NIH Public Access
Author Manuscript
Vision Res. Author manuscript; available in PMC 2013 June 01.
NIH-PA Author Manuscript
Published in final edited form as:
Vision Res. 2012 June 1; 62: 220–227. doi:10.1016/j.visres.2012.04.011.
Binocular retinal image differences influence eye-position
signals for perceived visual direction
Deepika Sridhar1 and Harold E. Bedell1,2
Harold E. Bedell: HBedell@Optometry.uh.edu
1College
of Optometry, University of Houston. 505 J Armistead Bldg, Houston, TX 77204-2020,
USA
2Center
for NeuroEngineering & Cognitive Science, University of Houston, Houston, TX
77204-4005, USA
Abstract
NIH-PA Author Manuscript
Correctly perceiving the direction of a visible object with respect to one’s self (egocentric visual
direction) requires that information about the location of the image on the retina (oculocentric
visual direction) be combined with signals about the position of the eyes in the head. The WellsHering laws that govern the perception of visual direction and modern restatements of these laws
assume implicitly that retinal and eye-position information are independent of one another. By
measuring observers’ manual pointing responses to targets in different horizontal locations, we
show that retinal and eye-position information are not treated independently in the brain. In
particular, decreasing the relative visibility of one eye’s retinal image reduces the strength of the
eye-position signal associated with that eye. The results can be accounted for by interactions
between eye-specific retinal and eye-position signals at a common neural location.
Keywords
Egocentric visual direction; visual suppression; eye-position; asymmetric vergence; heterophoria;
pointing
1. Introduction
NIH-PA Author Manuscript
The direction of visual objects with respect to one’s self is known as egocentric
(specifically, headcentric) visual direction (EVD). According to the Wells-Hering laws of
visual direction that were proposed approximately two centuries ago (Hering, 1879/1942;
Wells, 1818) perceived EVD results from a combination of the information about the
location of the image on the retina and extra-retinal signals about the position of the eyes in
the head:
© 2012 Elsevier Ltd. All rights reserved.
Corresponding author: Deepika Sridhar, College of Optometry, University of Houston, 505 J Armistead Bldg, Houston, TX
77204-2020, USA. DSridhar.2010@alumni.opt.uh.edu, Phone: +1 713 743 1993, Fax: +1 713 743 2053.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our
customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of
the resulting proof before it is published in its final citable form. Please note that during the production process errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Sridhar and Bedell
Page 2
NIH-PA Author Manuscript
In this formulation, the perceived retinal image location is provided by local sign
information, the direction with respect to the fovea that is associated with stimulation of a
specific retinal location (Helmholtz, 1910/1962; Lotze, 1886; Rose, 1999), and sensed eye
position. Eye-position information is obtained from extra-retinal eye-position signals, which
have been shown to derive from an internal representation of the efferent commands to the
eyes and from proprioceptive signals from the extraocular muscles (Bridgeman & Stark,
1991; Gauthier, Nommay & Vercher 1990; von Holst, 1954).
For perceived EVD to be accurate, the eye-position signals from each eye should be
weighted equally, as assumed by both Wells and Hering for observers with normal binocular
vision. However, subsequent reports indicate that between-eye differences in the weighting
of eye-position information occur in some observers (Simpson, 1992; Sridhar & Bedell,
2011). Therefore, in the general case,
where w1 and w2 are the weights given to the left- and right-eye’s position information,
respectively.
NIH-PA Author Manuscript
Similarly, the retinal component of perceived direction for a binocular visual target depends
to some extent on the characteristics of each eye’s retinal stimulus. Specifically, the
perceived direction of a disparate target is weighted more heavily toward the image with less
blur (Charnwood, 1949), higher luminance (Charnwood, 1949; Francis & Harwood, 1951;
Sridhar & Bedell, 2011), or more contrast (Ding & Sperling, 2007; Mansfield & Legge,
1996; Sridhar & Bedell, 2011), indicating unequal weighting of the retinal information from
the two eyes.
The Wells-Hering laws and modern restatements of these laws assume implicitly that the
retinal and the eye-position contributions to perceived visual direction are independent of
one another (Helmholtz, 1910/1962; Hering, 1879/1942; Ono, 1991; Walls, 1951; Wells,
1818). However, we reported recently that many individuals exhibit similar quantitative
between-eye differences in the relative weighting of the retinal and eye-position information
that contribute to perceived EVD (Sridhar & Bedell, 2011), raising the possibility that these
two sources of information are not independent.
NIH-PA Author Manuscript
Binocular rivalry suppression is a commonly encountered perceptual phenomenon in which
observers are unaware intermittently of part or all of the visual image that is presented to
one eye (e.g., Bharadwaj, O’ Shea, Alais & Parker, 2008; Blake & Logothetis, 2002; Breese,
1909; Helmholtz, 1910/1962; Müller, 1842; Le Conte, 1897; Schor, 1977; Smith, Levi,
Manny, Harwerth & White, 1994; Wheatstone, 1838; Wolfe, 1986). Subjects with
strabismus exhibit clinical suppression when one eye is deviated, to avoid the confusion that
would occur when dissimilar images fall on corresponding retinal points and diplopia when
similar images fall on non-corresponding retinal points (Cooper, Feldman & Pasner, 2000;
Hess, 1991; Holopigian, 1989; Schor, 1977; Serrano-Pedraza, Clarke & Reed, 2011;
Sireteanu, 1982; Smith et al., 1994; Steinbach, 1981; Travers, 1940). In observers without
strabismus, suppression is fostered by dissimilarity between the images in the two eyes, as
occurs for example when the image in one eye is blurred as a result of anisometropia (e.g.,
Heath, Hines & Schwartz, 1986; Humphriss, 1982; Liu & Schor, 1994; Pianta & Kalloniatis,
1998; Schor, Landsman & Erickson, 1987; Shors, Wright & Greene, 1992; Simpson, 1991).
The deviation of one eye has been reported not to influence perceived EVD in individuals
with constant strabismus, who typically behave as if they are unaware of both the retinal and
eye-position information from the deviated eye (Gauthier, Berard, Deransard, Semmlow &
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 3
Vercher, 1985; Mann, Hein & Diamond, 1979). These observations again raise the
possibility that retinal and eye-position signals are not independent.
NIH-PA Author Manuscript
In the studies described here, we test the assumption of the Wells-Hering laws that
independent retinal information and eye-position signals contribute to perceived EVD by
estimating the weighting of the eye-position signals from each eye under conditions of
unequal image visibility in the two eyes. The results demonstrate that the relative weighting
of extra-retinal eye-position signals for the two eyes of normal observers depends on the
relative visibility of the binocular retinal images. A parsimonious way to account for these
findings is to assume that retinal and extra-retinal information interact to determine
perceived EVD at a location in the brain where eye-specific signals exist.
2. General Methods
NIH-PA Author Manuscript
The first experiment investigated the relative weighting of the eye-position signals from the
two eyes on perceived EVD when the image from one eye was suppressed foveally. To
determine whether the perceptual suppression of one eye’s image is necessary for eyeposition signals to be altered, the second experiment compared the weighting of eye-position
information from the two eyes when the image in each eye was of different luminance. In
the third experiment, we assessed the weighting of eye-position signals when one eye was
occluded, and allowed to drift to its heterophoric position, i.e., the angular misalignment of
the occluded non-viewing eye with respect to the viewing eye. A fourth, control experiment
assessed whether the contribution of eye-position information to perceived EVD depends on
the velocity of the eye movements.
At least seven adults participated in each of the experiments described here. All observers
had best-corrected visual acuity of at least 20/20 in each eye and no history of abnormal
binocular vision or ocular motility. Each observer voluntarily provided written informed
consent in accordance with the tenets of the Declaration of Helsinki, after the experimental
protocol was reviewed by the University of Houston Committee for the Protection of Human
Subjects.
NIH-PA Author Manuscript
Perceived EVD was determined from normal observers’ open-loop pointing responses
(Barbeito & Ono, 1979; Bock & Kommerell, 1986; Ono & Weber, 1981; Sridhar & Bedell,
2011; Steinbach & Smith, 1981) to a fixation cross that was presented on the black
background of a 120-Hz frame-rate, gamma-corrected Clinton Monochrome Monoray
monitor. The stimuli were presented at a physical distance of 50 cm. Observers viewed the
monitor in a completely dark room through FE-1 ferro-electric goggles (Cambridge
Research Systems) that were attached to a stand with a chin and head rest. Synchronization
of the video-frames of the monitor with the ferro-electric goggles allowed for dichoptic
image presentation during binocular viewing (in experiments 1, 2 and 4), and for the
presentation of alternate video frames to one eye during monocular viewing (in experiment
3). The monitor is equipped with a DP 104 phosphor that decays to 0.1% of the peak value
in 0.6 ms. Consequently, none of the presented images persisted from one frame to the next.
For experiments 1, 2 and 4, different positions of one eye were produced by presenting that
eye’s image of the cross with different horizontal separations on the monitor from the cross
that was shown to the other eye (Figure 1). In these experiments, the eye that underwent a
change in its final position as the result of the asymmetric vergence demand with respect to
the plane of the monitor is referred to as the varying eye. The eye that undergoes no change
in position as a result of the vergence demand is referred to as the non-varying eye.
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 4
3. Experiment 1: Perceived EVD during binocular suppression
NIH-PA Author Manuscript
Nine observers participated in the first experiment, which compared the contribution of the
eye-position signal for the varying eye to the perceived EVD of a foveal target during and in
the absence of foveal suppression. The suppression of the varying eye’s image was induced
by blurring the image presented to that eye (Humphriss, 1982; Schor et al., 1987; Simpson,
1991), and detection thresholds were measured to determine the highest luminance at which
the blurred foveal target could be presented and still be suppressed (see section 3.2., below).
NIH-PA Author Manuscript
The luminance detection threshold was measured for a cross (39.7 arc min long, and 5.3 arc
min wide) and a 5.3 arc min square suppression check (presented approximately 21 arc min
from the center of the cross) presented to one eye at the center of the monitor for an
unlimited viewing time. Both the cross and the suppression check were filtered with a twodimensional Gaussian to simulate +1.25 D of blur during viewing with a 4 mm pupil (Smith,
Jacobs & Chan, 1989). The blurred cross and suppression check will be referred to
collectively as the blurred target (Figure 2). The detection threshold for the blurred target
was determined also during binocular viewing when a clear 2.5 cd/m2 cross with no
suppression check was presented concurrently to the other eye. For binocular detection
thresholds, the clear cross seen by the non-varying eye was at the center of the monitor and
the blurred target presented to the varying eye was positioned to stimulate 0, 8.55 deg of
convergence, or 2.85 deg of divergence with respect to the plane of the monitor. In both the
monocular and binocular viewing conditions, the observer adjusted the luminance of the
blurred target so that the suppression check just disappeared. The order of the four viewing
conditions (presentation of the blurred target to the right or left eye, for monocular and
binocular viewing), and the asymmetric vergence demand during binocular viewing was
randomized. Four separate detection thresholds were obtained for each monocular viewing
condition and averaged. The binocular detection threshold was estimated twice for each
vergence demand and averaged. The magnitude of foveal suppression was quantified as the
log ratio of the average binocular to the average monocular threshold. Because the size of
the target is small, we assume that the suppression measured in this experiment was limited
to the fovea.
3.1. Assessment of eye-position
NIH-PA Author Manuscript
To ensure that the vergence responses during binocular viewing were approximately equal
to the vergence demand in both the foveal-suppression and non-suppression conditions, in
this and in experiments 2 and 4, fixation disparity was measured for each vergence demand
using a pair of flashed dichoptic unblurred vertical Nonius lines (Ogle, Martens & Dyer,
1967), presented above and below a binocular fixation cross. During the foveal-suppression
condition, the cross seen by the varying eye was blurred, and during the non-suppression
condition the crosses seen by both eyes were clear. The horizontal offset of the Nonius line
seen below the fixation cross by the varying eye that is perceived to be in alignment with the
Nonius line seen above the fixation cross by the non-varying eye provided an estimate of the
fixation disparity.
3.2. Measurement of perceived EVD
The strength of the eye-position signals during the foveal suppression and non-suppression
conditions was inferred from the open-loop pointing responses to targets with different
asymmetric vergence demands. In the foveal-suppression condition, observers viewed the
blurred target with the varying eye and a clear cross (2.5 cd/m2 luminance) with the nonvarying eye (Figure 2). The clear cross was presented at the center of the monitor or at 1.15,
2.29, or 3.44 deg to the right or left of center. The blurred target was displaced on the
monitor with respect to the clear target to stimulate asymmetric vergence demands of 0,
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 5
NIH-PA Author Manuscript
2.85, 5.70, or 8.55 deg of convergence, or 2.85 deg of divergence. The luminance of the
blurred target was set one standard deviation lower than the observer’s binocular detection
threshold (mean = 0.24 cd/m2, see results for detection thresholds), to ensure that it was
completely suppressed. In the non-suppression condition, the blurred target was replaced by
a clear cross with a luminance of 2.5 cd/m2. The foveal-suppression and non-suppression
conditions, the eye that viewed the blurred target during suppression, the target location and
the asymmetric vergence demand were randomized from trial to trial.
On each trial, the observer pressed a button to confirm that the suppression check was not
visible when the clear and the blurred target were superimposed and to indicate that (s)he
was ready to point to the target. The shutters of the goggles then closed in front of both eyes,
and a dim horizontal scale, marked in 0.6 deg intervals and visible only to the experimenter,
replaced the pointing target on the monitor. The observer pointed to the remembered
location of the target using the index finger of the preferred hand and the experimenter noted
the pointing response to the nearest 0.3 deg. A homogeneous white screen was presented for
3 s before the next pointing target, to control dark adaptation and to facilitate the visual
recalibration of sensed eye position (Blouin, Amade, Vercher, Teasdale & Gauthier, 2002).
Each set of trials consisted of 75 pointing responses. The responses from at least two sets of
trials were combined for each observer.
NIH-PA Author Manuscript
Pointing errors were calculated as the difference between where the observer pointed on the
monitor and the location of the target that was seen by the non-varying eye. The constant
error was defined as the average pointing error (calculated separately for the fovealsuppression and non-suppression conditions) when there was no vergence demand with
respect to the plane of the monitor. Pointing errors were corrected for each observer’s
idiosyncratic constant error by subtracting the constant error from each pointing error. No
statistically significant (p > 0.05) difference was observed in the corrected pointing errors to
the target seen by the non-varying eye for the different locations on the monitor and the
corrected pointing responses were plotted against the asymmetric vergence demand,
separately for the right- and left eye-varying conditions and for the foveal-suppression and
non-suppression conditions. Straight lines were fit to the corrected pointing errors to
determine the strength of the eye-position signal for the varying eye in the presence and
absence of foveal suppression.
3.3. Results
NIH-PA Author Manuscript
Average monocular and binocular detection thresholds for the blurred target were 0.17 ±
0.11 (SD) and 0.34 ± 0.13 cd/m2. The average depth of suppression was −0.34 ± 0.21 log
units (across observers, range: −0.11 to −0.66 log units), in agreement with previous studies
in strabismic subjects with suppression, and in non-human primates with experimentally
induced strabismus (Holopigian, 1989; Wensween, Harwerth & Smith, 2001).
For the observers in this experiment, the fixation disparity was small, amounting to 8.95
(±5.78 SD) and 5.26 (±4.20) arc min in the foveal-suppression and non-suppression
conditions, respectively, indicating that in both conditions the asymmetric vergence
response was essentially equivalent to the vergence demand. Nevertheless, the difference in
fixation disparity between the two conditions reaches statistical significance (tdf=8 = 2.32; p
= 0.049). Across observers, the average variability of the pointing responses in the fovealsuppression and non-suppression conditions was 1.88 ± 0.57 and 1.66 ± 0.40 deg,
respectively, which do not differ significantly (repeated tdf=8 = 0.57, p = 0.58, 2-tailed).
Straight lines fit to each observer’s pointing responses as a function of the asymmetric
vergence demand specify the weighting afforded to the position signals for the varying eye
(Figure 3). According to the Wells-Hering laws of visual direction the expected slope of the
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 6
NIH-PA Author Manuscript
fitted lines is 0.5, which would indicate an equal contribution of the position signal for each
eye to perceived EVD. For the 9 observers tested in this experiment, the slopes of the fitted
lines were significantly shallower during the foveal-suppression condition than the nonsuppression condition, indicating a reduced weighting of the eye-position signal from the
suppressed eye (mean absolute values of slope = 0.45 ± 0.08 [SD] and 0.34 ± 0.13 in the
non-suppressed and suppressed conditions, respectively; repeated measures tdf=8 = 3.68, p =
0.006, 2-tailed; Figures 3a, b and 4).
4. Experiment 2: Perceived EVD for targets of unequal luminance
NIH-PA Author Manuscript
The second experiment determined how the relative luminance of the retinal images in the
two eyes influences the strength of the eye-position signal to perceived EVD. Eight
observers participated. In this experiment, the pointing targets presented to the two eyes
(including the suppression check for the varying eye) were unblurred and had either the
same luminance in both eyes, or the luminance of the cross and the suppression check
shown to the varying eye was 4 or 8 times lower than the cross shown to the non-varying
eye. The mean luminance of the two eyes’ images remained equal to 2.5 cd/m2 in all
viewing conditions. The strength of the eye-position signal for each luminance condition
was estimated as in section 3.2 by first calculating the corrected pointing errors and then
fitting straight lines separately to the corrected pointing errors made when the right or the
left eye was the varying eye.
4.1. Results
As expected, the observers in this experiment reported that the suppression check presented
to the varying eye always remained visible. Nevertheless, the slopes of the lines fit to the 8
observers’ pointing responses were significantly shallower when the varying eye viewed a
target that was 4 or 8 times lower in luminance than the non-varying eye, compared to when
both eyes viewed targets of equal luminance (Geisser-Greenhouse corrected post-hoc
comparisons following significant [F df=2,14 = 6.34, p = 0.023] repeated-measures ANOVA;
Fdf=1,14 = 5.55, p = 0.049 for 4 times difference in luminance; Fdf=1,14 = 12.18, p = 0.009,
for 8 times difference in luminance, Figures 3c, d and 4).
5. Experiment 3: Perceived EVD during occlusion of one eye
NIH-PA Author Manuscript
The third experiment measured the contribution of the eye-position signal to perceived EVD
during monocular viewing, i.e., in the complete absence of retinal information from one eye.
Open-loop pointing responses were obtained from 10 observers when either the right or left
eye viewed the target, or when both eyes viewed a target with no vergence demand with
respect to the plane of the monitor. The average pointing error during binocular viewing was
subtracted from the average pointing error during monocular viewing to indicate the position
information that was available from the occluded eye during monocular viewing.
As in experiments 1 and 2, above, accurate eye alignment during binocular viewing was
verified by fixation disparity measurements. During monocular viewing, the fixation cross
was visible to only one eye. Therefore, the physical offset between the flashed dichoptic
Nonius lines that produced the perception of vertical alignment specified the direction and
magnitude of the observer’s lateral heterophoria.
5.1. Results
During monocular viewing, the average angle of hetereophoria for the 10 observers was 2.10
± 2.17 [SD] deg (range = 0.12 to 7.18 deg) in the divergent direction. Because the position
of the occluded eye was not manipulated in this experiment, Figure 5 plots one data point
per observer, corresponding to the average difference between the pointing responses during
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 7
NIH-PA Author Manuscript
binocular fixation of the cross and during fixation when one eye was occluded. The straight
line fit to the observers’ average monocular pointing responses was constrained to have a yintercept of zero, based on the Wells-Hering prediction that no systematic pointing error
should occur in the absence of hetereophoria.* The best fitting slope is 0.27 ± 0.05, which is
significantly greater than a slope of zero (tdf=9 = 5.08, p = 0.006), but also is significantly
less than the value of 0.5 that is predicted by the Wells-Hering laws of visual direction (tdf=9
= 4.49, p = 0.003).
6. Target visibility and eye-position information
NIH-PA Author Manuscript
Figure 4 compares the average slopes of the pointing vs. eye-position functions for the
different conditions of relative target visibility. A comparison of the dark gray bars indicates
that no significant difference exists between the slopes fit to the observers’ pointing
responses in the non-suppression condition of Experiment 1 (central pair) and the equalluminance condition of experiment 2 (leftmost set of 3 bars; Welch’s two-sample tdf=14, =
0.996, p = 0.336, 2 tailed). Neither of these average slopes differ significantly from the
expected value of 0.5 (for the non-suppression condition, tdf=8 = 1.89, p = 0.095; for the
equal-luminance condition, tdf=7 = 0.36, p = 0.73). The average slope of the pointing vs.
eye-position function fell to 0.41 when the target shown to the varying eye was 8 times
lower in luminance than the target seen by the fellow eye (light gray bar in the left set of 3
bars), and to 0.35 when a dim, blurred target was presented to the varying eye and
perceptually suppressed (light gray bar in the central pair of bars). In the condition when one
eye was occluded, the strength of the eye-position signal was calculated as the difference in
pointing during monocular vs. binocular viewing, divided by the observers’ measured
heterophoria. Figure 4 (rightmost bar) shows that the contribution of the occluded eye’s
position signal to perceived EVD is on average less than half of the value expected from the
Well-Hering’s laws, with substantial between-observer variability.
7. Experiment 4: Perceived EVD and asymmetric vergence velocity
NIH-PA Author Manuscript
The velocity of the vergence eye movement that occurs following the occlusion of one eye is
typically slower (Barnard & Thomson, 1995; Kim, Granger-Donetti, Vicci & Alvarez, 2010;
Ludvigh, McKinnon, & Zaitzeff, 1964; Park & Shebilske, 1991; Peli & McCormack, 1983)
than vergence eye movements in response to horizontal image disparity during binocular
viewing (Maxwell, Tong & Schor, 2010; Semmlow, Hung & Ciuffreda, 1986). It is
therefore possible that the low velocity of the eye movement that occurs following the
occlusion of one eye, compared to the higher velocity of vergence during binocular viewing
is responsible for the decreased eye-position-signal weight for the occluded eye in
Experiment 3. To assess whether the velocity of asymmetric vergence influences estimates
of perceived EVD, nine observers viewed identical unblurred crosses (without suppression
checks) in the two eyes that either (1) appeared with one of several asymmetric vergence
demands with respect to the monitor, or (2) appeared initially with no vergence demand,
after which the target seen by the varying eye moved at a velocity of either 0.75 or 1.5 deg/s
to produce the same final amplitudes of asymmetric vergence. The velocities of 0.75 and 1.5
deg/s were chosen to approximately match the range of eye velocities reported previously
after the occlusion of one eye (Barnard & Thomson, 1995; Kim, Granger-Donetti, Vicci &
Alvarez, 2010; Ludvigh, McKinnon, & Zaitzeff, 1964; Park & Shebilske, 1991; Peli &
McCormack, 1983).
*If the regression line is not constrained to pass through (0, 0), the fitted intercept is −0.420, which does not differ significantly from
an intercept of zero (tdf=8= 1.92, p = 0.092).
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 8
NIH-PA Author Manuscript
When the target appeared initially with an asymmetric vergence demand, the observer either
pointed at the target immediately or after a delay of 9 s. This is the duration required to
achieve the largest vergence demand when the targets seen by both eyes appeared initially
with no vergence demand and then the target seen by the varying eye moved at a velocity of
0.75 deg/s. The different amplitudes, which are identical to those used in Experiments 1 and
2, and velocities of vergence demand and the varying eye changed randomly from trial to
trial.
7.1. Results
The observers’ pointing responses do not differ significantly among the different eyevelocity conditions in Experiment 4 (repeated-measures ANOVA, F3,18 = 0.127, p = 0.943,
Geisser-Greenhouse corrected). This outcome indicates that the reduced eye-position signal
for the occluded eye in the third experiment is unlikely to be attributable to the slow
dynamics of the phoric eye-movement response.
8. Discussion
NIH-PA Author Manuscript
We showed previously that normal individuals frequently exhibit between-eye differences in
the weighting of retinal and eye-position information and that, within individuals, the
weighting of these two types of information is correlated (Sridhar & Bedell, 2011). The
results presented here indicate that the relative visibility of the retinal images in the two eyes
influences the weighting of the eye-position signals for perceived EVD. Previously, Blouin
et al. (2002) determined that information about the perceived direction of binocular eccentric
gaze is approximately 20 percent weaker in darkness or with minimal visual stimulation,
compared to conditions in which the visual stimulation is more extensive. Considering the
results of our second experiment, it is not clear whether the reduction of eye-position
weighting that we found when a blurred, low luminance target is presented to one eye is
attributable to the suppression of that eye’s visual information, to the associated reduction of
target luminance (on average, by a factor of 10.6 times, compared to the luminance of the
target presented to the non-suppressed eye), or both.
NIH-PA Author Manuscript
The application of a common weighting factor for retinal and eye-position information and
the interaction between retinal information and eye-position signals would occur most
parsimoniously at a common neural site. However, the characteristics of the neural signals
that are required for these interactions to occur make the identification of a likely neural
location difficult. For the relative visibility of the two retinal images to influence the
weighting of eye-position signals, both the retinal and the eye-position signals at any
candidate site should be eye-specific. Eye-specific retinal information is available primarily
in neurons at lower levels of processing in the visual pathway (for example in the lateral
geniculate nucleus and in cortical areas V1 and V2 (Hubel & Wiesel, 1962; Zeki, 1978a),
whereas the neurons in higher visual areas carry primarily binocular information (Zeki,
1978b). On the other hand, neurons that combine eye-position and visual information are
either absent or minimal in areas V1 and V2, and are increasingly common in higher visual
areas, such as V3, V3A, LIP and MST (Balslev & Miall, 2008; Galletti & Battaglini, 1989;
Gur & Snodderly, 1997; Nakamura & Colby, 2002; Prevosto, Graf & Ugolini, 2009; Wang,
Zhang, Cohen & Goldberg, 2007). It remains unknown whether the neuronal eye-position
signals in any of these higher visual areas are eye-specific, as would be required for a
change in the weighting of eye-position signals to be implemented.
A recent psychophysical result indicates that eye-specific visual information must be
available in the higher visual areas, such as MT or MST, that process two-dimensional and
three-dimensional motion information (Rokers, Czuba, Cormack & Huk, 2011). Lehky
(2011) attempted to account for the availability of monocular visual information beyond the
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 9
NIH-PA Author Manuscript
neural locus of binocular visual combination by demonstrating that the visual information
available to the left and right eyes can in principle be recovered from the responses of
binocular visual neurons that exhibit different degrees of ocular dominance. Similarly, an
appropriate combination of neural signals that specify conjugate and convergence eye
positions (essentially reversing the strategy envisioned by Hering (Hering, 1868/1977, also
Minken, Gielen & Van Gisbergen, 1995) to generate oculomotor commands would allow
the brain to recover separate eye-position signals for the left and right eyes. Implementation
of these or similar neural strategies could account for our results that retinal and eye-position
information are not independent of each other, as has long been assumed, by allowing for
interactions to occur at a higher level of visual processing where visual and eye-position
signals are available concurrently.
NIH-PA Author Manuscript
Our finding that the relative visibility of the retinal images affects the weighting of eyeposition signals also addresses a long-standing controversy as to whether the brain takes the
position of the non-viewing eye into account during monocular viewing (Erkelens, 2000;
Gauthier, Nommay & Vercher, 1990; Helmholtz, 1910/1962; Hering, 1879/1942; Morgan,
1947; Ono & Weber, 1981; Park & Shebilske, 1991; Simpson, 1992; Wells, 1818). Both
Hering (1879/1942) and Helmholtz (1910/1962) believed that eye-position signals from the
non-seeing eye are taken into account when either eye is occluded. However, Helmholtz
predicted that eye position is taken into account more completely when the favored (or
dominant) eye is occluded and Walls (1951) argued that only the position of the dominant
eye is accounted for during occlusion. Most (but not all, see Walls, 1951) studies confirm
that the eye-position signals from an occluded eye contribute to the perceived EVD of a
target that is visible only to the unoccluded eye. In agreement with previous reports (Ono &
Weber, 1981; Park & Shebilske, 1991; Simpson, 1992), our data show that the contribution
of the eye-position signal from the non-viewing eye is not complete, and exhibits
considerable variability among observers. For example, Ono and Weber’s (1981) data
indicate a weighting of eye position for the occluded eye of 0.28 ± 0.28 in a sample of 19
observers, in contrast to an expected weighting of 0.5 based on the Wells-Hering’s laws.
This difference in eye-position weighting between binocular and monocular viewing
conditions is not accounted for by differences in the eye movements during binocular and
monocular viewing and raises questions about the quantitative contribution of eye-muscle
proprioception to signaled eye position, which has been evaluated previously only during
monocular viewing (Bridgeman & Stark, 1991; Gauthier, Nommay & Vercher, 1990).
Acknowledgments
NIH-PA Author Manuscript
We thank Drs. Saumil Patel and Scott Stevenson for helpful suggestions and Dr. Ying-sheng Hu for statistical
advice. This research was supported in part by Core Center grant, P30 EY07551 from the National Eye Institute.
References
Balslev D, Miall RC. Eye position representation in human anterior parietal cortex. Journal of
Neuroscience. 2008; 28:8968–8972. [PubMed: 18768690]
Barbeito R, Ono H. Four methods of locating the egocenter: A comparison of their predictive validities
and reliabilities. Behavior Research Methods & Instrumentation. 1979; 11:31–36.
Bharadwaj R, O’Shea RP, Alais D, Parker A. Probing visual consciousness: Rivalry between eyes and
images. Journal of Vision. 2008; 8:1–13. [PubMed: 19146280]
Blake R, Logthetis NK. Visual competition. Nature Reviews Neuroscience. 2002; 3:1–11.
Bock O, Kommerell G. Visual localization after strabismus surgery is compatible with the “outflow”
theory. Vision Research. 1986; 11:1825–1829. [PubMed: 3617525]
Barnard NA, Thomson WD. A quantitative analysis of eye movements during the cover test: a
preliminary report. Ophthalmic & Physiological Optics. 1995; 15:413–419. [PubMed: 8524567]
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 10
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Blouin J, Amade N, Vercher JL, Teasdale N, Gauthier GM. Visual signals contribute to the coding of
gaze direction. Experimental Brain Research. 2002; 144:281–292.
Breese BB. Binocular rivalry. Psychological Review. 1909; 16:410–415.
Bridgeman B, Stark L. Ocular proprioception and efference copy in registering visual direction. Vision
Research. 1991; 31:1903–1913. [PubMed: 1771774]
Charnwood JRB. Observations on ocular dominance. Optician. 1949; 116:85–88.
Cooper J, Feldsman J, Pasner K. Intermittent exotropia: stimulus characteristics affect tests for retinal
correspondence and suppression. Binocular vision & Strabismus Quarterly. 2000; 15:131–140.
[PubMed: 10893455]
Ding, J.; Sperling, G. Binocular combination: Measurements and a model. In: Harris, L.; Jenkin, M.,
editors. Computational vision in neural and machine systems. Cambridge, UK: Cambridge
University Press; 2007. p. 257-305.
Erkelens CJ. Perceived direction during monocular viewing is based on signals of the viewing eye
only. Vision Research. 2000; 40:2411–2419. [PubMed: 10915882]
Francis, JL.; Harwood, KA. Transactions of international optical congress. London: British Optical
Association; 1951. The variation of the projection center with differential stimulus and its relation
to ocular dominance; p. 75-87.
Galletti C, Battaglini PP. Gaze-dependent visual neurons in area V3A of monkey prestriate cortex.
Journal of Neuroscience. 1989; 9:1112–1125. [PubMed: 2703870]
Gauthier, GM.; Berard, PV.; Deransard, JL.; Semmlow, JL.; Vercher, JL. Adaptation processes
resulting from surgical correction of strabismus. In: Keller, EL.; Zee, DS., editors. Adaptive
processes in visual and oculomotor systems. New York: Pergamon Press; 1985. p. 185-189.
Gauthier GM, Nommay D, Vercher JL. The role of ocular muscle proprioception in visual localization
of targets. Science. 1990; 249:58–61. [PubMed: 2367852]
Gur M, Snodderly DM. Visual receptive fields of neurons in primary visual cortex (V1) move in space
with the eye movements of fixation. Vision Research. 1997; 37:257–265. [PubMed: 9135859]
Heath DA, Hines C, Schwartz F. Suppression behavior analyzed as a function of monovision addition
power. American Journal of Physiological Optics. 1986; 63:198–201.
Helmholtz, H. Helmholtz’s treatise on physiological optics. Vol. III. New York: Dover Publications,
Inc; 1910/1962. The direction of vision.
Hering, E. On the bifixation of the double eye. In: Bridgeman, B.; Stark, L., editors. The theory of
binocular vision. New York: Plenum Press; 1868/1977.
Hering, E. Spatial sense and movements of the eye. American Academy of Optometry; Baltimore:
1879/1942.
Hess R. The site and nature of suppression in squint amblyopia. Vision Research. 1991; 31:111–117.
[PubMed: 2006544]
Holopigian K. Clinical suppression and binocular rivalry in visual suppression: the effects of stimulus
strength on the depth of suppression. Vision Research. 1989; 29:1325–1333. [PubMed: 2635462]
Hubel DH, Wiesel TN. Receptive fields, binocular interaction and functional architecture in the cat’s
visual cortex. Journal of Physiology. 1962; 160:106–154. [PubMed: 14449617]
Humphriss D. The physiological septum. An investigation into its function. American Journal of
Physiological Optics. 1982; 59:639–41.
Le Conte, J. Sight. An exposition of the principles of monocular and binocular vision. D Appleton &
Co; New York: 1897.
Lehky SR. Unmixing binocular signals. Frontiers of Computational Neuroscience. 2011; 5:1–7.
Liu L, Schor C. The spatial properties of binocular suppression zone. Vision Research. 1994; 34:937–
947. [PubMed: 8160405]
Lotze, H. The intuitions of space. In: Ladd, GT., editor. Outlines of psychology, dictated portions of
the lectures of Hermann Lotze. Ginn & Co; Boston: 1886. p. 47-65.
Ludvigh E, McKinnon P, Zaitzeff L. Temporal course of the relaxation of binocular duction (fusion)
movements. Archives of Ophthalmology. 1964; 71:389–399. [PubMed: 14100765]
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 11
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Kim EH, Granger-Donetti B, Vicci VR, Alvarez TL. The relationship between phoria and the ratio of
convergence peak velocity. Investigative Ophthalmology & Visual Science. 2010; 51:4017–4027.
[PubMed: 20335602]
Mann VA, Hein A, Diamond R. Localization of targets by strabismic subjects: contrasting patterns in
constant and alternating suppressors. Perception & Psychophysics. 1979; 25:29–34. [PubMed:
432088]
Mansfield JS, Legge GE. The binocular computation of visual direction. Vision Research. 1996;
36:27–41. [PubMed: 8746240]
Maxwell JS, Tong J, Schor CM. The first and second order dynamics of accommodative convergence
and disparity convergence. Vision Research. 2010; 50:1728–1739. [PubMed: 20561972]
Minken AW, Gielen CC, Van Gisbergen JA. An alternative three-dimensional interpretation of
Hering’s equal-innervation law for version and vergence eye movements. Vision Research. 1995;
5:93–102. [PubMed: 7839614]
Morgan MW. The direction of visual lines when fusion is broken as in duction tests. American Journal
of Optometry & Archives of American Academy of Optometry. 1947; 24:8–12. [PubMed:
20280308]
Müller, J. On the action of the retina, optic nerve and sensation of vision. In: Baly, William, translator.
Elements of physiology. Taylor & Walton; London: 1842. p. 1162-1214.
Nakamura K, Colby CL. Updating of the visual representation in monkey striate and extrastriate cortex
during saccades. Proceedings of the National Academy of Science USA. 2002; 99:4026–4031.
Ogle, KN.; Martens, TG.; Dyer, JA. Oculomotor Imbalance in Binocular Vision and Fixation
Disparity. Philadelphia: Lea & Febiger; 1967.
Ono, H. Binocular visual direction of an object when seen as single or double. In: Regan, D., editor.
Vision and Visual Dysfunction Vol 9, Binocular vision. London: Macmillan; 1991. p. 1-18.
Ono H, Weber EU. Nonveridical visual direction produced by monocular viewing. Journal of
Experimental Psychology: Human Perception & Performance. 1981; 7:937–947. [PubMed:
6457113]
Park K, Shebilske WL. Phoria, Hering’s laws and monocular perception of direction. Journal of
Experimental Psychology: Human Perception & Performance. 1991; 17:219–231. [PubMed:
1826313]
Pianta MJ, Kalloniatis M. Characteristics of anisometropic suppression: simple reaction time
measurements. Perception & Psychophysics. 1998; 60:491–502. [PubMed: 9599998]
Peli E, McCormack G. Dynamics of cover test eye movements. American Journal of Optometry &
Physiological Optics. 1983; 60:712–724. [PubMed: 6624871]
Prevosto V, Graf W, Ugolini G. Posterior parietal cortex areas MIP and LIPv receive eye position and
velocity inputs via ascending preposito-thalamocortical pathways. European Journal of
Neuroscience. 2009; 30:1151–1161. [PubMed: 19735295]
Rokers B, Czuba TB, Cormack LK, Huk AC. Motion processing with two eyes in three dimensions.
Journal of Vision. 2011; 11:1–19.
Rose D. The historical roots of the theories of local signs and labeled lines. Perception. 1999; 28:675–
685. [PubMed: 10664763]
Schor CM. Visual stimuli for strabismic suppression. Perception. 1977; 6:583–593. [PubMed: 593791]
Schor C, Landsman L, Erickson P. Ocular dominance and the interocular suppression of blur in
monovision. American Journal of Optometry & Physiological Optics. 1987; 64:723–730.
[PubMed: 3688174]
Semmlow JL, Hung GK, Ciuffreda KJ. Quantitative assessment of disparity vergence components.
Investigative Ophthalmology & Visual Science. 1986; 27:558–564. [PubMed: 3957575]
Serrano-Pedraza I, Clarke M, Read JCA. Single vision during ocular deviation in intermittent
exotropia. Ophthalmic & Physiological Optics. 2011; 31:45–55. [PubMed: 21158884]
Shors TJ, Wright K, Greene E. Control of interocular suppression as a function of differential image
blur. Vision Research. 1992; 32:1169–1175. [PubMed: 1509709]
Simpson TL. The suppression effect of simulated anisometropia. Ophthalmic & Physiological Optics.
1991; 11:350–358. [PubMed: 1771072]
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 12
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Simpson, TL. Thesis (PhD). University of Houston; 1992. The vergence components and observerrelative direction.
Sireteanu R. Binocular vision in strabismic humans with alternating fixation. Vision Research. 1982;
22:889–896. [PubMed: 7135851]
Smith EL, Fern K, Manny RE, Harwerth RS. Interocular suppression produced by rivalry stimuli: a
comparison of normal and abnormal binocular vision. Optometry & Vision Science. 1994;
71:479–491. [PubMed: 7970564]
Smith G, Jacobs RJ, Chan DC. Effect of defocus on visual acuity as measured by source and observer
methods. Optometry & Vision Science. 1989; 66:430–435. [PubMed: 2771329]
Sridhar D, Bedell HE. Relative contributions of the two eyes to perceived egocentric visual direction
in normal binocular vision. Vision Research. 2011; 51:1075–1085. [PubMed: 21371491]
Steinbach M. Alternating exotropia: temporal course of the switch in suppression. Vision Research.
1981; 22:889–896.
Steinbach M, Smith R. Spatial localization after strabismus surgery: Evidence for inflow. Science.
1981; 213:1407–1409. [PubMed: 7268444]
Travers TÁB. The origin of abnormal retinal correspondence. British Journal of Ophthalmology. 1940;
24:58–64. [PubMed: 18169675]
Wang X, Zhang M, Cohen IS, Goldberg ME. The proprioceptive representation of eye position in
monkey primary somatosensory cortex. Nature Neuroscience. 2007; 10:640–646.
von Holst E. Relations between the central nervous system and the peripheral organs. British Journal
of Animal Behavior. 1954; 2:89–94.
Walls GL. The problem of visual direction. American Journal of Optometry & Archives American
Academy of Optometry. 1951; 28:173–212.
Wheatstone C. Contributions to the physiology of vision-Part the first. On some remarkable, and
hitherto unobserved, phenomenon of binocular vision. Philosophical Transactions of the Royal
Society. 1838; 128:371–394.
Wolfe J. Stereopsis and binocular rivalry. Psychological Review. 1986; 93:269–282. [PubMed:
3749398]
Wells, WC. Two essays, one upon single vision with the two eyes; the other on dew. Longman, Hurst;
Rees, Orms; Brown; Hurst, editors. Archibald Constable and Co; Edinburgh: Robinson and Co;
London, UK: 1818.
Wensween J, Harwerth R, Smith EL. Clinical suppression in monkeys reared with abnormal binocular
visual experience. Vision Research. 2001; 41:1593–1608. [PubMed: 11343724]
Zeki SM. The third visual complex of rhesus monkey prestriate cortex. Journal of Physiology. 1978a;
277:245–272. [PubMed: 418175]
Zeki SM. Uniformity and diversity of structure and function in rhesus monkey prestriate visual cortex.
Journal of Physiology. 1978b; 277:273–290. [PubMed: 418176]
NIH-PA Author Manuscript
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 13
Highlights
NIH-PA Author Manuscript
•
Eye-position and retinal information determine perceived egocentric direction.
•
It is assumed that eye-position and retinal information are processed
independently.
•
Here we show that these two sources of information interact with one another.
•
Monocular eye-position signals are likely to occur at higher cortical visual areas.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 14
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 1.
NIH-PA Author Manuscript
The Wells-Hering laws of visual direction predict that the perceived egocentric visual
direction (EVD) of a foveally viewed target should vary systematically with the positions of
both eyes, even if only one eye views the target. The three arrows in the left panel show the
predicted direction of a foveal target for three different positions of the left eye. In this
example, virtual targets farther and nearer than the plane of the monitor (i.e., at the gray
square and circle) are produced by shifting the location of the target that is seen by the left
eye (dotted gray crosses to the left and right of the cross seen by the right eye at the center of
the monitor). The thin solid and dashed lines represent the left and right eyes’ lines of sight,
which correspond to three difference asymmetric vergence demands. The right panel
illustrates how pointing responses are expected to change as a function of the left eye’s
position. If the position of both eyes is weighted equally, then the line fit to the pointing
responses is predicted to have a slope of 0.5. A shallower slope indicates that the position
signal for the varying left eye is weighted less in the computation of perceived EVD than the
position signal for the non-varying right eye. If the position of the right eye varies, then the
expected slope of the line fit to the pointing responses is −0.5.
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 15
NIH-PA Author Manuscript
Figure 2.
Stimuli for Experiment 1. The images of the cross and suppression check (below and right
of the cross) seen by the varying eye (left panel) were blurred. To measure monocular
detection thresholds, only the blurred image was presented to one eye. To measure detection
thresholds during binocular viewing, a clear cross (and no suppression check) was presented
concurrently to the other eye. Similarly, during the measurement of perceived EVD, the two
eyes viewed images that were presented dichoptically.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Vision Res. Author manuscript; available in PMC 2013 June 01.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Sridhar and Bedell
Vision Res. Author manuscript; available in PMC 2013 June 01.
Page 16
Sridhar and Bedell
Page 17
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 3.
Sample plots of corrected pointing errors vs. vergence demand (i.e., the position of the
varying eye) for the foveal-suppression and non-suppression conditions in experiment 1 (top
two panels, a and b) and for the conditions of different target luminance in the two eyes in
experiment 2 (bottom two panels, c and d). Triangles and circles in the top two plots indicate
the corrected pointing errors obtained when the left or the right eye position varies,
respectively. The best fitting lines to the corrected pointing responses as a function of the
asymmetric vergence demand are shown. The different symbols in the bottom two plots
indicate pointing responses for different interocular luminance ratios, when either the left or
the right eye position varies. Specifically, black unfilled, grey, and filled black circles
indicate the corrected pointing errors for 1:1, 4:1 and 8: 1 luminance ratios, respectively.
Bold black lines, grey lines, and dotted black lines indicate the best-fitting lines for 1:1, 4:1
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 18
and 8:1 luminance ratios, respectively. Symbol and line conventions in panel 3d are same as
those in panel 3c.
NIH-PA Author Manuscript
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 19
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 4.
NIH-PA Author Manuscript
Absolute values of the slopes fit to the pointing vs. eye-position data, averaged for all
observers and for changes in the position of the left and right eyes in the different
experimental conditions. Slope values are interpreted to indicate the eye-position weighting
for the varying eye. The leftmost 3 bars indicate averaged right- and left-eye absolute slope
values when both eyes viewed targets of equal luminance (dark gray bar), and when the
luminance of the target seen by one eye was lower by a factor of 4 or 8 times (stippled and
light gray bars, respectively). The middle pair of bars indicates the average right- and lefteye absolute slope values in the absence of suppression (dark gray bar) and during
suppression of the varying eye’s fixation target (light gray bar). The rightmost bar indicates
the averaged magnitude of the eye-position signal during occlusion of the left or the right
eye. The eye-position signal during occlusion was calculated for each observer as the
corrected pointing error divided by the lateral heterophoria, averaged for occlusion of the
left and right eyes. Error bars indicate ±1 SE of the mean.
Vision Res. Author manuscript; available in PMC 2013 June 01.
Sridhar and Bedell
Page 20
NIH-PA Author Manuscript
NIH-PA Author Manuscript
Figure 5.
Relationship between corrected pointing errors and lateral heterophoria. Each data point
represents the average corrected pointing error during occlusion of the left and right eyes for
one observer, with positive errors in the direction predicted by the Well’s-Hering laws. The
continuous and the dashed lines represent the best-fit line to the data (constrained to have a
y-intercept of 0) and the Wells-Hering slope prediction of 0.5.
NIH-PA Author Manuscript
Vision Res. Author manuscript; available in PMC 2013 June 01.