Interhemispheric interaction affected by computational complexity

Abstract

The present study investigated whether dividing information between the hemispheres becomes more advantageous to task performance as computational complexity increases. We hypothesized that interhemispheric processing would benefit performance especially for computationally complex tasks, whereas it would hinder performance for relatively simple ones. A letter-matching task was given to 23 subjects at three levels of computational complexity. Complexity was varied either by increasing the number of inputs to be processed or by the nature of the decision to be made. The results indicated that each of these manipulations of complexity influenced performance by making it more advantageous to have both hemispheres involved in processing rather than just one. Furthermore, the effects of each manipulation were separable.

Full text

Page 1

0028 3932;92 $5.00+0.00
,c’m 1992 Pergamon Prrs Lfd
NOTE
INTERHEMISPHERIC INTERACTION AFFECTED BY COMPUTATIONAL
COMPLEXITY
AYSENIL BELGER and MARIE T. BANICH
Department of Psychology and the Beckman Institute, University of Illinois at Urbana-Champaign, IL, U.S.A.
(Receit,ed 12 July 1991; accepted 20 M~J 1992)
Abstract--The
becomes
hypothesized
tionally
matching
varied either by increasing
made. The results indicated
by making it more advantageous
one. Furthermore,
present study investigated
more advantageous
that interhemispheric
complex tasks, whereas
task was given to 23 subjects at three levels of computational
the number of inputs to be processed
that each of these manipulations
to have both hemispheres
the effects of each manipulation
whether dividing
as computational
would benefit performance
information between the hemispheres
complexity increases.
especially
for relatively simple ones. A letter-
complexity.
or by the nature of the decision to be
of complexity influenced
involved in processing
were separable.
to task performance
processing
it would hinder performance
We
for computa-
Complexity was
performance
rather than just
INTRODUCTION
DESPITE
yet known about how the hemispheres
are important because
hemispheres [8, 241. One aspect of interhemispheric
increases or decreases the processing
since some suggest that interhemispheric
291) while others suggest it helps performance
Previously we [l] have reported
studies. In our studies we compared
hemisphere (within-hemisphere
hemisphere trials). For a physical-identity
to either of two others, we obtained
which subjects decided whether a lower-case
was more difficult because a decision required more than a physical comparison
advantage was observed. Further evidence in support of our hypothesis
within- and across-hemisphere trials on a physical-identity
task. In the summation task, subjects decided whether,
equalled 10 or more, and in the ordinal task subjects decided whether a digit was smaller in value than either of two
others. Both the summation and the ordinal tasks are more difficult than the physical-identity
require more processing than a simple physical comparison.
previous one, in that the two more difficult tasks yielded an across-hemisphere
physical-identity task yielded a within-hemisphere advantage.
across the hemispheres hinders performance, whereas at higher levels of difficulty it improves
In the present paper we wish to further elaborate and refine our notion of task difficulty, and to further test our
hypothesis. In all our previous studies we have manipulated
of the task, although we have not referred to it in this manner
conceptualized as being dependent on the number of transformations
be reached. For example, the name-identity task is computationally
because additional computations beyond perceptual identification
THE great number of studies investigating the functional
information
differences between the hemispheres
to yield a unified response. Understanding
are integrated and analyzed
processing that has received attention
power of the brain. The results emerging from such investigations
interaction hinders performance
(e.g. Refs [9. 10, 12, 13, 28, 31, 33, 34, 391).
that task difficulty is one factor that may explain some of the variance
performance when all items critical
trials) to performance when the information
task, in which subjects decided whether an upper-case
a within-hemisphere advantage.
letter had the same name as either of two upper-case
1401, little is
such effects
by both
integrate
nearly all types of information simultaneously
lately is whether it
are equivocal
(e.g. Refs 12, 3, 5, 11, 15, 16,23, 26,27,
across
to one for a decision
was divided between them (across-
were directed
letter was identical
For another task, a name-identity task, in
letters (and hence
of the items), an across-hemisphere
was obtained by comparing
task to that obtained on a summation
when a digit was added to either of two others, the sum
performance
and an ordinal
on
task because they
The results of this study were similar to those of the
advantage,
Thus, at lower levels of difficulty, dividing processing
whereas the easier
performance.
task difficulty by varying the computational
previously. Computational
an input must undergo
more complex than the physical-identity
must be performed
complexity
can be complexity
before a decision can
task
to reach a decision. Hence,
923

Page 2

we wish to modify
computationally
In all the experiments
the number of steps required
increasing computational
complexity makes interhemispheric
issue, we designed
stages ofinformation
of inputs to be encoded and compared
wished to determine
effects from complexity
has analogous effects on performance
vs decision-making),
advantage.
our previous
complex,
we have performed
hypothesis to state that interhemispheric processing is beneficial for more
rather than more difficult, tasks.
so far, computational complexity has been manipulated
question
processing all have the same effect, in that the
to task performance. As a first step in addressing
the effects of increasing computational
weexamine whether increasing complexity
to the target also yields an across-hemisphere
whether the complexity, as manipulated by an increase in the number of inputs, has separable
as manipulated by the number of steps required for a decision. Ifcomputational
for both these stages ofinformation
then increasing the complexity for either stage should
by increasing
to reach a decision about the inputs. One important
complexity at different levels ofinformation
processing advantageous
the present study to determine
processing. In particular,
that arises is whether
this
complexity
by increasing
advantage.
at different
the number
Furthermore, we
complexity
processing (e.g. encoding and comparison
result in an across-hemisphere
METHODS
Subjects
Twenty-three
23, with normal
questionnaire
were considered
right hand.
students (12 males, 11 females) from an introductory
or corrected-to-normal visual acuity served as subjects.
that determined lateral preference
right-handed if they performed
psychology course, ranging in age from 18 to
All were right-handed
a nail or throwing
as assessed by a
a ball. Subjects for items such as hammering
seven of the eight items with their right hand and wrote with their
Stimuli
In this experiment,
make-up of the stimulus arrays was the same as that employed by BANICH and BELGEK [l]. Briefly, items were placed
in a triangular array with a digit at the centre, whose report was used to control central fixation 1301. Half of the
trials consisted of “match” trials in which the bottom item matched one ofthe top two items (which never matched),
and the other half were “mismatch” trials. Half of the match trials had matching
half had matching items in opposite visual fields, The visual held of the bottom item was counterbalanced
within- and across-field match trials, as well as for mismatch
For the three-letter physical-identity task, the two top letters were presented
above fixation. The bottom letter appeared 1.4” below the fixation
eccentricities should allow for initial receipt by each hemisphere
nasotemporal connections (e.g. Ref [21]), and that any mildline
stereopsis (e.g. Ref. 1191). Each letter subtended a maximum
A, B, M, P, T, R, F, G, H and Q were used as stimuli since they appeared
employed.
The stimulus arrays for the five-letter physical-identity
identity task, except that four letters were displayed above the fixation point. The two letters appearing
fixation point in each visual field were positioned one above the other, to help minimize the use of a post-perceptual
strategy that might be related to reading, and so that the center ofeach pair would match the location
letters in the three-letter task. Hence, the bottom letter ofeach pair appeared
fixation point, while the top letter ofeach pair appeared 1.9’ above and 2.X. lateral from the fixation point. For half
the match trials, the matching item was the lower of the top two letters. and for half it was the upper of the two top
letters, The stimulus arrays presented during the five-letter name-identity
identity task in that the bottom letter was in lower-case. Figure I presents examples of the four match trial types for
each task and some sample mismatch trials.
The three-letter physical-identity task provides a baseline to ensure that the pattern of performance
the subjects in the present study is consistent with our previous results, in which we obtained
advantage for this task [l]. The five-letter physical-identity
complexity by manipulating the number of items that need to be encoded and compared
reached, will yield an across-hemisphere advantage, as has been found with tasks for which computational
complexity is increased at the decision level. Furthermore,
whether the effects ofcomplexity at two levels have separable
from that found for the five-letter physical-identity task.
three different stimulus arrays were used for each of the three different tasks. The basic
items in the same visual field, and
for both
trials.
2.8” lateral from midline, and 1.4’
point and 1.4’ lateral from midline. These
since most studies report little or no overlap in
fusing functions mainly
of 0.85” horizontally and 1.17’ vertically.
least confusable
for the purposes of
The letters
screen on the computer
task were similar to those for the three-letter physical-
above the
of the single
0.87-’ above and 2% lateral from the
task differed from the five-letter physical-
exhibited by
a within-hemisphere
whether increasing
before a decision can be
task allows us to determine the
the five-letter name-identity
effects. If so, the results for this condition
task allows us to determine
should differ
Subjects performed
counterbalanced
each task in three different sessions, with the order of administration
across subjects. During each session a subject’s head was stabilized
of the different tasks
on a chin rest. such that the

Page 3

NOTE
925
MATCH TRIALS
3-ITEM PHYSICAL IDENTITY TASK
Within LVF
A
G
2
A
Within LVF
A
R
G
C
2
A
Within LVF
Within RVF
A
G
2
G
Across b -LVF
A
G
3
G
S-ITEM PHYSICAL IDENTITY TASK
Within RVF
A
R
G
C
2
C
Across b-LVF
A
R
G
C
3
C
LVF
S-ITEM NAME IDENTITY TASK
Within RVF
A
R
G
C
2
c
Across b-LVF
A
R
G
C
3
c
SAMPLE MISMATCH TRIALS
RVF
Across b-RVF
A
G
6
A
Across b-RVF
A
R
G
C
6
A
Across b-RVF
A
R
G
C
6
a
A
R
G
C
A G A
R
G
C
A G
2
2
3 6
t
E
2
H
Fig. 1. The four match trials and examples of the mismatch trials for each of the three tasks
subject’s eyes were located 32.9 cm from a graphics display screen. At the start ofeach trial a central dot, upon which
the subject was told to fixate, appeared on the screen. A single button mounted
midline of the screen was located in front of the subject. Reaction time (RT) was the dependent
three-letter and five-letter physical-identity tasks the subjects were required to press the button if the bottom letter
matched any of the top letters, and to refrain from pressing the button
subjects were instructed to press the button if the bottom item had the same name as any of the four top items, and to
refrain from pressing the button if it did not (go/no-go RT).
As a control over central fixation, the subjects were required to report the center digit, and if they could not do so,
the trial was disregarded. Pilot work indicated that the pattern of results was the same regardless
report of the central item was required (Banich, unpublished
clearly. stimulus duration was set at 200 msec. This value was chosen since mean saccadic eye movements
have been found to range between 180 to 200 msec 17, 181, and since perceptual
both during an eye movement and for some 4&50 msec before and after the movements
As an additional control over fixation, at the beginning
blindspot was located, and a peripheral dot was placed 0.5’ from the edge. Thus, if a subject’s eyes deviated more
than 0.5’ from central fixation, the peripheral dots were seen Cl], Any trial on which the subject missed the central
digit or reported seeing either of the peripheral dots was disregarded
Within each session, subjects were first given two blocks
administred in four 56-trial blocks. Within each block, half of the trials were match and half were non-match.
Subjects responded with the left hand on two blocks and with the right hand on two blocks. The order of response
hand was counterbalanced across subjects. Data from trials were included
subjects made a correct button press, (2) the central item was reported correctly,
seen.
on a grip, aligned with the vertical
measure. During the
if it did not. For the name-identity task,
of whether or not
data). In order to allow subjects to encode all the items
in humans
reduced sensistivity is substantially
begin [25,41].
of a test session an edge of each subjects’s left and right
from subsequent
of 28 practice
analysis.
trials, followed by 224 test trials
in subsequent
and (3) the peripheral
analyses only if (1) the
dots were not

Page 4

926
NOT E
Stimuli were presented on a medium resolution
was under the control of a Hewlett-Packard
Hewlett-Packard
310 computer,
colorgraphics terminal, while trial presentation
reaction time. which also recorded
RESULTS
A three-way
five-item physical-identity,
RVF). and Hand w’as performed
[F (6, 132)=9.57,
advantage for the three-item
significant within-hemisphere
compared to those previously
task yielded a significant
[F (I. 132)= 10.97, Pt0.00251
physical-identity
items to be processed,
repeated measures
tive-item name-identity),
on mean RT. The results revealed a highly significant
P-~O.0001]. Planned comparisons
physical-identity task [F(l.
advantage for this task [l]. However.
obtained, they were found not to dither significantly.
across-hemisphere advantage
from the marginally
task (see Table I). Thus, increasing
yielded a shift to an across-hemisphere
analysts of variance with the factors
Trialtype (within LVF. within RVF, across b-LVF, across b-
of Task (three-item physical-identity,
Task by Trialtype
significant
interaction
indicated
22)=2.83,
a marginally
P~0.1 I], Previously,
when the results of the present study were
The five-item physical-identity
[F (1. 22)=9.79. P-x0.005],
significant within-hemisphere
computational complexity, as instantiated
advantage.
within-hemisphere
we have obtained a
which differed significantly
advantage for the three-item
by the number of
Table I. Mean RT and Percent Errors for the three-item
and five-item name-Identity
physical-identity,
tasks
five-item
physical-identity
Task
Trial type
Within
LVF
Within
RVF
Across
b-LVF
Across
b-RVF
Three-item physical-identity 600
(5)
759
(33)
895
(39)
6.59
(8)
782
(34)
928
(42)
656
(9)
752
(20)
X29
(25)
630
(8)
726
(27)
X23
(27)
Five-item physical-identity
Five-item name-identity
Percent errors are presented in parentheses
A significant
77.22, P<O.OOOl].
item physical-identity
on the five-item name-identity
compared to the five-item physical-identity
mean RT. Even calculated
name-identity task (I = 3.72, d.f. = 22. P <0.005).
number of steps required
across-hemisphere
Furthermore,
task [F (I. 132)~ 15.42, P<O.OOOl]. To ensure that the larger across-hemisphere
task did not result merely because ofthe elongated
task. we computed
in this manner, the across-hemisphere
Thus, increasing
for a decision also contributes to the magnitude
advantage
this advantage
was also obtained
was found to be significantly
for the five-item name-Identity
larger than that obtained
task [F (I. 22)~
for the five-
advantage
mean reaction time on this task as
advantage as a percentage
was significantly larger for the fve-item
computational complexity
of the across-hemisphere
the across-hemisphere
advantage
of
by increasing
advantage.
the
A repeated-measures
rates. This analysis revealed a main effect ofTask
accurate at performing
results also revealed a highly significant
which was perfectly compatible
absence of any speed-accuracy
advantage was larger for the five-item name-identity
that just missed conventional
patterns obtained for the RT and error rate analysis is important
high and might make one cautious
analysis of variance with the factors of Task, Trialtype
[F (2, 44)=80.90,
the three-item physical identity task. which was also found to yield faster responses.
Task by Trialtype interaction
with the reaction time data (see Table I). Thus, the error rate analysis indicated
trade-offs. Of importance, the error data also indicated
task than for the five-item physical-identity
levels of statistical significance [F (I, 132) = 3.63, P = 0.0591. The concordance
since error rates for some conditions
about drawing inferences from the RT data.
and Hand was performed
P<O.OOOl], indicating
on error
that subjects were more
The
[F (6, 132) = 7.86, P<O.OOOl], the pattern of
the
that the across-hemisphere
task, to a degree
in the
were rather

Page 5

NOTE
921
DISCUSSION
Our previous studies [I] have demonstrated
performance especially
demonstrates that increasing
comparison stages, also causes interhemispheric
the present experiment
human brain hinders
between the hemispheres
Our results are conststent
complementary to our own. In their paradigm.
hemisphere,
or divided evenly between the two. In the trials ofinterest,
orientation ancl half in an inverted one. Complexity
letters were presented and in other cases only two were. When four letters were presented,
items to one hemisphere and the two upright to the opposite hemisphere
unilateral trials on which the two upright and the two inverted letters were all presented
However, when the input load was decreased,
hemisphere and an inverted letter to the other, did not increase performance
same hemisphere. Hence, the advantage of dividing
higher-load condition.
In addition, the results of the present study indicate that manipulations
information processing (e.g. encoding vs decision-making)
by speed or accuracy of response. Indicating
hemisphere advantage was obtained on the five-letter name-identity
task. Although it might be tempting to consider the larger across-field
as reflecting a summation of the effects of the difficulty of encoding
data from our previous studies [I] suggest this is not likely to be the case. The size of the across-hemisphere
advantage obtained in the five-letter name-identity task in the present study, is approximately
obtained for the three-letter name-identity task in our previous study [I, experiment
Thus, although for a physical-identity task the additional
advantage, the additional inputs do not shift the pattern
complexity is added by the extra inputs is significant when the complexity
identity task), but inconsequential when the complexity
An alternative possibility is that the size of the across-hemisphere
which it does not exceed. Consistent with such an idea are some of our previous findings. We [I] had found that the
size of the across-hemisphere advantage was equivalent for both the summation
the ordinal task could be considered more difficult because mean RT for this task was significantly
the summation task. Thus. the added difficulty did not increase
possibly because it had reached a ceiling value. In our previous paper we speculated
that if a task surpassed a particular threshold of difficulty,
Furthermore, we suggested that the size of the across-hemisphere
the task, implying that the relationship between task complexity
was a step function. Clearly. the data obtained from the present study do not support
because the size of the across-hemisphere advantage was greater for the five-letter name-identity
letter physical-identity task. However, the finding that the across-field
summation and ordinal tasks (which seemed to differ in difficulty based on mean RT), and for the three- and five-
letter name-identity tasks (which differed in difficulty based on the number
possibility of a ceiling on the size of the across-hemisphere
An issue raised by the present study is why interhemispheric
complex tasks. The absence ofany cues prior to stimulus presentation
within-hemisphere trial is about to appear. eliminates the possibilty ofsubjects
of these conditions. Rather, the across-hemisphere advantage
efficiency of information processing when the critical information
compared to both. A model that might explain the results would be the horse-race
attempts to reach a decision as quickly as possible. Across-hemisphere
advantage because it would allow for a division of processing
or aspects of the task to be processed in parallel. For such parallel processing
must have two important characteristics. First, the load must be such that it can be distributed
hemispheres. One reason why interhemispheric processing
examined so far, is that each hemisphere is indeed capable
therefore, under higher levels of complexity the hemispheres
portion ofa task could be performed by only one hemisphere,
would not yield any gain in performance. Thus, we hypothesize
that dividing information
computations
by another means. namely adding
processing
clearly demonstrate that for relatively
performance, whereas for computationally
aids task performance.
with those of MEKOLA and
subjects had to name letters. all of which were directed either to one
between the hemispheres
in the decision
computations
to be beneficial to task performance.
simple tasks, interhemispheric
more complex
is beneficial to
The present study
to the encoding
Thus, the results of
processing
ones, dividing the information
when additional
complexity
are involved process.
and
in the
LIEDERMAN [32] who report findings that are
halfof the letters were presented in an upright
at the input level, because in some cases four
directing the two inverted
led to a performance
in the same visual-field.
condition, directing an upright
relative to presenting
between the hemispheres
was manipulated
advantage relative to
as in the two-item letter to one
them both to the
appeared only in the information
of task complexity
effects on performance,
were different, a larger across-
task than on the five-letter physical-identity
advantage in the five-letter name-identity
and difftculty of the decision. comparison
at different stages of
whether indexed have separable
that the effects of the two manipulations
task
with
equivalent to that
I], namely 10% of mean RT.
inputs cause a shift from a within- to an across-hemisphere
observed for a name-identity
of the decision is low (e.g. in the physical-
of the decision is high (i.e. for the name-identitv
advantage may have an inherent ceiling, past
task. It is as if whatevet
task).
and the ordinal task, even though
longer than for
advantage. the size of the across-hemisphere
that these results might imply
advantage would be observed.
might be independent
and the size of the across-hemisphere
the idea of a step function,
an across-hemisphere
advantage ofthe dilliculty of
advantage
than for the hve-
in size in the advantage was equivalent
of items to be encoded) raises the
advantage.
interaction would be advantageous
indicating whether an across-hemisphere
adopting different strategies for each
can only be ascribed to an inherent
is presented initially to only one hemisphere
model. in which each hemisphere
processing would lead to a performance
across the hemispheres, enabling different components
to occur, interhemispheric
during relatively
or a
difference in
as
processing
between
tasks we have
of the task. and,
the
may be beneficial
of performing
can share the load. We suspect that if a significant
distributing the information
that a task such as one involving a rhyme decision,
in all the complex
most components
between the hemispheres

Page 6

928
NUTF
which in right-handers
advantage.
The second characteristic
distinct, so that there is an isolation,
cross-talk or interference
evidence suggests that this faculty ofcallosal
time course of its myelinization
resources of the hemispheres
In sum, our results provide cvidcncc that interhemisphcric
The utility of any hypothesis
takes a step in this direction,
dificulty manipulated at the level of the decision process. We have demonstrated
instantiated in two very different manners,
performance. Exactly how difficulty at different levels of processing
indicate that under certain conditions they can have separable
comprehend fully the implications of interhemispheric
consideration the factors that modulate
is the sole province
investigating
required
of the left hemisphere
this issue.
is that the processing
shielding or independence
between the hemispheres,
insulation
120, 38. 421. Furthermore,
can be distinct (e.g. Refs [I71 and 1221).
1361, would not yield an across-hemisphere
We are currently
performed
of processing.
by each hemisphere
This characteristic
a function of the corpus callosum.
must be functionally
would minimize
and can be considered
increases with age 1311 and may be related to the maturation
results from dual-task
Recent
studies suggest the processing
interaction aids performance
for a wider array of findings. The present study
is not limited to the specific case of
that computational
when tasks are complex.
becomes greater when it can account
because it provides evidence that the hypothesis
complexity,
on task has similar effects on the influence of interhemispheric
interact is not clear, although
effects. Overall, the present results suggest that to
interaction for cognitive
these effects, task complexity being one of them.
processing
the present results
peformance one must take into
REFERENCES
I. BANICH, M. T. and BELGEK , A Interhemispheric
task’? C’orteu 26, 77-94. 1990.
2. BEAUMONT, J. G. and DIMONL). S. J. Transfer between the cerebral hemispheres
37,87~91. 1973.
3. BEAUMONT, J. G. and DIMOYD, S. J. Interhemispheric
handed subjects.
Aclu psycho/. 39, 97 104, 1975.
4. Boots, D. B. and EVELANU, D. C. Visual and phonetic codes and the process of generation
exp. Psychoi: Hum. Percept. Prtform. 9, 657-675,
5. BKADSHAW, J. L., NETTLETOX, N. C. and PATTEKSON, K. Identdication
and opposite visual fields. J. e.~p. Psycho/. 99, 4248,
6. CAKPENT~R, R. H. S. Mooements offhe Eyes. Pion, London,
7. COHEN, G. T/jr Psychology of Cognition. Academic
8. DAMASIO, A. R., DAMASIO, H. and VAIZHOESEN. G. W. Prosopagnosia:
mechanisms. !v’eurology, 32, 33 I 341, 1982.
9. DAVIS. R. and SCHMIT. V. Timing the transfer of information
335-346, 1971.
IO. DAVIS, R. and SCHMIT, V. Visual and verbal coding in the interhemispheric
347-356, 1973.
I I. DIMONI), S. J. Hemisphere function and immediate
12. DIMOND, S. J. Hemisphere function and word registration.
13. DIMONI), S. J. and BEAUMONT, G. Hemisphere
14. DIMONU, S. J. and BFAUMONT, G. Processing
hemispheres. Ur. J. Psycho/. 63, 509 514, 1972.
15. DIMOND, S. J. and BEAUMONT, G. Hemisphere
215-m278, 1974.
16. DIMOND, S. J.. GIBSON, A. R. and GAZZANI(;A, M. S. Cross-field
information.
IYeurop,s~~ho/o(/i(I
17. FRIEDMAN, A., POLSON, M. C. and DAMON, C. G. Dividing
performance trade-offs between rapid finger-tapping
Pe~fbrm. 14 (l), 60 68, 1988.
18. FUDIN, R. and MASTERSON, C. C. Note on duration
laterally shown targets.
Perwpt. mof. Skills 42, 223 -226. 1976.
19. HAKVEY, L. O., JK. Single representation ofthe visual midline in humans. Nrurops~~hologicr, 16,601~610,
20. HATTA, T. and MOKIYA. K. Developmental changes
information. Int. J. B&n>. Drr. 11 (4). 451456.
21. HAUY, F. Functional dissociation of the hemispheres
1978.
22. HELLIC;E, J. B. and Cox, P. J. Effects of concurrent
right visual fields, J. cup. Psychd Hum. Percept. Pwfiwm. 2, 210 221, 1976.
23. KLEINMAN, K. M. and LITTLE, R. W. Inter-hemisphere
subjects.
Nature 241, 55 57, 1973.
24. LASSEK:, N. A., IKGVAK, D. H. and SHINKOJ. F. Brain function and blood flow. S&nt.
interaction: How do the hemispheres divide and conquer a
in human learning. Acts p.sychol.
transfer of figural information in right and non-right
in letter matching. J.
1983.
ofmirror-reversed facial profiles in same
1973.
1977.
Press, London, 1977.
Anatomical basis and behavioral
between the hemisphere in man. Actu psyhol.
35,
transfer of information. Bruin 95,
memory. P.s)chon. So&l.
J. rup. Psychol. 87, l83~-186, 1971.
and color naming. J. cup. Psycho/. 96, 87 91. 1972.
in perceptual integration between
16, I1 1 112, 1969.
function
and within the cerebral
function and paired-associate learning.
Br. J. P.sq’c’hol. 65.
and within-lield integration of visual
IO, 379 3X I, 1972.
attention between
J. r~.up. P.yhol:
the hands and the head:
Hum. Prrwpf.
and verbal memory.
of stimulus exposure in experiments on covert scanning of
1978.
of hemispheric collaboration for tactile sequential
1988.
using fovea1 visual input. Neurop.~?c,holoyitr 16,725 733.
verbal memory on recognition of stimulus from the left and
transfer of meaningful Information in normal human
Am. 239, 50 59, 1978.

Page 7

NOTE
929
25. LATOR, P. L. Visual threshold
26. LEIBER, L. Interhemispheric
1982.
27. LIEL)ERMAN, J. Interhemispheric
28. LIEDERMAN, J., MEROLA, J. L. and HOFFMAN, C. Longitudinal
increases during early adolescence.
29. LIEDERMAN, J., MEROLA, J. L. and MARTINEZ, S. Interhemispheric
bilateral input. Neuropsycholoyiu
30. MCKEEVEK, W. F. and HULING, M. D. Lateral dominance
obtained with simultaneous bilateral
31. MEROLA, .J. L. and LIELXRMAN, J. Development
11841194, 1985.
32. MEROLA, J. L. and LIEDERMAN, J. The effects of task difficulty upon the extent to which performance
from between-hemisphere division of inputs. In?. J. Neurosci.
33. MILLER, L. K. Perceptual independence of the hemifields
298-312, 1981.
34. MILLER. L. K. Hemifield independence in the left-handed.
35. NORMAN, D. A. and BOBROW, D. G. On data-limited
1975.
36. RAYMAN, J. and ZAIUEL, E. Rhyming and the right hemisphere.
37. RAYNER, K. Fovea1 and parafoveal cues in reading.
pp. 149-162. Lawrence Erlbaum Associates, Hillsdale,
38. SALAMY, A. Commissural transmission: Maturational
39. SERENO, A. B. and KOSSLYN, S. M. Discrimination
of attention.
Neuropsycholqiu
29, 659-675, 1991.
40. SPERRY, R. W. Lateral specialization in the surgically
Program,
I‘. SCHMITT and F. WORDEN (Editors).
41. VOLKMANN, F. C., SCHICK, A. M. L. and RIGGS, L. A. Time course of visual inhibition
saccades. J. optic. Sm. Am. 58, 362Z369, 1968.
42. YAKOLEV, P. I. and LECOURS, A. The myelogenetic
Derelopment of’tke Brain in Early L+,
A. MINKOWSKI (Editor),
during eye movements.
effects in short term recognition
Vision Res. 2, 261 262. 1962.
memory for single words. Covrrv 18, 113 124.
interference during word naming.
ht. J. Neurosci. 30, 43 56. 1986.
data indicate that hemispherical
2, 183-201, 1986.
collaboration
independence
Dec. Neuropsychol.
in response to simultaneous
23, 673-683, 1985.
in tachistoscopic
9, 15-20, 1971.
in hemispheric
word recognition performances
input. Neuropsycholoyia
changes independence. Child Du. 56,
benefits
51, 3544,
in children
1990.
and adults. J. exp. Child Psychol.
32,
Bruin Lang. 20, 3343.
and resource-limited
1983.
Coqn. Ps~‘chol. 7,44&64, processes.
Brain Lany. 40, 89 105, 1991.
and Pe@wmm
NJ, 1978.
changes in humans.
within and between hemifields: A new constraint
In Attention VII, J. REQUIN (Editor),
Science
200, 1409 1410, 1978.
on theories
separated
MIT Press, Cambridge,
hemispheres. In The Neurosciencrs:
MA, 1974.
Third Stud)
during voluntary
cycles of regional maturation of the brain. In Reqiond
London, pp. 3- 65. Blackwell. 1967.