Vision Research 47 (2007) 1–7

www.elsevier.com/locate/visres

Stimulus Xicker alters interocular grouping during binocular rivalry

Tomas Knapen a, Chris PaVen b, Ryota Kanai c, Raymond van Ee a,¤
a
Helmholtz Institute, Utrecht University, Princetonplein 5, 3584 CC Utrecht, The Netherlands
b
Helmholtz Institute, Utrecht University, Heidelberglaan 2, 3584 CS Utrecht, The Netherlands
c
Division of Biology, California Institute of Technology, M/C 114-96, Pasadena, CA 91125, USA

Received 10 April 2006; received in revised form 7 August 2006

Abstract

When the two eyes are presented with suYciently diVerent stimuli, the stimuli will engage in binocular rivalry. During binocular
rivalry, a subject’s perceptual state alternates between awareness of the stimulus presented to the right eye and that presented to the left
eye. There are instances in which competition is not eye-based, but instead takes place between stimulus features, as is the case in Xicker
and switch rivalry (F&S). Here we investigate another such instance, interocular grouping, using a Diaz–Caneja type stimulus in conjunc-
tion with synchronous stimulus Xicker. Our results indicate that stimulus Xicker increases the total duration of interocularly bound per-
cepts, and that this eVect occurs for a range of temporal Xicker frequencies. Furthermore, the use of contrast-inversion Xicker causes a
decrease of total dominance duration of the interocularly bound percepts. We argue that diVerent Xickering regimes can be used to diVer-
entially stimulate lower and higher levels of visual processing involved in binocular rivalry. We propose that the amount of interocularly
combined pattern-completed percept can be regarded as a measure of the level at which binocular rivalry is resolved.
© 2006 Elsevier Ltd. All rights reserved.

Keywords: Bistable perception; Binocular rivalry; Interocular grouping; Stimulus rivalry; Flicker and switch rivalry; Shape perception

1. Introduction blind spot have been shown to correlate with perceptual

alternations during binocular rivalry (Tong & Engel, 2001).
There has been ongoing debate regarding the nature of Since the blind spot in V1 receives solely monocular aVer-
binocular rivalry, the process of perceptual alternation that ents, this is strong evidence for a monocular basis of binoc-
occurs when the two eyes view suYciently diVerent stimuli ular rivalry suppression.
(Blake & Logothetis, 2002; Leopold & Logothetis, 1999). There is, however, also evidence supporting the contrary
One subject of debate is whether suppression during binoc- hypothesis. Data from monkey physiology suggest that the
ular rivalry acts on eye-based representations or on higher suppression during binocular rivalry increases up the visual
level representations dependent on stimulus properties. hierarchy, with relatively little percept-dependent modula-
The Wrst view is substantiated by the fact that when an tion of cell activity in V1 (Leopold & Logothetis, 1996).
eye is suppressed, detection thresholds in a range of modal- Psychophysical examples of ‘stimulus’, or ‘pattern’ rivalry
ities are increased in the suppressed eye (Fox & Check, can be divided into two categories: interocular grouping
1968, 1972; Wales & Fox, 1970). Also, when the monocular and Xicker and switch (F&S) rivalry. Interocular grouping
half-image stimuli are switched between the eyes, percep- occurs when the stimuli used in binocular rivalry are spa-
tion will follow the switches as if suppression during rivalry tially non-uniform but can be recombined between the eyes
were based on the eyes’ images (Blake, Westendorf, & such that uniform shapes may alternate in subjects’ percep-
Overton, 1980). Furthermore, BOLD responses in the V1 tion (Kovacs, Papathomas, Yang, & Feher, 1996, but also
see Lee & Blake, 2004). Thus, perception alternates not
*
Corresponding author. between images projected into each eye, but between
E-mail address: [email protected] (R. van Ee). higher-level interpretations of the stimuli. F&S rivalry

0042-6989/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.visres.2006.09.007
2 T. Knapen et al. / Vision Research 47 (2007) 1–7

results from a stimulus presentation technique that is com- they are a result of eye-independent information. In F&S
posed of two operations, both of which are necessary to rivalry, the information from the two eyes is combined tem-
create F&S rivalry. The Wrst is to Xicker the stimulus, in an porally, leading to percepts that are based alternatingly on
on–oV regime, at frequencies of 15–20 Hz. Second, the stim- the right eye image and the left eye image and are in this
ulus’ monocular half-images are swapped between the eyes manner independent of the stimulated eye. In pattern com-
at 1.5 Hz, i.e. presentation periods are 333 ms in each eye. pletion dependent on interocular grouping, information
The eVect of the combination of these operations is that from the two eyes is combined spatially, resulting in per-
subjects perceive ‘normal’ binocular rivalry (Logothetis, cepts of shapes that obey, for instance, the principle of col-
Leopold, & Sheinberg, 1996), instead of the perceptual linearity more strongly than do the respective monocular
alternations at 3 Hz predicted by an eye-based suppression images. Both stimulus paradigms, however, elicit ‘pattern
hypothesis of binocular rivalry. Although dependent on a rivalry’ percepts that are independent of the monocular
limited range of stimulus parameters (Lee & Blake, 1999, half-images and thus can be used for the investigation of
but see Bonneh, Sagi, & Karni, 2001), this phenomenon is a the eye-dependence of binocular rivalry.
strong indicator that binocular rivalry can indeed occur Our stimuli combine characteristics of both of these
between representations of stimulus features, and can occur stimulus presentation paradigms and are, because they lack
independently of the stimulated eye. switches between the eyes, speciWcally suited for the investi-
We used a novel stimulus paradigm, consisting of a com- gation of the eVects of the temporal properties of the Xicker
bination of interocular grouping and the Xicker component component on interocular pattern combination. We con-
of F&S rivalry to investigate whether stimulus Xicker ducted three experiments designed to investigate the inXu-
increases interocular grouping. To this end, we used the ence of stimulus Xicker on the preponderance of interocular
stimulus Wrst conceived by Diaz–Caneja (1928), (translated pattern combination. After a Wrst proof of principle experi-
by Alais, O’Shea, Mesana-Alais, & Wilson, 2000), shown in ment, we altered temporal frequency of on–oV Xicker in a
Fig. 1A. This stimulus type (which is in our case composed second experiment and also changed the type of Xicker
of horseshoe-shaped gratings presented to each eye) has while leaving the frequency content identical in a third
been used previously, for instance to investigate the impor- experiment. We show that interocular pattern combination
tance of interhemispheric connections during rivalry increases due to stimulus Xicker; that this eVect is indepen-
(O’Shea & Corballis, 2005). Aside from perceiving eye- dent of temporal frequency; yet does depend on the on–oV
based binocular rivalry between the horseshoe shapes, sub- transients that accompany on–oV Xicker.
jects viewing this stimulus report percepts of full circles and
full line patterns. These percepts are the result of interocu- 2. Experiment 1: stimulus Xicker increases interocular
lar grouping and pattern completion. Note that these per- grouping
cepts diVer from F&S rivalry percepts in the way in which
2.1. Methods
A Stimuli
1.6˚ Six observers participated in the experiment, one of
1 2 which was aware of the hypotheses (author TK). All had
1.6˚ normal or corrected-to-normal vision. They viewed the
dichoptic stimuli, renderings of which are depicted in
Fig. 1A, through a mirror stereoscope at a viewing dis-
tance of 57 cm. The stimuli were presented on a 22 in.
LaCie monitor running at a resolution of 1600 £ 1200 and
B a refresh frequency of 75 Hz, driven by an Apple G4 com-
Percepts puter using custom software. The background was black
Pattern completed Piecemeal rivalry
(luminance 0.06 cd/m2), and a surrounding pattern (white,
1 2
luminance 71.9 cd/m2) of crosses together with concentric
a a
circles directly surrounding the stimulus provided ample
aid for correct binocular fusion. Stimuli were composed
b b of circular patches (diameter 1.6 degrees) of sine-wave
Monocularly based Monocularly based
luminance-modulated gratings, either concentric or linear.
Fig. 1. Stimuli and accompanying percepts. (A) Renderings of the stimuli
Grating contrast was set to 75% Michelson; spatial fre-
used in the experiments, each of which was presented to one eye while the quency of the gratings was 5.5 cycles/degree. Gratings
other stimulus was presented to the other eye; (1) The Diaz–Caneja type were bisected along the vertical meridian and recombined
horseshoe-shaped stimuli. (2) The circle and line stimuli. (B) Possible per- to produce the horseshoe shaped Diaz–Caneja stimuli.
cepts for stimuli from panel (A1) a, pattern completion percepts (reported Under Xicker conditions, stimuli were presented to both
as circle and line percepts) b, monocularly based percepts (reported as
horseshoe percepts) (A2) a, piecemeal rivalry percepts (reported as horse-
eyes for 2 frames, alternating with blanks of 2 frame dura-
shoe percepts) b, monocularly based percepts. (Reported as circle and line tions resulting in a symmetric 18 Hz Xicker which was in
percepts.) phase across both eyes.
 T. Knapen et al. / Vision Research 47 (2007) 1–7 3

The task was as follows: subjects reported percepts of complement the circle and line percepts, the fraction of
either circles or lines separately by pressing buttons on the these percepts is also an indicator of pattern completion
keyboard. The durations of buttonpresses were recorded during horseshoe stimulus viewing. If pattern completion
separately for each type of percept. Note that for circle and increases this should decrease the dominance of monocu-
line stimuli these are monocularly based percepts, whereas larly based horseshoe percepts for horseshoe stimuli, and
for horseshoe stimuli these are pattern-completed percepts vice versa. Four observers participated in this experiment,
(Fig. 1B). The order of trials in a session, and which eye in which each condition was tested once in one session, and
received which input were randomized; each condition was trial duration was 120 s. All further stimulus parameters
tested once per session, with a trial duration of 60 s. Each were identical to those of the original experiment.
subject completed two sessions, each of which contained
four trials. 2.2. Results
The use of interocular grouping stimuli could be consid-
ered controversial, as Lee and Blake (2004) have shown Fig. 2 shows the results of experiment 1. As a measure
that the report of complete percepts may be dependent on for the strength of pattern completion, we summed domi-
subjects’ reporting criterion. We have taken two measures nance durations of both circle and line percepts to give
to ensure that this is not the case in our experiments. First, the total time spent in pattern-completed percepts for
we used stimuli that are very simple compared to those horseshoe stimuli, and monocularly based percepts for
used in previous interocular grouping experiments (Kovacs the circle and line stimulus conditions. The fraction of
et al., 1996; Lee & Blake, 2004), leaving less room for crite- time spent in circle and line percepts from experiment 1
rion diVerences to have an impact on our results. Second, and identical conditions from experiment 2 is shown in
we conducted a control experiment to investigate whether Fig. 2A. Monocularly based circle and line percepts
any eVect of stimulus Xicker on total dominance durations (right pair of bars) were reported more than pattern-com-
could also be due to a change in response criterion. For pleted circle and line percepts (left pair of bars). For
instance, due to stimulus Xicker subjects could become less horseshoe shaped Diaz–Caneja stimuli, this fraction is
conservative in reporting circle and line percepts, thus caus- less than that reported by Ngo, Miller, Liu, and Pettigrew
ing the amount of circle and line percept to increase. Sub- (2000). This divergence may be due to spatial frequency
jects were asked to report left-open and right-open and duty cycle diVerences. The application of stimulus
horseshoe percepts. For circle and line stimuli, these per- Xicker increased the pattern-completed percept for the
cepts only occur during short periods of piecemeal transi- horseshoe stimulus condition. This increase of circle and
tions, while for horseshoe stimuli these are frequently line percept is absent for circle and line stimulus
occurring monocularly based percepts. As these percepts conditions.

A Continuous presentation B C
18 Hz Flicker Pattern completion ratio Pattern segregation ratio
Time in circle or line percept /

Flicker / Continuous

Flicker / Continuous

1
0.6
total trial duration

0.8
0.5
0.4 0.6
0.3
0.4
0.2
0.1 0.2

Stimuli
Stimuli Stimuli

Fig. 2. Data of experiment 1, illustrating the Xicker-induced increase in pattern completion. (A) Fraction of time subjects perceived circles and lines, which
in cases of horseshoe stimuli were pattern-completed percepts. This fraction is greater for circle and line stimuli than for horseshoe stimuli. The amount of
circle or line percept does not diVer between Xicker and non-Xicker conditions for the circle and line stimuli. The horseshoe stimuli, however, show a sig-
niWcant increase in the amount of circle and line percept as a consequence of stimulus Xicker. Error bars are §1 SEM. (B) Pattern completion ratios for all
individual subjects for each of the two stimulus types used. As a measure of the eVect of stimulus Xicker on pattern completion, we deWne the pattern com-
pletion ratio as the fraction of circle and line percept under Xicker conditions divided by the fraction of circle and line percept under continuous presenta-
tion conditions. This ratio is signiWcantly greater than unity for the horseshoe stimulus, indicating that stimulus Xicker increases the amount of pattern
completion. Each line connects data points from one subject, showing a consistent eVect in all subjects. (C) Pattern segregation ratio from the control
experiment in which subjects reported horseshoe percepts. Consistent with the pattern completion ratio, it is the fraction of time subjects perceived the
horseshoe conWguration under Xicker conditions divided by the fraction of this percept under continuous viewing conditions. These percepts complement
the circle and line percepts, so an increase in circle and line percepts should be reXected in a decrease in horseshoe percepts and consequently a decrease in
pattern segregation ratio. This occurs for horseshoe stimuli but not for circle and line stimuli, conWrming the increase in pattern completion found for cir-
cle and line percept reports.
4 T. Knapen et al. / Vision Research 47 (2007) 1–7

We deWne the pattern completion ratio as the fraction of from those normally required to induce monocular
time spent in circle and line percepts for the Xicker condi- rivalry, such as low duty cycle, large stimuli and low con-
tion divided by the fraction of time spent in circle and line trast.
percepts for the continuous presentation condition. Fig. 2B Six observers participated, Wve were subjects in the previ-
(points with error bars) shows these ratios, which are a ous experiment. Setup and stimuli were identical to those
measure of the eVect caused by Xicker, for both stimulus used in the previous experiment, but Xicker frequency was
conWgurations averaged over all subjects. The lines between varied. DiVerent Xicker frequencies were implemented by
the two conditions represent data from one subject each, varying the number of frames presented between successive
and show clearly that the eVect consistently occurs in all blanks each lasting one frame duration (13 ms). Thus,
subjects. Stimulus Xicker increases the amount of pattern- Xicker frequency was varied between 10 and 24 Hz in Wve
completed percept for horseshoe stimuli (p D .001, paired t- steps. Subjects completed one session during which trial
test), but does not increase the amount of monocularly duration was 60 s.
based percept for circle and line stimuli (p > .4).
The data from the control experiment in which subjects 3.2. Results
reported horseshoe percepts are shown in Fig. 2C. The pat-
tern segregation ratio is deWned as the fraction of time Fig. 3 shows pattern completion ratios for both circle
spent in horseshoe percepts during stimulus Xicker, divided and line stimuli and horseshoe stimuli, over a 10–24 Hz
by the fraction of time spent in these percepts during con- range of frequencies. Pattern completion ratios are greater
tinuous presentation. An increase in pattern completion than unity at all frequencies for horseshoe stimuli, whereas
would result in a decrease of the pattern segregation ratio the ratios for circle and line stimuli are »1. Thus, there is no
in the present control experiment, an eVect opposite to the eVect of stimulus Xicker at any of the frequencies tested for
change of pattern completion ratio in the previous experi- the circle and line stimuli, whereas for horseshoe-shaped
ment. The decrease in pattern segregation ratio for stimulus stimuli the eVect of stimulus Xicker is present at all frequen-
rivalry shown in Fig. 2C conWrms the increase in pattern cies. The diVerence between the two types of stimuli is
completion as a result of stimulus Xicker. The decrease in highly signiWcant (two-way ANOVA F(1, 4): 28.7, p < .10¡5).
horseshoe percept was signiWcant for horseshoe stimuli Subjects, when debriefed after the experiment, reported an
(p < .05, paired t-test), whereas the decrease in horseshoe increase in transparent percepts at lower Xicker frequencies.
percept that occurs for circle and line stimuli was not With less convincing percepts and more transition time, the
(p D .2). amount of time reported as monocularly based percept
diminishes for circle and line stimuli.
3. Experiment 2: interocular grouping is independent of More importantly, for horseshoe stimuli the pattern
Xicker frequency completion ratios do not show any trend across frequen-
cies, indicating that the increase in pattern completion as a
3.1. Methods
Pattern completion ratios
Counterphase Flicker / Continuous

DiVerent Xicker frequencies have successfully been

used to generate F&S rivalry (Bonneh et al., 2001; Lee &
Blake, 1999; Logothetis et al., 1996; Pearson & CliVord,
2005b). The range of frequencies at which it is possible to
diminish the importance of monocular images in binocu-
lar rivalry is highly informative, especially regarding the
relationship to physiological data, such as the diVerent
critical fusion frequencies (CFF) of diVerent cell popula-
tions (van de Grind, 1973). Therefore, we conducted an
experiment in which we varied the frequency at which
on–oV Xicker was deployed. The range of frequencies Frequency of flicker
tested was limited as for lower frequencies the transients
Fig. 3. Frequency-independence of the increase in pattern completion. The
as a result of stimulus Xicker become increasingly strong
black line represents data for horseshoe stimuli, the light gray line repre-
perceptually and induce perceptual switches (Kanai, sents data for the circle and line stimuli. Pattern completion ratio is
Moradi, Shimojo, & Verstraten, 2005). These stimulus- deWned as the total duration of pattern-completed percept during Xicker
induced alternations are not the subject of the present divided by the total duration of pattern-completed percept during contin-
paper, thus these frequencies were omitted. Higher fre- uous stimulation. Therefore, a value of one (indicated by the dashed line)
means there is no diVerence in pattern completion as a result of Xicker.
quencies of Xicker creates monocular temporal fusion,
Values represent means across four subjects, error bars are §1 SEM. It is
increasing the predominance of plaid-like percepts. Mon- evident that the increase in pattern completion occurs equally for all fre-
ocular rivalry could ensue under these conditions of quencies for the horseshoe stimuli. No such eVect occurs for circle and line
monocular temporal fusion, however, the stimulus char- stimuli, apart from an opposite eVect at lower frequencies which does not
acteristics used in the present study are quite diVerent reach signiWcant levels.
 T. Knapen et al. / Vision Research 47 (2007) 1–7 5

result of stimulus Xicker is not dependent on Xicker fre- Pattern completion ratio

Counterphase Flicker / Continuous

quency. 1

4. Experiment 3: counterphase Xicker decreases interocular

0.8
grouping
0.6
4.1. Methods

Pattern completion increases as a result of stimulus 0.4

Xicker and does so approximately equally at multiple fre-
quencies, as experiment 2 showed. Although insensitive to 0.2
changes in frequency, the eVect may be altered by changing
the type of Xicker used. To investigate this further, we mod-
iWed the Xicker regime to alternate not between presence
and absence of the stimulus, but between two counterph- Stimuli
ased presentations of the stimuli. Under this regime there
Fig. 4. Pattern completion ratios for counter-phase Xicker. Counterphase
are only local inversions of contrast but no global lumi-
Xicker decreases the amount of circle and line percept when compared to
nance modulations at a given Xicker frequency. In the pre- the non-Xicker condition for both stimulus conWgurations. This eVect is
sentation regime used here the patterns of circles, lines and signiWcant for the horseshoe stimulus. Thus, counterphase Xicker has
horse-shoe shapes are continuously presented, while the fre- opposite eVects on pattern completion when compared to on–oV Xicker at
quency content of the stimulus presentation is kept equal to the same frequency. Bars represent pattern completion ratios for both
types of stimuli. Error bars are §1 SEM.
that during on–oV Xicker. Moreover, counterphase Xicker
introduces a variation in stimulus phase over time, and
modulation of cellular responses due to varying spatial 5. Discussion
phase over time is a traditional method of separating diVer-
ent computational levels in primary visual cortex (i.e. sim- We conducted three experiments to investigate whether
ple and complex cells) (Hubel & Wiesel, 1962; Ringach, interocular grouping is increased by stimulus Xicker. Our
2004). Phase-invariant cells would not change their Wring results demonstrate that on–oV Xicker increases interocu-
behavior due to the diVerence between on–oV and counter- larly combined pattern completion (experiment 1), that it
phase Xicker. However, two separate pools of phase-sensi- does so independently of Xicker frequency (experiment 2),
tive neurons may respond to counterphase Xicker whereas and that this increase depends critically on the on–oV lumi-
only one would respond to on–oV Xicker. Therefore, count- nance transients that occur during stimulus Xicker (experi-
erphase Xicker may clarify at which neural level interocular ment 3).
grouping occurs. Based on our results, we propose that the amount of
The six observers of the Wrst experiment participated in interocularly combined pattern-completed percept can be
this counter-phasing experiment. Again, setup and stimuli regarded as a measure of the level at which binocular
were identical to those used in experiment 1. However, rivalry is resolved. We employ this measure to examine
instead of blanks, 2 frames of each cycle were used to pres- whether the temporal characteristics of stimulus presenta-
ent a contrast-inverted version of the stimulus that tion can alter the level at which binocular rivalry takes
appeared during the other two frames of the cycle. Thus, place. Local retinotopic factors are of paramount impor-
over time, the gratings composing the stimulus would tance for the occurrence of normal binocular rivalry (Carl-
counterphase at 18 Hz. Subjects completed one session, in son & He, 2004), and the alternations between binocular
which 4 conditions were sampled. Trial duration was 120 s. rivalry percepts do occur in a retinotopically ordered fash-
ion (Wilson, Blake, & Lee, 2001). However, binocular
4.2. Results experiment 3 rivalry suppression occurs at multiple levels in the visual
hierarchy, as it has been found that the depth of suppres-
Fig. 4 shows the results from experiment 3. For circle sion increases with increasing stimulus complexity (Ngu-
and line stimuli, no signiWcant diVerence in pattern comple- yen, Freeman, & Alais, 2003). In addition, the pattern that
tion ratio occurs as a result of counterphasing Xicker. There undergoes suppression alters the percept of the dominant
is, however, a marked eVect of counterphasing Xicker on pattern in a manner that implicates continuous interaction
the pattern completion ratio for horseshoe stimuli. The at multiple levels (Pearson & CliVord, 2005a). The fact that
decrease in pattern completion is signiWcant (paired t-test, rivalry can occur independently of the eye being stimulated,
p < .01). So, subjects perceived less pattern-completed as occurs in F&S rivalry, points directly to the existence of
shapes during counterphasing Xickering when compared to multiple rivalry stages (Lee & Blake, 1999; Logothetis et al.,
continuous stimulation, an eVect that is opposite to the 1996; Pearson & CliVord, 2005b). Stimulus Xicker is
eVect of on–oV Xicker on pattern completion ratios for thought to cause rivalry to bypass lower, monocular stages
horseshoe stimuli as found in experiment 1. and transpire at higher stages whose activity is independent
6 T. Knapen et al. / Vision Research 47 (2007) 1–7

of eye-of-origin information. A putative physiological ther using a combination of physiological and psychophys-
mechanism for this decrease of direct interocular inhibition ical experiments.
due to stimulus Xicker is a decrease in V1 neuron IPSP The data presented here provide evidence supporting the
amplitude at stimulation frequencies above 6 Hz (Kuhnt & hypothesis that binocular rivalry occurs at multiple stages,
Creutzfeldt, 1971, see also Wilson, 2003). A compelling and provide insights in the divisions between these diVerent
result by Bonneh et al. (2001) shows that the occurrence of stages. Further research employing the F&S stimulus or its
F&S rivalry is dependent on the size and structural coher- derivatives is very likely to be of use in investigating binoc-
ence of the stimulus used, under conditions of Wxed spatial ular rivalry, especially in physiological studies.
frequency. This strongly suggests that pattern-based repre-
sentations are mediated by neurons with larger receptive References
Welds, likely located higher in the visual processing stream.
These higher stages could consist of binocular cells in V1, Alais, D., O’Shea, R. P., Mesana-Alais, C., & Wilson, I. G. (2000). On bin-
but also constitute activity anywhere up the ventral path- ocular alternation. Perception, 29(12), 1437–1445.
way along which the perceptual modulation of neuronal Blake, R., & Logothetis, N. K. (2002). Visual competition. Nature Review
on Neuroscience, 3(1), 13–21.
activity increases (Fang & He, 2005; Leopold & Logothetis, Blake, R., Westendorf, D. H., & Overton, R. (1980). What is suppressed
1996). Interocular pattern completion is likely to be medi- during binocular rivalry? Perception, 9(2), 223–231.
ated by the higher stages at which rivalry may occur. The Bonneh, Y., Sagi, D., & Karni, A. (2001). A transition between eye and
increase in interocular pattern completion due to stimulus object rivalry determined by stimulus coherence. Vision Research,
Xicker accords well with the hypothesis that rivalry may 41(8), 981–989.
Carlson, T. A., & He, S. (2004). Competing global representations fail to
transpire at several levels simultaneously (Bonneh et al., initiate binocular rivalry. Neuron, 43(6), 907–914.
2001). Diaz–Caneja, E. (1928). Sur l’Alternation binoculaire [On binocular alter-
Previously, diVerent frequencies of 15–20 Hz were used nation]. Annalses d’Oculistique, 165, 721–731.
for the generation of F&S rivalry (Bonneh et al., 2001; Lee Fang, F., & He, S. (2005). Cortical responses to invisible objects in the
& Blake, 1999; Logothetis et al., 1996; Pearson & CliVord, human dorsal and ventral pathways. Nature Neuroscience, 8(10), 1380–
1385.
2005b). We show that in our stimulus paradigm, the fre- Fox, R., & Check, R. (1968). Detection of motion during binocular rivalry
quency of on–oV Xicker has no decisive inXuence on the suppression. Journal of Experimental Psychology, 78(3), 388–395.
increase in pattern completion within the range of ca. 10– Fox, R., & Check, R. (1972). Independence between binocular rivalry sup-
25 Hz. If extrapolated to F&S rivalry, our results could be pression duration and magnitude of suppression. Journal of Experi-
construed to indicate that there is no ‘sweet spot’ for the mental Psychology, 93(2), 283–289.
Gilbert, C. D. (1977). Laminar diVerences in receptive Weld properties
on–oV Xicker frequency as there is for the switch frequency of cells in cat primary visual cortex. Journal of Physiology, 268(2),
(Lee & Blake, 1999). 391–421.
We also show that the increase in interocular pattern Hubel, D. H., & Wiesel, T. N. (1962). Receptive Wleds, binocular interac-
completion as a result of stimulus Xicker is diminished and tion and functional architecture in the cat’s visual cortex. Journal of
even reversed when there are no blanks in stimulus presen- Physiology, 160, 105–154.
Kanai, R., Moradi, F., Shimojo, S., & Verstraten, F. A. (2005). Perceptual
tation (Fig. 4). Although the global patterns projected into alternation induced by visual transients. Perception, 34(7), 803–822.
the eyes are the same as in the other experiments, the local Kovacs, I., Papathomas, T. V., Yang, M., & Feher, A. (1996). When the
luminance and spatial phase change rapidly over time when brain changes its mind: interocular grouping during binocular rivalry.
the stimulus is counterphased. Geniculate and cortical sim- Proceedings of the National Academy of Sciences of the United States of
America, 93(26), 15508–15511.
ple cells are known to be sensitive to changing spatial
Kuhnt, U., & Creutzfeldt, O. D. (1971). Decreased post-synaptic inhibition
phase, and thus might change their Wring behavior as a in the visual cortex during Xicker stimulation. Electroencephaphy and
result of counterphase Xicker. These cells also have a ten- Clinical Neurophysiology, 30, 79–82.
dency (strong for geniculate cells, less so for simple cells Lee, S. H., & Blake, R. (1999). Rival ideas about binocular rivalry. Vision
(Hubel & Wiesel, 1962; Gilbert, 1977; Skottun & Freeman, Research, 39(8), 1447–1454.
1984)) to be more monocularly driven and have a higher Lee, S. H., & Blake, R. (2004). A fresh look at interocular grouping during
binocular rivalry. Vision Research, 44(10), 983–991.
critical Xicker frequency (CFF) than complex cells that Leopold, D. A., & Logothetis, N. K. (1996). Activity changes in early visual
have a lower CFF than the frequencies used in the present cortex reXect monkeys’ percepts during binocular rivalry. Nature,
stimulus (van de Grind, 1973). Also, complex cells would 379(6565), 549–553.
not respond to counterphase Xicker in a diVerent manner Leopold, D. A., & Logothetis, N. K. (1999). Multistable phenomena:
changing views in perception. Trends in Cognition Science, 3(7), 254–
when compared to on–oV Xicker given their lack of phase
264.
sensitivity. Logothetis, N. K., Leopold, D. A., & Sheinberg, D. L. (1996). What is rival-
Thus, assuming that the lower stages at which rivalry ling during binocular rivalry? Nature, 380(6575), 621–624.
occurs are stimulated twice every Xicker period, we inter- Ngo, T. T., Miller, S. M., Liu, G. B., & Pettigrew, J. D. (2000). Binocular
pret the decrease in pattern completion that results from rivalry and perceptual coherence. Current Biology, 10(4), R134–R136.
Nguyen, V., Freeman, A., & Alais, D. (2003). Increasing depth of binocular
the counterphase Xicker operation as a shift of the binocu-
rivalry suppression along two visual pathways. Vision Research, 43(19),
lar rivalry process to monocular stages, that would in this 2003–2008.
view consist of monocular simple cells in V1. However, this O’Shea, R. P., & Corballis, P. M. (2005). Visual grouping on binocular
interpretation is tentative, and should be investigated fur- rivalry in a split-brain observer. Vision Research, 45(2), 247–261.
 T. Knapen et al. / Vision Research 47 (2007) 1–7 7

Pearson, J., & CliVord, C. W. G. (2005a). Suppressed patterns alter vision van de Grind, W. A., Gruesser, O. J. & Lunkenheimer (1973). Temporal
during binocular rivalry. Current Biology, 15(23), 2142–2148. transfer properties of the aVerent visual system. In R. Jung (Ed.),
Pearson, J., & CliVord, C. W. G. (2005b). When your brain decides what Handbook of sensory physiology, Central processing of visual infor-
you see. Psychological Science, 16(7), 516–519. mation (pp. 431–573), Vol. VII/3A, Part A. Berlin: Springer Verlag.
Ringach, D. L. (2004). Mapping receptive Welds in primary visual cortex. Wales, R. F., & Fox, R. (1970). Increment detection threshold during bin-
Journal of Physiology, 558(3), 717–728. ocular rivalry suppression. Perception and Psychophysics, 8, 90–94.
Skottun, B. C., & Freeman, R. D. (1984). Stimulus speciWcity of binocular Wilson, H. R. (2003). Computational evidence for a rivalry hierarchy in
cells in the cat’s visual cortex: ocular dominance and the matching of vision. Proceedings of the National Academy of Sciences of the United
left and right eyes. Experimental Brain Research, 56, 206–216. States of America, 100(24), 14499–14503.
Tong, F., & Engel, S. A. (2001). Interocular rivalry revealed in the human Wilson, H. R., Blake, R., & Lee, S. H. (2001). Dynamics of travelling waves
cortical blind-spot representation. Nature, 411(6834), 195–199. in visual perception. Nature, 412, 907–910.