-
MADELEINE M. GROSSCalifornia State University, San Jose, San
Jose, California 95114
Hemispheric specialization forprocessing of visually
presentedverbal and spatial stimuli*,**
Two "same-different" reaction time experiments, analogous in
task demandsmade on the S, were designed to test laterality
differences in. perception. Tennormal right-handed Ss performed a
verbal task in which they decided whetheror not two three-letter
words belonged to the same conceptual class. Tendifferent Ss
performed a spatial task in which they decided whether two
16-cellmatrices with 3 blackened cells were identical. Reaction
times were found to besensitive to laterality differences in
perception. Verbal stimuli were processedfaster when presented in
the right visual field, and thus projected directly to theleft
cerebral hemisphere; spatial stimuli were processed faster when
presented inthe left visual field, and thus projected directly to
the right cerebral hemisphere.These results were analyzed in terms
of implications regarding hemisphericasymmetries for processing of
verbal and spatial material and the nature ofinterhemispheric
transfer of information.
A large body of literature, derivedfrom a number of diverse
researchmethodologies, lends support to theidea of differential
specialization ofthe two cerebral hemispheres in man.One such
apparent division offunction, suggested by studies ofpatients with
asymmetric cerebraldamage (e.g., Milner, 1968) or withsurgical
disconnection of the cerebralhemispheres (Gazzaniga, Bogen,
&Sperry, 1965; Gazzaniga & Sperry,1967), is that of
separation ofprocessing of linguistic andnonlinguistic information.
Perceptuallaterality studies in both the auditoryand visual
modalities on normal Sshave also supported the conclusionthat
linguistic processing and/orou tput appears to be
primarilylocalized in the left cerebralhemisphere, while the
righthemisphere is specialized fornonlinguistic (in the visual
modality,spatial) processing. This cerebraldominance conformation
is especiallypredominant in right-handed Ss, theoverwhelming
majority of whom have
*This research was supported by NIMHPredoct01'al Fellowship MH
49556-01 to theauthor and NIMH Grant NB-06501. Thispaper is based
upon a dissertation submittedby the author to the
PsychologyDepartment of Stanford University inpartial fulfillment
of the requirements forthe degree of Doctor of
Philosphy,"Hemispheric Specialization for theProcessing of Visually
Presented Verbal andSpatial Stimuli: A Reaction Time
Analysis."1971.
**1 would like to thank my advisor. Dr.Charles R. Hamilton. for
his helpfulsuaestions during all phases of thisresearch. Thanks are
also due to Sally P.Springer for being instrumental in
firstsuggesting the possibilities of a reaction timeanalysis of
hemispheric dominance (personalcommunication. 1969).
speech output centers in the leftcerebral hemisphere.
A number of previous lateralitystudies in normal Ss have used
ameasure of percent correct report. Inthe auditory modality, for
example,Kimura (1961) found that digitspresented to the right ear
wereidentified more accurately than thosepresented to the left ear
in a dichoticpresen tation. Conversely, Kimura(1964) found that
recognition withthe left ear was superior when the taskwas
nonverbal.
In the visual modality, lateralitystudies have utilized stimuli
presentedto either the left or the right visualhemifield in order
to producecortically lateralized presentations.The structure of the
human visualsystem is such that the initialprojection of stimuli
presented to theright of fixation (RVF) is solely to theleft
cerebral hemisphere, while that ofstimuli presented to the left
offixation (LVF) is to the right cortex.Bryden (1965) and Kimura
(1966), forexample, have reported a right-fieldadvantage for the
perception of verbalmaterial. Conversely, Kimura (1969)found that
location of a dot in spacewas superior in the left visual
field,while Schell and Satz (1970) found aleft-field superiority in
recognition ofa previously shown block design.
If the perceptual asymmetries foundin visual presentations are
interpretedas being indicative of a hemisphericlocalization of some
part of theperception of or response to thesestimuli, two
differentprocessing/output alternatives aresuggested. It may be
that all or somenecessary part of verbal processingmust take place
in the left hemisphere(at least in right-handed Ss), such that
those verbal stimuli presented to theRVF have the advantage of
moredirect access to the appropriateprocessing center.
Alternatively, verbalstimuli may be processed in the
righthemisphere, but (1) the processingmay be slower and not as
accurate,and/or (2) the verbal response mustalways be directed by
the lefthemisphere, permitting possibleinformation loss during the
callosaltransmission preceding output. Forspatial stimuli, direct
presentation tothe right hemisphere should yield aprocessing
advantage. Such anadvantage would be expected to bedecreased (if
the left hemisphere werealso capable of doing some
spatialprocessing) if verbal response wererequired.
Recently, several studies utilizing areaction time (RT) measure
ofhemispheric asymmetries wereundertaken. In the auditory
modality,Springer (1971a) found that motor RTto right ear verbal
targets in a dichoticsituation was shorter than that to leftear
targets. In the visual modality, anumber of studies (Filbey
&Gazzaniga, 1969; Gibson, Filbey, &Gazzaniga, 1970;
Klatzky, 1970;Moscovitch & Catlin, 1970) supportedthe notion
that hemisphericspecialization of function for thelinguistic VB
spatial processing modes isreflected in RT differences betweenthe
visual hemifields.
It was the object of the presentexperiments to utilize a RT
measure asa potentially more sensitive index ofdominance within an
individual than ispercent correct report, as well as toprovide more
extensive data about thenature of the information flowbetween
hemispheres than haspreviously been possible. Just as thetransfer
of information across thecallosum for processing and/or outputmay
yield a performance decrement inaccuracy, it could also yield a
slowerRT, attributable either to callosalcrossing time or to
information loss.By manipulating stimulus parameters(linguistic VB
spatial input) andresponse parameters (verbal VB motorresponse,
inter- vs intrahemisphericsensory-motor connections by meansof
visual field and responding handcombinations), it should be
possible tostudy the laterality effect in terms ofboth processing
and output functions.
In the present experiments, bothstimulus and response parameters
weremanipulated. Two experiments, one a"verbal" task and one a
"spatial" task,were performed, which madeanalogous task (decision
and response)demands upon the S. In each, Sperformed a
"same"-"different" RTtask. In the verbal task, S decidedwhether two
words presentedsimultaneously were in the same
Perception & Psychophysics, 1972, Vol. 12 (4) Copyright
1972, Psychonomic Society, Austin, Texas 357
-
category (both animals or both partsof the body) or in different
categories.In the spatial task, S decided whethertwo 16-cell
matrices with 3 cellsblackened were the same (all 3 blackcells in
register) or different. In eachexperiment, each S gave
verbalresponses, as well as manual responseswith both right and
left hands.
EXPERIMENT 1: VERBAL TASKIn Experiment I, Ss decided
whether two three-letter word stimuliwere in the same category
or indifferent categories.
MethodSubjects. The Ss were four female
and six male members of the Stanfordcommunity between the ages
of 18and 26, who were paid $1.75 persession. All Ss met the thrt!e
selectioncritera of (1) having normal orcorrected vision of at
least 20/22 ineach eye, as measured with a KeystoneOphthalmic
Telebinocular, with anacuity difference no greater thanone-half
step between eyes (e.g., 20/20and 20/20-), and having at least
100%stereopsis on the Navy eye-screeningtest scale, (2) being
right-handed asdefined by self-label and by the handwith which they
wrote when fillingout the experimental questionnaire,and (3) having
no history ofneurological disorder or speech defect.The 10 Ss
scored an average of 22.7 onthe Crovitz and Zener (1962)handedness
scale.
Stimuli. Two sets of eightthree-letter words were prepared,
oneset of animals and one of parts of thebody. Three-letter words
were chosenin order to preclude any potentialscanning effects [see
White (1969) fora discussion of the laterality literatureoriented
toward this explanation offield differences] . Studies byColtheart
and Merikle (1970), Krueger(1970), and Smith and Haviland(1972)
indicate that such words areperceived as units rather than
scannedfrom left to right. As an additionalsafeguard, the response
was dependentupon a "completed" perception ofboth words, since it
was necessary tocompare their "categories." The eightanimal words
were: ape, cat, cow, dog,elk, hen, pig, rat. The eight parts ofthe
body words were: arm, ear, eye,hip, jaw, leg, rib, toe.
Using these words, 64 stimulusconfigurations were designed.
Eachwas composed of two different words,one above the other, drawn
eitherfrom the same category or fromdifferent categories.
Thirty-two"same" units were designed; 16"same" units were drawn
from the"animal" group and 16 from the"body" group. Each word in
eachgroup appeared in four "same" pairs,
twice on top and twice on the bottom.. Thirty-two "different"
units weremade up, half with an animal word ontop, half with a
"body" word on top.Each word in each group appeared infour
"different" pairs, twice on topand twice on the bottom. Each of
the64 stimulus configurations appearedonce to the left and once to
the rightof fixation, making a total of 128stimulus configurations.
These 128stimuli were divided into four blocksof 32 stimuli; both
the right- andleft-field presentation of a stimulusappeared within
a block, randomizedwith the constraint that no more thanfour
consecutive correct responseswere identical. Ss received these
fourblocks in different assigned orders ineach session, with blocks
presented inreverse order one-half of the time.
The stimuli were made with black48-point Helvetica Medium
Letrasetletters (lAi in. high, subtending about1 deg 14 min of
visual angle at aviewing distance of 23 in.) mountedon white
construction paper. The twowords appeared, respectively, at 1
mmabove and below the fixation dot,which appeared in the
preexposurefield, centered 37 mm to the left orright of the
fixation dot (3 deg 34 minof visual angle). The words were from30
to 34 mm in length, subtendingfrom 2 deg 58 min to 3 deg 16 min
ofvisual angle. Therefore, the closestapproach to the fixation
point wasabout 2 deg.
Procedure. Experimental stimuliwere exposed in a Gerbrands
two-fieldmirror tachistoscope. Centered in onefield was a black
fixation dot; thisfield was always on except when teststimuli were
being presented, and Swas instructed to maintain fixation onthe
center dot. The luminance of thefixation field was 5.9 fL; that of
thetest field was 6.2 fL. The E signaledthe onset of a test by
saying "now."Stimulus duration was 150 msec, topreclude shifts in
fixation.
Reaction time was measured to thenearest millisecond by an
electronicdecade counter (Hunter KlocKounter),which was activated
simultaneouslywith stimulus onset and was stoppedby S's response.
In verbal responsesessions, S spoke into a microphonethat activated
a voice relay (Trans-VoxModel Vox-I) that stopped the clock;half
the Ss responded "it" to thesame stimuli and "It" to different
stim-uli, while half did the reverse. In motorresponse sessions, S
pushed a levergrasped between the thumb andforefinger up or down,
which stoppedthe clock by means of a mechanicalclosing. Those Ss
who verbalized "it"for "same" pushed the lever down for"same" and
up for "dif!erent" stimuli;those who verbalized "It" for "same"did
the reverse.
Each S served in a practice session,during which he became
familiar withthe stimuli and methods ofresponding. Each then served
in eightexperimental sessions, consisting of sixpractice trials
followed by 128 teststimuli. The Ss were instructed torespond as
quickly as possible,consistent with accuracy. Stimuli onwhich
errors were made werereinserted into the stimulus sequence2-5
presentations later. Order ofnature of response (verbal or
manual)for sessions was counterbalancedwithin a S and across Ss. In
the motorsessions, right and left hands were usedin alternate
blocks; the alternation wascounterbalanced within a S and
acrossSs.
ResultsMean RTs for each visual field (right
or left), response mode (verbal, righthand, or left hand), and
response("same" or "different") werecomputed for each session.
Thesemeans excluded error or repeat trials;however, the error rate
for each S waslow and did not seem to differ withregard to visual
field of thepresentation. The mean error rate was4.5%; the range
over Ss was 1.2% to11.5%. In addition, extreme RTs weredropped.
(The drop rule used entailedexamining the distribution of RTs
forall "same" stimuli and all "different"stimuli within a session
separately. If a200-msec gap was found in thedistribution, numbers
that wereseparated from the main body of thedata by the gap were
dropped.Number of dropped stimuli did notdiffer with respect to
side ofpresentation.) The means so obtainedfor each session were
averaged acrosssessions, yielding 12 means for each S.The pattern
of means obtainedaveraged across Ss is shown in Table 1.
An analysis of variance of thesemean RTs (Field by Response
Modeby Response, within Ss)was performed. The analysis ofvariance
showed that RT for thisverbal processing task was faster forthe RVF
than for the LVF, 1,063.2and 1,093.9 msec, respectively[F(l,9) =
11.56; P < .01]. All of the10 Ss showed faster RT to RVFstimuli,
a "right-field effect."
Verbal response was slower thanthat of either the right or left
hand,1,145.0 msec vs 1,048.5 and1,042.2 msec, respectively[F(2,18)
= 11.32; p < .001], whileRTs for right and left hand did
notdiffer significantly [t(9) = 1.005, n.s.]."Same" responses were
faster than"different" responses, 1,042.5 and1,114.7 msec,
respectively[F(l,9) = 15.36; p < .01]. TheResponse Mode by
Responseinteraction was also significant
358 Perception &£ Psychophysics, 1972, Vol. 12 (4)
-
Table 1Overall Mean Reaction Time (MWiMconds) for Each Field by
Respon..
Mode by Respon.. Combination: Verbal T"
ResponseField
Mode Response Ri&ht Left
Right Hand Same 981.6 1023.8 1048.5Different 1081.1 1107.5
Left Hand Same 979.6 1011.1 1042.2Different 1068.2 1110.1
Verbal Same 1115.8 1142.8 1145.0Different 1153.2 1168.1
1063.2 1093.9
Table 2Percent Errors for Each Field by Response
Mode Combination: Verbal Task
ResultsMean RTs for each visual field (right
or left), response mode (verbal, right
5.24.65.0
4.43.34.4
Field
Right Left"(Percent) (Percent)
ResponseMode
Right HandLeft HandVerbal
matrices differed from each other bytwo black cells and had one
black cellin common; Group 3 matrices had noblack cells in common.
The matricesused are shown in Fig. 1.
Using the 12 matrices, 48 stimulusconfigurations, 12 "same" and
36"different," were designed. Each wascomposed of either two of the
samematrix, one above the other ("same"),or of two matrices from
the samegroup, one above the other("different"). Each of the 12
possible"same" configurations appeared threetimes to both the right
and left offixation. Each matrix in each groupwas paired with all
other matrices inthat group in both the top and bottomposition to
form the "different"stimuli. Each different configuratiunappeared
once to both the left andright of fixation. The 144
stimuluspresentations were divided into fourblocks of 36 stimuli;
both the right-and left-field presentations of astimulus appeared
within a block. Theassigned block order for each sessionfor each S
in Experiment 2 matchedan order assigned to an S inExperiment
1.
Multiple copies of the 12 matriceswere made by an offset
printingprocess. The 144 stimuli were made bycutting out individual
matrices andaffixing them to white constructionpaper. The matrices
appeared,respectively, 1 mm above and below afixation dot that
appeared in thepreexposure field, centered 36 mm tothe left or
right of the fixation dot(3 deg 28 min at a viewing distance of23
in.). The matrices were1% x 1% in., subtending a visual angleof 3
deg 8 min x 3 deg 8 min at aviewing distance of 23 in. The
closestapproach to the fixation dot was,therefore, approximately 2
deg.
Procedure. The apparatus andexperimental procedure were
identicalto those of Experiment 1. Each Sserved in eight seuion,
during whichRT measures for the 144 test stimuliwere obtained. For
each S, thesequence of the nature of responseused in a given
session was matched toa sequence assigned to an S inExperiment
1.
EXPERIMENT 2: SPATIAL TASKIn this experiment, as in the
verbal
experiment, Ss made a"same"-"different" decision regardingthe
stimulus presentation. In this case,the decision was whether two
16-cellmatrices with 3 blackened cells wereidentical (all 3
blackened cells incorrespondence) or not ("different").It was felt
that this task would beappropriate for the spatial condition inpart
because Sekuler and Abrams(1968) reported that Ss presented witha
similar task seemed to perform a"template match." In an effort
toemphasize the spatial "Gestalt" natureof the task, Ss were
instructed to "tryto take into account the total
stimulusconfiguration" when making theirresponses.
summed across Ss, between the handby field combinations of the
motorresponse sessions, [x 2 (3) = 5.84, n.s.]or between the
response mode by fieldcombinations, summing across allsessions [x 2
(3) = 4.77, n.s.]. Thepresent results indicate that evidencefor
hemispheric asymmetry, asindicated by RT differences, may befound
even under conditions in whichno significant difference in error
rate isfound. This suggests that RT mayprove to be a more sensitive
measureof hemispheric specialization than isthe more traditional
percent correctmeasure.
MethodSubjects. The Ss were five male and
five female members of the Stanfordcommunity between the ages of
18and 27, who were paid $1.75 persession. All Ss met the
selectioncriteria (vision, right-handedness, andno speech defects)
described inExperiment 1. The 10 Ss scored anaverage of 20.8 on the
Crovitz andZener (1962) handedness scale. Noneof the Ss had served
in Experiment 1.
Stimuli. Three groups of four16-cell black-and-white matrices
with3 cells blackened' were designed.Group 1 matrices differed from
eachother by one black cell and had twoblack cells in common; Group
2
[ F ( 2 ,1 8) = 4.0 1 ; P < .05 ] ; the"different" response
took relativelylonger than "same" for the motorresponse modes than
for the verbal(93 rnsec VB 31 msec). This difference,however, is
significant at only the .05level, one-tailed, and it does notappear
to yield any informationrelevant to hemispheric specialization.
No other interactions weresignificant. The Field by ResponseMode
interaction failed to reachsignificance, indicating no
differentialfield effect depending upon presumedhemispheric locus
of response (forexample, if the RT difference for theright VB the
left visual field had beengreater for the verbal response
andright-hand response than for theleft-hand response, a
statisticallysignificant difference would have beenobtained). A t
test performed on RTdifferences for RVF vs L VF usingmotor
responses alone (motorresponse field difference) showed
nodifferential field effect for the right vsleft hand [t(9) = .13,
n.s.}. A t testbetween combined right-hand andleft-hand motor field
differences andverbal field differences was alsononsignificant
[t(9) =.29, n.s.},indicating that the right-field RTadvantage is
the same with both motorand verbal responses. The Field byResponse
interaction also failed toreach significance, indicating that
theright-field superiority is maintainedregardless of whether S is
making a"same" or a "different" response.
Because the traditional measure ofthe laterality effect has
beendifference in percent correct reportbetween fields, several
analysesregarding error rate for differentconditions were
performed. Percenterrors summed across Ss for eachcondition is
shown in Table 2. A t testfor correlated means was performedon the
differences between thenumber of errors made in the RVF vsthe LVF
for each S (mean error rate of4.1% in the RVF VB 5% in the
LVF).This difference between fields did notreach significance (t(9)
=1.86, n.s.},In addition, chi-square testa showed nosignificant
difference in errors,
Perception & Psychophysics, 1972, Vol. 12 (4) 359
-
Field
Right Left
843.0 827.1 866.6910.6 886.8
837.7 830.6 869.7912.4 898.1
1007.0 980.4 1020.91068.8 1037.4
928.2 909.9
9
II
10
12
Response
SameDifferent
SameDifferent
SameDifferent
GROUP 3
ResponseMode
Right Hand
Left Hand
Verbal
Experiment 1, a t test indicated nodifference in the magnitude
of theleft-field advantage obtained with rightvs left hand [t(9)
=.83, n.s.] or forverbal vs motor response [t(9) = 1.13,n.s, ).
These data, comparable to theresults obtained in the
verbalexperiment, support the hypothesisthat at least some part of
all spatialprocessing must take place in the rightcerebral
hemisphere, since the locus ofoutput did not affect the magnitude
ofthe field difference obtained. TheField by Response interaction
alsofailed to reach significance, indicatingthat the left-field
superiority ismaintained regardless of whether S ismaking a "same"
or a "different"response.
Several analyses regarding error ratefor different conditions
wereperformed. Percent errors summedacross Ss for each condition
are shownin Table 4. A t test for correlatedmeans was performed on
thedifference between the number oferrors made in the LVF vs R VF
foreach S (mean error rate of 5.3% in theLVF vs 5.5% in the RVF).
Thisdifference between fields did not reachsignificance [t(9) =
.47, n.s.},
In addition, as for Experiment 1,chi-square tests showed no
significantdifferences in errors summed acr088 Ssfor right and left
visual fields and rightand left hands [x' (3) -= 1.12, n.s.] orfor
right and left visual fields andverbal and motor response [x' (3)
=2.70, n.s.). As in Experiment I, then,evidence for hemispheric
asymmetry,as indicated by RT, was found in theabsence of
significant differences in
differ significantly [t(9) = .27, n.s.}. error rate.As in the
verbal task, "same" A secondary analysis was performedresponses
were faster than "different" in order to examine the notion
(LeVY,responses, 887.6 msec and 950.5 msec, 1969) that it is the
Gestalt nature ofrespectively [F(I,9) = 6.98; p < .05). the task
for which the right
No interactions were significant. hemisphere is specialized. If
this wereThe Field by Response Mode so, the number of blackened
cells ininteraction failed to reach significance, register should
not significantly affectindicating no differential field effect the
RT for the "same"-"different"depending upon presumed hemispheric
judgment, since simultaneous checkinglocus of response (e.g., if
verbal and of all cells would be made in aright-hand response
behaved similarly "template match." The average RT forand showed a
smaller left-field effect each type of "different" judgmentthan did
left-hand response). As with (zero, one, or two blackened cells
inthe right-field advantage found in register) and for the "same"
judgment
Table SOverall Mean Reaction Time (Milliseconds) for Each Field
by Response
Mode by Response Combmation: Spatial Task
7
5
6
8
GROUP 2
Fig. 1. Matrices used in spatial task.
2
3
4
GROUP I
hand, or left hand), and response("same" or "different")
werecomputed for each session for each S.As for Experiment 1, these
meansexcluded error trials. Mean error ratewas 5.3%; the range over
Ss was 2.9%to 10.1%. Extreme RTs were droppedaccording to the same
criterion used inExperiment 1. Means so obtained foreach session
were averaged acrosssessions, yielding 12 means for each S.The
pattern of means obtainedaveraged across Ss is shown in Table
3.
An analysis of variance of thesemean RTs (Field by Response
Modeby Response, within Ss)was performed. The analysis ofvariance
showed that RT for thisspatial task was faster for the LVFthan for
the RVF, 909.9 and928.2 msec, respectively[F(1,9) =16.83; P <
.01). All of the10 Sa showed faster RT to LVFstimuli, a "left-field
effect."
Verbal response was slower thanthat of either the right or the
lefthand, 1,020.9 msec vs 866.6 and869.7 m s e c ,
respectively[F(2,18) = 15.98; P < .001), whileRTs for right and
left hand did not
360 Perception & Psychophysics, 1972, Vol. 12 (4)
-
Table 4.Percent Enors for Each Field by Resporue
Mode COlDblnation: Spatial T"
Comparison ofVerbal and Spatial Tasks
In order to determine whether themagnitudes of the field
differencesobtained were the same for both theverbal and spatial
tasks, the average ofthe motor response (average of rightand left
hands) and verbal responsefield differences was obtained for eachS
in each experiment. A two-samplet test was performed on the
absolutemagnitude of field differences for thetwo tasks (28.3 msec
for the verbaltask, 19.8 rnsec for the spatial task).The absolute
magnitude of the fielddifference did not differ
significantlyaccording to the nature of the task[t(18) =1.02,
n.s.},
In general, then, the magnitude offield differences obtained for
each taskwas the same for verbal or manualresponse modes and "same"
or"different" responses. The field
was obtained for each visual field foreach session for each S.
These scoreswere then averaged across sessions.The relationship of
RT (averaged over10 Ss) to number of cells in registerfor each
visual field is shown in Fig. 2.
As was indicated by the..same..·..different.. dichotomy in
theanalysis of variance for thisexperiment, "same" RT was leaa
thanany of the "different" RTs. The slopeof the line for the
"different" RTs wasfitted by the least-squares method foreach S for
each visual field. Asignifican( linear component wasfound for
increasing RT withincreasing number of cells in registerfor both
the right visual field [slope =51.5; t(9) .. 6.20; p < .001] and
forthe left visual field [slope = 47.4; t(9)=4.6; p < .01]. The
slopes for right valeft visual fields, compared with acorrelated
means t test, were notsignificantly different [t(9) .. .95,n.a ].
These results are not in accordwith those of Sekuler and
Abrams(1968), since they did not obtain any"same"·"different"
dichotomy, butobtained instead a linear trend (withsmall slope) for
increasing RT withincreasing number of cells in register,regardless
of whether the appropriateresponse was "same" or "different."The
results of the present analyses donot suggest the employment of
asolely Gestalt processor.
"SAMEllCELLS
more consistent with the time requiredfor a few synaptic
transmisslons. Suchstudies measured the difference in RTfor simple
visual stimuli (which could,presumably, be proceaaed in
eitherhemisphere), which were presented tothe visual field
ipsilateral (uncrossedconnections) or contralateral
(crossedconnections) to the responding hand.RT differences on the
order of2·6 msec were found. The magnitudeof RT differences
obtained in thepresent experiment (an average of24 msec) is not
consistent with thesemeasures of simple callosaltransmiaaion
time.
Even if it were poaaible for theobtained latency differences to
beconsistent with all previous measuresof callosal transmission
time, however,it is neceaaary to account for the dataregarding
inferior recognition of verbalmaterial in the LVF (Kimura,
1966;Bryden, 1965) and of spatial materialin the RVF (Kimura, 1969;
Schell &Satz, 1970). A required callosaltransmission would
cause longerlatency of response, but, if thetransmiaaion were
complete in detail,should not cause inferior performancefor stimuli
initially projected to thenonspecialized hemisphere. If,however,
the corpus callosum had alimited channel capacity, this
inferiorperformance would be explained, sinceinformation loaa
during transmissioncould decrease accuracy. This effectwould be
most pronounced in thoselituations where stimulus informationwas
already minimal, as when stimuli
e----e RIGHT VISUAL FIELD.. - .. LEFT VISUAL FIELD
• 923.8• 905.1
2,
//
~{022.6/
//~Y: 47.4 X + 917.6
Y: 51.5 X +934.9///-964.2
// ·944.4/ 946.0
927.9
"DIFFERENT"NUMBER OF BLACKENED
IN REGISTER
o,
900
1050
UIVE1000UJ2F=z 950o
~IJJa:::
Fig. 2. Reaction time (spatial task) for each visual hemifield
as a function ofnumber of blackened cells in register.
differences for the two tasks were inopposite directions
(right-fieldsuperiority for the verbal task,left-field superiority
for the spatialtask), but were not significantlydifferent in
absolute magnitude. Thetwo field differences, then, canreasonably
be attributed to the sameunderlying phenomenon. A summaryof the RTs
obtained in the twoexperiments is shown in Table 5.
DISCUSSIONIt might be asked whether the
between-field RT differences obtainedin this study are
consistent withmeasures of callosal transmission time.The magnitude
of field differencesobtained in these experiments isconsistent with
someelectrophysiological studies (Bremer,1968; Grafsstein, 1959;
Teitelbaum,Sharpless, & Byck, 1968), which haveahown that
excitation originatingexclusively in one hemisphere
takesapproximately 10 msec (primarypositive wave) to 35 maec
(secondarynegative wave) to cross the callosumand its related
synapses to theopposite hemisphere. This consistencywould suggest
that differences in RTfor verbal va spatial tasks for the twovisual
hemifields are attributable tothe extra time required for a
callosaltransmission of stimulus informationfor processing. Other
behavioralstUdies (Poffenberger, 1912; Berlueehi,Heron, Hyman,
Rizzolatti, & Umilt8,1971), however, indicated a muchshorter
callosal transmission time,
4.85.15.6
Field
5.65.66.4
Right Left(Percent> (Percent>
ResponseMode
Right HandLeft HandVerbal
Perception & Psychophysics, 1972, Vol. 12 (4) 361
-
Table 5Overall Mean Reaction Time (Milliseconds): Comparison of
Verbal and Spatial Tasks
Resjronse .fodeField by
Task Field Motor* Verbal Task Means
Right 1027.6 1134.5 1081.0Verbal Left 1063.1 1155.5 1109.3
Verbal Means 1045.3 1145.0 1095.2
Right 875.9 1032.9 954.4Spatial Left 860.4 1008.9 934.6
Spatial Means 868.2 1020.9 944.5
*Mean of the right and left hand
are presented at very short durations.Such a callosal
transmissionexplanation, then, would explain bothdifferences in
error rate and latencydifferences between visual hemifields.
Such an inherent limitation inchannel capacity is indicated
bybehavioral (Myers, 1962) andelectrophysiological (Berlucchi
&Rizzolatti, 1968) studies in cats. Theseexperiments suggest
that the effect ofstimuli projected via the corpuscallosum is not
identical to that of thesame stimuli projected to a givenhemisphere
via geniculocorticalpathways.
Furthermore, Buchsbaum and Fedio(1970) report a greater
stability (tworesponses to the same stimulus have ~higher
correlation with each Ot}l":l."~ .,.evoked response activity in
humans tostimuli projected via the direct visualpathways to each
hemisphere than tostimuli taking secondary, indirectpathways. It
seems very likely,therefore, that the latency differencesobserved
in the present tasks areattributable to both callosaltransmission
time and to increasedlatency of response accompanyinguncertainty
attendant upon a"degenerate" stimulus transmission. 1
Geffen et al (1971) have suggestedthat processing of a given
stimulusmay be performed, although moreslowly, by the hemisphere
which is notspecialized for that stimulus type.They found a shorter
RT in the LVFfor perception of faces with a motorresponse, but not
with a verbalresponse. Geffen et al interpretedthese data as
indicative of aright-hemisphere superiority(quickness) for this
spatial task, whichwas "cancelled" because of thenecessity for
verbal response (initiatedsolely by the left hemisphere ).
Thepresent experiment yielded aright-hemisphere superiority for
thespatial task regardless of whetherverbal or manual response
wasrequired; the magnitude of this fielddifference did not differ
significantlyfor verbal vs motor response. The dataof the present
experiment are not inaccord with Geffen's conclusion thatthe
nonspatial hemisphere can performspatial processing, although
more
slowly; rather, they indicate that all orat least some part of
all spatialprocessing must take place in the righthemisphere. It is
possible that thematrix task in the present experimentwas a more
difficult spatial task thanthe "faces" task or was more
strictlydependent upon spatial cues, such thatprocessing had to
take place in theright hemisphere, while processing ofthe faces
might take place in the lefthemisphere as well.
The present data can be mostparsimoniously accounted for
bymodels that postulate that all or partof the processing of verbal
stimulimust take place in the left cerebralhemisphere, while all or
part of theprocessing of spatial stimuli must takeplace in the
right cerebral hemisphere.It might then be expected that,
inaccordance with the results ofBradshaw and Perriment (1970),
theright hand would be faster for a verbaltask (left hemisphere),
while the lefthand would be faster for a spatial task(right
hemisphere). (This "handeffect" hypothesis is distinct from aHand
by Field interaction, theimplications of which have beenpreviously
discussed.) It is interesting,however, to note that the
differencesin RTs for the two hands in thepresent experiment,
althoughnonsignificant, were exactly theopposite of that proposed
by a modelthat assumes motor output directedexclusively from the
hemispheredominant for a given task (in order toavoid an additional
callosaltransmission for output). The lefthand was faster for the
verbal task,while the right hand was faster for thespatial task.
The total number oferrors for all Ss for each hand in eachtask were
in the same direction as thereaction time differences,
althougherror differences are also notsignificant. Rizzolatti et al
(1971 )reported RT differences and error ratedifferences for the
two hands that arein the same direction as those in thepresent
experiment, although theseauthors did not discuss this
trend.Klatzky and Atkinson (1971) found ageneral left-hand
superiority in a RTtask, which was lessened if thenecessary
processing was spatial rather
than verbal. Although none of theresults described are
statisticallysignificant, consistent results such asthese might
suggest that motor output(possibly for both hands) is beingdirected
by the hemisphere that is notperforming the processing, but
is,instead, somehow monitoring for theresult (a trme-sharing
system, of sorts).
The results of this experiment alsoprovide some suggestions
regarding thenature of a "spatial" task. Rizzolattie t al (1971)
mentioned thatright-hemisphere superiority for therecogni tion of
faces (their spatial task)might be related to some peculiarity
ofphysiognomy as a visual pattern. Inthe present experiment,
aright-hemisphere superiority wasfound for processing of
complex,nonface spatial stimuli, suggesting thatit is the necessary
spatial coding(nonlinguistic nature of the task) thatis the
relevant factor.
Levy (1969) has suggested that theright hemisphere may be
specializedfor Gestalt processing. Bradshaw andWallace (1971),
however, foundevidence for a serial model for theprocessing of
faces, and facial stimulihave been shown to yield aright-hemisphere
advantage. Similarly,in the present spatial task, no evidencewas
found to support the hypothesisof the employment of a
Gestaltprocessor, even though, under otherexperimental conditions
(Sekuler &Abrams, 1968), a similar matrices taskwas found to be
Gestalt in nature. (Itshould be noted, however, thatSekuler and
Abrams got better"Gestalt" results with two-cell stimulithan with
four-cell stimuli. It may bethat the three-cell stimuli of
thepresent experiment are too difficult tobe processed in an
entirely Gestaltmanner.) In the present experiment,the highly
significant linear trend forincreasing RT with increasing numberof
cells in register for "different"stimuli is not suggestive of a
Gestaltprocess. In spite of mitigating factorsthat might serve to
explain whyprocessing was not Gestalt in nature inthe present
matrices task, it is stillvalid to say that the data
obtainedprovide no evidence to indicate thatright-hemisphere
superiority must bedependent upon the Gestalt nature ofthe task.
Rather, it seems sufficientthat processing for the task
utilizesspatial cues. It is, however, possiblethat further
experiments will indicatethat Gestalt tasks produce a
morepronounced laterality effect.
The results of the presentexperiment, then, definitely support
amodel of hemispheric specializationfor processing (as distinct
from justoutput), the left hemispherespecialized for verbal tasks
and theright hemisphere specialized for spatial
362 Perception & Psychophysics, 1972, Vol. 12 (4)
-
tasks. The results also indicate nosignificant effect of
lateralization ofright-hand vs left-hand motor output,but there are
some aspects of the datathat suggest, although they are
notstatistically significant, that the locusof motor output may
vary with taskand output demands.
REFERENCESBERLUCCHI, G., HERON, W., HYMAN,
R., RIZZOLATTI, G., &< UMILTA, C.Simple reaction times of
ipsilateral andcontralateral hand to lateralized visualstimuli.
Brain, 1971,94.419-430.
BERLUCCHI, G., &< RIZZOLATTI, G.Binocularly driven
neurons in visualcortex of split-chiasm cats. Science. 1968,159,
308-310.
BRADSHAW, J. L., &< PERRIMENT, A. D.Laterality effects
and choice reactiontime in a unimanual two-finger task.Perception
&< Psychophysics, 1970, '1,185-18 2
BRADSHAW, J. L. &< WALLACE, G.Models for the processing
andidentification of faces. Perception &<Psychophysics,
1971,9, 443-448.
BREMER, F. PhYsiology of the corpuscallosum. Research
Publications of theAssociation for Nervous &<
MentalDiseases, 1958, 36, 424-448.
BRYDEN, M. P. Tachistoscopic recognition.handedness, and
cerebral dominance.Neuropsychologia. 1966,3. 14J.
BUCHSBAUM, M., &< FEDIO, P.Hemispheric differences in
evokedpotentials to verbal and nonverbal stimuliin the left and
right visual fields.Physiology &< Behavior. 1970, 5,
207-210.
COLTHEART, M., &< MERIKLE, P. M. Arebriefly exposed
letter rows processedfrom left to right? Updated version ofpaper
presented at the meeting of theCanadian Psychological
Association,Winnipeg, March 1970.
CROVITZ, H. F., &< ZENER, K. A grouptest for assessing
hand and eyedominance. American Journal ofPsychology, 1962. 75,
271·276.
FILBEY, R. A., &< GAZZANIGA, M. S.Splitting the normal
brain with reactiontime. Psvchonomic Science, 1969, 17.335-336.
GAZZANIGA. M. S. BOGEN, J. E., &<SPERRY, R. W.
Observations on visualperception after disconnection of thecerebral
hemispheres in man. Brain, 1965.88. 221-236.
GAZZANIGA, M. S., &< SPERRY, R. W.Language after section
of cerebralcommissures. Brain, 1967, 99, 131-148.
GEFFEN, G., BRADSHAW, J. L., &<WALLACE, G.
Interhemispheric effectson reaction time to verbal and
nonverbalvisual stimuli. Journal of ExperimentalPsychology,
1971,87,415-422.
GIBSON, A. R., FILBEY, R., &<GAZZANIGA. M. S.
Hemisphericdifferences as reflected by reaction time.Federation
Proceedings, 1970, 29, 658.(Abstract)
GRAFSSTEIN, B. Organization of callosalconnections in
supra-sylvian byrus of cat.Journal of Neurophysiology, 1959,
22,504-515.
KIMURA, D. Cerebral dominance and theperception of verbal
stimuli. CanadianJournal of Psychology, 1961, 15,166-171.
KIMURA, D. Left-right differences in theperception of melodies.
Quarterly Journalof Experimental Psychology, 1964, 166.355-358.
KIMURA, D. Dual functional asymmetry ofthe brain in visual
perception.Neuropsychologia, 1966, 4, 275·285.
KIMURA, D. Spatial localization in left andright visual fields.
Canadian Journal ofPsychology, 1969, 23, 445-468.
KLATZKY, R. L. Interhemispheric transferof test stimulus
representations inmemory scanning. Psychorromfc Science,1970, 21,
201-203.
KLATZKY, R, L., &< ATKINSON, R. C.Specialization of the
cerebral hemispheresin scanning for information in
short-termmemory. Perception &< psychophysics,1971, 10,
335-338.
KRUEGER, L. E. Visual comparison in aredundant display.
Cognitive Psychology,1970, I, 341-357.
LEVY, J. Possible basis for the evolution oflateral
specialization of the human brain.Nature, 1969, 224, 614-615.
MILNER. B. Visual recognition and recallafter right
temporal·lobe excision in man.Neuropsychologia, 1968, 6,
191·209.
MOSCOVITCH, M., &< CATLIN, J.Interhemispheric
transmission ofinformation: Measurement in normalman. Psychonomic
Science, 1970, 18,211-212.
MYERS, R. E. Transmission of visualinformation within and
between thehemispheres: A behavioral study. In V. B.Mountcastle
(Ed.), Interhemisphericrelations and cerebral dominance.Baltimore:
Johns Hopkins Press. 1962.Pp.51-73.
POFFENBERGER, A. T. Reaction time toretinal stimulation with
special referenceto the time lost in conduction throughnerve
centers. Archives of Psychology.1912,23,1-73. ,
RIZZOLATTI, G., UMILTA, C., &<BERLUCCHI, G. Opposite
superioritiesof the right and left cerebral hemispheresin
discriminative reaction time tophYSiognomical and alphabetical
material.Brain, 1971, 94, 431-442.
SCHELL, B., &< SATZ, P. "Nonverbal"visual half-field
perception andhemispheric asymmetry. Proceedings,78th Annual
Convention of the AmericanPsychological Association. Miami,
1970.
SEKULER. R. W., &< ABRAMS, M. Visualsameness: A
choice-time analysis ofpattern-recognition processes. Journal
ofExperimental Psychology, 1968. 77,232-238.
SMITH, E. E., &< HAVILAND, S. E. WhYwords are perceived
more accurately thannon words: Inference vs unitization.Journal of
Experimental Psychology,1972, 92, 59-64.
SPRINGER, S. P. Ear asymmetry in adichotic detection task.
Perception &<Psychophysics, 1971a. 10, 239-241.
SPRINGER, S. P. Lateralization ofphonological processing in a
dichoticdetection task. Unpublished doctoraldissertation, Stanford
University, 1971b.
TEITELBAUM, H .• SHARPLESS, S. K., &<BYCK, R. Role of
somatosensory cortexin interhemispheric transfer of tactilehabits.
Journal of Compuative &<PhYsiological Psychology, 1968,
66.623'632.
WHITE, M. J. Laterality differences inperception: A review.
PsychologicalBulletin, 1969, 72, 387-405.
NOTE1. Support for the idea that the latency