DIRECTIONAL SCAN AND RIGHT-LEFT
DISCRIMINATION IN RIGHT AND LEFT WANDERS
REVA SUSAN TANKLE
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Dr. Ken Heilman is largely responsible for my growth and
development during my years as a graduate student. I would like
to thank him for staying with me and sticking by me through it all.
I am most appreciative of the support given to me by Dr. Carol
Van Hartesveldt. She made it possible for me to pursue my interest
in neuropsychology and remain a student in physiological psychology.
Dr. Dawn Bowers has invested much time in helping me to produce
a conceptually and linguistically sound dissertation for which I
am truly grateful. Dr. Eileen Fennell, Dr. Ira Fischler, and
Dr. Keith Berg have been ideal committee members. I would like to
thank each of them for their individual contributions to this
dissertation and to my education in general.
I appreciate the help given to me by my friends, David Juras,
Kathleen Shannon and Greg Galbicka in preparing this dissertaion.
My best friend and husband, John McArthur, is responsible for keeping
me sane and even smiling through a lot of rough days and nights. I
deeply appreciate his support and the extraordinary amount of time he
has invested in me and this dissertation.
Finally, a special thank-you to all the faculty and staff in
the Department of Neurology. I have truly enjoyed working with you.
TABLE OF CONTENTS
ACKNOWLEDGMENTS ................................... iii
LIST OF TABLES .......... ....................... vii
LIST OF FIGURES ............................... viii
CHAPTER I HANDEDNESS ......... ................... 1
Introduction to Handedness .......... 1
Incidence of Right and Left Handedness 1
Assessment of Handedness .... .......... 3
Early Theories of Right and Left
Handedness ........ ............... 4
Localization of Function ..... ........... 7
Historical Review ....... ............ 7
Handedness and Localization of Function . 10
Generalized Cognitive Defect. ........ ....22
Familial vs. Nonfamilial Left Handedness 25
Genetic Models of Handedness .......... ... 26
Neuroanatomical Asymmetries ......... .. 30
Predicting Localization of Function from
Handedness ...... .................. ... 32
Motor and Language Asymmetry .... ....... 32
Handposture and Lateralization of
Function ..................... 33
Degree of Handedness and Lateralization
of Function ..... ......... ....... ... 36
Conclusion ...... .................. ... 40
CHAPTER II EVIDENCE IN SUPPORT OF THE DIRECTIONAL SCAN
HYPOTHESIS ...... ................... .... 41
Mirror Reading in Right and Left Handers . 41
Directional Tendancies in Drawing ......... ... 50
Directional Tendancies and Eye Movements . 53
DIRECTIONAL SCAN HYPOTHESIS AND RIGHT-
LEFT DISCRIMINATION .. ........... ....56
Introduction ......... .......... 56
Asymmetrical Nervous Systems and the
Directional Scan Hypothesis .......... 56
The Development of Right-Left
Discrimination ..... ................ 57
Right-Left Discrimination in Adults ...... .. 59
Disturbances in Right-Left
Discrimination ..... ................ 62
Right-Left Discrimination and Directional
Scanning ....... ................... 65
Summary ....... .................... 67
STATEMENT OF THE PROBLEM .... ........... 68
Introduction .... ............
Directional Scan in Right and Left
Handers ........... .......
Right-Left Discrimination on Self.
Right-Left Discrimination in Space
Manikin Figures .. ..........
Self-Space ... .............
METHODS ........ .....................
Introduction ..... ...........
Subjects ..... ..............
Part I: Direction of Scan .......
Purpose ....... ........
Procedure .... ............
Part II: Right-Left Discrimination
and in Space ... .............
I. Simple-Self (Right-Left Discrimination
on Self) ....... ................
II. Hemispace-Self (Right-Left
Discrimination on Self). ..........
III. Manikin Figures (Right-Left
Discrimination in Space) ... .........
IV. Self-Space (Right-Left Discrimination
on Self and in Space) .... ...........
V. Baseline Responding
Overall Order of Presentation of the Tasks
Analysis ........ ................
CfLPTER VI Results ......... ........ 83
Introduction ................ 83
Part I: Direction of Scan .. ........ .. 83
Introduction ..... ............. .... 83
Reaction Times .... .............. ... 83
Error Rates ..... .............. ... 92
Part II: Right-Left Discrimination on
Self and in Space ... ................ 92
Introduction ....... ................ 92
I. Simple-Self (Right-Left Discrimination
on Self) ................ 94
II. Hemispace-Self (Right-Left Discrimin-
ation on Self) ... ............. ..97
III. Manikin Figures (Right-Left Discrim-
ination in Space) ... ........... .. 100
IV. Self-Space (Right-Left Discrimination
on Self and in Space) ............ i. 101
CHAPTER VII DISCUSSION ...... .................. ... 118
Part I: Directional Scan .. ......... 118
Part II: Right-Left Discrimination on
Self and in Space ....... ... 121
Introduction .... ............... .... 121
I. Simple-Self (Right-Left Discrimination
on Self. ........................... 125
II. Hemispace-Self (Right-Left Discrimin-
ation on Self) ... ............. ... 128
III. Manikin Figures (Right-Left Discrim-
ination in Space .... ........... ... 129
IV. Self-Space (Right-Left Discrimination
on Self and in Space) ............ ... 131
General Discussion .... ............ ... 134
Overall Discussion ... ............. ... 135
A EXAMPLES OF STIMULI USED IN PART I: DIRECTIONAL SCAN 140
B EXAMPLES OF STIMULI USED IN MANIKIN FIGURES CONDITION. 143
C EXAMPLES OF STIMULI USED IN SELF-SPACE CONDITION . 144
BIBLIOGRAPHY .......... ......................... 145
BIOGRAPHICAL SKETCH ........ ..................... 160
LIST OF TABLES
1 OBSERVED (0) AND EXPECTED (E) FREQUENCY OF 88
SUBJECTS WITH LONGER DR AND LONGER DL REACTION
TIMES (CHI-SQUARE ANALYSIS)
2 FREQUENCY OF RIGHT, LEFT AND MIXED HANDERS WITH 90
LONGER DR AND LONGER DL REACTION TIMES WITH PHI
AND CHI-SQUARE VALUES
3 DIRECTIONAL SCAN: SUMMARY OF ANALYSIS OF VARIANCE 93
OF MEAN ERRORS ON "SAME," "DR" AND "DL" TRIALS
4 SUMMARY OF ANALYSIS OF COVARIANCE OF MEAN REACTION 98
TIMES IN HEMISPACE-SELF CONDITION
5 SUMMARY OF ANALYSIS OF VARIANCE OF ERROR RATES IN 99
6 SUMMARY OF ANALYSIS OF VARIANCE ON MEAN REACTION 102
TIMES IN MANIKIN FIGURES CONDITION
7 SUMMARY OF ANALYSIS OF VARIANCE ON ERROR RATES 104
IN MANIKIN FIGURES CONDITION
8 MEAN REACTION TIMES IN SELF-SPACE CONDITION 106
9 SUMMARY OF ANALYSIS OF COVARIANCE ON THE MEAN 111
REACTION TIMES OF THE SELF-SPACE CONDTION TRIALS
WITH CONSISTENT COMMANDS
10 SUMMARY OF ANALYSIS OF COVARIANCE ON THE MEAN 112
REACTION TIMES OF THE SELF-SPACE CONDITION TRIALS
WITH INCONSISTENT COMMANDS
11 MEAN ERROR RATES IN THE SELF-SPACE CONDTION 115
12 SUMMARY OF THE ANALYSIS OF VARIANCE OF THE MEAN 116
ERROR RATES ON THE SELF-SPACE CONDITION TRIALS
WITH CONSISTENT COMMANDS
13 SUMMARY OF THE ANALYSIS OF VARIANCE OF THE MEAN 117
ERROR RATES ON THE SELF-SPACE CONDITION TRIALS
WITH INCONSISTENT COMMANDS
LIST OF FIGURES
1 DIRECTIONAL SCAN: MEAN REACTION TIMES ON DR 86
AND DL STIMULI FOR RIGHT, LEFT AND MIXED
2 MEAN REACTION TIMES OF MALES AND FEMALES IN THE 96
3 MEAN REACTION TIMES OF MALES AND FEMALES IN THE 103
MANIKIN FIGURES CONDITION
4 MEAN REACTION TIMES OF LEFT, MIXED AND RIGHT 108
WANDERS TO DOUBLE CROSSED COMMANDS AND TO
UNCROSSED COMMANDS IN THE SELF-SPACE CONDITION
5 MEAN REACTION TIMES OF LEFT, MIXED AND RIGHT 109
WANDERS TO MIXED COMMANDS AND TO DOUBLE CROSSED
COMMANDS IN THE SELF-SPACE CONDITION
6 MEAN REACTION TIMES OF LEFT, MIXED AND RIGHT 110
WANDERS TO CONSISTENT AND INCONSISTENT COMMANDS
IN THE SELF-SPACE CONDITION
7 MEAN REACTION TIMES OF MALES AND FEMALES IN THE 113
Abstract of Dissertation Presented
to the Graduate Council of the University of Florida
in Partial Fulfillment of the Requirements for the Degree of
Doctor of Philosophy
DIRECTIONAL SCAN AND RIGHT-LEFT
DISCRIMINATION IN RIGHT AND LEFT HANDERS
Reva Susan Tankle
Chairperson: Carol Van Hartesveldt
Co-Chairperson: Kenneth M. Heilman
Major Department: Psychology
It has been postulated that the hemispheric lateralization
of motor engrams in man contributes to the direction used to scan
the environment. It is predicted that strongly lateralized right
handers use a consistent left-to-right direction of scan. Less
strongly lateralized left and mixed handers are predicted to use
an inconsistent direction of scan. Further, differences in lateral-
ization may result in variable abilities when discriminating right
and left on oneself and in space. The strong lateralization in right
handers is predicted to confer an advantage such that right handers
should be better than left or mixed handers when making discriminations
on themselves and in space.
The Directional Scan hypothesis was tested by comparing subjects'
reaction times to two pictures presented successively. The second
picture was either the same as the first, or had a difference on the
left (DL) or right (DR) side of the picture. If the subjects were
using a left-to-right scan, then the differences on the left
should be perceived faster than the differences on the right.
If no consistent scan is being used then the reaction times to DR
and DL trials should not differ.
The results of this study support the hypothesis that the
handedness groups use different scanning directions. Right handers
were more likely to have shorter reaction times to DL than DR
suggesting a left-to-right scan. The left handers showed a trend
toward shorter reaction times to DR than DL indicating a right-to-
left scan. The mixed handers appeared to use an inconsistent
direction of scan.
The ability to make right-left discriminations on self and in
space was tested with a variety of methods. Generally, there was
no evidence of differences between the handedness groups when making
discriminations on self. There was partial support for the prediction
that right handers would be better than left handers when making
discriminations in space. There were significant sex differences
with both types of discriminations. The results are discussed in
terms of lateralization of function in males and females, visuo-
spatial processing and directional scan.
TO MY MOTHER, WHO SHOWED ME THE WAY.
TO MY FATHER, WHO SHOWED ME HOW.
Introduction to Handedness
Incidence of Right and Left Handedness
For centuries philosophers and psychologists have noted that
humans almost always develop a manual preference for one hand. Plato
attributed the underdevelopment of the other hand to the "stupidity of
mothers and nurses" (Hecaen and De Ajuriaguerra, 1964) who carried chil-
dren pressed against their bodies and, thereby, limited the use of one
hand. Plato, however, failed to observed that children were most often
carried with their right hand pressed against the body (Harris, 1980).
Despite this, right-hand preference is predominant.
The predominance of right handedness is evident from man's earli-
est recordings including cave drawings, Egyptian art work depicting
warriors with swords in their right hand (Dennis, 1958) and Old Test-
ment writings with reference to the use of the right hand for prayer.
Archaeological findings include a predominance of tools and weapons
designed for right-handed use (Harris, 1980).
These same sources provide early evidence that some left handed-
ness was observed. For example, the Old Testament refers to the 700
left-handed soldiers of Benjamin.
EstUates of the percentage of right and left handedness have
varied over the years from as low as 1 percent left handedness (Hasse &
Debner, 1914) to as much as 30 percent left handedness (Wile, 1934).
Coren and Porac (1977) maintain that incidence of right and left
handedness has remained constant for at least 5000 years. They
collected 1180 art works spanning 5000 years which included some
representation of a unimanual tool or weapon. In 92.6 percent of
the art work, right-hand use was depicted. This was constant across
all historical eras and geographical regions studied. They suggest
that this evidence supports a physiological or heritability compo-
nent in handedness.
One problem with using this type of historical evidence for the
incidence of handedness is that it is not known whether the artist's
rendering is an accurate estimate of right and left handedness at
that time. For example, even if the incidence of left handedness
rose as high as 30 percent, the majority of individuals would still
be right handed. The artist may be more likely to characterize his
figures as the majority rather than according to the actual percent-
ages. Therefore, although the Coren and Povac study gives some evi-
dence for a consistency in handedness, which suggests a physiological
basis for handedness, it does not rule out the influence of social
and cultural factors on the fluctuating percentage of right and left
The study of a number of cultures has revealed considerable vari-
ability in the incidence of left handedness. For example, among
Japanese children the incidence has been estimated at 11.54 percent
(Komai & Fukuoka, 1934), among Greek children 10.35 percent (Pelacanos,
1969) and among the Swedish 5 percent (Beckman & Elston, 1962).
Dawson (1972, 1974) studied various cultures and reported the follow-
ing incidences of left handedness: Temne-Arunta-3.4 percent, Chinese
in Hong Kong-l.5 percent, Australian aboriginese-10.5 percent and
Eskimos-ll.3 percent. Hardyck, Goldman and Petrinovitch (1975)
reported on the assessment of 7688 children in the United States and
found that 91.3 percent were right handed and 8.7 percent were left
In a recent study, Brackenridge (1981) calculated the variation
in handedness in Australia over a 90-year period. He found an in-
crease in left handedness from 2 percent to 13 percent during this
period. The increase was attributed to relaxation of cultural taboos
against left handedness.
Assessment of Handedness
An important issue in handedness research concerns the methods
used to assess an individual's preferred hand, One method involves
using an individual's self-reported hand preference. Though right-
handed subjects' self-report and behavioral performance are highly
correlated, this is not the case with left handers. It has been
reported that 12 14 percent of left handers are actually better on
motor tasks with their right hand and 23 28 percent are equally
skilled with both hands (Benton, Myers & Polder, 1962; Satz, Achenbach &
Fennell, 1967). A second method of assessing handedness directly
measures the subject's performance on unimanual tasks. A number of
the cross-cultural studies mentioned above used behavioral assessment.
A consistent problem with this method is the variable criteria used
in deciding how many left-handed behaviors constitute left handedness.
For example, in Pelecanos' (1969) study of Greek children, five
behavioral measures were administered twice. A left hander was de-
fined by at least 2 of 10 actions performed by the left hand. In
Komai and Fukuoka's (1934) study of Japanese children, those children
who used either hand were classified as left handed.
A final method of handedness assessment to be considered is the
use of a questionnaire. Bryden (1977) used a factor-analysis tech-
nique on two standard handedness questionnaires (the Crovitz-Tener
Test, 1962, and the Edinburg Handedness Inventory or Oldfield Test,
1971). In each case, he found a factor for what can be described as
handedness. There were also additional factors that he considered
"somewhat idiosyncratic" (p. 662). The reliability and validity of
handedness questionnaires are generally high (Coren & Povac, 1978),
although certain items (for example, placement of top hand on broom
when sweeping) have been questioned (Raczkouski, Kalat & Nebes, 1974).
Since much of today's research involves a large number of subjects,
the questionnaire method is most often used. The evidence supports
this method as an acceptable assessment technique.
Early Theories of Right and Left Handedness
Theories to explain the dominance of one hand over the other
date back to the time of Plato. Harris (1980) provides an excellent
review of the early explanations for handedness. A number of these
hypotheses focus on the contribution of body and brain asymmetries.
The sinus inversus totalis hypothesis is based on the asymmet-
rical position of the organs in the body. Due to the position of
the liver and other organs, the right side-of the body was judged to
be heavier, and, therefore, dominant. Left handedness was attributed
to a total reversal of the position of the internal organs (sinus
inversus totalis) resulting in a heavier and, therefore, dominant
left side. A second body asymmetry considered was the postulated
difference in the blood supply to the two sides of the body (subcla-
vian artery differences). Since the right subclavian artery is
approximately one inch closer to the heart than the left artery, it
was presumed that blood pressure on the right side of the body was
greater than on the left. The greater right pressure was associated
with right handedness. Left handers, it was believed, had a reversed
distribution of the blood vessels, such that the left subclavian
artery was closer to the heart, providing more blood to the left side
of the body. When these theories were empirically tested with post-
mortem examination of right and left handers, no support for either
sinus inversus totalis or subclavian artery differences in left
handers was found.
The study of brain structural-asymmetry differences between right
and left handers largely arose from Broca's (1861) findings on aphasia
following damage to the left third-frontal convolution. Though the
right area appeared anatomically larger, he reported that the left
area was more dense, (i.e., had greater number of gyri that may indi-
cate greater relative mass). Initially, Broca assumed that left handers
were the opposite of right handers with language areas in the right
hemisphere. However, following reports of aphasia in left handers
following left hemisphere damage, this position was abandoned.
In the mid-1800'sit was postulated that manual asymmetry was
due to greater blood supply to the motor area in the hemisphere contra-
lateral to the dominant hand. This hypothesis was based on the posi-
tion of the carotid arteries relative to the aortic arch. The right
carotid arises indirectly from the aortic arch (via the brachiocephalic
branch), whereas, the left carotid arises directly from the aortic
arch. It was assumed that this direct connection on the left provided
greater blood supply to the left hemisphere. It was postulated that
in left handers the position of the arteries was reversed, so that
the right, rather than the left, carotid arose directly from the aor-
tic arch, providing greater flow to the right hemisphere. However,
as early as 1898, it was observed that the anterior communicating
artery pooled blood supply and the distribution to each hemisphere
was equal. Also, there is no evidence that left handers have a dif-
ferent distribution of cerebral blood vessels than do right wanders.
In general, attempts to understand the predominance of right handed-
ness (and the occurrence~of some left handedness) from body and brain
structural and blood supply differences have been unable to demon-
strate any consistent differences between right and left wanders.
A final hypothesis to be considered is Carlyle's (1871) Warfare
Shield Theory. He proposed that warriors carried their shields in
the left hand in order to protect their heart, thus, leaving the
right hand for sword wielding. Use of the right hand for this pur-
pose reinforced right-handed manual superiority, The major problems
with this theory are its failure to (a) explain the fairly consistent
occurrence of left handedness despite shield carrying and (b) consider
that weapons (usually right handed) were used long before warriors
were equipped with shields (Harris, 1980).
From both cultural and psychological perspectives, left handed-
ness has been strongly associated with negative factors. Until
recently, stringent measures were taken to discourage the develop-
ment of left handedness in children showing early signs of left pre-
ference. Early psychoanalytic theories considered left handedness
to be a neurotic symptom characterized by a stubborn and willful per-
sonality often marked by sexual confusion (i.e., homosexuality)
(Harris, 1980). Left handedness has also been associated with cog-
nitive defects (Gordon, 1920), criminality (Lombrosop 1903), epilepsy
(Hicks & Kinsbourne, 1976a), stuttering (Moore, 1976) and autism
Localization of Function
Theories of localization of functions to specific brain areas
date back to the work of Aristotle and Galen. The ventricular chara-
bers were considered the important structures in intellectual and
mental functioning. The focus on the flow of fluid through the ven-
tricles continued for almost 1000 years. Descartes, in the late 17th
century, chose the pineal gland as the "seat of the soul" due to its
location relative to the ventricular chambers and its potential for
interaction with the fluids moving between the chambers (Walsh, 1978).
In the decades that followed, focus shifted from attempts to
identify one vital organ for mental functioning to attempts to delin-
eate the distribution of functions throughout the brain. Gall was
instrumental in elevating the cortex from relative unimportance to
the source of higher intellectual abilities (Walsh, 1978). Gall, in
this theory of phrenology, attempted to localize such functions as
self-esteem, benevolence and parental love in various cortical areas.
It was postulated that development of these areas resulted in protru-
sions in the skull, so that an experienced phrenologist, by examining
an :individual's skull, could divine that person's personality and
During the mid- to late-1800's, two major schools of thought
prevailed. On apparent opposite sides of the fence, the "localiza-
tionists" and the "antilocalizationists" attempted to explain how
the brain carried out mental functioning. The "localizationists,"
supported largely by the work of Broca (1861) and Wernicke (1874),
considered certain structures and areas within the brain to be neces-
sary for particular functions. Broca and Wernicke were specifically
involved in identifying areas within the left hemisphere that re-
sulted in aphasia following discrete cortical lesions. Most important,
according to Geschwind (1974), was Wernicke's application of the
brain's anatomy (i.e., fiber tracts) to aphasia, such that, one was
not only able to account for existing syndromes, but also was able
to predict the existence of aphasic syndromes that had not as yet been
documented. Wernicke proposed that the phrenologist's attempt to
localize complex mental functioning was incorrect. Rather, he per-
ceived the brain as a system of primary (perceptual and motor) and
association fiber pathways from which complex intellectual activity
The "antilocalizationist" position was strongly supported by
Flourens (Walsh, 1978). He proposed that mental functioning was
dependent on the whole brain and not localizable to specific brain
areas. This holistic approach was held by many eminent neurologists;
Marie, Von Monakow, Head and Goldstein. However, Geschwind (1974) in
his review of the position of the "antilocalizationists," and of
Goldstein in particular, concluded that their position was not all
that divergent from the localizationist theory. Basically, "antilo-
calizationists" acknowledged thatspecific defects in cognition
could occur from discrete focal lesions.
The nonlocalizationist position was further investigated by
Lashley (1929). Based on his research with rats, he proposed the
Principle of Mass Action. This principle states that performance on
a complex task is more dependent on the amount of cortex remaining
after surgical ablation than the location of the remaining cortex.
He suggested that complex memories are diffusely represented in the
brain; the more complex the memory is, the more diffuse the represen-
tation. Although ablations of cortical areas in rats may have sup-
ported his Principle of Mass Action, discrete cortical lesions in
humans result in generally predictable behavioral deficits.
Luria (1973) and his Russian colleagues managed to unify much of
the divergence of the "localizationist" and "nonlocalizationistl posi-
tions. His "functional systems" incorporated the importance of dis-
crete brain areas' participation in mental functioning within the
higher order interaction of many brain areas.
Handedness and Localization of Function
The cerebral organization of right and left handers has been an
area of controversy since Broca's (1865) initial reports of aphasia
following left hemisphere damage. Though Broca initially assumed
that left handers had language represented directly opposite to right
handers, the observation of crossed aphasia (.aphasia in left handers
following left hemisphere damage) necessitated a re-evaluation of
this position (Bramwell, 1889; Chescher, 1936). Later observations
indicated that there are real differences between right and left
handers, although not in the manner suggested by Broca.
In right handers the localization of functions to the two cere-
bral hemispheres has been extensively studied. The general conceptu-
alization is that the left hemisphere is specifically organized or
programmed for language processing, whereas, the right hemisphere's
abilities are related to visuo-spatial type processing. Naturally,
this is a very simplified description of each hemisphere's ability,
and much research has been conducted to identify the components of
the information processing capacities of each hemisphere. (For an
in-depth review of this area, see Bowers, 1978,)
There are two major methodologies used to determine the extent
of lateralization of function within the brain-damaged population:
1) sodium amytal and 2) estimates from aphasia. Sodium amytal injec-
tions are used as a presurgical diagnostic tool in order to determine
the hemisphere which is dominant for language. A neurosurgeon will
use this information during surgery and attempt to avoid excising
critical speech areas. The barbiturate is injected via the carotid
artery and selectively anesthetizes a hemisphere. It is possible to
observe those behaviors which are depressed while a particular hemi-
sphere is inactivated It has been reported that in right handers
a left-sided injection depresses language output (Branch, Milner &
Rasmussen, 1964) and a right-sided injection interferes with singing
ability (Gordon & Bogen, 1974). The results with left handers sug-
gest that approximately 15 percent of the left handers have bilateral
speech representation (Milner, 1973). However, great care must be
taken in applying these findings to a normal population, The amytal
procedure is an invasive and dangerous procedure that is only used
in medically justified situations. The patients tested are often
epileptics who have long-standing neurological problems (Levy, 1974).
The validity of using these results to understand localization in
normails is questionable. This problem will be discussed in more
A second method used to localize language function is based on
the incidence of aphasia following right- and left-hemisphere damage.
Satz (1979) has reviewed the literature on left wanders and language
localization and summarized the previous findings into three possible
models. The Variable Unilateral model proposes that left handers
have either right or left hemisphere, and the majority (66 percent)
have left hemisphere language. The Variable Unilateral and Bilateral
model estimates that 70 percent of the left handers have left-sided
language, 15 percent have right-sided language and 15 percent have
bilateral language representation. Finally, the Bilateral and Variable
Unilateral model proposes that 40 percent of the left handers have
left language, 20 percent have right language and 40 percent have bi-
lateral representation. Using the incidence of aphasia following
right- and left-hemisphere damage reported in 12 studies, Satz demon-
strated that the Bilateral and Variable Unilateral best fit the data
provided. This finding is consistent with other reports that left
handers 1) are more likely to become aphasic following damage to
either hemisphere and 2) are more likely to recover from aphasia
(Goodglass & Quadfasel, 1954; Subirana, 1953; Hecaen & Sauguet, 1971;
Hecaen, De Agostini & Monzon-Montes, 1981).
A final method for studying language representation is the admini-
stration of electro-convulsive therapy (ECT). Though these studies
use neurologically intact individuals, it is necessary to remember
that the reason for their receiving ECT is usually severe depression,
and this separates them from the truly normal subject. In an early
study, Cohen, Noblin, Silverman and Penick. (1968) assessed depressed
patients on verbal and nonverbal tasks pre- and post-ECT. They found
that left-sided ECT disrupted performance on verbal tasks and right-
sided ECT disrupted performance on nonverbal tasks. Bilateral ECT
resulted in disruption of performance on both tasks. Warrington and
Pratt (1973) observed disruption in language following right- and
left-sided ECT in left handers. In those psychiatric patients with
no demonstrable organic disease, 70 percent had language disrupted
with left ECT, 23 percent with right ECT and only 7 percent lost
language with both right and left ECT. This finding is not consistent
with Satz's prediction of 40 percent left, 20 percent right and 40
percent bilateral speech. The divergence of the methodologies to
assess language localization, however, makes comparisons difficult.
Controversy still remains concerning the actual percentages of left
handers with right, left and bilateral language representation.
Research with normal populations
Attempts to explain normal brain function with clinical data are
helpful but fraught with problems. First, many of the patients
studied have had long-term neurological problems (e.g., epileptic
patients). It is not known to what extent these patients' brains
have undergone reorganization in order to compensate for the neuro-
logical damage. Consequently, comparisons of neural and functional
organization between patients and normal, intact individuals are
questionable. Second, when an area of the brain is damaged and no longer
functioning, it is not possible to study what was the function of
that area. Rather, it is only possible to study how the remaining
intact brain is compensating for the loss of a particular area.
Therefore, since the function of a particular area can not be direct-
ly determined in a brain-damaged patient, extrapolations about func-
tion to a normal, intact individual must be made cautiously. Finally,
the occurrence of recovery of function must be considered. Following
brain damage, some functions will recover to various degrees. It is
possible that other areas close to the damaged area can take over the
functions. It is also possible that homologous areas in the nondamaged
hemisphere can provide some of the functions previously executed by
the damaged hemisphere. Regardless of which is in fact the case, when
patients are tested at varying points posttrauma, it is not possible
to determine whether intact behaviors were never affected by the
damage, or whether they have already been compensated for, and func-
tion has been recovered.
A number of lateralization assessment techniques have been
developed over the past 20 25 years to study hemispheric asymmetries
in normals. The Pathway Transmission Model (Kimura, 1961) is gener-
ally used to account for the findings with these techniques. The two
major postulates of this model are (a) the hemispheres differentially
process verbal and nonverbal material and (b) the contralateral sensory-
input and motor-output hemispheric pathways are more direct than the
ipsilateral pathways for the visual, auditory and the tactile modal-
ity. Therefore, for example, if verbal material, which is differen-
tially processed by the left hemisphere, is presented auditorily to
both ears, then the information presented to the right ear will be
conveyed more directly via the contralateral pathways to the left hemi-
sphere. The ipsilateral verbal input (left ear right hemisphere)
must make an additional crossing of the corpus callosum (right
hemisphere left hemisphere) and is more indirect. Therefore, recog-
nition of the material presented to the right ear should be superior.
The following sections deal separately with each of the major tech-
niques used and the findings with right and left handers.
Visual half field (VHF). The VHF technique allows for channeling
visual input directly into a single hemisphere. For example, a stimu-
lus presented to the left visual field (LVF) is passed via ipsilateral
(right temporal) and contralateral (left nasal) pathways to the right
hemisphere. The stimulus must be presented for less than 150 200
milliseconds to prevent time for eye movements to occur and to insure
presentation to only one visual field. Generally, this research is
conducted using a tachistoscope which allows for very short presenta-
The early research using the VHF technique found that word recog-
nition was better when the stimuli were presented to the RVF than to
the LVF (Miskin & Forgays, 1952; Heron, 1957). This finding was attri-
buted to a postexposure (i.e., after the stimulus was removed) direc-
tion of scan consistent with left-to-right reading strategies. Other
investigators preferred to view the RVF effect as evidence for cere-
bral dominance and the Pathway Transmission model. That is, the words
were better processed by the left hemisphere, and the contralateral,
RVF left hemisphere connections were more direct than the ipsilateral,
LVF left hemisphere connections. Support for this position came
from comparing VF effects when English and Hebrew words were presented
vertically. Barton, Goodglass and Shai (1965) removed the effect of
reading strategies with vertical presentation. Both American and
Israeli subjects showed a significant RVF effect (Israelis with English
and Hebrew; the Americans with English only). Therefore, despite the
fact that Hebrew is normally read from right to left, when presented
vertically, a RVF effect is evident. This supports the Pathway Trans-
mission model. Bryden (1965) reasoned that the scanning necessary
for recognition of a whole word would not be necessary with single
letters. Therefore, if the RVF effect with words was due to scanning,
no VF effect should be evident with single letters. However, contrary
to the prediction from the scanning hypothesis, there was a RVF effect
for single letters. This supports the cerbral-dominance explanation
for the RVF effect.
Bryden (1973) compared right-and left-hander's tachistoscopic-
recognition of letters, forms (nonverbal task) and dot localization
(Kimura, 1969). Although Kimura had found a significant LVF on dot
localization with right handers, in Bryden's study neither handedness
group showed a significant field effect. There was also no signifi-
cant field effect on the forms test. Only in the letter recognition
condition did both left and right handers show a significant RVF
effect. The RVF superiority was significantly greater for the right
handers. In other words, left wanders showed less asymmetry in VF
recognition (suggesting more bilateral representation of language).
Zurif and Bryden (1969) also reported that on tachistoscopic letter-
recognition right handers demonstrate a significant RVF effect, where-
as, left handers showed no right-left asymmetry.
The VHF technique has also been used to determine if spatial
abilities are lateralized differently in right and left wanders.
Using a facial recognition task (generally considered to be a right-
hemisphere task), 'Moscovitch, Scullion, and Christie (1976) reported
that right handers showed greater recognition for faces presented
tachistoscopically to the LVF than to the RVF. This suggests superior
right-hemisphere processing for faces in right handers. The left
handers showed no consistent visual field asymmetry, suggesting a
more bilateral representation for processing faces in left handers.
It has also been reported that facial emotions (emotionality being
considered a right-hemisphere function) are better recognized by
right handers in the LVF than the RVF, whereas, left handers show
no consistent group asymmetries (Heller & Levy, 1981).
Dichotic listening (DL). In this procedure, different auditory
stimuli are presented simultaneously to both ears. Kimura's (1961)
early dichotic research with patient groups revealed that verbal
stimuli (numbers) presented to the ear opposite the hemisphere domi-
nant for language were more efficiently recognized than verbal stimuli
presented to the ear ipsilateral to the dominant hemisphere. Zurif
and Bryden (1969) demonstrated a significant right-ear asymmetry (REA)
with normal subjects with both free and ordered recall. Others
report reduced ear differences in left handers compared to right
handers (Satz et al., 1965; Fennell et al., 1978; Curry, 1967;
Curry & Rutherford, 1967).
Before continuing the discussion of neuropsychological techniques,
it is necessary to briefly consider the reliability and validity of
the VHF and dichotic methods. Since both are considered to demonstrate
language localization, a high degree of correlation between them would
be expected. The research findings are equivocal. Bryden (.1966) and
Zurif and Bryden (1969) did not find a significant correlation between
VHF and DL measures. Hines and Satz (1974), however, found a signifi-
cant VHF-DL correlation for right handers but not for left handers.
In a recent study, Fennell, Bowers and Satz (1977a) tested right
wanders on a VHF and DL task in four daily sessions. Although they
did not find a significant correlation between the measures, the
"directional concordance rates between ear and visual half-field
scores rose dramatically from 25 percent between Days 1 and 2 to
31.3 percent between Days 3 and 4" (p. 69). This study also demon-
strated a high degree of reliability over days of the dichotic task.
The VHF reliability was significant only after the first session.
Fennell, Bowers and Satz (1977b) repeated this experimental pro-
cedure with a group of left-handed subjects. The left handers showed
a high degree of reliability on the DL task and reliability on the
VHF task after the first session. There were no significant cross-
modal (DL-VHF) correlations. However, the concordance rates were
not as consistent as was demonstrated with right handers. The per-
formance of the left handers was more variable over days than the
right handers' performance.
Satz (1977) has questioned the validity of using these measures
to categorize an individual as either r or left-brain dominant
for language. In the normal right-handed population, it is estimated
that at least 95 percent of the individuals are left-brain dominant
for language. The DL procedures generally report that 70 percent of
the subjects show a REA, while 30 percent show an LEA. Thus, while
only 5 percent of the subjects are expected to have right-brain
language dominance, 30 percent show anLEA. Using a Baysean analysis,
the author demonstrated that the probability of misclassifying a
subject into the right-brain dominant group based on anLEA is 90
percent. The author recommends that this error can be minimized by
"construing this phenomenon (LEA in many right handers) as an example
of less complete lateralization of speech in the left hemisphere"
(p. 221). In sum, although these techniques are generally reliable
and relatively simple noninvasive procedures for assessing laterali-
zation, there remain questions regarding their validity.
Lateral eye movements (LEM). Kinsbourne (.1970) proposed that
when a cerebral hemisphere was activated during a cognitive task, the
eyes should be driven conjugately away from the activated hemisphere.
He found that during left-hemisphere verbal processing, right-handers'
eyes deviated to the right. Conversely, during right-hemisphere
spatial processing, the eyes deviated to the left (Kinsbourne, 1972).
Verbal and spatial processing was induced by a series of questions.
For example, the verbal questions included explanation of proverbs
(e.g., Explain "Where there is a will, there is a way.") and defini-
tions (e.g., What word is a definition of a "large mammal that lives
in oceans?"). The spatial questions included geographic directions
(e.g., "Where is Saskatchewan relative to Manitoba?") and nonverbal
visualizations (e.g., "Visualize a man with a tuxedo and a top hat,").
Later studies both supported (Cur & Harris, 1975; Kocel, Galin &
Ornstein, 1974) and failed to support (Erlichmann, Weiner & Baker,
1974; Tankle, 1976) this methodology for measuring hemispheric acti-
vation. More recently, Gur and Gur (1980) compared conjugate eye
movements of right and left handers while answering the verbal and
spatial questions devised by Kinsbourne. They reported that right
handers moved their eyes to the right with verbal questions and to
the left with spatial questions. This finding replicated Kinsbourne's
(1972) original report. The left wanders, however, showed no consistent
movements in either direction with either the verbal or spatial ques-
tions. The authors propose this as evidence for more bilateral repre-
sentation of function in left handers.
Competition tasks. Kinsbourne and Cooke (1971) proposed that if
two behaviors share the same brain area for execution (shared functional
space) then a task which sets these behaviors in competition should
result in a depression of one or both of the behaviors. The authors
reported that,when subjects were required to simultaneously balance a
dowel and speak, there was a significant decrease in balancing with the
right (but not the left) hand. This is consistent with the competition
hypothesis, since the motor area for the right hand and the speech area
are both in the left hemisphere and relatively close to each other.
Bowers et al. (1977) replicated this finding with a variety of verbal
tasks (verbal fluency, reading aloud, reading silently) and finger tap-
ping. However, they found that a nonverbal task (facial learning)
depressed finger tapping bilaterally and not just the left-hand tapping
as would be predicted from this model.
Hicks, Provenzano and Rybstein (1975) tested left-handed subjects
on a dowel balancing and verbal output task. In contrast to the asym-
metrical decrements obtained with right handers, the left handers showed
a bilateral decrease in dowel balancing during verbal-production. This
finding provides further support to the position that left handers do
not have language as strongly left-hemisphere lateralized as right
Tactile tasks. The visuo-spatial superiority of the right hemi-
sphere has been studied using dichaptic tasks, in which blindfolded
subjects are required to feel objects, shapes or forms presented to
either (unimanual) or both (bimanual) the right and left hands. The
rationale for this procedure is similar to the dichotic listening
method. That is, if two objects are simultaneously manipulated, the
more direct left-hand to right-hemisphere pathways will result in a
recognition advantage for the left hand.
Right-hemisphere superiority has been demonstrated for processing
of geometric figures in commissurotomized patients (Franco & Sperry,
1977), tactual exploration-of random shapes in different orientations
in normal subjects (Dodds, 1978) and Braille reading in blind individu-
als (Hermelin & O'Connor, 1971). Witelson (1974) compared the perfor-
mance of right-handed children's right and left hands on a dichaptic
nonsense-shape task. She found a left-hand recognition superiority
when the left hand (but not the right hand) was used to point out the
correct shape from a visual array. Nilsson, Glencross, and Geffen
(1980) repeated Witelson's study with adults. They did find a signifi-
cant left-hand effect. However, in contrast to Witelson's finding, the
difference was significant only with right-hand identification of the
visual target. In a second experiment with right and left handers, they
used an ordered recall. That is, subjects received a pretrial cue which
indicated which hand was to be identified first. The right handers
showed no hand asymmetry. The left handers, however, showed a signifi-
cant effect for order of report. The hand reported first showed greater
recognition. The implication of this finding for understanding differ-
ences in cerebral organization of right and left handers is unclear.
However, it does indicate that this method, as the others, does not.pro-
vide a clear or even consistent picture of tactile asymmetries.
Generalized Cognitive Defect
The higher incidence of left handedness among epileptic and retard-
ed groups has suggested to some investigators that left handers as a
group may suffer from impaired cognition. Levy (1969) proposed that
bilateral representation of speech in left handers will result in their
showing inferior performance on tasks requiring spatial processing when
compared to right handers. She attempted to demonstrate this using the
Wechsler Adult Intelligence Scale (WAIS). A small number of right and
left handers were tested on the Verbal and Performance scales. The
results indicated that, although there wasno significant difference be-
tween the groups on the Verbal scale, there was a significant difference
between them on the Performance scale. The left handers reportedly did
poorer than the right wanders on the Performance scale.
The basis of Levy's proposed visuo-spatial deficit in left handers
is an evolutionary perspective. That is, she suggests that as language
functions evolved they came into conflict with perception (presumably
the major function of both hemispheres before language). The dominance
of language functioning resulted in lateralization of visuo-spatial pro-
cesses to the mute right hemisphere. This idea, according to Levy,
"suggest(s) that people with total language centres or even with partial
language competency in both hemispheres would perform relatively poorer
in tasks of perceptual function" (p. 615). The conceptual leap from
bilateral speech representation to poorer visuo-spatial abilities is
unwarranted. It is equally likely that bilateral speech would result
in increased visuo-spatial abilities (due to bilateral spatial abilities)
or a decrease in verbal ability (assuming bilaterality confers a disad-
A major methodological flaw in this study was the use of the WAIS
Performance scale as a measure of spatial abilities. This scale con-
sists of a variety of perceptual-type tests, each of which may require
a different type of processing. The WAIS Performance IQ score is not
generally considered to be a pure indicator of right-hemisphere pro-
cessing ability. Other problems with this study include 1) relatively
small sample size (10 left and 15 right wanders), 2) the subjects
(graduate science students) may not be representative of the general
population of right and left handers and 3) there is no mention of how
handedness was determined. Although this study was replicated by Miller
(1971), other attempted replications with larger samples failed to find
differences between right and left handers on the WAIS (Hardych et al.,
1976; Newcombe &Ratcliff, 1973; Briggs et al., 1980).
Fennell, Satz, Van den Abell, Bowers and Thomas (.1978) reasoned
that, if left handers had poorer spatial-processing abilities than right
handers, this deficit should be apparent on Gestalt-type spatial tasks
such as the Block Design (subtest from the WAIS-Performance) and the
visual-spatial subtest from Thurstone's Primary Mental Abilities (PMA)
battery. They tested a sample of high school students and college stu-
dents. There was no significant difference between right and left
handers on these measures of visuo-spatial abilities in either sample.
Partial suDport was found for the hypothesis that left handers have more
bilateral representation of language, whereas right handers have
language strongly left-hemisphere lateralized. Using a dichotic listen-
ing paradigm, they reported that only in the college sample did the
left handers have less clear left-hemisphere language dominance.
Recently, Bradshaw, Nettleton and Taylor (1981) tested a largenum-
ber of subjects (96) in order to determine the contribution of family
history on WAIS verbal and nonverbal tachistoscopic performance. Their
results suggested that 1) left handers, particularly familial, had sig-
nificantly lower WAIS performance scores, 2) tachistoscopic asymmetries
were weakest in nonfamilial left handers and 3) there was no difference
in degree of lateralization between familial and nonfamilial right
These results are consistent with Levy's (1969) finding that left
handers have lower Performance IQ's. However, further analysis indi-
cated that there was no relationship between low Performance IQ and
extent of right-hemisphere language (determined by tachistoscopically
produced asymmetries). That is, according to Levy's hypothesis, those
subjects whose right-hemisphere language reduces spatial capabilities
should perform more poorly on the Performance IQ scale. The data are
not consistent with this hypothesis. No new explanation for left handers'
lower Performance IQ is given.
There is evidence that left handers are actually better than right
handers on some spatial tasks. Herrmann and van Dyke (1978) presented
two nonsense-stimulus patterns to right and left handers. The patterns
were at 0', 450, 90', or 1350 rotations. The task was to decide if the
two patterns were the same or different regardless of any changes in
rotations. The subject's reaction time to respond was recorded. The
results revealed that overall the left handers were significantly faster
than the right handers. Also, as the orientation discrepancy between
the two patterns increased both groups took longer. However, this dis-
crepancy affected the right handers (longer RT) significantly more than
the left handers. Therefore, on this test of spatial capacity, left
handers are not inferior to right handers as Levy's model would predict.
In fact, left handers actually performed better than right handers.
The right hemisphere's apparent involvement in spatial processing
has suggested to some investigators that left handers may actually have
more direct access to this hemisphere's abilities Peterson and Lansky
(1974) found that the incidence of left wanders among the faculty and
students in architecture was higher than would be predicted from the
normal population (29.4.percent and 18 percent respectively). Further,
they found that in maze solving (mazes of the students' own design) the
right handers made more errors than the left handers. A higher inci-
dence of left handers than would be expected has also been noted among
art students (Mebert & Michel, 1980) and musicians (Byrne, 1974).
Familial vs. Nonfamilial Left Handedness
The research on the influence of familial left handedness on right
and left handers has been reviewed by Bradshaw (1980). Generally, the
research in this area has yielded equivocal findings. There are reports
of more bilateral representation in familial left handers (Hines & Satz,
1971; Zurif & Bryden, 1969) and findings of more bilaterality among non-
familial left handers (Higginbottam, 1973; McKeever & Van Deventer,
1977a, 1977b). There is also evidence that right wanders with close
relatives who are left handed may be lateralized differently than right
handers with no left-handed relatives. That is, right handers with a
close left-handed relative may show more bilaterality (similar to left
handers) than right handers with no close left-handed relatives (Hines &
Satz, 197]; McKeever, Van Deventer & Suberi, 1973; Springer & Searleman,
1978). Failure to replicate this finding has been reported by Hines
and Satz (1974) and Briggs and Nebes (1976). Comparisons of right
handers with and without left-handed relatives with aphasia indicates
that those with left-handed relatives show better recovery of function
(Hecaen & Sauguet, 1971; Luria, 1970).
Genetic Models of Handedness
The genetic approach to handedness attempts to explain the relative
frequencies of right and left handedness with genes determining which
hand will be dominant. An early estimate of the probablities of left-
handed offspring indicated that the probabilities are 0.02 for two right-
handed parents, 0.17 for one left-handed parent and 0.46 if both parents
are left handed (Chamberlain, 1928). These data support a genetic model.
However, the environmental contibuting factors must also be considered.
Left-handed parents may design and reinforce a left-handed environment.
Therefore, the nature and nurture factors cannot be totally separated
(Springer & Deutch, 1981).
A model based on the Mendellian principles of genetics postulates
that right handedness is determined by a dominant gene and left handed-
ness by a recessive gene. According to this model, two left-handed
parents should produce 50 percent right and 50 percent left-handed chil-
dren. This prediction is not consistent with the observation that two
left-handed parents produce slightly more right-handed children (proba-
bility estimated at 0.54).
Annett (1964) has proposed that there is a single allele for right
handedness ("right shift") but not left handedness. An individual with-
out a handedness gene has the potential to be either right handed, left
handed, or quite likely ambidextrous. In these cases, handedness will
be determined by random environmental influences. A second source of
left handedness, according to Annett, is birth trauma which can shift
a predisposed right hander to left handedness. Annett found that this
model best explained the data collected by Chamberlain (1928) on the
incidence of left handedness in a large sample.
Morgan and Corballis (1978) expand Annett's model to include the
association between handedness and cerebral lateralization, Since they
prefer a nongenetic explanation for handedness and lateralization, they
refer to the direction handedness or lateralization takes as a "shift."
If an individual has the "shift," then handedness (right) and brained-
ness (left) are codetermined. If,however, there is no right shift, then
"handedness and cerebral lateralization are-established independently
and at random" (p. 275). The authors propose that 12 percent of the
population lack the right shift and approximately 10 percent of the
population are left handed. Since handedness is randomly determined
without the "shift," 6 percent of left handers lack the right shift.
Half of the 6 percent should be right brained for speech, and the
other 3 percent should be left brained for speech. They propose that
about 4 percent of the population are left handed due to a pathological
state. According to Satz (1972), brain trauma is more likely to result
in a shift in handedness than a shift in language dominance. Therefore,
this 4 percent of pathological left handers can be classified as left
brained for speech. These figures are consistent with some reports
from amytal (Milner, 1973) and ECT (Warrington & Pratt, 1973) procedures.
In addition, this model may be useful in understanding those left
handers who have bilateral representation of language. A percentage
of those with no right shift may be more diffusely lateralized for both
handedness and speech.
Levy and Nagylaki (1972) have proposed a genetic model which in-
cludes no random environmental influences. They suggest there are four
alleles at two loci. One pair of alleles was postulated to determine
which hemisphere is language dominant and which is the preferred hand.
The second pair of alleles was expected to determine if the language-
dominant hemisphere controls the ipsilateral or contralateral hand.
These authors applied their model to Rife's (1940) family data rather
than to Chamberlain's (1928). While they consider Annett's use of
Chamberlain's data inappropriate due to inconsistencies in his data,
Annett has stated that Levy and Nagylaki are exaggerating the problems
with Chamberlain's data. The controversy continues. Furthermore,
Annett (1978) criticizes Levy and Nagylaki's model on the grounds that
"it requires four independent gene mutations, (and) it treats handed-
ness as a right or left dichotomy without provisions for mixed handers"
Morgan and Corballis (1978) consider two problems in the Levy and
Nagylaki model. First, motor control is generally considered to be
contralateral, and there is no conclusive evidence for ipsilateral con-
trol. Second, they object to division of handedness genes into a right-
The issue of whether there are actually genes which determine di-
rection of asymmetry is unresolved. Morgan (1977) points out that in-
heritance is determined not only by genes, but also from the cell (oocyte)
with its highly complex spatial structure. He states, "there is evidence
that the spatial information carried in this cell is vital in determining
the form of the developing embryo." He argues that brain and hand asym-
metries are probably oocytic and not due to different genes. The basis
for this argument comes from the consideration of the development of a
number of species (starfish, frog, chickens) and the tendancy for the
left-right asymmetries in formation. Also, Morgan contends there is no
evidence that direction of asymmetry is coded in the DNA of any organism.
Levy (1977) disagrees with this point and cites as examples the coiling
direction of the snail, Limnaea peragra, which is determined by a single
gene and where a right coil is dominant, and the rotational direction of
the abdomen of the fruit fly.
In summary, according to Morgan, there is no evidence for genetically
determined directionality. Therefore, human handedness is probably not
determined by gene differences, but rather by oocytic mechanisms. On the
other hand, Levy finds evidence supporting directionality in genes and
uses this as a basis for a model to explain human handedness.
In an attempt to delineate the genetic and environmental factors in-
fluencing behavior, research has often turned to the study of twins.
Monozygotic twins share identical genetic makeup. Therefore, differences
they exhibit are predicted to be due only to environmental variability.
Dizygotic twins are as genetically alike as any nontwin siblings. How-
ever, they share the intrauterine environment and possibly similar post-
natal environments. Levy and Nagylaki (1972) studied the incidence of
left handedness among mono- and dizygotic twins and found it to be sig-
nificantly higher than in singletons. An alternative to the twin method
of studying the genetic-environmental question is the method of compar-
ing offspring with biological parents and with step- or adoptive parents.
Hicks and Kinsbourne (1976b) have used this method to compare the
within-family handedness of 1252 students. In this group, 1101 had two
biological parents, 108 had one biological and one stepparent and 43 had
only one biological parent. The correlations revealed a significant re-
lationship between the handedness of the biological parent and the off-
spring, but no such relationship between the offspring and the stepparent.
This suggests that the handedness of the offspirng is determined by gene-
tic factors. Morgan and Corballis (1978), point out that the average
years these subjects spent with the stepparent was 7.24 years (standard
deviation=3.12) and their average age was 20.18 years (standard deviation
=2.47). It is likely that they were not influenced by the handedness of
the stepparent during the first years of life, which may be critical in
the determination of handedness (Gessell & Ames, 1947). Therefore, ac-
cording to Morgan and Corballis, this cross-fostering study provides no
evidence for the genetic determinants of handedness.
The study of neuroanatomical asymmetries has provided some interest-
ing findings on the brain structure of right and left wanders. These
studies have reported findings that are consistent with the behavioral
In a review of the early studies of neuroanatomical asymmetries,
von Bonin (1962) cited studies which compared the weight of each hemi-
sphere (presumably right handers' brains). The results were equivocal,
with some studies finding that the left hemisphere was heavier, while
others reported that the right hemisphere was heavier. This line of
inquiry is no longer pursued. Rather, investigators are now looking
for asymmetries of certain structures or areas within the cortex.
Geschwind and Levitsky (1968) measured the planum temporale (part
of Wernicke's area) in a sample of 100 human brains that had come to
autopsy. They found that the planum temporale was larger on the left
side in 65 percent of the brains and larger on the right in 11 percent
of the brains. These findings have been replicated with adult and in-
fant brains (Wada, Clark & Hamm, 1975; Wittelson & Pallie, 1973).
The comparisons of the neuroanatomical asymmetries between right
and left handers have shown that these groups do differ in the extent
of neuroanatomical asymmetries. The results indicate that right hand-
ers' brains show significant asymmetries in the parietal operculum
(LeMay & Culebras, 1972; Hochberg & LeMay, 1975), the frontal lobe
(LeMay, 1977) and the occipital lobe (LeMay, 1977). Left handers'
brains generally did not show any asymmetries.
The functional significance of these neuroanatomical findings is
purely speculative. The increased size of an area does not directly
imply greater functional capacity. Furthermore, it has been reported
that Broca's area is actually smaller in the left hemisphere than the
right hemisphere (LeMay, 1976; Wada et al., 1975). It is, however, pos-
sible that the asymmetries seen in the right handers are related to
right handers' lateralization of function. The lack of right-left asym-
metries in left handers may reflect a less lateralized system.
Predicting Localization of Function from Handedness
An important focus of much of the handedness research is the at-
tempt to predict how cortical functions are lateralized within different
handedness groups. In the following sections, a number of these ap-
proaches will be considered.
Motor and Language Asymmetry
Making predictions regarding language lateralization in right hand-
ers is quite reliable. Almost all right handers have language localized
in the left hemisphere. Liepmann (1908) proposed that the left hemi-
sphere of right handers also contained motor engrams. These engrams are
the neural representations of skilled movements. He suggested that left-
hemisphere lesions disconnected the engrams from the primary motor area,
rendering the patient apraxic (unable to perform skilled movements).
Support for left hemisphere motor engrams comes largely from lesion stud-
ies. If, however, normal subjects have left motor engrams, it should be
reflected in within-subject right and left hand differences on acquisi-
tion of motor skills. This has been demonstrated by Hicks (1974) with
letters and Taylor and Heilman (1977) with button pressing in sequence.
Heilman (1979) summarized the studies by stating:
(B)oth studies showed not only that the right hand learns
faster than the left, but also that, once a skill is ac-
quired, the right hand retrieves this engram better than does
the left hand (p. 453).
Although language function and the motor engrams are usually both lo-
cated in the left hemisphere of a right hander, they can be dissociated
into the separate hemispheres (Heilman, Gonyea & Geschwind, 1974).
In left handers, the localization of motor engrams is less clear.
Heilman (1979) observed that left handers are rarely apraxic. Also, if
apraxia is seen following right hemisphere damage (which is generally
an infrequent combination), the patient is usually left handed (Hecaen &
Sauguet, 1971). Heilman, Coyle, Gonyea and Geschwind (1973) have pre-
sented a case report of a left-handed patient who apparently had lan-
guage functions localized to the left hemisphere and the motor engrams
localized to the right hemisphere.
This literature is generally consistent with the hypothesis that
left handers are likely to have functions localized differently from
right handers. The motor engrams for skilled movements in left handers
are more likely to be in the right hemisphere or bilaterally represent-
ed, while the right handers are more likely to show left-hemisphere en-
grams. However, with the left handers, it is difficult to make predic-
tions about lateralization of either motor engrams or language laterali-
zation from knowing about handedness alone.
Handposture and Lateralization of Function
Levy and Reid (1976) proposed a novel, and they claim, valid method
to predict lateralization of function in left handers. It involves sim-
ple observation of the handposture used by left handers while writing.
Left wanders generally adopt one of two possible handpostures. The
noninverted position is the same posture used by virtually all right
wanders. The hand is positioned below the writing line and the pen
faces the top of the page. The inverted (or hooked) posture is charac-
terized by a bent wrist, with the hand above the writing line and the
pen facing toward the bottom of the page. Most, but not all, left hand-
ers can be identified as inverters or noninverters. There are some left
handers who have devised some unusual handpostures that are both unique
Levy and Reid (1976) compared inverted and noninverted left handers
on recognition of verbal (syllables) and spatial (dot localization) stim-
uli presented tachistoscopically. Noninverters demonstrated a left vis-
ual field superiority with verbal stimuli, while inverters showed a right
visual field superiority for verbal stimuli. They concluded that invert-
ed writers have language localized in the left hemisphere, whereas, non-
inverted writers have language localized in the right hemisphere.
McKeever (1980) contends that methodologicalflaws are apparent in
the Levy and Reid study. Their method of subject classification was
deemed "highly questionable" (p. 441).
Ss were classified for both visual and spatial dominance
by the higher half-field superiority magnitude obtained
on one of the two tasks. Thus, if a S was RVF superior
on both tasks, but his verbal RVF superiority was greater
than his spatial RVF superiority, he was classified as
left hemisphere dominant for language and right hemisphere
dominant for spatial processes. (p. 441)
Other researchers have also reported limited support for the Levy
and Reid position. Smith and Moscovitch (1979) and Herron (1980) have
found differences only with presentation to the visual system, but no
differences with dichotic presentations of language. They suggest that
the differences between inverters and noninverters may be some anomoly
in the visual system that is unrelated to language localization.
Beaumont and McCarthy (1981) used the dichotic listening procedure
with a large sample of right and left handers (70 subjects in each
group). They found that with verbal stimuli (digits) right handers
showed the expected right ear superiority. Left handers as a group
showed no right ear advantage. Analysis for handposture revealed no
difference between left-handed inverters and noninverters. Interesting-
ly, the performance of those right handers who used an inverted posture
(n=10) resembled that of the left handers (no asymmetry) more than the
noninverted right handers (REA). The authors conclude:
It would appear that whatever the validity of the Levy
and Reid hypothesis concerning writing posture and the
divided visual field advantage, it cannot be extended
into the auditory modality (p. 471)
There is, however, some question of the validity of the model with-
in the visual modality. Levy and Reid contended that those noninverted
left handers with contralateral motor pathway and right-hemisphere lan-
guage had left hemisphere specialization for spatial processing. Using
a dot localization task, they obtained results supporting this hypothe-
sis. Lawson (1978) used a tachistoscopically presented facial recogni-
tion task with right and left handers. Overall, the right handers show-
ed a LVF superiority, while the left handers had no visual field prefer-
ence. There was no difference between inverted and noninverted writers.
However, there was a borderline significant (p=.052) Sex X Hand X Hand-
posture interaction. The left inverted males did show a LVF superiority.
Furthermore, the right inverted males were more likely to show no visual
field preference. These findings are consistent with Levy and Reid's
prediction. However, among females, this relationship was reversed.
The reason for this Sex X Hand X Handposture difference is unclear.
The author suggests that "matters are more complicated than Levy pro-
poses" (p. 211).
Degree of Handedness and Lateralization of Function
In much of the early handedness research, an individual was class-
ified right or left handed on the basis of writing hand alone. It has
become increasingly clear, however, that manual preference may be char-
acterized by many activities and is not simply represented by writing
hand. Thus, many individuals will write with one hand and perform many
other activities with the other hand. Therefore$ as was discussed in
a previous section, most current handedness research measures the
strength or degree of a subject's hand preference by either question-
naire or actual performance measures.
There is some evidence that strength of handedness is a discrimi-
nating factor between groups on at least spatial type tasks. Nebes
(1971b) used the Arc-Circle (Part-Whole) test to compare right and left
handers. The subject is required to tactually explore a part of a cir-
cle and match it to one of three whole circles (of different sizes)
visually available. An earlier study by Nebes'(1971a)with commisuroto-
mized patients indicated the left hand performed this task better than
the right, suggesting that the right hemisphere was more directly in-
volved in this tactuo-spatial task. With normal right and left handers,
Nebes found that the performance of left handers was poorer than that
of right handers, regardless of the hand used to execute the task.
There was no difference between hands in either group. He attributed
this finding to an inefficient right-hemisphere organization among left
handers. Two replications (Hardyck, 1977; Kutas, McCarthy & Donchin,
1976) failed to find this difference. Nebes' findings were considered
to be a sampling error.
In a recent replication of Nebes' study, Thomas and Campos (1978)
considered the issue of degree of handedness and performance on spatial
tasks. They noted that, whereas Nebes used only self report for deter-
mining handedness, the replication studies were very careful in choosing
only strongly left- and right-handed subjects as determined by assessing
performance on 13 unimanual tasks. These authors propose the Nebes'
"sampling error" was in choosing a number of subjects who were not
strongly left handed. They repeated, essentially, Nebes' experimental
procedure and included right and left handers who showed a strong hand
preference and those who were "mixed." Their results supported the de-
gree of handedness proposal; both strongly right and left handers per-
formed better than mixed right or left handers.
The relationship between strength of handedness and lateralization
of functions has been studied using the VHF and DL techniques with vary-
ing results. Bryden (1973) found that with VHF presentation of letters,
significantly more right handers showed a RVF effect than left handers.
Further, the strong left handers tended to show better LVF recognition,
whereas the performance of weaker left handers fell in between the
strong right and left handers. Gilbert (1977), however, found that when
verbal (letters) and nonverbal (faces) stimuli were presented to the
right and left visual fields, there was no difference between strong and
weak right and left handers.
Knox and Boon (1970) used Kimura's (1961) digit dichotic task in
its original form and with modifications to make recognitions more dif-
ficult (white noise andinterruptions). They stringently assessed hand
and side preference using behavioral measures with 80 percent concor-
dance of sidedness as criteria. While right handers showed a consistent
REA, the performance of the left handers was related to task difficulty.
With the simpler task, they showed no ear asymmetry. However, in the
difficult condition, they showed a significant LEA. Dee (1971) reports
very different findings. He compared strong right and left handers and
moderate left handers on a verbal (words) and nonverbal (melodies) DL
task. On the verbal task, he found that both strong right and left hand-
ers had a REA (83 percent and 81 percent, respectively). The moderate
left handers showed no group significant ear asymmetries (31 percent had
a REA). All groups showed an LEA for the melodies, and there were no
The reasons for these equivocal findings are unknown. However, it
may be due to 1) differences in the dichotic tasks or 2) differences in
the criteria for handedness groups. Which method is appropriate for
assessing the relationship between strength of handedness and laterali-
zation of function remains undetermined.
In general, the Pathway Transmission Model, previously described,
has been accepted as the mechanism underlying the findings in dichotic
and visual field studies. However, this explanation has been questioned.
Morais and Bertelson (1973) found that when spatial cues are added to a
dichotic task (e.g., perceived versus actual source of input) ear asym-
metry was modified. They considered ear asymmetry to be the result of
differential distribution of "spatial attention" rather than stronger
Recently, Warrington and Pratt (1981) have applied this concept to
understanding the differences between right and left handers on dichotic
tasks. They suggest that the consistency of sidedness (right or left)
rather than brainedness (language laterality) is a better predictor of
ear asymmetry on a dichotic listening task (digits). They tested de-
pressed patients receiving ECT. Consistency of sidedness was determined
by 1) handedness on the Oldfield Questionnaire and 2) the ear preference
stated and demonstrated when holding a phone and taking a written mess-
age. Language laterality was determined by the extent of dysphasia fol-
lowing ECT to the right and left hemispheres. All five right handers
demonstrated left-hemisphere language, yet only one showed a contrala-
teral REA on the dichotic task (one hadan LEA, while the remaining three
showed no ear asymmetry). Four of the five right handers had a left ear
preference. The dichotic testing revealed that six of the eight left
handers hadan LEA. Therefore, left handers with left language and a
left ear preference were more likely to demonstrate a LEA than the right
handers (the majority showed no ear asymmetry). The authors conclude
that "there is a closer correspondance between strength and consistency
of lateral preferences and ear advantages than between hemisphere lan-
guage laterality and ear advantage. ." (p. 195).
It is unclear why these subjects' DL performance was different from
the previously reported findings with right and left handers (right hand-
ers showing a REA, and left handers showing no consistent ear superiority).
This study indicates that caution should be used when inferring language
laterality from a DL task.
The authors suggest that the difference between right and left
handers lies in their ability to "distribute spatial attention" (p. 195).
It appears that left handers are more tied to the left, whereas the
right wanders can distribute their attention more equally to the right
This review of the major areas of handedness research indicates
that many of the issues are as yet controversial. There are conflicting
estimates of the incidence of right and left handers and the stability
of handedness estimates over many centuries. The best method of deter-
mining handedness is unclear. Though the questionnaire method is the
simplest with large samples, Hicks (personal communication) determined
that the major variable that predicted handedness was the hand that was
best able to learn new motor tasks (determined using factor analysis).
The explanations for the underlying basis of handedness range from gene
to cell organization to environmental mechanisms. Support has been pro-
vided for all these possibilities. Finally, the study of the relation-
ship between handedness and localization of function has generated much
research. The most definite conclusion that can be drawn from the liter-
ature is that right and left handers are different. Exactly how they
differ remains undetermined. The most consistent finding is that right
handers show strong lateralization of functions, whereas left handers
are more likely to have functions bilaterally represented.
EVIDENCE IN SUPPORT OF THE DIRECTIONAL SCAN HYPOTHESIS
Mirror Reading in Right and Left Handers
The phenomenon of mirror reading was of interest to educators and
psychologists in the first part of this century, as it related to read-
ing problems and brain injury. Recently, Tankle and Heilman (1981) have
reopened this area of research from a neuropsychological perspective.
In mirror reading, words are read from right-to-left, rather than
in the usual left-to-right direction. The letters are in a reverse
right-left orientation (e.g., a "b" is seen as a "d"). A word written
in mirror spatial orientation can be read by a normal reader by viewing
the word in a mirror (Critchley, 1927).
Investigations of reading strategies have consistently noted the
tendency for normal children up to age 7 or 8 to confuse mirror-image
letters like "b," "d"' and "p," "q" (Davidson, 1935). Confusion with
mirror-image spatial configurations (e.g., "<" for >') has also been
observed in children (Aaron & Malatesha, 1974; White, 1977).
Spontaneous mirror-reading errors occur in certain populations.
It has been reported that dyslexic children make mirror-image letter
errors past the age at which children with no reading disabilities make
these errors (Critchley, 1970; Money, 1962). Mirror reading has been
observed in a patient with probable left-hemisphere damage (Heilman,
Howell, Valenstein & Rothi, 1980) and appears to be more common among
such patients who are left handed (Carmichael & Cashman, 1932).
Mirror reading may be induced by brain dysfunction, but since mir-
ror reading occurs more often in left handers, differences in the corti-
cal organization of right and left handers may predispose the left hand-
er to mirror read after brain damage. The purpose of the following ex-
periment was to determine if this predisposition could be demonstrated
with "normal" subjects.
In an initial experiment, a tachistoscope was used to investigate
the ability of normal right and left handers to mirror read. Subjects
were presented words written or printed in normal or mirror image across
the midline. They were instructed to press a microswitch when they
could read the word and to say it aloud as quickly as possible. It
was suspected that, due to the inherent difficulty in mirror reading,
the subjects would have more difficulty reading mirror script than mir-
ror print. Since there were concerns about ceiling and floor effects
with one or the other of these stimuli, both mirror print and mirror
script words were included.
If left wanders are more predisposed than right wanders to read
mirror words, this ability should be reflected in faster reaction times
and fewer errors. There were 48 subjects (24 right handed and 24 left
handed). Within the two groups, half of the subjects were men and half
were women. Handedness was determined by preferred writing hand.
The stimuli were concrete nouns containing three, four, five, or
six letters (e.g., mat, doll, glass, and cookie). Four types of writing
were used--normal print (NP), normal script (NS), mirror print (MP), and
mirror script (MS). Therewere 20 trials of each type (NP, NS, MP, MS).
The order of presentation of the 80 stimuli was randomized. The stimuli
were presented to the subjects in an Iconix tachistoscope. Subjects
sat with their head placed in a rubber mask that fit around the eyes
in order to keep the head steady. A fixation dot was presented in the
center of the screen for 2500 msec, followed by presentation of a word
stimulus arranged horizontally and bisecting the center of the screen.
Analysis was performed on two sets of data: a) reaction times from
presentation of the stimulus until identification of the word and b)
error scores. Error analysis of the mirror print condition was not pos-
sible due to an extremely low number of errors. The analysis of the
MP reaction times revealed a significant difference between right and
left wanders in the distribution of reaction times (X =-6.2, p<.03).
The sum of the ranks of the left handers was lower than that of the
When reading mirror s left handers made fewer errors than
right handers. Left handers also tended to read mirror script words
more quickly than right handers, although this trend was not signifi-
cant (X =-2.89, p<.l). The failure of the mirror script condition to
show more than a trend may have occurred because data from those trials
in which the subject erred were excluded. Since in this condition the
handers made significantly more errors than the left handers, fewer data
points were obtained from the right handers for reaction-time analysis.
Excluding these presumably higher reaction times might have masked the
true difference between right and left handers. Another reason for the
failure of the MS reaction time to reach significance might relate to
differences in the level of difficulty of these conditions. The higher
overall error rates in the mirror script condition indicate that reading
mirror script is more difficult than reading mirror print. With the
more difficult mirror script words, the left handers made fewer errors
than the right handers. In the easier mirror print condition, in which
errors were rare, the difference between the right and left handers be-
came apparent only in the reaction time scores.
Although this study demonstrated that left handers are more capable
of successfully reading mirror words, it is not known what is different
about the organization of their brain that gives them this advantage.
At least two explanations are possible.
Orton (1928) proposed that visual images in each hemisphere are mir-
ror images of each other. In right handers, the language dominant left
hemisphere would contain the images or engrams of words in their usual
left-to-right orientation, whereas the right hemisphere would contain
their mirror image. Orton was mainly interested in comparing normal
and dyslexic readers. He proposed that in the normal reader the right
hemisphere's mirror engrams were suppressed. For dyslexics, however,
the left hemisphere suppression of the right-hemisphere engrams was pos-
tulated to be incomplete or absent. This would result in an intrusion
of mirror lettering in writing and reversal of mirror letters in reading
(e.g., "d" for "b").
Orton did not specifically make predictions about the visual images
in the hemispheres of left handers. However, his theory could be ap-
plied to left wanders. As discussed in the previous chapters, a per-
centage of left handers have left-hemisphere language, some have right
language and some have bilateral representation (the proposed percent-
ages vary; the actual percentages are unknown). For those left handers
with left-hemisphere language, one might expect, according to Orton's
hypothesis, that their right hemisphere contains the mirror images of
the engrams in the left hemisphere. This is the same prediction made
for right handers. Since the right hemisphere in most right handers
and in left handers with left language is virtually mute these individ-
uals must access both hemispheres to read a mirror word. That is, they
must access the right hemisphere in order to identify the mirror word
and then access the left in order to verbalize the word. Left handers,
with right and bilateral speech may be able to respond without involv-
ing the left hemsiphere speech mechanism. Therefore, if left handers
only need to access one hemisphere to respond while right handers
must access both hemispheres, then left handers may be able to respond
(verbalize a mirror word) more quickly than right handers.
An alternative explanation for the left handers' superiority with
mirror words is based on the possible strategy used to process the
information. To read mirror script or mirror print successfully, one
must reverse the direction of scan from left-right to right-left. The
difference between right and left handers could be due to differences
in their ability to shift from a left-right to a right-left direction
of scan. The underlying basis for this difference may be related to
the differences in the asymmetries in the nervous systems of right and
The Directional Scan hypothesis proposes that right and left hand-
ers adopt different visual scanning strategies when confronted with
virtually any stimulus. If the stimulus is long or large, the scan
may be reflected in saccadic movements of the eyes. With a small stim-
ulus, (e.g., a single letter or short word) the scan may be more "atten-
tional." In other words, the eyes may not move over the stimulus, but
attention is cognitively directed in a specified direction (Posner,
Nissen & Ogden, 1978).
The Directional Scan hypothesis has evolved from the consideration
of the ability to make right-left discriminations on one's self and in
space. Ernst Mach (1894) observed that a symmetrical visual system
could not tell right from left, because there would be no reference
point. The fact that man can distinguish right from left, even in the
presence of a symmetrical stimulus, suggests that an asymmetry does
exist somewhere in the nervous system. Corballis and Beale (1970) sug-
gested that the ability to discriminate between right and left is based
on a motor asymmetry or a response asymmetry. However, two dissociable
factors may be responsible for discriminating right from left (Heilman
et al, 1980). We may use a limb or response asymmetry to tell right
from left on ourselves, while a preferred direction of exploration may
be used for discrimination in space. That is, right handers prefer the
right hand, because they have motor engrams in their left hemisphere
Liepmann, 1908). This preferred hand may be used as a reference for
the remaining parts of our body. For discriminations in space, however,
we may use a different strategy. Instead of having a preferred limb as
a reference, we may have a preferred direction of action or exploration.
It is possible that the strategies used for discrimination on self
and in space are related. Benton (1959) noted that ability to
discriminate on self develops before discrimination in space. There-
fore, the limb asymmetry possibly responsible for right-left discrimi-
nation on self contributes to the development of a preferred direction
of exploration, possibly used in making right-left discriminations in
When one reads, the direction of scan remains constant. Thus, one
may learn the directions and positions-in space by assessing their tem-
poral relationships. For example, when we are taught to read and write
English, we learn to scan from left to right. We may also learn that a
"d" is a convex line (c) followed by a straight line (1) and that the
letter "b" is a straight line followed by a concave line. As one reads
and scans from left to right and comes upon a convex line followed by a
straight line, it will be labelled as a "d." To mirror read, one must
overcome this left-to-right scanning pattern.
If a system that is highly asymmetrical is more likely to have a
consistent direction of scan, then it would be expected that right hand-
ers should show a consistent left-right direction of scan. Left handers,
however, being more symmetrical, would be more likely to scan either
left-right or right-to-left.
Two alternative explanations for left wanders' superiority with
mirror words have been presented: the mirror engram hypothesis and the
Directional Scan hypothesis. Different predictions can be made by each
of the alternatives regarding the expected performance of right and left
handers on a mirror-reading task in which mirror words are tachisto-
scopically presented to the right and left visual fields.
If right and left handers both have mirror engrams in their right
hemisphere, then both right and left handers should show a left visual-
field superiority for naming mirror words rather than the typical right
visual-field superiority that is seen with normal words. In addition,
since right handers, unlike left handers, may have to use both hemi-
spheres before they can verbally respond to mirror words, overall they
might take longer to respond and might make more errors. If, however,
the difference lies in direction of scanning strategies, then a RVF
superiority should be found among left but not right handers. Since
subjects are more likely to scan away from the fixation dot (11arcum &
Finkel, 1963), mirror words projected into the LVF would more likely be
scanned right-to-left than words presented into the right visual field.
Although words in the right visual field would most naturally be scanned
left-to-right, this would conflict with reading mirror words. There-
fore, a left visual-field superiority would be expected for both handed-
ness groups. Furthermore, if left handers can more easily reverse direc-
tions, they may be more capable of saccading to the right and scanning
right-to-left those words in the RVF and, thereby, be superior to right
wanders in reading mirror words in the right visual field.
To test these hypotheses, a VHF experiment was conducted, in which
three-letter common words (e.g., put, boy, dog) were projected to either
the right or left visual field for 200 msec (Tankle & Heilman, 1981.).
The subject was required to state the word aloud as quickly and accurate-
ly as possible. A microphone in front of the subject was hooked into a
Voice Activated Relay which stopped an electronic millisecond timer when
the subject spoke. Analyses were carried out on reaction-time data.
The results indicated that both right and left handers more rapidly de-
tected mirror words projected to the LVF than the RVF. Furthermore,
although there were no differences between the two groups in their reac-
tion times to LVF stimuli, the performance of left handers was superior
to that of the right handers when the stimuli were projected to the RVF.
The results of this study are only partially consistent with the
predictions from Orton's (1928) dominance hypothesis. That is, reaction
times to mirror-print words were found to be significantly shorter in
the left than the right visual field. An alternative explanation for
this finding comes from Harcum and Finkel (1963), who also reported a
LVF superiority for reading tachistoscopically presented mirror words.
They interpret their findings as evidence for a "sequential visual pro-
cess, the beginning and end of which is determined in part by character-
istics of the stimulus pattern" (p. 234). Since the natural tendency is
to read away from the fixation point from the beginning to the end of
the word, normal words are more accurately perceived in the RVF and
mirror words in the LVF. When one is required to read mirror words in
the RVF, it is necessary to overcome or be free of this natural scanning
The prediction from Orton's hypothesis that left handers would show
greater left visual-field superiority than right wanders was not substan-
tiated. Tankle and Heilman interpreted this to indicate that the dif-
ference found in the first experiment was not due to differences in the
degree of lateralization of speech output mechanisms between right and
left handers in association with having mirror engrams in the right hemi-
sphere. It was found in the VHF experiment that left handers had shorter
reaction times than right handers when mirror words were presented to
the right visual field. This finding is consistent with the hypothesis
that left handers' superiority in reading mirror words may be related
to their superior ability to reverse their scanning direction from a
left-right pattern to a right-left pattern.
Isseroff, Carmon and Nachshon (1975) also conducted a VHF experi-
ment with mirror words. They presented mirror and normal words to eight
presumably right-handed subjects. They found a RVF effect for both nor-
mal and mirror words. Tankle and Heilman (1981), however, found a LVF
superiority for mirror reading in right handers. It is not clear in the
Isseroff et al. (1975) study whether mirror print or mirror script was
used. Tankle and Heilman (1981) found that results differed for mirror
print and mirror script. Two methodological differences between these
studies were a) Isseroff et al. (1975) presented each word twice (once
to the RVF and one to the LVF. Tankle and Heilman (1981) presented each
word only once. b) Isseroff et al. used a 100 msec exposure time,while
Tankle and Heilman used a 200 msec exposure. Although a 200 msec pre-
sentation should preclude eye movements, it is possible that subjects
actually engaged in some degree of scanning. This may account for the
right handers' overall LVF effect. It does not explain the differences
between the right and left handers.
Directional Tendancies in Drawing
Further support for the Directional Scan Hypothesis presented in
the previous section comes from the literature on "Directionality."
In a review of the area, Dreman (1977) defines directionality as "the
tendancy of a movement tQ pursue a characteristic course.. (p. 125).
He divided the previous research into two groups: 1) those studies which
suggest innate directionality trends and 2) those which suggest environ-
mental influences on directionality. Innate directionality is consider-
ed to be due to the fact that abductive movements (away from the body)
are easier than adductive (towards the. body) (Brown, Knauft & Rosenbaum,
1948). Therefore, when a right hander draws a straight line from left-
to-right, this is an abductive movement and would be expected to be the
preferred direction. The left hander, however, is making an abductive
movement when drawing from right-to-left. In a task where subjects were
required to fill in horizontal lines between two vertical lines (rungs
on a ladder), Reed and Smith (1961) found that 35 out of 50 right hand-
ers moved left-to-right, while 43 out of 50 left handers moved right-to-
left. If direction of line drawing was influenced by reading strategies,
then no difference would have been observed between right and left hand-
ers. Weiss (1969) observed 311 Israeli school children while the stu-
dents were drawing selected Bender-Gestalt figures. Despite their right-
left reading (Hebrew) strategy, he found that the right handers (n=279)
drew from left-to-right significantly more than the left handers (n=32).
Finally, Rice (1930) observed children drawing a diamond. The right-
handed children who drew the diamond in segments (not continuous lines)
showed a strong preference for left-to-right motions. The left handers
tended to draw from right-to-left. These studies suggest that direction-
ality in at least line and figure drawing is independent of direction of
Research suggesting that reading direction does influence direc-
tionality is largely developmental. Ghent-Braine (1967) and Weiss
(1971) studied Israeli children who learned to read right-to-left. They
found that the younger (elementary school) children were more likely to
draw from right-to-left, whereas the college students preferred the
left-to-right direction. These authors suggest that the innate direc-
tion (left-to-right) may come into conflict with reading direction
(right-to-left) at younger ages. By adulthood, the innate mechanisms
reach their full strength. They did not analyze for handedness differ-
Weiss (1969) also compared 50 Israeli university students with 50
American students on directionality with the Bender-Gestalt figures.
He found that the majority of subjects drew from left-to-right, and that
there was no difference between the Israeli and American groups. He
concluded that in adults directionality in drawing is independent of
Dreman (1974) extended Weiss' study to include right- and left-
handed Americans and Israelis. He observed the subjects while drawing
the Bender-Gestalt. He found that all right handers, regardless of
whether they were American or Israeli, preferred a left-to-right direc-
tion (88 percent and 92 percent, respectively). The left wanders' per-
formance was more variable. The American left handers also preferred
the left-to-right direction (78.8 percent), whereas the Israeli left
handers preferred right-to-left (71.2 percent). Similar findings have
been reported by Shannon (1979). It has been suggested that right hand-
ers' preference for the left-to-right direction is biologically determined.
Left handers, however, appear to be more environmentally influenced
(Americans preferred a left-to-right direction, whereas Israelis pre-
f erred a right-to-left direction).
The results of these studies are consistent with the Directional
Scan hypothesis. The right handers show a consistent left-to-right di-
rectionality in drawing. The left handers seem to be influenced by fac-
tors in the environment, such as reading direction. It is possible that
the lateralization differences between right and left handers are respon-
sible for the varying effects of the environment. That is, in right
wanders, where the L-R direction of scan is determined by the high de-
gree of cortical asymmetry, the environmental factors (reading right-to-
left) have no effect on preferred direction of scan. However, among
left handers, the lack of a cortically induced direction of scan results
in susceptibility to environmental factors. For this reason, left-
handed Israelis show a right-to-left directional preference. Although
supportive of the Directional Scan hypothesis, the directionality liter-
ature supplies no direct evidence for the hypothesis that right handers
are tied to a consistent left-to-right scan, while left handers can move
equally well from left-to-right and right-to-left.
Directional Tendancies and Eye Movements
Investigation of the latency of rightward and leftward saccades has
revealed that subjects (predominantly right wanders) make faster moves
to the right than to the left (Rayner, 1978). Pirozzolo and Rayner
(1.980) propose two explanations to account for this finding. First, the
faster right saccade may be influenced by the left-to-right direction
of reading in English. The authors state that "this pattern of eye
movement behavior (left-to-right when reading English) may influence
the length of time needed by the left hemisphere to program and launch
a rightward saccadic eye movement" (p. 225). In other words, experience
with reading with a left-to-right move facilitates the left hemisphere's
motor response, resulting in faster right than left saccades. At this
time, there is no direct evidence for this hypothesis. It would be pos-
sible to determine the influence of reading direction by comparing the
right and left latencies of English readers with the latencies of Hebrew
or Arabic readers. If reading direction is influencing the speed of the
saccades, then Israelis and Arabs, who read from right-to-left, should
exhibit faster leftward than rightward saccades.
An alternative hypothesis is suggested which considers a structural
factor which is independent of reading direction. "Specifically, it has
been suggested that the left cerebral hemisphere, which generates right-
ward saccades, is the more efficient neural substrate in visuo-motor
tasks. ." (p. 225). Pirozzolo and Rayner (1980) attempted to deter-
mine which alternative hypothesis correctly explains the data by compar-
ing the performance of right and left handers. The rationale for this
experiment is that, although all subjects read in a consistent direction,
there may be underlying differences in the cerebral organization of
right and left handers. Therefore, if faster rightward scans are due to
left-to-right reading, no difference in rightward and leftward latency
would be expected between right and left handers. If, however, struc-
tural or cerebral organization differences are responsible, then left
and right handers may perform differently. In their experiment, sub-
jects were required to fixate straight ahead until a stimulus word or
symbol (asterisk) appeared at one of 8 locations to the right or left.
The subject's task was to look at the stimulus as quickly as possible.
In the word condition the subject was instructed to say the word aloud.
The results revealed that right handers moved significantly faster to
the right than the left (176 msec and 192 msec, respectively). The
left handers, however, showed no asymmetry (188 msec to the left and
186 msec to the right).
These findings do not support the reading direction hypothesis.
The faster rightward saccades of right handers are apparently not due
to experience in reading in a particular direction. The results do
lend support to the structural hypothesis. The authors state that the
results with right wanders "strongly suggest an underlying difference
in sensori-motor organization between the left and right hemispheres"
(p. 228). It is possible that the greater brain symmetry in many left
wanders results in their ability to move equally fast to the right and
In summary, the study of directional tendancies in drawing and the
study of eye movements are consistent with the Directional Scan hypo-
thesis. These results suggest that, in fact, right handers are tied to
a left-to-right scan (or, at least, scan left-to-right faster), where-
as left handers move equally well in either direction.
DIRECTIONAL SCAN HYPOTHESIS AND RIGHT-LEFT DISCRIMINATION
The Directional Scan hypothesis has received indirect support from
a number of research paradigms. This was discussed in depth in Chapter
II. Further support for this hypothesis could be obtained with Infra-
red Oculography. With this method, the direction of the eye movements
can be observed while a subject reads mirror and normal words, and it
could be determined if left handers do, in fact, adopt a right-to-left
scan with more ease than do right handers. However, this equipment is
not available at this time. Therefore, an alternative methodology has
been designed to test for the hypothesized differences between right and
left handers' direction of scan preference. It will be discussed in
Asymmetrical Nervous Systems and the Directional Scan Hypothesis
The research presented in Chapter I suggests that right wanders
generally show more asymmetries than left handers. Mach (1894) proposed
that a symmetrical system could not tell right from left, because there
would be no internal reference point. More recently, Corballis and
Beale (1976) have proposed that the ability to discriminate right from
left arises from a child's "developing laterality, including handedness
and cerebral lateralization" (p. 159). Failure to develop lateraliza-
tion would result in an individual unable to discriminate between right
and left. It is, however, likely that all individuals have some degree
of cerebral lateralization. Therefore, differences in right-left dis-
crimination abilities between individuals of varying degrees of cerebral
lateralization are likely to be express in terms of relative ease or
difficulty in making the distinction.
If a group of individuals can be classified as being less asymme-
trical than another group, this should be reflected in differential abil-
ities on a task requiring discrimination between right and left. Speci-
fically, as the literature suggests that left handers are less asymme-
trical (i.e., have more bilateral representation) than right handers,
they may perform more poorly than right handers on a task requiring
right-left discrimination. However, right-left discrimination may not
be a unitary function. That is, it may have separate components.
The Development of Right-Left Discrimination
The discrimination between right and left has been separated into
two different categories: personal right-left discrimination (i.e., on
self) and extrapersonal discrimination (i.e., in space).
Benton (1959) reviewed the early research on the development of
right-left discrimination and the loss of this ability following brain
damage. The methods generally used to assess discrimination on self
and in space include the following tasks: 1) simple identification on
self of a body part (e.g., "Which is your right hand?"), 2) double com-
mands on self that are uncrossed (e.g., "With your right hand, point to
your right eye.") or crossed (e.g., "With your left hand, point to your
right ear.") and 3) simple and double commands on a person facing the
subject (e.g., "With your right hand, point to my left eye."). The
developmental findings can be summarized by the following points:
a) the ability for personal right-left discrimination develops earlier
than extrapersonal right-left discrimination; b) by age 12, performance
reaches adult level and is nearly perfect (at least as measured by
errors produced); c) children with higher IQ scores show right-left
discrimination ability earlier than those with lower scores; d) there
are no sex differences; and e) there is a small but significant correla-
tion between strength of handedness and right-left discrimination abil-
ity. Stronger handedness is associated with better right-left discri-
mination (Benton & Menefee, 1957).
In a more recent study, Lacouisiere-Page (1974) tested children on
seven variables which were suspected to be important in the development
of right-left discrimination. The variables were age, visuo-motor coor-
dination (tested by having the children draw a diamond), verbal and non-
verbal intelligence (WAIS vocabulary subtest and Raven Progressive Ma-
trices), handedness (Harris Test of Lateral Preference), space relations
(Thurstone's PMA), and conceptualization of the human figure (Goodenough's
Draw-A-Man). All subjects were given these tasks and Benton's right-
left discrimination task. Overall, the Benton R-L scores correlated
.82 with vocabulary, .80 with age, .79 with space relations and .76
with Draw-A-Man. A separate analysis was done for each type (.self vs.
space) of right-left discrimination. Discrimination on self correlated
.78 with age, .71 with vocabulary (when age is partialled out; r=.29),
.71 with space relations, .71 with Draw-A-Man and .60 with Draw-A-
Diamond. Discrimination in space correlated .46 with vocabulary and
PMtA scores (reduced to .35 and .38, respectively, when age is partialled
out). The author concludes that:
On the whole, the results of this study do concur with
Benton's finding. It seems indeed that the development
of verbal intelligence is concurrent with the evolution
of the right-left concept on one's self and on others,
while age and conceptualization of the human figure are
connected with the application of these labels on one's
self. (p. 116)
Right-Left Discrimination in Adults
Olson and Laxar (1973) studied how subjects processed the words
"right" and "left." They designed an experiment in which subjects had
to make true/false judgments regarding the word "right" or "left" and a
position of a light. For example, the word "right" would be presented
directly in front of the subject. Following this, a light would go on
either on the right or left side of the screen. If the right light went
on, there was concordance, and the correct response was true. If the
left light went on, there was discordance, and the response should be
false. They found that the subjects (right handers) were faster when
the target word was "right" rather than "left." This was particularly
evident in the true condition. The authors interpret this finding as
reflecting right handers' dominant use of the right hand, which sets up
a strong right-left dichotomy. Since right is the preferred direction,
the word "right' is processed more quickly. The procedure was repeated
with a group of left handers (Olson & Laxar, 1974). The results reveal-
ed no asymmetry in processing "right" and "left" in left handers. The
authors used writing hand alone to determine handedness. It is likely,
therefore, they tested a greater number of mixed than strong left hand-
ers. It is possible that a more ambidextrous group processes the words
"right" and "left" with equal speed, whereas the right handers are
faster with the word "right." It would be interesting to separate the
left handers into strong and mixed groups to determine if, indeed,
strong left handers respond more quickly to the word "left."
A number of studies of right-left discrimination in adults have
used drawings of right and left body parts. For example, the subjects
are shown pictures of a hand, foot or eye and asked to determine if it
is the right or left hand, foot or eye. Silverman, Adevai and McGough
(1966) administered a series of these body parts to right- and left-
handed male college students. They found that left-handed males per-
formed significantly poorer on this task than right-handed males.
Culver (1969) found this method to be reliable within subject. He also
found that performance on this task correlated with the "Rod and Frame"
tasks (considered to be a perceptual task). In his sample of 157 right
handers and 23 left handers, there was no significant difference between
Wolf (1973) sent out questionnaires to physicians and their spouses.
It asked them to report if they had trouble discriminating right and
left all the time, frequently, occasionally, rarely or never. Responses
were received from 408 males and 382 females. The results revealed that,
while 17.5 percent of the females reported right-left confusion all the
time or frequently, only 8.8 percent of the men did so. The difference
was significant. Wolf concluded that women were more likely to have
Bakan and Putnam (1974) point out the subjective nature of the Wolf
study and that the results may reflect the women's greater tendancy to
report the discrimination difficulties. Consequently, they tested 400
undergraduate students using the body parts method discussed above.
There were 123 right-handed males, 228 right-handed females, 28 left-
handed males and 21 left-handed females. The results indicated a signi-
ficant sex difference among right handers, where the females subjects
made more errors than the males subjects (8.02 and 4.75, respectively,
out of 32 trials). The difference between male and female left hand-
ers was in the same direction, but did not reach significance. Over-
all, there was no difference between right and left handers.
In summary, the results of studies of body part identification and
handedness are equivocal. The finding of sex differences is relatively
more consistent. The authors of the above cited literature consider
the body identification task as a test of right-left discrimination.
It is not clear, however, if deciding whether an abstracted body part
is right or left sided is the same as observing a whole person and mak-
ing a decision about the right and left side.
Cooper and Shepard(1975) investigated one possible strategy that
subjects might use to identify a right or left hand in a picture. The
stimuli were pictures of hands spread out with the palm or the back of
the hand facing up. The hands were presented at 6 different rotations
(00, 600, 120, 180', 2400 and 300'). There were two conditions: a)
the subject received no pretrial information about the hand's orienta-
tion or which side was facing up and b) partial information was given
to the subject.
The results indicated that response time was shorter when the sub-
jects had prior information as compared to no information. The authors
suggest that, when there is no prior information, subjects must con-
struct a visual image of the picture and maneuver their hands into the
position of the stimulus. By matching the image with their hand, sub-
jects can determine if the hand is right or left. With some prior know-
ledge, less information needs to be considered at the time of decision,
and, therefore, the decision should be made more quickly.
It is possible that this rotational strategy is used by both right
and left handers to decide right or left on an abstracted hand and right
and left on a whole object. It is also possible that this strategy is
only used in a situation that is different from the normal way a hand
is usually seen (i.e., attached to a body). Therefore, the studies on
right-left discrimination and handedness cited above may not be indica-
tive of subjects' performance on a discrimination task that includes a
whole person or object.
Disturbances in Right-Left Discrimination
Benton (1959) considers the disorders of right-left discrimination
to be the result of a breakdown in the foundation or the determinants of
discrimination abilities. Simple R-L discrimination (lateral responses)
on self requires the ability to differentiate and identify right and
left body parts. This arises from a child's developing body awareness.
Benton states that "since the (right-left) gradient is a cognitive spa-
tial organization of tactual-proprioceptive and visual elements, it is
not surprising that patients showing evidence of a defective gradient
are usually found to have disease of the parietal or parieto-occipital
areas" (p. 151). Many of the reported cases involve patients with
bilateral lesions. There are cases reported following unilateral le-
sions (usually in the left hemisphere). The research on localization
of lesions and discrimination abilities will be considered in greater
detail below in order to first conclude this discussion of Benton's
work. The right-left discrimination in more complex tasks (crossed and
double commands on self and on others) is, according to Benton, deter-
mined by the development of the verbal concepts "right" and "left."
Support for this comes from the developmental studies, which indicate
that those children with higher IQ scores (and presumably higher verbal
skills) perform better on right-left discrimination tasks. Benton argues
that disruption of crossed discrimination on self and in space is attri-
butable to a disorder in symbolic comprehension of the terms "right"
and "left." He claims that this is often associated with an aphasia.
The relationship between right-left discrimination disorders and
other disturbances is still controversial. Gerstmann (1940) noted that
right-left disorientation occurs along with acalculia, agraphia and
finger agnosia from focal lesions in the area of the left angular gyrus.
Later studies suggested that this tetrad of symptoms resulted from
larger and more diffuse left-hemisphere lesions (Hermburger, De~lyer &
Reitan, 1957). Other deficits noted to occur frequently with the
Gerstmann Syndrome are constructional apraxia and aphasia (Warrington &
In a recent case report, Roeltgen, Sevusch and Heilman (1981) de-
scribe a patient who exhibits the four Gerstmann Syndrome symptoms with
no accompanying deficit of apraxia or aphasia. The CT scan localized
the lesion in the inferior parietal lobule (mainly the angular gyrus
with some involvement of the supramarginal gyrus). The authors suggest
that, although infrequently seen, the Gerstmann Syndromecan occur when
the lesion is restricted to the inferior parietal lobule.
Semmes, Weinstein, Ghent and Teuber (1963) tested a group of 76
war victims with varying extent and placement of head injuries. The
purpose of the study was to localize the areas of the brain which, when
damaged, resulted in disorientation on self and in space. Orientation
on self was tested by showing the patient a diagram of the human body
,(front and back) with numbers marking a number of body parts. The
patient was asked to point to his body parts that corresponded to the
numbered body parts on the diagram. The test of orientation in space
involved the patient walking on paths guided by a map. "The map was
not always in the proper orientation with respect to the room and a
series of translations of its co-ordinate system to the true directional
co-ordinates had to be effected by the subject" (p. 753). The results
showed that both self and space discriminations were impaired by left
posterior damage. Ability to discriminate on self was impaired mainly
by left-hemisphere anterior lesions, whereas discrimination in space
was more distrubed by right-posterior lesions. Their subjects were also
tested on a number of other tasks which ruled out aphasia, sensory dis-
turbance (patients with and without visual field defects) and visual
agnosia (pattern discrimination and visual problem solving) as a corre-
lating factor. They did find a relationship between right-left discri-
mination on self and a sorting task (similar to the Wisconsin Card Sort).
They suggest that both of these behaviors are anterior lobe functions.
More recent research has supported the finding that right-left discri-
mination in space is disrupted by posterior lesions. However, these
studies contend that bilateral and not unilateral lesions result in
impairment (Butters & Barton, 1970; Ratcliff & Newcombe, 1973).
A major problem with most of the above cited studies is that they
used subjects who suffered from penetrating missile wounds. First, the
extent and location of the wounds, even within a hemisphere,was quite
varied. Second, in the Semmes et al. (1963) study, some of the men were
WWII victims (10-15 years post injury) while others were injured during
the Korean War (2 years post injury). The effect of years of recovery
is unknown, but it is possible that certain behaviors which had been
impaired had been compensated for over the years.
In summary, while the results are not conclusive, it is probable
that right-left discrimination in space is a posterior brain function
(most likely parietal lobes), whereas discrimination on self may be a
more anterior function.
Right-Left Discrimination and Directional Scanning
It has been proposed (Heilman et al, 1980; Tankle & Heilman, 1981)
that the mechanism for right-left discrimination on one's self and in
space can be dissociable. For discrimination on one's self, one may
use a limb or response asymmetry. The greater the limb or response
asymmetry is, the better one's ability to discriminate between right
and left. This is due to the fact that the limb asymmetry gives a strong
reference point to discriminate between the right and left hand.
Right wanders typically show manual bias favoring the right hand.
That is, they use their right hand for most motor behaviors (writing,
throwing, eating, etc.). Left handers are more likely to show equal
manual abilities with both hands (Benton, 1962). As has been suggested
by Heilman et al. (1980),right handers may have motor engrams strongly
lateralized to the left hemisphere, whereas left handers may have more
bilaterally represented motor engrams.
Thus, if it is the case that limb asymmetry does in fact enable one
to have a reference point for discrimination on one's self, then several
predictions can be made. Strongly asymmetrical right handers should be
better able to discriminate between right and left on themselves than
are left handers. Since left handers may be less asymmetrical, they may
have greater right-left confusion on themselves. Although one may use a
limb asymmetry ("this is my right hand, so this must be right hemispace")
for discrimination in space, it is possible that one also uses a pre-
ferred direction of scan or exploration. The research discussed in pre-
vious sections suggests that, while right handers show a left-right scan
preference, left handers are less likely to have a preferred direction.
When one consistently scans the environment in a left-right direc-
tion, what one sees first can be correctly identified as the left side,
and what one sees second can be identified as the right side. Using
this strategy, one can easily distinguish between right and left. If,
However, there is no consistent direction of scan, then right and left
in space may be confused.
The brain mechanism for a left-right scan preference is not as
clear as for right-left discrimination on self. It is possible that a
similar motor asymmetry in right handers is responsible for the scan
preference. As Pirozzolo and Rayner (1980) pointed out, a structural
brain difference may be responsible for right handers' faster left-right
saccades and left handers' equal latencies to both sides. This struc-
tural difference may be a motor mechanism.
Right handers with this asymmetrical motor mechanism may have a
stronger left-right scan preference, while left handers' system may be
more symmetrical. Therefore, on a task requiring right-left discrimina-
tion in space, it would be expected that left handers would have right-
Right and left handers appear to differ in their degree of cerebral
lateralization. It may be that differences in the asymmetry of motor
engrams contribute to their difference in directional scan preferences.
Further, the Directional Scan hypothesis suggests that these groups will
differ in their ability to discriminate right and left on themselves and
Specifically, left handers, being more symmetrical, may have an in-
consistent direction of scan, and, therefore, be more likely to confuse
right and left. On the other hand, right handers, being more asylmetri-
cal, may have a consistent left-right direction of scan preference and
be more capable than left handers in discriminating right and left on
themselves and in space.
STATEMENT OF THE PROBLEM
The first hypothesis to be tested in the following experiments is
that right and left handers adopt different patterns in scanning the
environment. These preferential scanning patterns are related to the
degree of motor asymmetry shown by right and left handers. That is,
right handers with strongly asymmetrical (left hemisphere) motor engrams
prefer to scan in a left-to-right direction. On the other hand, left
handers with more bilaterally represented motor engrams are equally
likely to scan in either direction.
The second hypothesis to be tested is that one's preferential motor
asymmetry also contributes to the individual's ability to make right-
left discriminations on self and in space. It is proposed that an indi-
vidual who is strongly lateralized motorically and uses one hand domi-
nantly will be better able to discriminate right from left on him/herself.
The ability to make right-left discriminations in space is tied to the
use or the absence of a consistent direction of scan. That is, if one
scans the environment from left-to-right, then that which is seen first
is on the left, and that which is seen second is on the right. It is
proposed that this preferential direction of scan is an effective stra-
tegy for discriminating right and left in space. Failure to use a con-
sistent scan necessitates alternative strategies such as use of body
position and rotation in relation to object in space (as Cooper &
Shepard (1975) suggest). Whether these alternative strategies require
different processing time is unknown. It would appear, however, that
a simple scan of a stimulus and a subsequent response would require less
time than maneuvering one's body to match the stimulus and then being
able to respond. Therefore, it is predicted that right handers, who
generally are more strongly lateralized for motoric functioning, will
have performance superior to left handers on tasks requiring right-left
discrimination on self and in space.
Directional Scan in Right and Left Handers
In Part I, the Directional Scan hypothesis will be tested. The
rationale for the methodology chosen is based on the following reason-
ing. If an individual consistently scans in a left-to-right direction,
what is on the left side of a picture will be perceived more quickly
than what is on the right side of a picture. If, however, an individual
scans from right-to-left as easily or as often as left-to-right, on the
average, the left side and right side of a picture will be perceived
equally as fast.
This hypothesis will be tested using a "Same-Different" paradigm.
That is, subjects are required to view two stimuli pictures. The second
picture is either identical to the first or different from the first.
The difference between the two pictures involves a salient detail occur-
ring on either the right or left side of the picture.
If subjects are using a L-R scan, it would be expected that the
reaction times to the differences on the left will be shorter than the
reaction times to differences on the right. An inconsistent scan should
result in similar reaction times to right- and left-sided differences.
Furthermore, there may be differences between right and left handers'
reaction times on the "Same" trials. It is possible that an inconsis-
tent scan would result in longer reaction times than a consistent scan.
For example, with a consistent left-to-right scan, both the first and
second picture would be scanned the same way. If the subject is creat-
ing a mental image of the first picture, then a left-to-right scan over
the second picture should match the visual image. With an inconsistent
scan, the first and second pictures may be scanned in different direc-
tions. The visual images are not so easily matched. Therefore, it may
take longer to respond "Same" when an inconsistent scan is used.
In Part II, there are a number of conditions which are designed to
assess subjects' ability to discriminate right and left on self and in
Right-Left Discrimination on Self
If the strength of one's limb response asymmetry affects performance
on right-left discrimination on self, then there should be differences
between right and left handers' performance on tasks of discrimination
ability. That is, right handers, with a strong limb asymmetry, should
perform better than left handers, who are generally more ambidextrous.
In this condition, right and left handers' ability to make simple
right-left discrimination on self will be measured. The subjects will
receive a visually presented command saying either "RIGHT" or "LEFT."
Their task will be to press a button under their right or left index
finger (right hand for "RIGHT" command and left hand for "LEFT" command).
It is predicted that left handers will take longer than right hand-
ers to make right-left discriminations on self. They may also make more
errors than right handers. This result would be consistent with the
response asymmetry explanation as the underlying mechanism for making
right-left discriminations on self.
It is possible that there will be no difference between the groups
on this task. This may be due to a ceiling effect. That is, this task
may be too simple to tap the differences between right and left handers.
Hemispace can be defined by drawing an imaginary line down the
exact midline of an individual. To the right of the line is right hemi-
space and to the left is left hemispace. Benton (1959) indicated that
crossing the midline in a pointing discrimination task ("Using the right
hand, point to your left ear.") is more difficult than lateral (same
side) pointing. It is possible that crossing the midline to press a
button in response to a verbal command will be more difficult than lat-
eral responses (as in Condition I).
In this condition, the subjects will respond to two-word visually
presented commands ("RIGHT-RIGHT," "LEFT-LEFT," "RIGHT-LEFT," and "LEFT-
RIGHT"). The first word indicates which hand should be used, and the
second word indicates which button (on the right or left side of the
subject) should be pressed.
As in condition I, performance on this task should be facilitated
by a strong limb asymmetry. That is, Ss with a limb asymmetry should
be faster in choosing the hand with which to respond and the hemispace
in which to respond.
Therefore, the predictions for this condition are similar to those
of the previous condition. That is, left handers are expected to per-
form more poorly than right handers. This may be reflected by left
handers having longer reaction times than right handers, and, possibly,
the left handers will make more errors than right handers.
The trials of this condition can be divided into two levels of
difficulty. The uncrossed trials (RIGHT-RIGHT and LEFT-LEFT) may be
less difficult and may be susceptible to a ceiling effect. However, on
the more difficult crossed trials (RIGHT-LEFT and LEFT-RIGHT), the dif-
ferences between right and left handers may be more apparent.
Right-Left Discrimination in Space
It has been proposed that when making right-left discriminations
in space, subjects may use a preferred direction of scan or exploration.
Further, the use of a preferred direction of scan is dependent upon the
degree of asymmetry in the nervous system, such that the greater the
asymmetry, the more consistent the direction of scan. Therefore, it is
expected that the greater the asymmetry and consistency of scan, the
better the ability will be to make right-left discriminations in space.
In this condition, the subject will see a manikin figure with a
black disk marking the right or left hand. The manikin is depicted as
either facing forwards or backwards. The subject's task is to state
whether the diskis on the manikin's right or left hand.
As in condition II, there are 2 levels of difficulty. When the
manikin is facing backwards, the subject is required to make an un-
crossed right-left discrimination. When the manikin is facing forwards,
the right-left discrimination is crossed. In both conditions, a consis-
tent direction of scan should facilitate performance. Therefore, it is
expected that right handers will perform better than left handers.
If a subject does not use a consistent direction of scan, an alter-
native strategy to discriminate right and left may be to figuratively
project themselves into the figure and rotate themselves relative to the
manikin. This strategy may be adopted by those Ss without a consistent
direction of scan (left handers). This strategy may take longer and
enhance the differences between right and left handers' reaction time.
This condition was designed to include right-left discrimination
both on self and in space. The subjects first see a two-word command
("RIGHT-RIGHT," "LEFT-LEFT,","RIGHT-LEFT," or "LEFT-RIGHT") followed by
a manikin figure (without a disk on the hands). The first word tells
them which hand to use (discrimination on self), and the second word
tells them which button to press relative to the manikin (discrimination
in space). For example, "RIGHT-LEFT" means "with your right hand, press
the button in front of the manikin's left hand."
Again, there are 2 levels of difficulty: crossed (manikin facing
forwards) and uncrossed (facing backwards) discrimination. As with the
previous conditions, it is predicted that right wanders will perform
better than left wanders.
The purpose of this experiment was to test the hypotheses that
(1) right and left handers differ in their preferred direction of scan,
and (2) right and left handers differ in their ability to discriminate
between right and left on themselves and in space.
The subjects were University of Florida undergraduate students.
There were 47 subjects: 16 strong right handers (8 males, 8 females),
17 strong left handers (9 males, 8 females), and 14 mixed left handers
(6 males, 8 females). Handedness was determined by the Briggs-Nebes
Handedness questionnaire. On the basis of this questionnaire, subjects
were classified into right (scoring from +9 to +24), left (-9 to -24),
or mixed (+8 to -8) handedness groups. The following information was
also obtained on the questionnaire: (1) the subject's age and sex and
(2) the number of first order left-handed relatives (parents and sib-
Part I: Direction of Scan
The purpose of this condition was to determine if left and right
handers show different vocal reaction times to pictures when changes
are made on the right or the left side. If they scan the pictures
using different strategies, then their reaction times may differ,
Specifically, it is expected that right handers will detect changes on
the left side of the picture faster than on the right side of the pic-
ture. Left handers are expected to respond equally fast to right- and
The stimuli were cartoons adapted from the "Heathcliff" and
"Marmaduke" series. The cartoons were traced directly onto tachisto-
scope cards (see Appendix A).
Each trial consisted of two cartoons that were presented 500 msec
apart. On the trials where the appropriate response was "same," the
two cartoons were identical. On those trials that were "different,"
a salient object or person on the right or the left was omitted in the
second picture. The differences between the two pictures were quite
obvious. On half of the "different" trials, the difference was on the
right side of the picture (DR),and on half, the difference was on the
left side of the picture (DL). The DL, DR, and Same trials were ran-
domly presented. There were a total of 80 trials (40 same and 40 differ-
The subject was seated in front of an Iconix Tachistoscope. The sub-
ject's head was placed in a rubber mask that fits around the eyes in
order to keep the head steady. The first picture was presented for 2
seconds. The screen was then darkened for 500 msec. This was followed
by presentation of the second picture. The second picture remained on
the screen until the subject made a response. The subject was instructed
to state into the microphone as quickly and accurately as possible
whether the two pictures were the same or different. The microphone
was connected to a Voice Activated Relay which was activated at the
onset of the stimulus presentation. The subject's voice stopped an
electronic millisecond timer, and the experimenter noted the subject's
reaction time and response.
Part II: Right-Left Discrimination on Self and in Space
The following conditions are designed to assess the ability to
make right-left discriminations on self and in space. The performance
of right and left handers will be compared.
I. Simple-Self (Right-Left Discrimination on Self)
The "Simple-Self" condition is a test of subject's responses to
simple directional ("RIGHT" or "LEFT") commands. The purpose of this
condition was to determine whether right and left handers differed in
their ability to make right-left discriminations on themselves as
measured by reaction times or error rates.
The stimuli were the words "RIGHT" and "LEFT" printed on tachisto-
scope cards. To avoiding confounding the task with reading direction,
the words were printed vertically in the middle of the screen.
The subjects were seated looking into the tachistoscope with their
hands positioned on the table in front of them. They were seated slight-
ly back from the screen (approximately 12 inches) in order to see both
the stimulus and their hands. Their index fingers were placed on
switches located on either side of their body at approximately 5 inches
from midline. Their right finger was on the right side and their left
finger on the left side.
The subjects were instructed to respond as quickly and accurately
as possible to the commands "RIGHT" and "LEFT" by pressing the key cor-
responding to the visual command. The command remained on the screen
until the subject terminated the trial by pressing a switch. The switch
press stopped an electronic millisecond timer. The experimenter record-
ed the subject's response and the time elapsed from the onset of the
command until the subject responded'.
There were a total of 80 trials with 40 "RIGHT" and 40 "LEFT" com-
mands. The right and left trials were randomized. Intertrial interval
was about 5 seconds.
II. Heispace-Self (Right-Left Discrimination on Self)
The "Hemispace-Self" condition is a more difficult test of right-
left discrimination on self than the previous condition. In this con-
dition, the subject was required to make two decisions: a) which hand
(right or left) to use; and b) which side of body midline (right or
left) to make the manual response. The purpose of this condition was
to determine if right and left handers' reaction times or error rates
differ on a more complex task of right-left discrimination on self.
The stimuli were words printed on tachistoscope cards. The words
were printed vertically in the middle of the screen. The words were
"RIGHT-RIGHT," "LEFT-LEFT," "RIGHT-LEFT," or "LEFT-RIGHT."
As in condition I, the commands were presented in the tachisto-
scope, and the subject was to respond by depressing a switch. The sub-
ject was seated slightly back from the tachistoscope, and the switches
were located approximately 5 inches from the midline on each side of
the body. Between trials, the subject's hands were placed on the table
directly in front of him or her.
On each trial, the subjects were given two commands. One command
indicated which hand to use in responding (i.e., right hand or left
hand), and the second indicated which side of body midline (right or
left) on which to respond. For example, the command "RIGHT-LEFT" indi-
cated that the subject should respond with their right hand by depress-
ing the switch located to the left side of their body midline. There
were four types of trials ("RIGHT-RIGHT," "LEFT-LEFT," "RIGHT-LEFT,"
and "LEFT-RIGHT") with 20 trials of each type randomized for a total of
III. Manikin Figures (Right-Left Discrimination in Space)
The "Manikin Figures" condition was designed to test the ability to
make right-left discriminations in space. The purpose of this condition
was to determine if right and left handers' reaction times or error rates
differ when making right-left discriminations in space.
The stimuli were manikin figures based on those used by Benson and
Gedye (1963) and Ratcliff (1979). The manikin figures were depicted as
facing forwards or backwards. The figure was depicted as facing for-
wards by the presence of facial features and buttons down the front of
the torso. The figure was depicted as facing backwards by the absence
of facial features and buttons. On the manikin, either the right or
left hand was marked by a black disk (see Appendix B).
The subjects were seated looking into the tachistoscope with their
heads in a rubber mask in order to keep their heads steady. They were
instructed to fixate on a center fixation point which remained on the
screen for 1 second. The manikin figure was presented in the center of
the screen. The figure remained on the screen until the subject termi-
nated the trial by stating whether the disk was on the manikin's right
or left side. The subjects spoke their responses into a microphone
which was connected to a Voice Activated Relay. An electronic timer
was activated at the onset of the stimulus presentation, and the subject's
voice stopped the timer.
There were a total of 60 trials with 30 right and 30 left identifi-
cations. Half of each were with the manikin facing forwards, and half
were with the manikin facing backwards.
IV. Self-Space (Right-Left Discrimination on Self and in Space)
In this condition, aspects of discrimination on self and in space
were combined in one task. The purpose of combining both self and space
discrimination tasks was consideration of the possibility that the tasks
described above would be too simple to tap right-left discrimination
differences in normal handedness groups. Since this task appears to be
more difficult to perform, it may show the expected differences. If
this is the case, however, discrimination in space and on self will be
confounded. The results will not indicate whether it is discrimination
in space, on self or a combination of both that is causing the differ-
The stimuli were the two-word commands used in condition II and
the manikin figures used in condition III. In this condition, the
manikin's hands were not marked by a black disk. Rather, a normal
hand was drawn on the manikin (see Appendix C).
The subjects were seated slightly back from the tachistoscope with
their hands placed on the table directly in front of them. The switches
were approximately 5 inches on either side of their body.
The subjects were required to make a right-left discrimination on
the manikin using a motor response. Thus, the subjects received com-
mands such as "with your right hand press the button in front of the
manikin's left hand (presented as "RIGHT-LEFT" in the tachistoscope).
There were four types of trials ("RIGHT-RIGHT" (RR), "LEFT-LEFT" (LL),
"RIGHT-LEFT" (RL), and "LEFT-RIGHT" (LR)) with 10 trials of each type
randomized across trials for a total of 40 trials. In half of the pre-
sentations, the manikin was facing forwards (F), and in half, the mani-
kin was facing backwards (B). Therefore, there were 8 types of trials:
FRR, FLL, FRL, FLR, BRR, BLL, BRL, and BLR. The types of trials can be
categorized by the discriminations necessary to respond correctly. That
is, FRR and FLL each require a crossed discrimination in space (manikin
forward) and a crossed discrimination on self (pressing the button on
the side of body opposite to the hand used). The FRL and FLR trials
require a crossed discrimination in space and an uncrossed discrimina-
tion on self (pressing the right button with the right hand). The BRR
and BLL trials involve an uncrossed discrimination on self and in space,
while the BRL and BLR require an uncrossed discrimination in space and
a crossed discrimination on self.
V. Baseline Respondin
It is possible that the speed of the motor responses will be differ-
ent if the subject uses the preferred or nonpreferred hand, regardless
of the verbal command presented. For example, a right wander may be
faster to the command "RIGHT" simply because the right hand is dominant.
Therefore, this baseline condition was intended to determine if in fact
there are differences between reaction times of the preferred and non-
preferred hand. If there are differences, then this condition can be
used as a co-variate with the other conditions requiring a motor response.
The stimulus consisted of a simple star (*) presented at midline.
The subjects were seated slightly back from the tachistoscope.
Their hands were positioned on the table in front of them. Their index
fingers were placed on the switches located at approximately five inches
A fixation dot was presented for 1 second. This was followed by
the stimulus. The subjects were instructed to respond to the star with
a button press as quickly as possible.
Half of the subjects in each handedness group were instructed to
respond with their right index finger during the first half of the ses-
sion and with their left index finger during the latter half of the
session. The other half of the subjects in each handedness group re-
sponded in the reverse order. A total of 20 right and 20 left trials
were given to each subject.
Overall Order of Presentation of the Tasks
The subjects were tested in two sessions separated by at least one
week. Half of the subjects received the Baseline, Simple-Self, Hemispace-
Self, and Self-Space conditions in Session I and the Manikin Figures and
Direction of Scan conditions in Session II. The remaining subjects were
tested in the reverse order. Each session lasted approximately one hour.
The analyses were conducted separately on each condition's reaction
times and error rates. On the between group analyses, the variables
were handedness (right, left and mixed) and sex. On the within handed-
ness group analyses, the variable was sex.
The analyses were performed on two sets of data: (a) reaction times
(RT) and (b) error rates (ER). In each condition, the reaction times
for each type (e.g., "RIGHT" and "LEFT") were averaged to yield one mean
RT of each type for each subject. The analyses of each condition will
be presented separately.
Part I: Direction of Scan
This condition was designed as a "Same" "Different" paradigm.
The subjects viewed 2 pictures that were either identical (same) or dif-
ferent. The different trials had either a difference on the left side
of the picture (DL) or a difference on the right side of the picture
The reaction time data were initially analyzed using a parametric
within-group ANOVA with handedness group (right handed, left handed, or
mixed handed) as the between-group factors and sex as the within-group
factor. The results of the analysis revealed no significant differences
between the mean RTs of same, DR and DL trials in any of the handedness
groups. However, examination of the individual subjects' mean RTs to
the stimuli suggested that there were consistent within-group differ-
ences. Furthermore, there appeared to be a high degree of variability
both within and between groups. Therefore, further analyses were con-
sidered to be appropriate.
In order to determine if a nonparametric statistic was the appro-
priate method for further analyzing the present data, the data were
analyzed to determine if the reaction times were normally distributed.
This was computed separately for the right, left and mixed handers for
each type of trial (i.e., Same, DR, and DL). Although the ANOVA proce-
dures are considered robust enough to deal with data that are not truly
normally distributed, the limits of this "robustness" are never clear.
Furthermore, if a nonparametric statistic reveals significant differ-
ences where an ANOVA failed to demonstrate any differences, then it is
reasonable to assume that some of the assumptions necessary for an ANOVA
are being violated.
The SAS Univariate procedure computes a 'W' statistic and tests the
null hypothesis that the scores are normally distributed. Therefore,
with a high probability (p) value the null hypothesis is not rejected
and the distribution is assumed normal. When the p-value is low, the
null hypothesis is rejected, and it suggests that the distribution is
Analysis of the mean reaction times (RTs) of the Same responses in-
dicated that the means were normally distributed in the right (W=.94,
p=.44), left (W=.94, p=.36), and mixed (W=.95, p=.61) handed groups.
Therefore, analysis of the Same trials was performed using a parametric
ANOVA. The mean RTs in the DL trials were found not to be normally dis-
tributed in all handedness groups (right handers W=.91, p=.02; left
handers W=.90, p=.0l; and mixed handers W=.91, p=.02). However, the
mean RTs in the DR condition were normally distributed. Therefore,
when comparisons were being made between DR (normally distributed) and
DL (not normally distributed), nonparametric statistics were used.
The predictions made for this condition involve within-group com-
parisons. That is, it was predicted that right handers' reaction times
to DR stimuli would be longer than reaction times to DL stimuli. The
left handers (including strong and mixed handers) were expected to show
no differences between DR and DL stimuli. Figure 1 shows the mean RT
performance with DR and DL stimuli.
The Wilcoxon Matched-Pairs Signed-Ranks Test was chosen as the ap-
propriate statistic to compare the within-group differences between
mean RTs to the DR and DL stimuli. The analysis involved computing the
difference between the mean DR and the mean DL reaction times. The dif-
ferences were then ranked and signed (negative or positive) according
to the sign of the difference score. The positive ranks were summed (T)
and compared to the sum that would be expected if half the ranks were
positive and half were negative.
The analysis of the right handers on DR DL revealed that the dis-
tribution of the ranks was significantly different from chance (T=110,
p=.025). Thus, the right handers were more likely to have a positive
difference between DR and DL. This suggests that the mean RTs to the
DR were longer than to the DL.
The difference between mean RTs to DR and mean RTs to DL was not
significant among the left handers (T=57, p=.10). Further analysis was
conducted to determine if left handers' mean RT to the DL was longer
DIRECTIONAL SCAN: MEAN REACTION TIMES ON DR AND DL STIMULI
FOR RIGHT, LEFT AND MIXED HANDERS
than to the DR. The results were not significant. However, the T-score
(91) was close to the critical T (101). Therefore, there is a trend
suggesting that left handers, in contrast to right handers, had longer
mean RTs to DL than to DR.
The analysis of the mixed left handers revealed that the distribu-
tion of the ranks did not differ significantly from chance. This was
the case for both DR DL (T=47, p<.10) and DL DR (T=58, p<.10).
There were no sex differences in any of the groups.
Within each of the handedness groups, the difference between mean
RTs to DR and DL was highly variable between subjects. That is, one
subject might have a difference of 8 milliseconds, while another subject
might have a difference as high as 80 milliseconds. Therefore, a Chi-
square analysis was performed to determine if the number of subjects
who had longer mean RTs with the DRs was different than the number of
subjects with longer mean RTs on the DLs. The observed frequencies for
longer RTs with DR (right handers=ll, left handers=6, mixed handers=7)
and with DL (right handers=5, left handers=ll, mixed handers=7) were
compared with the expected frequencies. The results revealed a signifi-
cant difference (X =3.85, p<.05) (see Table 1). Since the mixed handers
were equally split between longer DR and longer DL, the significant X
is due entirely to the difference between right and left handers. More
right handers were likely to take longer with the DR (11 of the 16 subjects),
while more of the left handers were likely to take longer with the DL
stimuli (11 of the 17 subjects).
The degree of relationship between handedness and preferred direc-
tion was tested using the (phi) coefficient. In this analysis,
OBSERVED (0) AND EXPECTED (E) FREQUENCY OF SUBJECTS
WITH LONGER DR AND LONGER DL REACTION TIMES
X = 3.85
p < 0.05
handedness (right, left and mixed) and preferred direction (longer DR,
longer DL) are treated as dichotomous variables. The analysis results
in a X value that indicates if handedness is correlated with a prefer-
red direction. Since the i-coefficient is a 2 X 2 design, three compar-
isons were made: right handers vs. left handers; right handers vs. mixed
handers; mixed handers vs. left handers. The results of each comparison
are presented in Table 2. The comparison between right and left handers
revealed a significant (X =11.04, p
ness and preferred direction. The right handers were more likely than
the left handers to be longer on DR, and the left handers were more
likely than the right handers to be longer on DL. The comparison be-
tween right and mixed handers was significant (X =-5.73, p<.05) and indi-
cated that mixed handers were more likely than the right handers to be
longer on DL, and the right handers were more likely than the mixed hand-
ers to be longer on DR. Finally, the comparison between mixed and left
handers also showed a significant relationship between handedness and
preferred direction (X =4.6, p<.05). The mixed handers were more likely
than the left handers to be longer on DR, and the left handers were more
likely than the mixed handers to be longer on DL.
In sum, these results suggest that the relative frequency of longer
RTs to DR is greater in right handers than mixed handers, and greater in
mixed than left handers (right>mixed>left). Furthermore, the relative
frequency of longer RTs to DL is greater in left handers than mixed hand-
ers, and greater in mixed handers than right handers (left>mixed>right).
The mean RTs for DR and DL presented in Figure 1 suggest that there
are between-group differences in mean RTs to DR and to DL. It appears
FREQUENCY OF RIGHT, LEFT AND MIXED HANDERS
WITH LONGER DR AND LONGER DL REACTION TIMES
WITH PHI AND CHI-SQUARE VALUES
X = 4.60