BRAIN
AND
LANGUAGE
3, 47-71
(1976)
On the Representation
of Language in the Right
Hemisphere of Right-Handed
People’
MORRIS MOSCOVITCH~
Department
of Psychology,
Erindale
College,
University
of Toronro
Laterality experiments using reaction time were conducted to assess the performance of the right hemisphere of normal people on verbal tasks. The results
show that if the task calls for pictorial encoding of visually presented verbal material, then the right hemisphere’s performance is superior to that of the left. When
the task calls for linguistic analysis, the minor hemisphere displays no aptitude in
dealing with the task. The latter finding is at variance with data from split-brain
research. To reconcile these differences, it was proposed that language functions,
though represented in the right hemisphere of normal people, are functionally
localized in the left. When the control which the left hemisphere exerts over the
right is weakened or removed, e.g., by commissurotomy,
right hemisphere language is released. The application of this model to other neurological phenomena
is briefly discussed.
INTRODUCTION
The demonstration that the right, minor hemisphere exhibited a wide
range of linguistic skills was one of the more revealing aspects of studies
conducted on a group of patients with surgical disconnection of the cerebral hemispheres. Sperry, Gazzaniga, Bogen and their collaborators (for
reviews see Bogen, 1969a,b; Gazzaniga, 1970; Sperry, 1968; Sperry,
Gazzaniga & Bogen, 1969) found that the minor hemisphere can comprehend some spoken and written language. In contrast to the dominant,
left hemisphere, the minor hemisphere is almost mute and its comprehension is limited to spoken and written nouns, some phrases, and very
simple sentences. Even so, the minor hemisphere has reasonably sophisticated verbal concepts for the material it comprehends. For example, it
understands that a “match” is “used to light fires,” that a “ruler” is a
“measuring instrument,”
and that a “glass” is a “liquid container.”
’ A paper based on these experiments was presented at the Eastern Psychological Association in Washington, 1973. Experiment 3 was part of a doctoral dissertation in psychology at the University of Pennsylvania which was supported, in part, by NSF Grant GB
8013 to Paul Rozin. The rest of the research was supported by an N KC of Canada grant to
the author. I thank Dr. Paul Rozin, my advisor, whose encouragement and suggestions at
all stages of this work were invaluable. I also thank Jerre Levy and Danny Klein for
their many helpful comments and criticisms of earlier versions of this paper.
Z Requests for reprints should be sent to Morris Moscovitch,
Psychology Department,
Erindale College, University of Toronto, Clarkson, Ontario, Canada MSP 1P8.
47
Copwght
0 IY76 by Academic
Press, Inc.
All rights of reproduction
in any form reserved.
48
MORRIS MOSCOVITCH
These findings led Sperry and his associates to propose that language is
represented in both hemispheres, though more strongly in the left. This
does not mean that the, correspondence between the right hemisphere’s
linguistic performance and its linguistic capacity will be as close in other
people as it is in split-brain patients. For example, when the interhemispheric pathways are intact, they may allow the left hemisphere to
influence the right and so obscure the right hemisphere’s linguistic skill.
This point was emphasized by Sperry et al. (1969) and by P. Milner
(1970) in explaining the surprisingly poor linguistic performance of some
aphasic people with normal, healthy right hemispheres, and by Moscovitch (1972, 1973) in discussing the linguistic performance of the right
hemisphere of normal people. According to this view, it is possible that
language may be represented more or less equally in the right hemisphere of all people, but may appear to be functionally
more lateralized
in some than in others.
To test this hypothesis, a series of experiments was begun a few
years ago to assess the linguistic performance of the right hemisphere of
normal people (Moscovitch, 1972, 1973; Moscovitch & Catlin, 1970).
The results of these experiments suggested that the linguistic performance of the right hemisphere of normal people did not seem to match its
true linguistic capacity as established by the split-brain research. The
present experiments extend the earlier observations and confirm the
impression that whatever the linguistic competence of the right hemisphere might be, language in normal people is functionally localized in
the left hemisphere.3
The general design and logic were similar to the previous experiments.
Stimuli were presented to the right or left visual field and the subject responded with either his left or right hand. Differences in reaction time
(RT) between visual fields indicate which hemisphere is superior at executing the task at hand. Thus, reaction time to stimuli that require
linguistic analysis favor the right visual field-left hemisphere while those
that require only pictorial-spatial analysis favor the left visual field-right
hemisphere (Cohen, 1972; Geffen, Bradshaw & Wallace, 1971; Geffen,
Bradshaw & Nettleton, 1973; Gibson, Filby & Gazzaniga, 1970;
Klatzky, 1970; Klatzky & Atkinson, 1972; Moscovitch, 1972, 1973;
3 It should be. emphasized at the outset that it is not the purpose of the paper to determine whether the right hemisphere’s capacity to comprehend and to convey verbal information is “linguistic” in the sense that linguists, of whatever stripe, use this term. The purpose, rather, is to determine why some of those capacities are reflected in performance in
some situations but not in others. These capacities, limited as they are, are termed
“linguistic” because they seem necessary for dealing with linguistic utterances such as
words, phrases and sentences. Questions regarding the properties of right hemisphere language and their congruence with the properties of normal language are left for later
research (but see General Discussion).
LANGUAGE
AND THE RIGHT HEMISPHERE
49
Rizzolatti, Umilta & Berlucchi, 1971). In addition, by manipulating
stimulus-response connections it is possible to determine the relative
contribution of a number of factors to the reaction-time difference.
Those factors include hemispheric efficiency in executing the task at
hand, interhemispheric transmission time, and attention. For example, if
reaction times favor the same visual field by a constant amount regardless of responding hand, it indicates that some or all of the task must be
mediated by the hemisphere contralateral to that visual field. If, however, reaction times favor the visual field on the same side as the responding hand or if the size of the reaction time difference changes with
responding hand, it suggests either that both hemispheres perform that
particular task equally well or that one does so more efficiently than the
other (see Moscovitch, 1973, for a more detailed exposition). This kind
of analysis should enable us to evaluate the performance of each hemisphere on some linguistic tasks.
EXPERIMENT
1
In an earlier experiment, Moscovitch (1972) found that when subjects
matched a visually presented probe letter to a letter presented binaurally
two seconds earlier, RTs favored the left visual field for a left-hand
response and neither field for a right-hand response. There are two ways
to interpret this result. Either the minor hemisphere compared the two
letters along a linguistic dimension, i.e., by matching their names, or the
subject converted an ostensibly linguistic task to one that can be solved
using only pictorial skills. The subject could have accomplished the
latter by evoking a visual representation of the acoustically presented
letter and matching the test and set letters according to shape.
Experiment 1 tested whether the hemispheres compared the letters
linguistically or pictorially. On half the trials the subject saw a probe or
test letter that is different from the acoustic set letter, but resembles it
either pictorially, e.g., E-F, or acoustically, e.g., E-D. A match based on
shape or “physical” (Posner, 1969) features should make it more difficult for the subject to discriminate between pictorially similar pairs
than between acoustically similar pairs and should be reflected in longer
RTs to the pictorially similar pairs. If the subject uses a linguistic strategy and compares the names of the two letters, then acoustically similar
pairs will yield longer RTs.
METHOD
Subjects.Sixteenright-handed students at Erindale College participated in the experimerit. None was aware of the exact purpose of the experiment.HaIf the subjects
responded with the right hand and the other half with the left. Within each of these two
groups, the subjects were matched so that test letters that appeared in one visual field for
one subject appeared in the opposite field for another subject.
50
MORRIS MOSCOVITCH
Apparatus.
Ready signals and memory sets were recorded on tape (33 l/3 speed) and
delivered to the subject via earphones. The test letters were presented by two rear-screen
stimulus projectors (Industrial Electronic Engineers, In-line Readouts, Series 80). The
projectors were at the rear of a black panel. The faces of the projectors were flush with the
panel and located 3.2 cm on either side of a fixation point in the center of the panel.
Presentation of the test letters started a timer (Monsanto) that was accurate to 10m2msec.
Closure of microswitches that were out of view but situated on a table directly before the
subject stopped the timer. The timer, the tape recorder, the control panel, and the experimenter were all in a room separate from the subject.
Procedure.
The subject wore earphones and sat facing the panel which was about 75 cm
away. He rested his hand on a table and placed his index finger between two microswitches. All auditory input was presented binaurally and at a comfortable sound level.
When he heard the word “ready,” the subject fixated on a small white hole in the middle of
the panel. Two seconds later he heard a single letter (set) followed 2 set later by a white
letter (test) that appeared for 100 msec, approximately 6’ peripherally in either the right or
left visual field. Half the subjects pulled on a microswitch with their index finger if the test
and set letters were the same (positive set, “same” judgement) and pushed on a microswitch if they were different (negative set, “different” judgement). The mode of response
was reversed for the other half. To promote speed and accuracy the subject received
$32.00 at the start of the experiment from which $0.01 was deducted for every 100 msec it
took him to respond and $0.25 for every mistake he made. Subjects whose error rate consistently exceeded 5% were dropped from the experiment, All RTs greater than 1000
msec were discarded. These occurred on less than 1% of the trials.
The events between the presentation of a ready signal and the subject’s response marked
a trial. The first day was a practice day during which the subject ran through two sessions
of 40 trials each. On the following three days, the subject received a total of 480 trials,
which were divided into IO sessions of 48 trials each. Each session was in turn divided into
6 blocks of 8 trials. The subject was unaware of this division. A block contained two negative and two positive trials in both the right and left visual fields. The trials were randomized within blocks, and the blocks within sessions, so that no two sessions were alike.
Over an entire session the negative trials within each visual field were randomized so that
on half the trials the test-set pairs were pictorially similar and the other half were acoustically similar. A trial occurred every 10 set making each session last 8 min. Unless the
subject complained of fatigue, he received 4 sessions a day and 2 sessions on the last day.
The subject was encouraged to take 5- 10 min breaks between sessions. During part of the
break the experimenter told the subject approximately how much money he was earning.
The instructions given the subject followed the steps of the procedure just described.
The purpose of the experiment or any other information which might bias the results were
withheld from the subject until the experiment was completed.
Before being admitted to the experiment, each subject was pretested for acuity on a
Snellen chart. Pretests for handedness required him to sign his name on a “subject sheet”
and to reach quickly for an object. The tests were conducted without the subject’s knowing
their true purpose. Unless a subject was judged to be right-handed on the pretests and had
about equal satisfactory eyesight in each eye, he was not accepted into the experiment.
Following the experiments the subject was questioned in detail to determine whether
he was indeed right-handed. If he showed signs of ambidexterity his data were discarded.
Stimuli.
The acoustic set letters were C, E, X, and Y. For “same” judgements (positive
set) each acoustic letter was followed by its visual counterpart. The negative set consisted
of 4 pairs of visual test letters, each member of the pair being matched for either pictorial
or acoustic similarity with one of the letters from the acoustic set. Thus, for “different”
judgements, C was always followed by either 0 or T: E, by F or D: X, by Z or S: and Y,
by V or 1. The test letters were all in capitals. Each letter was 5.4 cm high and, depending
LANGUAGE
AND
THE
RIGHT
51
HEMISPHERE
on the letter, I .h to 6.0 cm wide. The letters appeared white on a black background and at
an average luminance of I9 foot lamberts.
RESULTS
The results are summarized in Table 1 and aspects of them are highlighted in Fig. 1. Separate analyses of variance were conducted for
“same” and “different” judgements. For “same” judgements, a two-way
analysis of variance with one between-factor (response hand) and one
by response
within-factor (visual field) showed a significant visual3eld
hand interaction [F (1,14) = 4.80, p < .05], RTs significantly favoring
the left visual field (LVF) for a left-hand response by 23 msec
(r(7)= 3.25, p < .Ol), but favoring the right visual field (RVF) by an
insignificant 5 msec (r(7) = .21, p > . I) for a right-hand response. No
other significant effects were found.
For “different” judgements. a three-way analysis of variance with one
between (response hand) and two within (visual field and similarity)
factors showed only a significant main effect of similarity, [F
(1,42) = 67.69; p < .OOl] pictorially similar pairs yielding larger RTs
than acoustically similar pairs, the difference being 28 msec. Overall
RTs for “different” judgements favor the RVF by 6 msec, but the effect
is not significant.
The number of errors was small (see Fig. 1). The difference in the
number of errors between visual fields for “same” judgements was
small. For “different” judgements, the error data parallel the RT data,
the subjects making many more incorrect responses when the set-test
pairs were pictorially similar, 64, than when they were acoustically similar, 23, regardless of response hand or visual field.
DISCUSSION
Both RT and error data attest to the greater difficulty the subject
experienced in discriminating between pictorially similar as compared to
TABLE
CONSECUTIVE
I
LETTER-MATCHING
TASK”
Different
Same
LVF
Left hand
Right hand
500(26)
459 (20)
RVF
523 (37)
454 (22)
RVF
LVF
A
P
A
P
529(10)
525 (2)
560(17)
549(13)
527 (5)
510 (6)
550(13)
542(21)
n Mean RT (msec) of correct responses averaged over subjects and broken down by response hand, visual field (LVF vs RVF), type ofjudgement (same vs different) and similarity
of set-test pairs (acoustic (A) vs pictorial (P)). The number in parentheses next to the RT
designates the number of errors in that condition.
52
MORRIS MOSCOVITCH
TOTAL NO.
of
ERRORS
AVERAGE
RT8
in m*ec.
LEFT
VISUAL FIELD
Rl6ffT
VfSUAL FIELD
FIG. I. Mean RT and total errors for “different” judgements in the consecutive lefrer
matching rusk averaged over subjects and response hand, and broken down by visual field
and similarity (pictorial (visual) vs acoustic). Dotted column, visual similarity; diagonally
lined column, acoustic similarity.
acoustically similar pairs regardless of whether they responded with
their right or left hand or whether the test item appeared in the right or
left visual field. These results suggest that the subjects used pictorial
strategies to compare ostensibly linguistic material, and would explain
why RTs favor the LVF for a left-hand response in the “same” condition. These results are consistent with a great deal of neurological and
behavioral evidence showing the superiority of the right hemisphere on
pictorio-spatial tasks (Milner, 1971). The results also confirm earlier
reports (Cohen 1972; Geffen et al., 1973; Gibson, Dimond & Gazzaniga, 1972; and Klatzky & Atkinson, 1972) that pictorial matches
between letters or words favor the right hemisphere. It should be noted,
however, that the left hemisphere can also operate in a pictorial modality. Results showing that right-hand responses slightly favor the RVF
and that pictorial similarity causes the same difficulty when the visual
test letter appears in the right visual field support this idea. Whether the
left hemisphere’s style of operation in the pictorial modality is similar to
that of the right hemisphere cannot be determined from these data.
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
53
It is difficult to interpret why RTs for “different” judgements do not
conform to the pattern of RTs for “same” judgements. RTs for “different” judgements slightly favor the right visual field, regardless of
response hand. One way to explain the anomaly is to assume, as others
have (see Egeth & Blecker, 197 1), that in comparison to the “sameness”
detector, the “difference” detector is relentless in its examination of the
features of the test stimulus. Because of this, the subject will occasionally consult both hemispheres as a precaution before responding
“different,” but rely on the judgement of either hemisphere alone when
responding “same.” Consistent with this interpretation are the longer
RTs for “different” judgements. It is interesting to note that so far investigators have found discrepancies between “same” and “different”
judgements with regard to laterality only for stimuli such as letters that
can be encoded easily either pictorially or linguistically (Cohen, 1972;
Egeth & Epstein, 1972; Moscovitch, 1972) and may, therefore, engage
both right and left hemisphere processes.
EXPERIMENT 2
The right hemisphere’s performance in the first experiment depended
on its having an internal visual representation of the acoustic letter. It
could have acquired this image in one of two ways. The first is that the
minor hemisphere, as well as the dominant, analysed the acoustic stimulus into its constituent phonemes, determined the name of the letter, and
used this linguistic information to generate the corresponding visual
image. The second possibility is that linguistic analysis of the acoustic
stimulus occurs only in the dominant hemisphere, the results of which
are transferred to the minor side before the test letter appears. The second experiment was designed to eliminate one of these possibilities,
thereby adding information about the minor hemisphere’s linguistic performance.
In Experiment 2, the acoustic set and visual test letters were presented simultaneously. It was hypothesized that if the minor hemisphere
treats the acoustic and visual letters as linguistic stimuli, then it should
find it more difficult to distinguish between letters that share phonemic
features than it did in the previous experiment because now the acoustic
trace will have no time to fade and weaken before the test letter appears.
In other words, the difference in RT between pictorially and acoustically
similar pairs obtained when the set-test pairs were presented successively (Expt 1) should become smaller now that they are presented
simultaneously (Expt 2). If, however, the minor hemisphere treats the
set and test letters as nonlinguistic, acoustic and pictorial stimuli, respectively, it should perform exactly as it did in the successive presentation
condition. The dominant hemisphere, being sensitive to linguistic fea-
54
MORRIS MOSCOVITCH
tures, should show the predicted change in performance regardless of
the behavior of the minor side.
METHOD
Subjects. Sixteen right-handed students at Erindale College participated in the experiment. Half the subjects responded with the right hand, the other half with the left. The
apparatus, stimuli, and procedure were identical to those used in Experiment I except that
the acoustic set letter and visual test letter were presented simultaneously instead of successively. The onset of the acoustic stimulus activated a Schmitt trigger that started the
delivery of the test stimulus.
RESULTS
The results are summarized in Table 2 and aspects of them are
highlighted in Fig. 2. Separate analyses of variance were conducted for
“same” and “different” judgements. For “same” judgements, a two-way
analysis of variance with response hand as the between factor and visual
field as the within factor showed no significant main effects or interactions. For “different” judgements, a three-way analysis of variance with
response hand as the between factor and visual field and similarity as the
within factors showed a significant main effect of similarity, pictorially
similar test-set pairs yielding longer RTs than acoustically similar pairs,
595 to 568 msec. [F (1,42) = 57.00, p < .OOl], and a significant visual
field by similarity interaction, the difference in RT between pictorially
and acoustically similar pairs being larger when the test letter appears in
the LVF, (36 msec) than when it appears in the RVF (18 msec) [F
(1,42) = 5.03; p < .05].
As in Experiment 1, very few errors were made. The error data, however, parallel the RT data (see Fig. 2). More errors were made in distinguishing pictorially similar from acoustically similar pairs, 76 to 33. Also
the difference between pictorial and acoustic confusions was greater in
the LVF (34) than in the RVF (9).
SIMULTANEOUS
TABLE
2
LETTER
MATCHING
TASK=
Same
LVF
Left hand
Right hand
521(23)
529(34)
Different
RVF
520(24)
530(20)
LVF
RVF
A
P
A
P
%X2(11)
544 (5)
617(33)
580(17)
599(lO)
547 (7)
614(16)
568(10)
(2Mean RT (msec) of correct responses averaged over subjects and broken down by response hand, visual field (LVF vs RVF), type of judgment (same vs different) and similarity
of set-test pairs (acoustic (A) vs pictorial (P)). The number in parentheses next to the RT
designates the number of errors in that condition.
LANGUAGE
AND THE RIGHT HEMISPHERE
LEFT
VISUAL FIELD
s5
Rl6HT
VISUAL FIELD
FIG. 2. Mean RT and total errors for “different” judgements in the simuhneous lettermatching task averaged over subjects and response hand, and broken down by visual field
and similarity (pictorial (visual) vs acoustic). Dotted column, visual similarity; diagonally
lined column, acoustic similarity.
DISCUSSION
The results show that the right and left hemisphere execute the same
matching task in different ways. When the test stimulus projects to the
right hemisphere the difference in RT between pictorially and acoustically similar items is significantly greater than when it projects to the
left. The minor hemisphere compares the items along a pictorial dimension, as it did in Experiment 1. The dominant hemisphere also compares
the letters primarily along a pictorial modality, but its sensitivity to
linguistic features makes it more likely than the minor hemisphere to
confuse pairs whose names sound alike and less likely to confuse pairs
that look alike. This interpretation would account for the significant
visualfield by similarity interaction and is consistent with the error analysis. The difference between the number of false positive responses
made to pictorially similar pairs as compared to acoustically similar pairs
is greater in the minor than in the dominant hemisphere.
Figure 3 emphasizes the difference in processing strategies that the
hemispheres assumed in Experiments 1 and 2. When a 2-set delay is introduced between the presentation of the acoustic set letter and the
visual test letter, both hemispheres relied on a pictorial strategy as
56
MORRIS MOSCOVITCH
( visual
wows
CONSECUTIVE
)
I acrrscic
40
wrors
j
30
30
10
( visual
-
CONSECUTIVE
RT )
40
SIMULTANEOUS
30
LVF
RVF
LVF
RVF
FIG. 3. Comparison of the difference in RT and total errors between pictorially (visually)
and acoustically similar test-set pairs (pictorial (visual) minus acoustic) for the right and
left visual field in the consecutive and simultaneous conditions of the letter-matching task.
Dotted column, right visual field: diagonally lined column, left visual held.
evidenced by the identical RT differences of 28 msec between visually
and acoustically similar pairs regardless of the visual field in which the
test stimulus appeared. In the second experiment, the simultaneous
presentation of the set and test stimulus enhanced the interfering effects
which the trace of the acoustic stimulus produced. The interference,
however, was selective, affecting the dominant but not the minor hemisphere. This suggests that only the dominant hemisphere is aware of the
linguistic relation between the acoustic and the visual stimulus. It further
suggests that the normal minor hemisphere is dependent on the dominant one for the linguistic analysis necessary in the identification of the
name of the letter from the acoustic stimulus and, ultimately, for the generation of the visual representation of that name. Whether the dominant
hemisphere transmits this information to the minor one in a visual code
or in another non-linguistic code cannot be determined from the present
experiment.
Another interpretation may explain the minor hemisphere’s failure to
exhibit linguistic skills in the simultaneous matching task. It may be that
the minor hemisphere attends to the linguistic features of acoustic stimuli if it is not distracted by additional information. The present experi-
LANGUAGE
AND THE RIGHT HEMISPHERE
57
ment, however, requires it to process acoustic and visual information
simultaneously and under these circumstances it may relinquish the
linguistic processing load to the dominant hemisphere. The next experiment was designed to choose between the two interpretations.
EXPERIMENT 3
The design of the experiment was similar to Experiment 1. The subject compared a visual test letter with a binaural set letter heard two seconds earlier, except that now he indicated whether the test letters had the
same terminal phoneme as the set letter. If, for example, the set letter
was “B” and the test was “G,” he responded “same” since both letters
end in “ee”; if the test, however, was “M” he responded “different”
since it does not end in “ee.” Successful performance in this task
requires that the subject attend to linguistic features, namely, the
phonemes in the letter’s name. Matching on the basis of pictorial characteristics would be useless unless the two letters were identical. Finally,
to meet the objection raised in Experiment 2, the acoustic stimulus is
presented 2 set prior to the visual stimulus, allowing the minor hemisphere to examine the test stimulus for the appropriate linguistic features
without having to simultaneously analyze the acoustic stimulus.
On one sixth of the trials for which the correct response was “same”
the test letter and set letter were identical. This condition replicated
Experiment 1 exactly and was included to see if subjects matching for
terminal phonemes treat identical pairs, the letters of which share all features, pictorial and linguistic, the same as they do the letters of end-alike
pairs which have only a terminal phoneme in common. This condition
helps test the sufficiency of Kinsboume’s (1970, 1973) attention model
of perceptual asymmetry (see Discussion).
METHOD
Subjects. Sixteen right-handed students at the University of Pennsylvania participated in
the experiment. Half responded with the left hand, half with the right.
Procedure
and stimuli. The procedure and stimuli were identical to the one used in the
previous experiments except for the following: (I) the subject responded “same” when the
name of the letter he saw (test letter) ended in the same sound as the letter he heard (set
letter) and “different” when they ended in a different sound. (2) The test letters comprising
the positive set were B, D, G, T, V, Z, those comprising the negative set were F, L, M, N,
S, X. The auditory set consisted of single letters drawn randomly from the positive set. All
the letters in the positive set end in “ee” and all the letters in the negative set begin with
“eh.” On one sixth of the positive trials the test and set letters were identical (identical
condition). On the remaining five sixths of the trials, the test letters had only a terminal
phoneme in common with the set letter (end-alike condition).
RESULTS
The results are summarized in Table 3 and aspects of them are
highlighted in Fig. 4. The RTs for “same” judgements were analyzed by
58
MORRIS MOSCOVITCH
TABLE 3
PHONEME
COMPARISON
TASK”
Different
Same
RVF
LVF
Left hand
Right hand
I
E
I
E
481(10)
491 (0)
519(16)
535(16)
479(10)
507 (3)
510(26)
525(12)
LVF
RVF
544 (28)
551(10)
542(3 1)
546(10)
a Mean RT (msec) of correct responses averaged over subjects and broken down by response hand, visual field (LVF vs RVF), type of judgment (same vs different) and similarity
of set-test pairs for “same” responses (identical (I) vs end-alike (E)). The number in parentheses next to the RT designates the number of errors in that condition.
a three-way analysis of variance, with response hand as the between
factor and similarity (identical vs end-alike) and visual field as the within
factor. The results of the analysis showed the following significant effects: similarity, identical RTs being 33 msec faster than end-alike RTs
[F (I ,42) = 68.50; p < .OOl], similarity by visual Jield interaction, identical RTs favoring the LVF, and end-alike RTs favoring the RVF [F
(1,42) = 4.10; p < .05]. The end-alike and identical conditions were
then analyzed separately. A two-way analysis of the end-alike condition,
with response hand as the between and visual field as the within factor,
( B.M 1
DIFFERENT
560
_-BP”
540
t
E
( B.G1
L,”
RT
in ma.
520
F
610
621
( B.B1
IDENTICAL
500
460 E
544
END-ALIKE
a..”
AVERAGE
546
:::
.::..:::::::
.:.:::::‘:
:.:...;,
::;::::;.
.:,j;;
g::;:+
:j,j
j;:;::
.j,::
_j:
j:
yj,:
j::
:::::
.:
.:
,’
:::;
;;:,y;
j:j::
EL
463
466
516
,i:::!i:::
": ',i::,:;
: ,y;:.::
:/: .: ::::.
.: ,.,.:..
i.;,: .$
;. ....I..
j::j,:: :
j:+ ;:;
:;:,;;:y:"
.:..:,,..,'
,:,::.::'::
. .. )...
y:;:;::,:
.:.:.;,,
::'j ii';; j
;:::ji:,::
..:..
::.,,. ::,::
:,::::::: ::
Ij:::: '. :
.,,:,j: :.j
::.:::::.::
. ., ,.::
1:.:.:..::jj
m
FIG. 4. Mean RT in the phoneme-comparison
task averaged over subjects and response
hand and broken down by visual field and similarity of set-test pairs (identical vs end-alike
vs different). Dotted column, right visual field; Diagonally lined column, left visual field.
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
59
showed that there was a significant visual field effect, RTs being 10 msec
shorter in the RVF [F (1,15) = 7.06; p < .025] and no significant
response hand by visual Jield interaction. A similar analysis of the identical condition showed no significant main effects or interactions.
Although RTs for “different” judgements slightly favored the RVF,
analysis of variance yielded no significant main effects or interactions.
DISCUSSION
Two main conclusions emerge from this experiment. The first is that
perceptual asymmetries are determined, as all three experiments
suggest, not by the stimulus material per se but by the cognitive processes that are emphasized in analyzing the information. The second is
that the minor hemisphere of normal people displays little, if any, of the
linguistic abilities that it might possess. The discussion that follows
will elaborate on these conclusions.
Cognitive Processes Determine Perceptual Asymmetries
Although subjects were instructed to respond only to the presence or
absence of a common terminal phoneme, RTs of “same” judgements to
test letters in the identical condition were significantly shorter than RTs
to the letters in the end-alike conditions. This suggests that different cognitive processes were used in the two conditions. In the identical condition, pictorial comparison of the test letter to a visual representation of
the set letter leads to correct responses. In addition, the pictorial strategy
has the advantage of eliminating the necessity of phonemically analyzing
the name of the test letter, a linguistic process that the end-alike condition calls for. The results also show that the visual field that is favored
depends on the process that is emphasized. Where a pictorial analysis is
sufficient and more expedient, as in the identical condition, RTs will
favor, though not significantly, the visual field projecting to the right
hemisphere. This is consistent with the results of the first two experiments. In the end-alike condition, however, the emphasis on linguistic
processes engages the left hemisphere and results in shorter RTs to test
letters presented in the right visual field.
Similar results were obtained by Cohen (1972) and Geffen et al.
(1972) in a modified Posner task. Subjects saw a pair of letters which
were either identical, e.g., “aa” or “AA,” or which had the same name,
e.g., “aA” and, in both conditions, were required to indicate whether the
two letters had the same name. The pairs appeared in either the right or
left visual field. The identical condition favored the left visual field
whereas the name-alike condition favored the right visual field. Also,
RTs to identical pairs were 80-l 30 msec shorter than to name-alike
pairs. The relatively smaller difference of 33 msec between identical and
60
MORRIS
MOSCOVITCH
end-alike RTs in our experiment could be accounted for by the low
probability of occurrence of identical pairs. Since RT is inversely proportional to the probability of an event, RTs were probably inflated for
identical pairs and reduced for end-alike pairs.
The results of these experiments also provide evidence about whether
linguistic and pictorial comparisons proceed serially or in parallel with
each other. One possibility is that the comparisons are handled serially,
such that the stimuli are first examined for identity on a pictorial dimension and are inspected for phonemic similarities only upon completion of
the pictorial stage. If this were the case RTs would favor the left visual
field even in the end-alike condition since the left visual field advantage
associated with pictorial processes would be maintained over all the succeeding stages. That RTs favored opposite visual fields in the two conditions suggests that the phonemic and pictorial comparisons proceed in
parallel with each other.
Minor
Hemisphere
Pe$ormance
on Linguistic
Tasks
A number of interpretations can account for the minor hemisphere’s
inferior performance in the end-alike condition of the phoneme matching
task. One interpretation is that when competing with the dominant hemisphere, the minor one does not engage in linguistic processes. Thus, all
test letters must travel to the dominant side for phonemic comparison.
Assuming that visual field differences in RT largely depend on the time
it takes to transmit, receive, and interpret information transferred
between hemispheres, then according to the above interpretation, RTs
should favour the right visual field by identical amounts no matter which
hand the subject uses to respond, i.e., no matter which hemisphere emits
the response. If, however, each hemisphere independently processes
linguistic information but the dominant does so more quickly than the
minor, RTs will still favor the right visual field, but the difference would
vary with response hand. For a dominant hand response, the time advantage gained by the accessibility of motor output pathways on the
dominant side will be added to the linguistic processing advantage of the
dominant hemisphere, whereas for a minor hand response the response
initiation advantage rests with the minor hemisphere and so must be subtracted from the dominant’s advantage for processing linguistic information. In short, if the minor hemisphere displays the requisite linguistic
skill, then a significant visualfield by response hand interaction would be
observed.
Examination of Table 3 shows that in the end-alike condition RTs
favor the right visual field by equal amounts regardless of response hand.
This favors the interpretation that in normal people the minor hemisphere does not contribute to the execution of tasks requiring linguistic
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
61
skill. Similar results were obtained in every reaction time study that has
come to the author’s attention (Geffen et al., 1971; Gross, 1972;
Klatzky, 1970; Klatzky & Atkinson, 1972; Rizzolatti et al., 197 1;
Springer, 197 1, 1972). A memory scanning study by Klatzky and Atkinson (1972) argues especially well against the notion that both hemispheres process linguistic information, but the minor is less efficient than
the dominant. Their subjects indicated whether the name of a picture
presented in the right or left visual field began with the same letter as
one of the two to five letters they held in memory. If the minor hemisphere is merely less efficient than the dominant at processing linguistic
information, then its performance relative to that of the dominant’s
should deteriorate with increases in the size of the memory set. This
would be reflected in a corresponding increase in the size of the RT difference favoring the right visual field. The difference, however, remains
constant over memory-set size regardless of the responding hand.
Although the evidence from these studies minimize the importance of an
efficiency interpretation of the hemispheres’ performance on linguistic
tasks, the possibility remains that the minor hemisphere processes
linguistic information so slowly that a dominant hemisphere decision initiates a motor response on the minor or dominant side before a similar
signal arrives from structures in the minor hemisphere. Even if this were
true, it would still reinforce the conclusion that for all practical purposes
linguistic behavior reflects only dominant hemisphere processes.
Before discussing the implications of this conclusion for models describing the localization of language, one last problem must be explored.
Although studies using reaction times are consistent in that they each
show visual field differences in the predicted direction, it is puzzling that
these differences vary from study to study. The differences range from 2
to 50 msec and the variation as well as the absolute values are greater
than expected if RT differences measured the time it took information to
cross the corpus callosum. There are a number of ways of accounting for
the variation. What is being measured is the time it takes to transmit,
analyze, and encode information across the hemispheric pathways and
nof the time it takes a nerve impulse to cross the cerebral commissures.
Assuming that stimulus information is degraded with transmission
across the cerebral commissures and that the receiving hemisphere must
clear up and interpret the transmitted signal, then it is likely that the
more complex or faint the effective signal is, the greater the RT differences will be. By and large, this is the case. When measuring RTs for
crossed and uncrossed sensory-motor projections to “simple” stimuli
such as a flash of light or a touch on the skin, the differences are small
and in the range of 2-20 msec (Blinkov and Arutyunova (and references in Russian), 1966; Bradshaw & Perriment, 1970; Efron, 1963;
62
MORRIS MOSCOVITCH
Jeeves, 1969; Kerr, Mingay & Elithorn, 1963; f’ofienberger, 1912).
The only exception is a study by Filby and Gazzaniga ( 1970) who
found differences of about 40 msec, but they required subjects to detect the presence or absence of a small dot. If, however, the attributes
of the stimulus to which a subject must respond are complex, such as
when letters, faces, pictures, slanted lines, speech, and random square
matrices are used, then the differences obtained are in the range of
lo-50 msec (Cohen, 1972; Geffen ef al., 1971, 1972; Gibson et al.,
1970; Gross, 1972; McKeever & Gill, 1972; Moscovitch, 1970, 1972;
Rizzolatti et al., 197 1; Springer, 197 1, 1972; Umilta, Frost & Hyman,
1972).
Kinsbourne (1970, 1972) offered an alternative interpretation: RT differences to stimuli presented in the right or left sensory fields reflect attentional asymmetries about the midline rather than communication
across hemispheric transmission lines. The more linguistically or pictorially complex the effective stimulus is, the more it selectively activates
the left or right hemisphere, respectively, which in turn causes the subject to direct more of his attention to the sensory field contralateral to
the activated hemisphere. As a result, stimuli presented to that field will
be perceived better and with a shorter latency than when they are
presented to the opposite field. The distinguishing feature of Kinsbourne’s attention model is that it explains perceptual asymmetries on
the basis of attentional biases induced by hemispheric activation prior to
the presentation of the stimulus, whereas the traditional access model on
which most of the theorizing in this paper is based claims that perceptual
asymmetries result from post-stimulus effects; namely, the relative
access which the effective stimulus trace has to the hemisphere specializing in its analysis. The attentional factors explain a variety of results
that cannot be incorporated into the access model (see Kinsbourne,
1972) and undoubtedly contribute to some of the variance found in RT
studies. The data from the present experiment, as well as that collected
by other investigators, indicate that an activation-attention model is not
sufficient to account for all the observed laterality effects-unless additional assumptions are made.4 Selective activation of one of the hemispheres could be eliminated by randomly presenting two types of stimulus material, one which engages left hemisphere processes and one
which engages right hemisphere processes. According to the attention
model, this should eliminate the corresponding perceptual asymmetries.
This prediction was not borne out in the present study, or in Cohen’s
(1972) or in Geffen et al.‘s (1972). In these experiments, the subject in4 Kinsboume (1972) modified his attention model to account for post-stimulus effects. In
its present form, however, there is little to distinguish the attention model from the access
model regarding this explanation of post-stimulus effects.
LANGUAGE
AND THE RIGHT HEMISPHERE
63
dicated whether a pair of letters had the same name (Cohen and Geffen
et al.) or had the same terminal phoneme (Moscovitch). RTs to pairs of
letters that were identical and could be matched on the basis of physical
characteristics favored the left visual field whereas RTs to pairs that had
only a name or terminal phoneme in common favored the right visual
field, despite the fact that the two conditions were presented randomly
and so could not selectively prime one of the hemispheres prior to the
presentation of the stimu1us.5 Although pre-stimulus attentional factors
may influence the magnitude of the differences obtained, they cannot account for these results as adequately as an access model.
GENERAL
DISCUSSION
The paper began with the question “How does the normal minor
hemisphere perform on linguistic tasks ?” More specifically, does its performance reflect the linguistic ski11it demonstrated in split-brain studies?
These are questions concerning the functional organization of the
normal cerebra1 hemispheres. The results of Experiments 2 and 3, as
well as of similar experiments conducted by other investigators, suggest
that the functional organization of the cerebral hemispheres of normal
people is not the same as that of split-brain patients.” This does not
imply that if it were possible to examine each cerebra1 hemisphere in
isolation in normal people it would reveal properties that are different
from those of the split-brain hemispheres. Rather, it means that a new
organization may emerge when the hemispheres interact with each other
such that some properties or capacities are functional, are realized in
behavior, while others, though present, may be obscured or inhibited.
Thus, although the opportunity was present for the linguistic skills of the
minor hemisphere to emerge, the linguistic behavior of normal people, as
opposed to that of split-brain patients, reflected only the performance of
5 In the phoneme matching experiment the end-alike pair appeared five times as often as
the identical one and, as a result, a slight “activation” effect was observed. RT difference-s
to identical pairs were smaller than in Experiment 1, although they continued to favor the
left visual field.
I/ Levy (personal communication) claims that there is no inconsistency between the splitbrain and normal data because the tasks chosen to test the linguistic skill of the normal
minor hemisphere require phonological recoding of the stimulus. The split-brain minor
hemisphere can semantically decode words to derive their meaning, but it cannot phonologically recode them in order to rhyme one word with another (Levy, 1973). This criticism, however, applies only to the phoneme matching task. It does not apply to Experiment 2 or to the name matching task of Cohen (1972) and Geffen et a/ (1973). In those
studies, subjects are asked not whether two letters, such as “aA,” sound alike, but rather
whether they have the same semantic referrent- the name “A”- a task that involves the
type of semantic decoding that the split-brain minor hemisphere should be able to execute.
One sure way of determining the validity of this claim is to present these tasks to Splitbrain patients.
64
MORRIS
MOSCOVITCH
the dominant hemisphere. The subjects behaved “as if” the minor hemisphere did not possess any linguistic skill. Language, though possibly
represented in the right hemisphere of normal people, appears to be
functionally localized in the left.
This view of cerebra1 dominance forms one of the principal components of a model of cerebral organization. The mode1 states that, allowing for individual differences, the right hemisphere’s limited linguistic
skills are present to an equal degree in all right-handed people, be they
split-brained, aphasic, or normal. These skills, however, may be obscured and made nonfunctional by the linguistically dominant left hemisphere unless its influence over the right can be weakened or removed.
This model, termed the model offunctional locafization, was initially
proposed to account for the apparent discrepancy between normal and
split-brain data concerning the right hemisphere’s linguistic performance
(Moscovitch, 1973). In this it succeeds. Commissurotomy removes
some of the major pathways via which the left hemisphere can influence
the right, thereby permitting the right hemisphere’s linguistic capacities
to emerge.7 When the commissures are intact and the hemispheres
healthy, as they are in normal people, the left hemisphere suppresses the
right one’s linguistic performance.
The model, however, also provides a conceptual framework for integrating various neurological phenomena that do not fit comfortably
into any current theories. A detailed discussion of the model’s application to these phenomena appeared elsewhere (Moscovitch, 1973). We
will, therefore, restrict ourselves to listing some phenomena that may be
better understood if they are seen as resulting from the release of right
hemisphere language from left hemisphere control. These phenomena
include (a) speech comprehension and production following dominant,
left hemispherectomy in adulthood (Burklund, 1972; Crocket & Estridge, 1951; French, Johnson & Brown, 1955; Hillier, 1954; Smith,
1966; Smith and Burklund, 1966; Zollinger, 1935). (b) Lateralization of
language in the right hemisphere following early injury to the language
structures of the left (see Basser, 1962, for review). (c) The locus of
aphasic speech in the healthy right hemisphere of some right-handed
individuals following damage to the left hemisphere in adulthood (Kins7 Even in split-brain patients, however, the left hemisphere does not readily relinguish its
dominance of the right hemisphere in the linguistic realm. If the hemispheres conflict with
each other regarding linguistic decisions, it is usually the left hemisphere that directs the
split-brain patient’s responses (Levy et al., 1971, Mimer, Taylor & Sperry, 1968) and in
some cases even causes the patient to neglect information presented to the sensory halffield projecting to the right hemisphere (Kinsboume, 1973; Teng and Sperry, 1973). These
findings suggest that the dominant hemisphere can influence the activity of the minor one
and gain control over the subject’s behavior via pathways other than the cortical commissures (Ojemann & Ward, 1971; Penfield & Roberts, 1959).
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
65
bourne, 197 1). (d) Verbal comprehension in people whose left hemisphere is anesthetized by intra-carotid injections of sodium amytal
(Milner, 1967; Rossi & Rosadini, 1967; Wada & Rasmussen, 1960). (e)
Reduction of the magnitude of right-ear superiority in verbal dichotic listening tests following left-hemisphere lesions (Oxbury & Oxbury, 1969;
Schulhoff & Goodglass, 1969; but see Sparks, Goodglass & Nickel,
1970, for contradictory results). (f) Some recently reported findings, as
yet not fully substantiated, may also merit investigation from the point of
view of our model. These findings concern the relative sparing of ideographic, as opposed to phonological, writing systems in some kinds of
aphasic syndromes (Sasanuma & Fujimura, 1971); the relative sparing
of sign language, as compared to finger spelling and speech, in deaf people with left hemisphere damage (Critchley, 1938; Sarno, Swisher &
Sarno, 1969, plus references, Tureen, Smolik & Tritt, 195 1; for opposite
results see Douglas & Richardson, 1959); the moderate, initial success
some investigators have had in teaching global aphasics to communicate
using an artificial, nonverbal language similar to that developed by
Premack (1971) for chimpanzees (Velletri-Glass, Gazzaniga & Premack, 1973; Errol Baker, Aphasia Research Center, Boston, lecture
delivered at L’Hopital Hotel-Dieu, Montreal.)8
The findings mentioned in point (f) present special problems of interpretation. First, the particular forms of nonverbal communication that
are spared may be dependent on processes that have little in common
with those used in natural language. Second, even if nonverbal and
verbal forms of communication share some common processes, it may
be that the nonverbal form is simpler in the sense that the link between
symbol and meaning is more direct than it is for speech and for phonological writing systems. Consequently, some nonverbal communication
may be spared not because it is mediated by the right hemisphere, but
because its relative simplicity allows it to be mediated by the remaining
healthy tissue of the damaged left hemisphere. These considerations
make the investigation of nonverbal communication particularly interesting both for the linguist and the neuropsychologist.
Other language-related disturbances, such as global aphasia, wordblindness, and word-deafness in individuals with healthy, normal minor
hemispheres may occur if the left hemisphere continues to suppress the
verbal activity of the right. As mentioned in the introduction, others
(Sperry et al., 1969; Milner, 1970) offered a similar explanation but they
considered the pathology on the dominant side as the cause of the minor
hemisphere’s poor linguistic performance. According to our model, the
” A finding related to those mentioned in point (f) concerns the selective loss of the ability to read abstract words while the ability to read concrete words is retained (Alan
Baddeley, Colloquium presented at the University of Toronto, November, 1974).
66
MORRIS
MOSCOVITCH
pathological situation in which a malfunctioning dominant hemisphere
interferes with the activity of a healthy minor one is merely an unfortunate extension of the normal condition.
Finally, the model of functional localization may lead to a clearer understanding of the development of lateralization of function. This matter
will be discussed in greater detail because it illuminates some of the
points mentioned earlier and because it underscores the potential the
model has for directing future research.
Anatomical asymmetries between the hemispheres have been found in
infants suggesting that the predisposition for the development of left
hemisphere linguistic dominance may already be present at birth. It is
possible that during the first 5 or 10 years of life, the dominant hemisphere magnifies this initial advantage in two ways: (1) by acquiring language faster than the minor side (2) by suppressing the verbal behavior
of the minor side in proportion to its (the dominant’s) rapidly accumulating linguistic skills.g Some of the verbal competence that the
minor hemisphere possessed in childhood remains. Some language functions, however, may deteriorate through disuse while others may
perhaps be lost as a result of the differentiation of potentially linguistic
structures into those serving nonlinguistic functions. Pathology of the
dominant hemisphere may reduce its ability to acquire language and to
control the minor one, leaving the minor hemisphere relatively free to
develop and practice its verbal skills which may result in the minor
hemisphere assuming the dominant role.
If true, this account predicts that early damage to, or absence of, interhemispheric pathways should also lead to greater language development on the minor side. There have been a number of recent reports of
laterality tests on patients with callosal agenesis (Bryden & Zurif, 1970;
Ettlinger, Blakemore, Milner & Wilson, 1972; Sperry, 1970). The results
of these tests are difficult to interpret because there are often other cerebral abnormalities accompanying this disorder. Nonetheless, it is interesting to note that language functions are imperfectly lateralized in this
population, suggesting that both hemispheres comprehend language and
may, perhaps, even speak (Sperry, 1970). Reports of more elaborate
9 Alternatively,
it is possible
that the left hemisphere
is dominant
for language from infancy. What appears as the development
of the lateralization
of language to the left hemisphere may merely be the progressive
loss of plasticity
of the right hemisphere.
Beyond a
critical
period. the right hemisphere
may be unable to acquire
relatively
normal language.
Thus, language becomes functionally
localized
to the left hemisphere
not because it takes
time for lateralization
to develop but because it takes time for the right hemisphere
to lose
its plasticity.
The corpus callosum
may help regulate this process by conveying
information from the left hemisphere
which either causes the right hemisphere
to develop normally
or which causes it to develop into the linguistically
dominant
hemisphere
should any damage occur on the left (see below).
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
67
tests determining whether language functions are represented equally in
both hemispheres are just beginning to appear (Saul & Scott, 1973).
It was suggested earlier that lateralization of language to the left hemisphere may lead to the differentiation of potentially linguistic structures
on the right into those serving nonlinguistic functions. This leads to the
prediction that the lateralization of language on the left precedes the lateralization of nonlinguistic functions on the right. It also predicts, as
some investigators have already found (Levy, 1973), that if language
develops on the right it may take over some structures that would normally mediate nonverbal function and, thereby, lower the person’s performance on nonverbal tasks.
There is another approach to the problem of lateralization of function.
Rather than speculate, as we did, about the processes whereby some undetermined differences between the two hemispheres lead to the lateralization of function, one can inquire into the nature of those differences. What special properties does the left hemisphere have which
cause it to become linguistically dominant? Opinion on this matter has
been divided. One group believes, simply, that language is lateralized in
the left hemisphere because it is rich in the special purpose mechanisms
necessary to deal specifically with phonological and syntactic information that is peculiar to natural speech (Liberman, 1974). A second group
sees the problem in much broader terms. To them, language, be it conveyed by speech, gestures, or pictorial symbols, is the product of highly
generalized cognitive operations. Because these operations are the special province of the left hemisphere, language is lateralized to that side.
The right hemisphere, on the other hand, operates in a mode that is
incompatible with the processes necessary for language. As yet, the
descriptions of these two opposing cognitive modes is rather vague. The
right hemisphere has been conceived as a “concrete, spatial synthesizer”
that perceives information as a “meaningful Gestalt” (Levy, 1973) and
can be expected to process information in parallel (Cohen, 1973). Its
mode of thinking has been described as appositional and creative
(Bogen, 1969b; Bogen & Bogen, 1969), nonlinear and intuitive (Ornstein, 1972). Such cognitive traits, it is claimed, are incompatible with
the rational, abstract, propositional, temporally-ordered and, therefore,
linear and serial mode of thinking which characterizes the left hemisphere and on which language depends. These notions are similar in
spirit, and sometimes in content, to those of Jackson (1878), Head
(1926), and Goldstein (1948) who believed aphasia resulted not only
from the impairment of mechanisms that evolved to deal specifically
with speech, but also from a deficiency in “symbolic formulation and
expression” and abstract reasoning. By examining the language that is
presumed to be mediated by the right hemisphere, one may be able to
68
MORRIS
MOSCOVITCH
decide between these two points of view. If the former is correct, right
hemisphere language should merely be an impoverished version of that
of the left. If the latter, more general view, is correct, right hemisphere
language should differ qualitatively from that of the left because the underlying properties of the two languages, their rules, not their external
symbols, should reflect the cognitive operations peculiar to the two
hemispheres. Current research on split-brain and hemispherectomized
patients, as well as on aphasic patients being taught to communicate in
artificial languages, should indicate if there is anything unique about
right hemisphere language.
REFERENCES
Basser, L. S. 1962. Hemiplegia of early onset and the faculty of speech with special reference to the effects of hemispherectomy. Brain, 85,427-460.
Blinkov, S. M., & Arutyunova, A. S. 1966. Reaction time and conducting pathways of the
brain. Neuropsychologia,
4, I77- 182.
Bogen, J. E. 1969a. The other side of the brain. 1. Dysgraphia and dyscopia following cerebral commissurotomy. Bulletin of the Los Angeles NeurologicalSocieties,
34, 73-105.
Bogen, J. E. 1969b. The other side of the brain. II. An oppositional mind. Bulletin of the
Los Angeles Neurological
Societies, 34, 135-l 62.
Bogen, J. E., & Bogen, G. M. 1969. The other side of the brain. 111.The corpus callosum
and creativity. Bulletin of the Los Angeles Neurological Societies, 34, 19 I-220.
Bradshaw, J. L., & Perriment, A. D. 1970. Laterality effects and choice reaction time in a
unimanual two-finger task. Perception and Psychophysics, 7, 185- 188.
Bryden, M. P., & Zurif, E. B. 1970. Dichotic listening performance in a case of agenesis of
8, 371-377.
the corpus callosum. Neuropsychologia,
Burklund, C. W. 1972. Cerebral hemisphere function in the human: fact versus tradition.
In W. Lynn Smith (Ed.) Drugs, Development and Cerebral Function. Springfield, IL:
Charles C Thomas. Pp. 8-36.
Cohen, G. 1972. Hemispheric differences in a letter classification task. Perception and
Psychophysics, 11, 137-142.
Cohen, G. 1973. Hemispheric differences in serial versus parallel processing. Journal of
Experimental
Psychology,
97, 349-356.
Critchley, M. 1938. Aphasia in a partial deaf-mute. Brain, 61, 163- 169.
Crockett, H. G., & Estridge, N. M. 1951. Cerebral hemispherectomy: A clinical, surgical,
and pathologic study of four cases. Bulletin of the Las Angeles Neurological Society,
16, 71-87.
Douglass, E., & Richardson, J. C. 1959. Aphasia in a congenital deaf-mute. Brain, 82,
68-80.
Efron, R. 1963. The effect of handedness on the perception of simultaneity and temporal
order. Brain, 86, 26 l-284.
Egeth, H., & Blecker, D. 1971. Differential effects of familiarity on judgements of sameness and difference. Perception and Psychophysics, 9, 321-326.
Egeth, H., & Epstein, J. 1972. Differential specialization of the cerebral hemispheres for
the perception of sameness and difference. Perception and Psychophysics,
12,
218-220.
Ettlinger, G., Blakemore, C. B., Mimer, A. D., & Wilson, J. 1972. Agenesis of the corpus
callosum: A behavioural investigation. Brain, 95, 327-346.
Filby, R. A., & Gazzaniga, M. S. 1969. Splitting the normal brain with reaction time.
Psychonomic
Science, 17, 335-336.
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
69
French, L. A., Johnson, D. R., Brown, 1. A., & von Bergen, F. B. 1955. Cerebral hemispherectomy for control of intractable convulsive seizures. Journal of Neurosurgery,
12, 154-164.
Gazzaniga, M. S. 1970. The Bisected Brain. New York: Appleton-Century-Crofts.
Geffen, G., Bradshaw, J. L., & Nettleton, N. C. 1972. Hemispheric asymmetry: Verbal
and spatial encoding of visual stimuli. Journal of Experimental Psychology, 93, 25-3 1.
Geffen, G., Bradshaw, J., & Wallace, G. 197I. Interhemispheric effects on reaction time to
verbal and nonverbal stimuli. Journal of Experimenfal Psychology, 85, 4 15-422.
Geschwind, N. 1970. The organization of language and the brain. Science, 170, 940-944.
Gibson, A. R., Dimond, S. J., & Gazzaniga, M. S. 1972. Left field superiority for word
matching. Neuropsychologia,
10, 463-466.
Gibson, A., Filby, R. A., & Gazzaniga, M. S. 1970. Hemispheric differences as reflected
by reaction time. Federation Proceedings, 29, 658 (abstract).
Goldstein, K. 1948. Language and Languuge Disturbances. New York: Grune and
Stratton.
Goodglass, H., & Kaplan, E. 1963. Disturbances of gesture and pantomime. Brain, 86,
703-720.
Gross, M. M. 1972. Hemispheric specialization for processing of visually presented verbal
and spatial stimuli. Perception and Psychophysics, 12, 357-363.
Head, H. 1926. Aphasia and Kindred Disorders of Speech. Cambridge: Cambridge University Press.
Hillier, W. 1954. Total left cerebral hemispherectomy for malignant glioma. Neurology, 4,
718-721.
Jackson, H. 1878. On the affections of speech from disease of the brain. Brain, 1,
304-330.
Jeeves, M. A. 1969. A comparison of interhemispheric transmission times in acollosals
and normals. Psychonomic Science, 16, 245-246.
Kerr, M., Mingay, R., & Elithorn, A. 1963. Cerebral dominance in reaction time
responses. British Journal of Psychology, 54, 325-336.
Kinsbourne, M. 1970. The cerebral basis of lateral asymmetries in attention. Acta Psychologica, 33, 193-201.
Kinsbourne, M. 1971. The minor cerebral hemisphere as a source of aphasic speech.
Archives
of Neurology,
25, 302-306.
Kinsbourne, M. 1973. The control of attention by interaction between the cerebral hemispheres. In Sylvan Kornblum (Ed.) Attention and Performance IV. North Holland
Publishing Co. Pp. 239-256.
Kinsboume, M. 1973. Interaction and competition between the hemispheres in the control
of attention. Paper presented at the Eastern Psychological Association, Washington,
DC.
Klatzky, R. L. 1970. Interhemispheric transfer of test stimulus representation in memory
scanning. Psychonomic Science, 21, 20 l-203.
Klatzky, R. L., & Atkinson, R. C. 1971. Specialization of the cerebral hemispheres in
scanning for information in short-term memory. Perception and Psychophysics, 10,
335-338.
Levy, J. 1971. Cerebral asymmetries as manifested in split-brain man. Paper presented at
the Third Annual Cerebral Function Symposium on Hemispheric Disconnection and
Cerebral Function at Oregon State University. Oregon State University Press. In
press.
Levy, J., Nebes, R. D., & Sperry, R. W. 1971. Expressive language in the surgically separated minor hemisphere. Cortex, 8, 49-58.
Levy, Trevarthen, C., & Sperry, R. W. 1972. Perception of bilateral chimeric figures following hemispheric deconnexion. Brain, 95, 6 1-78.
70
MORRIS
MOSCOVITCH
Liberman, A. M. 1974. The specialization of the language hemisphere. In F. 0. Schmitt
and F. G. Worden (Ed%) The Neurosciences: Third Study Program. Cambridge, MA:
M.I.T. Press. Pp. 43-56.
McKeever, W. F., and Gill, K. M. 1972. Interhemispheric transfer time for visual stimulus
information varies as a function of the retinal locus of stimulation. Psychonomic Science, 26,308-3
IO.
Milner, B. 1967. Discussion on Cerebral Dominance in Man. In C. H. Millikan and F. L.
Darley (Eds.) Brain Mechanisms Underlying Speech and Language. New York:
Grune and Stratton. Pp. I77- 184.
Milner, B. I97 1. Interhemispheric differences and psychological processes. British Medical Bulletin,
27, 272-277.
Milner, B., Taylor, L., & Sperry, R. W. 1968. Lateralized suppression of dichotically
presented digits after commissural section in man. Science, 161, 184-185.
Milner, P. 1970. Physiological
Psychology. New York: Holt, Rinehart, and Winston.
p. 258.
Moscovitch, M. 1972. A choice reaction time study assessing the verbal behavior of the
minor hemisphere in normal, adult humans. Journal of Comparative and Physiological
Psychology,
80, 66-74.
Moscovitch, M. 1973. Language and the cerebral hemispheres: Reaction-time studies and
their implications for models of cerebral dominance. In P. Pliner, L. Krames, and T.
M. Alloway (Eds.) Communication and Affect: Language and Thought. New York:
Academic Press. Pp. 89-126.
Moscovitch, M., & Catlin, C. 1970. Interhemispheric transmission of information: Measurement in normal man. Psychonomic Science, 18, 2 11-2 13.
Ojemann, G. A., & Ward, A. A., Jr. 1971. Speech representation in ventrolateral thalamus. Bruin, 94, 669-680.
Ornstein, R. E. The Psychology ofConsciousness.
San Francisco: W. H. Freeman and Co.
Oxbury, J. M., & Oxbury, S. M. 1969. Effects of temporal lobectomy on the report of
dichotically presented digits. Cortex, 5, 3- 14.
Penfield, W., & Roberts, L. 1959. Speech and Brain Mechanisms. Princeton, NJ: Princeton University Press.
Premack, D. I97 1. Language in chimpanzee. Science, 172,808-822.
Poffenberger, A. T. 1912. Reaction time to retinal stimulation with special reference to the
time lost in conduction through nerve centres. Archives of Psychology, 13, l-73.
Posner, M. 1. 1969. Abstraction and the process of recognition. In G. H. Bower and J. T.
Spence (Eds.). The Psychology ofLearning
and Motivation.
New York: Academic
Press, pp. 43- 100.
Rizzolatti, G., Umilta, C., and Berlucchi, G. 197 I. Opposite superiorities of the right and
left cerebral hemispheres in discriminative reaction time to physiognomical and alphabetical material. Brain, 94, 43 l-442.
Rossi, G. F., & Rosadini, G. 1967. Experimental analysis of cerebral dominance in man.
In C. H. Millikan and F. L. Dorley (Eds.). Brain Mechanisms Underlying Speech and
Language: New York: Grune & Stratton.
Sarno, J. E., Swisher, L. P.. and Sarno, M. T. 1969. Aphasia in a congenitally deaf man.
Cortex, 5, 398-4 14.
Sasanuma, S., & Fujimura, 0. I97 I. Selective impairment of phonetic and non-phonetic
transcriptions of words in Japanese aphasic patients: Kana vs. kanji in visual recognition and writing. Cortex, 7, I-1 8.
Saul, R. E., & Scott, P. S. 1973. Compensatory mechanisms in agenesis of the corpus
callosum. Neurology, 23(4), Abstract BT 68.
Schulhoff, C., & Goodglass, H. 1969. Dichotic listening, side of brain injury, and cerebral
dominance. Neuropsychologia,
7, I49- 160.
LANGUAGE
AND
THE
RIGHT
HEMISPHERE
71
Smith, A. 1966. Speech and other functions after left (dominant) hemispherectomy.
Journal
of Neurology,
Neurosurgery,
and Psychiatry,
29, 467-47 1.
Smith, A., & Burklund, C. W. 1966. Dominant hemispherectomy: Preliminary report on
neuropsychological sequelae. Science, 153, 1280- 1283.
Sparks, R., Goodglass, H., & Nickel, B. 1970. lpsilateral versus contralateral extinction in
dichotic listening. Cortex, 6, 249-260.
Sperry, R. W. 1968. Hemispheric deconnection and unity in conscious awareness. American Psychologist,
23, 723-733.
Sperry, R. W. 1970. Cerebral dominance in perception. In F. A. Young and D. B.
Lindsley (Eds.) Early Experience in Visual Information Processing in Perceptual and
Reading Disorders. Washington, DC: National Academy of Sciences. Pp. I67- 178.
Sperry, R. W., Gazzaniga, M. S., & Bogen, J. E. 1969. Interhemispheric relationships: The
neocortical commissures; syndromes of hemispheric disconnection. In P. J. Vinken &
G. W. Bruyn (Eds.) Handbook of Clinical Neurology, Vol. 4. Amsterdam: North
Holland Publishing Co.
Springer, S. 197I. Ear asymmetry in a dichotic detection task. Perception and Psychophysics, 10, 239-241.
Springer, S. 1972. The effect of contralateral noise on the perception of speech: A reaction
time analysis. Paper presented to the Western Psychological Association, Portland,
WA.
Teng, E. L., & Sperry, R. W. 1973. Interhemispheric interaction during simultaneous bilateral presentation of letters and digits in commissurotomized patients. Neuropsychologia,
11, 13 I-140.
Tureen, L. L., Smolik, E. A., & Tritt, J. H. I95 I. Aphasia in a deaf-mute. Neurology, 1,
237-,249.
Umilta, C., Frost, N., & Hyman, R. 1972. Interhemispheric effects on choice reaction
times to one-, two-, and three-letter displays. Journal of Experimental Psychology, 93,
198-204.
Velletri-Glass, A., Gazzaniga, M. S., & Premack, D. 1973. Artificial language training in
global aphasics. Neuropsychologia,
11, 95- 103.
Wada, J. 1969. Paper presented at 9th International Congress of Neurology, New York.
See Geschwind, N. (1970).
Wada, J., & Rasmussen, T. 1960. lntracarotid injection of sodium amytal for the lateralization of cerebral dominance: Experimental and clinical observations. Journal of Neurosurgery, 17, 266-282.
Zollinger, R. 1935. Removal of left cerebral hemisphere: Report of a case. Archives
Neurology
and Psychiatry,
34, 1055-1062.
of