaze shifts (Gur & Harris, 1975; Kinsbourne, 1972: Kocel, Ga-

lin, &; Ornstein, 1974), this paradigm is replete with critical scoring

and methodologic problems, and any finding, are probably confounded b:,

the left-right nature of the reading process. Furthermore, Erlichmiann

and coworkers (Erlichmann, Weiner, & Baker, 1974), in perhaps the most

carefully controlled series of these studies, as well as several "un-

published" reports (Lewis, 1973; Tankle, 1975) have been unable to rep-

licate earlier findings that verbal questions induce more gaze shifts

to the riqht and spatial questions more gaze shifts to the left.

"Cognitive set" is a hypothetical construct which has been most ex-

tensively addressed by Bruner (1957) and incorporated by Kinsbourne into

his model of perceptual asyinetries. It refers to the probabilistic

expectations that stimuli will be processed according to a particular







strategy. It is based on the assumption that perception and cognition

are selectively organized so that "...new experiences can be assimilated

into ordered categories that are meaningful and functionally useful to

the individual" (Bruner, 1957, p. 128). According to Kinsbourne (1970a,

1975), every S comes into an experimental situation with an implicit

set of expectations, an implicit cognitive set. A cognitive set can be

experimentally induced by "pre-task" instructions, as well as by task

performance itself. In the latter, stimulus presentation presumably

alters the individual's probabilistic expectations about the processing

of subsequent stimuli, thereby inducing a particular cognitive set.

At least two issues regarding cognitive set are in need of further

clarification. One relates to whether a cognitive set induced by "pre-

task" instructions is equivalent to that induced by task performance it-

self. These appear to be different phenomena within the framework of

Kinsbourne's model, since "pre-task" instructions and cueing have not

been found to alter perceptual asymmetries, whereas "priming" tasks

do appear to affect perceptual asymmetries.

The second issue concerns incorporating "cognitive set" into some

neurophysiological model of arousal-attention, since Kinsbourne postu-

lates that it sets up and maintains hemispheric activation. Because

"cognitive set" implies more than a transient phenomenon, one might in-

terpret it as roughly corresponding to a form of "tonic" (i.e., long-

lasting) activation. This contrasts with the phasic activational re-

sponse that is "stimulus bound" (i.e., it appears with stimulus presen-

tation and disappears shortly after stimulus onset). With performance

of some verbal task, stimulus presentation should produce a brief phasic

activational response of the left hemisphere, as well as a more "long-







lasting" or tonic response of the left hemisphere. Although a somewhat

simplistic formulation, the latter might represent a neurophysiological

analogue of cognitive set as defined by Kinsbourne. This problem has

not been experimentally addressed.

Perhaps the most critical assumption of Kinsbourne's model is.that

stimuli or tasks that require specialized hemispheric processing selec-

tively activate the hemispheres. Support for this postulate has been

indirectly inferred from the changes in perceptual asymmetries that are

induced by verbal and nonverbal priming tasks (See Table 7). Although

there are contradictory findings as to whether the cognitive set hypoth-

esis explains perceptual asymmetries, these studies do not specifically

and directly address the hypothesis that verbal and nonverbal stimuli

asymmetrically activate the hemispheres.

In summary, Kinsbourne's attentional model suffers serious problems,

both in terms of equivocal findings for the "cognitive set" hypothesis

and in terms of many of the assumptions underlying this attentional model.

Clearly, the notion of hemispheric attentional balance can be eliminated

entirely. Likewise, the construct "cognitive set" is ambiguous, espe-

cially with respect to how it corresponds to neurophysiological mecha-

nisms. Furthermore, the most critical of all Kinsbourne's assumptions,

i.e., that verbal and nonverbal materials asymmetrically activate the

hemispheres, has not be adequately assessed.

Notes

'Evidence in favor of the first and third postulates of the path-
way transmission model of perceptual asymmetries (i.e., cortical pro-
cessing differences and intercallosal transfer of information) has been
robust. Not everyone has accepted, however, the structural account of
the priveleged access of information to the contralateral hemisphere
(Postulate 2). The visual system poses no problem for this assumption
since stimuli in each VHF project directly and exclusively to only one







hemisphere. The problem is the auditory system where each hemisphere
receives direct auditory inputs from both ears via ipsilateral and con-
tralateral pathways.
To account for the bilateral wiring of the auditory system, Kimura
(1961, 1967) proposed that under conditions of dichotic competition,
the ipsilateral ear-hemisphere pathways were rendered essentially non-
functional and suppressed by either afferent or central mechanisms. Be-
cause of this "ipsilateral occlusion," it was argued that the contralat-
eral ear-hemisphere connections were functionally prepotent during di-
chotic listening.
While there is little question that dichotic competition is suffi-
cient to produce perceptual asymmetries, there is considerably disagree-
ment as to whether dichotic competition, and ipse facto ipsilateral oc-
clusion is actually necessary. In fact, it does not appear to be essen-
tial since RT measures of laterality have demonstrated monaural ear ad-
vantages with quite simple verbal tasks (Bever, Hurtig, & Handel, 1976;
Catlin & Neville, 1976; Catlin, Vanderveer, & Teicher, 1976; Fry, 1974;
Hayden & Spellacy, 1973; Kallman, 1977; Morais & Darwin, 1974; Studdert-
Kennedy, 1972). These laterality findings with monaural tasks have led
some investigators to seriously question the necessity of postulating
the occurrence of ipsilateral occlusion.
An empirical test of this assumption would entail obtaining electro-
physiological responses of ipsilateral and contralateral sensory pathways
during both dichotic and monaural stimulation. This has not been done.
Rather, electrophysiological research conducted thus far has measured
cortical evoked responses, from which inferences are made about which
pathways (contralateral or ipsilateral) were used (Majkowski et al, 1971;
Matsumiya et al., 1972; Mononen & Seitz, 1977). Such findings are equiv-
ocal at best in terms of interpretation, as are recent findings with
split-brain patients. Milner, Taylor, and Sperry (1968), Sparks and Gesch-
wind (1968), and more recently Zaidel (1974) have shown that, while split-
brain patients perform equally well with the left and right ears on mo-
naural identification of digits and letters, their dichotic performance
reveals massive, often total left ear loss. Although these findings
have been interpreted as support for "ipsilateral occlusion," they, in
fact, could be used equally well to support alternative mechanisms, in-
cluding the attentional model proposed by Kinsbourne (1970a).













STATEMENT OF THE PROBLEM

The hypothesis that verbal and nonverbal stimuli asymmetrically

activate the hemispheres can underlie various attentional models of

perceptual asymmetries, whether or not it is the model specifically

outlined by Kinsbourne (1970a, 1975). Consequently, the first purpose

of the present study was to experimentally address this hypothesis.

One way of testing this hypothesis would involve obtaining electrophys-

iological measures of activation subsequent to the presentation of ver-

bal and nonverbal stimuli. Since, however, the correlations among var-

ious electrophysiological measures of arousal are not robust, an alter-

native and perhaps more functional approach involves the employment of

more overt behavioral measures of activation-attention. This is parti-

cularly relevant when one considers that there is sometimes a dissoci-

ation between EEG indices of activation and actual behavioral manifesta-

tions of arousal. In light of thee correlational difficulties, a be-

havioral measure of activation-attention was chosen (i.e., RT) to test

the hypothesis that verbal and nonverbal stimuli asymmetrically activate

and prepare the hemispheres for action.

One major problem inherent in behaviorally assessing this hypothesis

is selecting a paradigm in which findings cannot be interpreted equally

well according to both a pathway-transmission and an attentional model.

The paradigm that was chosen was an analogue of that used by Lansing et

al. (1959), who measured RTs to stimuli that were sometimes preceded by

a warning stimulus (WS). These investigators found that faster RTS







obtained in the WS condition, were correlated with EEG desynchroniza-

tion and suggested that the WS served the function of alerting and pha-

sically arousing the S.

In the present study, manual RTs were obtained to a "neutral" light

flash that was preceded by either a verbal or nonverbal WS. Both the

light flash and WS were presented at central fixation. The light flash

stimuli were considered "neutral" because previous findings have indi-

cated that they are processed equally well by both the left and right

hemispheres (Filbey & Gazzaniga, 1969). Earlier studies have also demon-

strated that digital movements are mediated by the contralateral hemi-

sphere, i.e., right hand-left hemisphere, and that there are hemispheric

processing differences for verbal (i.e., words) and visuospatial (i.e.,

faces) stimuli (Gazzaniga, 1970).

Consequently, if verbal WS asymmetrically activate and prepare the

left hemisphere for action, then manual responses initiated by that hemi-

sphere should be faster than responses initiated by the right hemisphere.

According to a material specific activation hypothesis, RTs by the right

hand should be faster than left hand RTs when verbal WS are used. Con-

versely, left hand RTs should be faster than right hand RTs when nonver-

bal WS are used. Thus, an activational-attentional model would predict

RT asymmetries with this paradigm.

What would a pathway-transmission model predict? As discussed pre-

viously, this model postulates that perceptual asymmetries arise when

verbal and nonverbal stimuli have the most direct access to the "target"

hemisphere. In the present paradigm, the WS are presented at central

fixation such that both hemispheres have immediate and direct access to

the information conveyed by the WS. However, responses are not made to








the verbal and nonverbal WS, but to the "neutral" light stimulus that

follows the WS. Thus, according to a pathway-transmission model, re-

sponses to a centrally presented neutral stimulus should be equally rapid

for the left and right hand. This model does not in any straight for-

ward manner take into account any material specific effects of the WS

that precedes the RT stimulus. Consequently, the pathway-transmission

model predicts that no RT asymmetries should occur with this paradigm,

whereas as attentional model predicts that RT asymmetries should occur.

Specifically, the following predictions are made with respect to

the hypothesis of material specific hemispheric activation. If verbal

WS asymmetrically activate and prepare the left hemisphere for action,

then RTs to a neutral stimulus will be faster for the right than left

hand (Hypothesis I). Conversely, if nonverbal WS asymmetrically acti-

vate and prepare the right hemisphere for action, then RTs to a neutral

stimulus will be faster for the left than right hand (Hypothesis II).

Reaction times to a neutral stimulus will be faster with verbal than

nonverbal WS when the right hand initiates a response (Hypothesis III).

Finally, when responses are initiated by the left hand, RTs will be faster

with nonverbal than verbal WS (Hypothesis IV).

If it is found that verbal and nonverbal WS produce RT asymmetries

as predicted b\ the attentional hypothesis, this would suggest that the

US induced phasic as)nTmetric hemispheric activation. A critical ques-

tion concerns the extent to which this asynmnetric activation is temporally

related to the WS and represents the second purpose of the present study.

To address how long this asymmetric activation might persist, unwarned

RT trials were randonly, embedded among those trials that were immediately

preceded b,, a verbal or nonverbal IS. If material specific hemispheric







activation is "stimulus bound" and "short lasting," then no RT asymme-

tries should occur for the unwarned trials embedded within the context

of a verbal or nonverbal task. If, however, this asymmetric activation

is "long lasting" and "tonic," then a similar pattern of RT asymmetries

should occur for both the warned and unwarned trials. Kinsbourne's

(1970a, 1975) cognitive set hypothesis would also predict that warned

and unwarned trials within a session should exhibit similar patterns

of RT asymmetries. The verbally/nonverbally warned trials would repre-

sent a "priming" task which should induce a verbal/nonverbal cognitive set.

It is possible that asymmetric hemispheric activation might persist

throughout the session and even extend beyond the session itself. How-

ever, this "tonic" activation might attenuate immediately at cessation

of the verbal and nonverbal RT tasks. To determine whether asymmetric

hemispheric activation extends beyond the verbal and nonverbal sessions,

blocks of unwarned RT trials were presented after the completion of the

main verbal and nonverbal tasks. Again, similar patterns of RT asymme-

tries should occur for the post-session unwarned RT trials and the with-

in-session warned trials, if asymmetric hemispheric activation is rela-

tively long lasting. Within the framework of Kinsbourne's model, the

main verbal and nonverbal RT tasks would constitute "priming" tasks.

Kinsbourne would predict that the subsequent neutral task (i.e., the

post-session trials) should be affected by the priming task and should

exhibit RT asymmetries similar to those obtained on the "priming" task.

The third purpose of the present study concerns the distinction be-

tween involuntary selective attention and voluntary selective attention.

It is not known, for example, whether asymmetric hemispheric activation

can be induced by the presentation of any verbal or nonverbal stimulus







that does not require the S to make some discriminative judgement about

stimulus categorization (attention) and response initiation (intention).

If a response-linked discriminative judgement is not necessary, then a

simple RT paradigm should be sufficient for eliciting RT asymmetries.

In this case, one could argue that asymmetric activation is character-

ized as a form of involuntary selective attention, in which the individ-

ual has little volitional control over a biologically innate activational

response (i.e., OR) to verbal and nonverbal stimuli.

If, however, a response-linked decisional judgement is critical for

inducing asymmetric hemispheric activation, then RT asymmetries should

not occur with a simple RT paradigm. More complex RT tasks, in which the

WS conveys discriminative information about what type response should be

made to the RT stimulus, would be required before RT asymmetries could

be obtained. In this case, one might argue that asymmetric hemispheric

activation is characterized as a form of voluntary selective attention.

In order to distinguish between these two possibilities (i.e., vol-

untary versus involuntary), two experimental conditions were administered.

For both conditions, verbal and nonverbal WS were used and manual responses

were made to a neutral light stimulus. Similarly, both conditions con-

tained embedded unwarned trials and were preceded and followed by blocks

of unwarned Pre-session and Post-session trials. One condition, however,

consisted of a simple warned RT paradigm, in which individuals were in-

structed to respond on every trial. The other condition was a "go-no go"

RT paradigm, in which half the WS designated that a response should be

made to the RT stimulus and the remaining WS indicated that no response

should be made to the RT stimulus. Thus, the simple RT paradigm involved

no discriminative decision concerning response ;election, whereas a

response-linked judgement was requited in the "go-no go" condition.












METHOD

Subjects

The Ss were 32 right-handed college students, 16 females and 16

males, who ranged in age from 18 to 27 yrs (Mean = 21 yrs). All Ss

were right-handed according to self-report and participated in this

study as partial fulfillment of an undergraduate psychology course.

Half the Ss (eight males and eight females) were randomly assigned to

one experimental condition and the remaining Ss to a second experimental

condition.

Apparatus

In both experimental conditions, RTs were obtained to a binaurally

presented tone of 250 ms. This tone was preceded at random intervals

(500-1500 ms) by a visual WS of 500 ms. The WS were projected at cen-

tral fixation via a tachistoscope (Polyietric, Model V-1459-B) onto a

rear-view screen. At random intervals following the WS offset, a com-

puter generated pure tone (500 Hz) was binaurally presented via Senn-

heiser stereo headphones at 60 db. To mask any distracting sounds, a

low level of white noise (30 db) was continuously presented through the

headphones, with the tone stimulus superimposed upon it. A BRS Foringer

module was programmed to control stimulus durations, interstimulus inter-

vals, and six sec intertrial intervals. Tone onset triggered a digital

timer that was stopped when a manual response key was depressed by the

index finger of either the right or left hand. The response key was

placed on a table in front of S and located either 30 cm to the left or








to the right of the S's midline. Throughout the sessions, the Ss sat

at a table 130 cm in front of the screen with their heads positioned in

a commercial chin and forehead rest.

Warning Stimuli

Two classes of WS were used: verbal and nonverbal. The WS were

presented in a blocked design so that only verbal or nonverbal WS were

used during any one session. The verbal WS consisted of eight three-

letter concrete nouns that were set in upper-case English letters. Each

letter was vertically arrayed on a slide and appeared white against a

darker grey-black background. The eight words were randomly assigned

to one of two sets of stimuli. One set included the words BAG, DOT, PEN,

and GUM (Set A), and the other set included FLY, DOG, BEE, and LEG (Set B).

The nonverbal WS consisted of two sets of four faces each. The

faces were obtained from a college yearbook and selected in such a way as

to minimize verbal coding (i.e., no glasses, no facial hair, hairlines

obscured, all faces smiling and oriented in the same direction). One

set consisted of four male faces (Set C) and the other set consisted of

four female faces (Set D).

Go-No Go Condition

The 16 Ss assigned to this condition participated in four separate

testing sessions given one week apart and lasting approximately one hr

each. Two of the sessions were "verbal" in that words were used as WS.

In one verbal session, the words from Set A were used and in the other

verbal session the words from Set B were used as WS. The remaining two

sessions were "nonverbal" and the faces from Set C or Set D were used

as WS. Within each session, the WS were randomized and occurred equally

often across trials. Half the Ss received the two verbal sessions first

and half the two nonverbal sessions first.







Go and Tonic Trials

In each session, two of the four stimuli within each WS set were

assigned as "go" stimuli and two as "no go" stimuli. The "go" WS sig-

nalled that the S should depress the response key at tone onset, and

the "no go" WS signalled that no response should be made to tone onset.

The stimulus pairs, assigned as "go" or "no go" WS, were counterbalanced

across Ss and conditions.

At the beginning of each session, the Ss were trained to discrimi-

nate between the "go" and "no go" WS. The designated WS were shown for

two min to the Ss who were instructed to remember them. This was fol-

lowed by a block of 64 practice trials in which the Ss were given feed-

back about their correct and incorrect responses. Incorrect responses

were of two types including: (a) those in which no response was made

to the RT stimulus when it was preceded by a "go" WS (i.e., omissions);

and (b) those in which responses were made to the RT stimulus when it

was preceded by a "no go" WS (i.e., commissions).

The practice trials were followed by two blocks of 64 experimental

trials, with a five min rest interval between the two blocks. Within

each block of trials, there were 26 warned "go" trials and 26 warned

"no go" trials. For the remaining 12 trials in each block, no WS pre-

ceded the tone. These unwarned trials ("tonic" RTs) were randomly em-

bedded among the warned trials, and Ss were instructed to respond to

these. Thus, Ss responded to the RT stimulus when it was preceded by

either the "go" WS (Go RT) or no WS (Tonic RT).

Two sessions were completed by the right hand (one verbal and one

nonverbal) and two sessions by the left hand. Each S, therefore, re-

ceived all combinations of Hand X Type US (i.e., Right hand-Verbal WS,








Left hand-Verbal WS, Right hand-Face WS, Left hand-Face WS). Hand order

was counterbalanced across sessions and Ss (RLLR or LRRL).

Pre-Post Trials

Prior to the beginning of the verbal and nonverbal RT tasks, each

S performed 20 simple RTs to a binaurally presented tone. These trials

were unwarned and had intertrial intervals ranging from five to 10 sec.

Half the trials were completed by the right hand and half by the left

hand (RLLR or LRRL). At the end of each verbal and nonverbal session,

this identical procedure was repeated. Thus, both "Pre" and "Post" sim-

ple RTs were obtained for each hand during each verbal and nonverbal

session.

Simple RT Condition

The 16 Ss assigned to this condition participated in two separate

testing sessions (one verbal and one nonverbal) given one week apart

and lasting approximately one hr each. In the verbal session, either

the words from Set A or Set B were used as WS and in the nonverbal ses-

sion, faces from Set C or Set D were used as WS. The WS sets were

counterbalanced across Ss. Within each session, the WS were randomized,

but occurred equally often across trials. Half the Ss received the

verbal session first and half the nonverbal session first.

Go and Tonic Trials

Unlike the Go-No Go Condition, Ss assigned to the simple PT task

were instructed to respond on every trial. This eliminated any differen-

tial decision as to respond, based on the information conveyed by the

WS. Consequently, every trial was a "go" trial.

Each session consisted of 64 p-ractice trials that were followed by

two blocks of 64 experimental trials. A five min rest interval occurred








between the two blocks of trials, and hand order was counterbalanced

across blocks (RL or LR). Within each block, 12 unwarned trials were

randomly embedded among the 52 warned trials. The Ss were instructed

to depress the response key at tone onset, whether the trials were

warned or unwarned ("tonic" RT).

Following the completion of the verbal or nonverbal trials, the

Ss were shown an array of 16 words/faces arranged on a 20.3 X 27.9 cm

piece of white paper. The Ss were asked to select those items which

had been used as WS during that particular session. The number of cor-

rect choices was recorded. This was done in order to determine whether

the Ss had differentiated among the WS used during a session.

Pre-Post Trials

As described in the Go-No Go Condition, blocks of simple unwarned

RT trials were also administered prior to and following each session.

Each S was given 20 simple RT trials prior to the session and again after

the session was completed. These trials were unwarned, and half were

completed by the left hand and half by the right hand (RLLR or LRRL).

Thus, both "Pre" and "Post" simple RTs were obtained for each hand du-

ring each session.

Analyses

The RTs from each experimental condition (Go-No Go and Simple) were

analyzed in terms of mean RT in ms. Reaction times that exceeded 1000 ms

were defined as omissions and RTs shorter than 50 ms were defined as an-

ticipations. In the Simple RT Condition, the mean number of omissions

and anticipations was computed. In the Go-No Go Condition, the mean num-

ber of anticipations, omissions, and commissions was computed.








Within each condition, the three dependent variables of major in-

terest were: (a) RTG ("go" RTs to the tone that was preceded by the WS);

(b) RTT unwarnedd "tonic" RT trials that were embedded among the warned

RTG trials); and (c) Pre-Post RTs unwarnedd trials prior to and following

the verbal or nonverbal RT tasks).

Since it was felt that intrinsic differences in motor agility be-

tween the left and right hands might disproportionately affect the di-

rection of RTs, baseline RTs were determined for each hand. This was

done by computing the RT means of each hand from the unwarned Pre-ses-

sion trials. For example, in the verbal session where right hand re-

sponses were measured, the baseline RT was computed from the mean of the

10 Pre-session trials. These baseline RTs were used as covariate values

in the subsequent analyses.

For each experimental condition, two separate analyses of covariance

were performed for the two dependent variables, RTG and RTT. Within-S

factors were Hand (Right versus Left), Type WS (Words versus Faces),

Block (1 versus 2), and Covariate. The Pre-Post RTs were analyzed using

an analysis of variance with Hand (Right versus Left), Time (Pre versus

Post), and Type Intervening Task (Verbal versus Nonverbal) as within-S

factors.













RESULTS

The results are presented in two sections. The first section deals

with the RTs obtained in the Go-No Go Condition and generally addresses

the hypothesis that material specific hemispheric activation represents

a form of voluntary selective attention-intention. The second deals

with RTs obtained from the Simple RT Condition and addresses the hypothe-

sis that material specific hemispheric activation can be characterized

as a form of involuntary selective attention.

If the predicted RT asymmetries are found only in the Go-No Go Con-

dition, this would suggest that material dependent activation is volun-

tary. If, however, the predicted RT asymmetries are also obtained in

the Simple RT Condition, this would indicate that material specific

hemispheric activation is involuntary. Since the distinction between

these two possibilities can be most easily drawn by analyzing the two

RT conditions separately, the data are presented in this way.

The overall mean RTs from the Go-No Go and Simple RT tasks (col-

lapsed across Hand, Type WS, and Block) are presented in Table 9. As

can be seen, the RTs from the Go-No Go Condition are approximately 10 ms

faster than those obtained from the Simple RT task. More importantly,

however, is that the RTs from the Go-No Go Condition are much more vari-

able. The source of this variability is unclear and possibly reflects

a number of factors including S sampling differences, paradigm differences

(four sessions versus two sessions), or task differences directly re-

lated to the different cognitive requirements of a go-no go versus a







simple RT procedure. It was this differing variability that further

prompted separate analyses of the two experimental tasks.

TABLE 9

MEAN REACTION TIMES IN THE GO-NO GO AND
SIMPLE RT TASKS


Go-No Go Condition Simple RT Condition

M* 229.6 ms 238.9 ms

S.D. 56.6 ms 28.9 ms


*Means are adjusted for the covariate.


Analyses of the Go-No Go Condition

The overall error rate (anticipations + omissions + commissions/to-

tal number of trials) was 3.8%. Consequently, the RT means of the Hand X

Type WS X Block matrix for each S are based on approximately 96% of the

administered trials. Three separate analyses (RTG, RTT, and Pre-Post RTs)

were performed and the results of each are discussed separately below.

Go RTs (RTG)

The results of the analysis of covariance of the "go" RTs are pre-

sented in Table 10. A significant covariate effect was found (F1, 4 =

10.7, p< .01), which justifies the use of the covariance procedure. The

mean covariate values for each session are presented in Table 11.

The Hand X Type WS interaction was also significant (F1,104 789

p4.01) and is depicted in Figure 1. Post hoc comparisons, using Dun-

can's procedure, indicated that: (a) RTs by the right hand were signi-

ficantly faster when the WS was a word, rather than a face (Word = 202.8 ms,

Face = 240.5 ms, p .05); (b) With word WS. right hand RTs were faster







TABLE 10

SUMMARY OF ANALYSIS OF COVARIANCE
OF THE MEAN "GO" REACTION TIMES IN THE GO-NO GO CONDITION


Source df SS F p


Covariate
Hand
Warning Stimulus (WS)
Block
Hand X WS
Hand X Block
WS X Block
WS X Block X Hand
Within-Subject
Error

TOTAL


34239.61
7759.51
2682.19
2329.80
25253.11
2350.20
4793.73
712.92
342581.79
332749.04

845668.92


*p .01


**Not significant


TABLE 11


RIGHT AND


MEAN COVARIATE VALUES OF THE
LEFT HANDS FOR VERBAL AND NONVERBAL SESSIONS
IN THE GO-NO GO CONDITION


Verbal Sessions Nonverbal Sessions

Right Hand 225.9 ms 215.9 ms

Left Hand 235.1 ms 236.4 ms


10.70
*2.47
.83
.72
7.89
.73
1.49
.22
7.13


.0015*
NS**
NS
NS
.0059*
NS
NS
NS



























F- -
LU



nc

L- C)
C c10






T,
-- ._
I-



LLJ
/ O


C)



CD3
I. 1-

C: II .I
W-


I-CD





-oC o <

0- = C



2. cm
\ I C
C.J




O O O O


su u N-
W \LLU
\SU Ur It








than left hand RTs (Right = 202.8 ms, Left = 246.0 ms, p <.05); and

(c) For the left hand, there were no RT differences between word and

face WS (Word = 246.1 ms, Face = 227.0 ms). Similarly, there were no

between-hand differences in RT when faces were used as WS (Right = 240.5 ms,

Left = 227.0 ms).

The means and Duncan's post hoc test of significance for the Hand X

Type WS interaction are presented in Tables 12 and 13. All remaining

effects and interactions were nonsignificant.

Tonic RTs (RTT)

The results of the analysis of covariance for the RTT data are pre-

sented in Table 14. A significant covariate effect was found (F1,104

8.01, p<.01). The mean covariate values for each session are identical

to those used in the RTG analysis and are again found in Table 11.

All remaining effects and interactions were nonsignificant. The

Hand X Type WS interaction approached, but did not reach significance

(F1,104 = 3.12, p = .08; See Figure 2). The trend of this interaction,

which was in the same direction as that found with the RTG data, lends

some support to the hypothesis of a "tonic" carry over effect from the

warned to the unwarned trials. The means of this interaction are pre-

sented in Table 15.

Pre-Post RTs

The results of the analysis of variance of the Pre-Post data are

presented in Table 16. Significant main effects were found for both

Time (F,233 = 6.26, p< .01) and Type Intervening Task (F1,233 = 5.89,

p<.01), The means for these effects are presented in Tables 17 and 18.

Reaction times prior to the intervening task (Pre = 227.0 ms) were fas-

ter than those following the intervening task (Post = 234.6 ms), and








TABLE 12

MEAN REACTION TIMES OF RIGHT AND LEFT HANDS
WITH WORD OR FACE WARNING STIMULI IN THE GO-NO GO CONDITION





Right Hand 202.9 ms* (45.4 ms)** 240.5 ms (100.4 ms)

Left Hand 246.1 ms (73.1 ms) 221.8 ms (89.5 ms)


*All means are adjusted for the covariates.
**Standard deviations are in parentheses.




TABLE 13

DUNCAN'S POST HOC COMPARISONS OF MEAN REACTION TIMES
OF RIGHT AND LEFT HANDS WITH WORD OR FACE WARNING STIMULI


Mean. Differences

Ordered Means 1 2 3 4

202.9 ms 24.9 37.6* 43.2*

227.8 ms 12.7 18.3

240.5 ms 5.6

246.1 ms


*p( .05 (Two-tailed test of significance)







TABLE 14

SUMMARY OF ANALYSIS OF COVARIANCE
OF THE MEAN "TONIC" REACTION TIMES IN THE GO-NO GO CONDITION


Source df SS F p

Covariate 1 77886.19 8.01 .0056*
Hand 1 6.55 .00 NS**
Warning Stimulus (WS) 1 820.12 .08 NS
Block 1 9449.98 .97 NS
Hand X WS 1 30338.67 3.12 .0800
Hand X Block 1 11279.09 1.16 NS
WS X Block 1 1064.16 .11 NS
WS X Block X Hand 1 1512.79 .16 NS
Within-Subject 15 1073449.08 7.36
Error 104 1010943.11

TOTAL 127 2387710.73


*p< .01

**Not significant







TABLE 15

MEAN REACTION TIMES OF RIGHT AND LEFT HANDS
TO AN UNWARNED TONE EMBEDDED AMONG VERBAL OR NONVERBAL WARNED
TRIALS IN THE GO-NO GO CONDITION



Verbal Trials Nonverbal Trials

Right Hand 278.3 ms* (120.6 ms)** 314.5 ms (140.1 ms)

Left Hand 309.7 ms (135.7 ms) 283.9 ms (144.2 ms)


*All means are adjusted for the covariates.

**Standard deviations are in parentheses.





80














z

oa o L




O



-z/
o



w-


m u

I -









0 O ,
o uW
\- ]




v \C







o00 (0 U
'S I- LU
5 _






SUJ ui U-
*su ui jty


/


Q /
/


0
OJ
N'









TABLE 16

SUMMARY OF ANALYSIS OF VARIANCE
OF THE MEAN "PRE-POST" REACTION TIMES IN THE
GO-NO GO CONDITION


SOURCE df SS F p

Hand 1 4.78 .00 NS*
Time 1 370957.12 6.26 .013**
Intervening Task (IT) 1 349059.41 5.89 .016**
Hand X Time 1 70523.44 1.19 NS
Hand X IT 1 926.44 .02 NS
Time X IT 1 2306032.06 38.93 .001**
Hand X Time X IT 1 50934.84 .86 NS
Within-Subject 15 2104569.21 23.69
Error 233 13800542.56

TOTAL 255 37994849.89


*Not significant
**p / .05








TABLE 17

MEAN "PRE-POST" REACTION TIMES
PRIOR TO AND FOLLOWING THE MAIN TASK IN THE GO-NO GO CONDITION



Pre-Session Post-Session

M 227.0 ms 234.6 ms

S.D. 35.2 ms 41.0 ms


TABLE 18

MEAN "PRE-POST" REACTION TIMES
WITH EITHER VERBAL OR NONVERBAL INTERVENING TASK
IN THE GO-NO GO CONDITION


Verbal Nonverbal

M 234.5 ms 227.2 ms

S.D. 40.1 ms 36.9 ms








overall RTs performed with the nonverbal intervening task (Mean = 227.1 ms)

were faster than those performed with the verbal intervening task (Mean =

234.5 ms).

The Time X Type Intervening Task interaction was also significant

(F1,233 = 38.9, p<.01) and is depicted in Figure 3. Post hoc compari-

sons (Duncan's procedure) revealed that: (a) When the intervening task

was nonverbal, there were no RT differences between Pre- and Post-trials

(Pre = 232.8 ms, Post = 221.4 ms); (b) In contrast, RTs following the

verbal task were significantly slower than those preceding the verbal

task (Pre = 221.2 ms, Post = 247.8 ms, p< .05); (c) During the Post-

trials, RTs following the nonverbal task were significantly faster than

those following the verbal task (Nonverbal = 221.4 ms, Verbal = 247.8 ms,

p<.05); and (d) During the Pre-trials, there were no differences in RTs

prior to the verbal task, when compared to those prior to the nonverbal

task (Verbal 221.2 ms, Nonverbal 232.8 ms). The means of this interaction

and Duncan's post hoc test of significance are presented in Tables 19

and 20.

Anticipations, Omissions, and Commissions

The mean numbers of anticipations, omissions, and commissions for

each S were 3.0, 7.0, and 5.9, respectively. The frequency of each of

these error types across each Hand X Type WS condition are presented in

Tables 21, 22, and 23. Since the occurrence of error types across the

Hand X Type WS categories was not independent for each S, one basic as-

sumption underlying the Chi-Square test was violated. Consequently, the

Cochran Q-test (Seigle, 1956) for related dichotomized data was used.

Results of three separate Cochran Q-tests revealed that: (a) the

frequency of anticipations was different across the Hand X Type WS




































C.


7




0


7


*sw Ul u _!


IC
le
/
0/
I/





C

I.-


,I
/ ^


U)



U)
0
4-





















a_


0
i-
"-


--

CO
o





L-
QO





CZ:

o
i







czln
o-











1 L
00








LU
CDW





U-


















czl
0 I


1








TABLE 19

MEAN REACTION TIMES PRIOR TO AND FOLLOWING
EITHER A VERBAL OR NONVERBAL INTERVENING TASK IN THE GO-NO GO CONDITION


Pre-Session Post-Session

Verbal Intervening 221.2 ms (30.8 ms) 247.8 ms (44.1 ms)
Task

Nonverbal Intervening 232.8 ms (39.5 ms) 221.4 ms (33.6 ms)
Task


*Standard deviations are in parentheses.






TABLE 20

DUNCAN'S POST HOC COMPARISONS OF MEAN REACTION TIMES
PRIOR TO AND FOLLOWING A VERBAL OR NONVERBAL INTERVENING TASK
IN THE GO-NO GO CONDITION


Ordered Means 1 2 3 4

221.2 ms -.2 11.6 26.6*

221.4 ms 11.4 20.4*

232.8 ms 15.0

247.8 ms


*p< .05 (Two-tailed test of significance)








TABLE 21

FREQUENCY OF ANTICIPATIONS
BY RIGHT AND LEFT HANDS WITH VERBAL AND NONVERBAL
WARNING STIMULI IN THE GO-NO GO CONDITION


Verbal Nonverbal

Right Hand 24 11

Left Hand 5 9


TABLE 22

FREQUENCY OF OMISSIONS
BY RIGHT AND LEFT HANDS WITH VERBAL AND NONVERBAL
WARNING STIMULI IN THE GO-NO GO CONDITION


Verbal Nonverbal

Right Hand 14 38

Left Hand 11 49


TABLE 23

FREQUENCY OF COMMISSIONS
BY RIGHT AND LEFT HANDS -JITH VERBAL AND NONVEPBAL
WARNING STIMULI IN THE Gn-riO ,0 CONDITION


Verhal


N. n ,rb= 1


Right Hand

Left Hand


V b al








categories (Q = 14.7, p<.01). With verbal WS, more anticipations were

made by the right than left hand. However, there were no between-hand

differences with nonverbal WS; (b) The frequency of omissions differed

across the Hand X Type WS categories (Q = 16.7, p< .001). More omissions

were made with nonverbal than verbal WS; and (c) The frequency of com-

missions also differed across the Hand X Type WS categories (Q = 15.2,

p <.001). More commissions were made with nonverbal than verbal WS.

Analyses of the Simple RT Condition

The overall error rate (anticipations + omissions/total number of

trials) was .7%. Consequently, the RT means of the Hand X Type WS X

Block matrix for each S are based on approximately 99% of the adminis-

tered trials.

Go RTs (RTG)

The results of the analysis of covariance of -the "go" RTs are pre-

sented in Table 24. A significant main effect for Type WS was found

(F1,104 = 12.5, p <.05). Reaction times with face WS were faster than

those with word WS (Face = 229.3 ms, Word = 247.8 ms; See Table 25).

All remaining effects and interactions were nonsignificant.

Since it was possible that the Ss in the Simple RT Condition may

have responded to the word/face WS in a nondiscriminatory manner, this

was tested by determining the Ss'accuracy in correctly identifying, from

an array of 16 faces/words, the particular WS that had been used during

a session. At the completion of the verbal session, all 16 Ss were 100%

accurate in identifying the four word WS. Following the nonverbal ses-

sion, 13 Ss were 100% accurate and three Ss were 75% accurate in iden-

tifying the target face WS.








TABLE 24


SUMMARY OF ANALYSIS OF COVARIANCE
OF THE MEAN "GO" REACTION TIMES IN THE SIMPLE RT


CONDITION


Source df SS F p

Covariate 1 523.35 .63 NS*
Hand 1 339.39 .40 NS
Warning Stimulus (WS) 1 10491.43 12.51 .0001**
Block 1 38.32 .61 NS
Hand X WS 1 510.21 .05 NS
Hand X Block 1 788.72 .94 NS
WS X Block 1 304.44 .36 NS
Hand X Block X WS 1 49.05 .06 NS
Within-Subject 15 150028.95 11.63
Error 104 87156.74

TOTAL 127 538447.25


*Not significant
**p / .01





TABLE 25

MEAN "GO" REACTION TIMES
WITH WORD OR FACE WARNING STIMULI
IN THE SIMPLE RT CONDITION



Verbal Nonverbal

M* 247.4 ms 22.3 ms

S.D. 62.8 ms 52.3 ms


*Means are adjusted for the covariate







Tonic RTs (RTT)

The results of the analysis of covariance of the RTT data are pre-

sented in Table 26. There were no significant main effects or interac-

tions. The Condition effect approached, but did not reach significance

(F1,104 = 3.3, p = .07). Thus, there was a trend for the unwarned trials

embedded within the nonverbal task to be faster than those trials embedded

within the verbal task (Nonverbal = 301.9 ms, Verbal = 334.9 ms, See

Table 27).

Pre-Post RTs

The results of the analysis of variance of the Pre-Post RTs are pre-

sented in Table 28. A significant Hand X Time interaction was found

(F1,105 = 7.3, p< .01) and is depicted in Figure 4. Post hoc compari-

sons, using Duncan's procedure, revealed that: (a) RTs by the right hand

remained unchanged across Pre- and Post-trials (Pre = 245.s ms, Post =

252.4 ms), regardless of the type of intervening task; (b) In contrast,

left hand RTs were significantly faster during the Post-trials when com-

pared to left hand RTs during the Pre-trials (Post = 234.9 ms, Pre =

262.4 ms, p (.05); (c) During the Pre-trials, right hand RTs were fas-

ter than left hand RTs (Right = 245.3 ms, Left = 261.4 ms, pZ4.05); and

(d) Conversely, left hand RTs were faster than right hand RTs during the

Post-trials (Left = 234.9 ms, Right = 252.4 ms, p, .05). The means of

this interaction and post hoc tests of significance are presented in

Tables 29 and 30.

Anticipations and Omissions

The mean number of anticipations and omissions for each S were 2.0

and .5, respectively. The frequency of each of these error types across

each Hand X Type WS category are presented in Tables 31 and 32. The




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs