Continuous flash suppression: Known and unknowns

Abstract

Studies utilizing continuous flash suppression (CFS) provide valuable information regarding conscious and nonconscious perception. There are, however, crucial unanswered questions regarding the mechanisms of suppression and the level of visual processing in the absence of consciousness with CFS. Research suggests that the answers to these questions depend on the experimental configuration and how we assess consciousness in these studies. The aim of this review is to evaluate the impact of different experimental configurations and the assessment of consciousness on the results of the previous CFS studies. We review studies that evaluated the influence of different experimental configuration on the depth of suppression with CFS and discuss how different assessments of consciousness may impact the results of CFS studies. Finally, we review behavioral and brain recording studies of CFS. In conclusion, previous studies provide evidence for survival of low-level visual information and complete impairment of high-level visual information under the influence of CFS. That is, studies suggest that nonconscious perception of lower-level visual information happens with CFS, but there is no evidence for nonconscious high-level recognition with CFS.

Full text

1 23
Psychonomic Bulletin & Review
ISSN 1069-9384
Psychon Bull Rev
DOI 10.3758/s13423-020-01771-2
Continuous flash suppression: Known and
unknowns
Ali Pournaghdali & Bennett L.Schwartz

1 23
Your article is protected by copyright
and all rights are held exclusively by The
Psychonomic Society, Inc.. This e-offprint is
for personal use only and shall not be self-
archived in electronic repositories. If you wish
to self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at link.springer.com”.

THEORETICAL REVIEW
Continuous flash suppression: Known and unknowns
Ali Pournaghdali
1
&Bennett L. Schwartz
1
#The Psychonomic Society, Inc. 2020
Abstract
Studies utilizing continuous flash suppression (CFS) provide valuable information regarding conscious and nonconscious
perception. There are, however, crucial unanswered questions regarding the mechanisms of suppression and the level of visual
processing in the absence of consciousness with CFS. Research suggests that the answers to these questions depend on the
experimental configuration and how we assess consciousness in these studies. The aim of this review is to evaluate the impact of
different experimental configurations and the assessment of consciousness on the results of the previous CFS studies. We review
studies thatevaluated the influence ofdifferent experimental configuration on the depth of suppression with CFS and discuss how
different assessments of consciousness may impact the results of CFS studies. Finally, we review behavioral and brain recording
studies of CFS. In conclusion, previous studies provide evidence for survival of low-level visual information and complete
impairment of high-level visual information under the influence of CFS. That is, studies suggest that nonconscious perception of
lower-level visual information happens with CFS, but there is no evidence for nonconscious high-level recognition with CFS.
Keywords Visual consciousness .Nonconscious perception .Continuous flash suppression .Neural correlates of consciousness
The neural correlates of consciousness
1
(NCCs
2
) and the dif-
ferences between conscious and nonconscious perception are
topics of interest in cognitive neuroscience. There are a variety
of methods and techniques that are used to manipulate con-
scious and nonconscious perceptions, ranging from methods
that target attentional processes, like inattentional blindness
(Mack, 2003) and attentional blink (Martens & Wyble,
2010), to the methods that manipulate conscious perception
at the perceptual level, like visual masking (Bachmann &
Francis, 2013) and binocular rivalry (Tong, Meng, & Blake,
2006). One of the newest methods, which has been used ex-
tensively in the recent years and is the focus of the current
review, is continuous flash suppression (CFS; Tsuchiya &
Koch, 2005).
In CFS, the presentation of a series of masks
(Mondrians
3
) to the dominant eye suppresses the percep-
tion of a low-contrast target presented to the nondominant
eye. This can be achieved using three different display
setups: red–blue or red–green anaglyph glasses, mirror
stereoscopes and prism goggles (Carmel, Arcaro,
Kastner, & Hasson, 2010). CFS can be used to dissociate
conscious perception from visual processes irrelevant to
awareness. Because CFS has several advantages over oth-
er suppression methods, it has become a popular method
in the field of visual awareness. These advantages include
longer duration of invisibility than other similar methods,
such as binocular rivalry and flash suppression (Tsuchiya,
Koch, Gilroy, & Blake, 2006), and reliable suppression
from the onset of target (e.g., Tsuchiya & Koch, 2005).
Despite these advantages of CFS for examining conscious
and nonconscious perception, there are still many unan-
swered questions regarding this method, and the neural
basis of suppression with CFS is still unknown.
Moreover, different researchers have used different exper-
imental configurations to suppress the target stimulus and
3
Because of the resemblance of these masks to the work of the artist Piet
Mondrian, they are named Mondrians. We, however, use the terms
Mondrians,masks,andmaskers interchangeably to refer to this class of
stimuli.
1
In this review, we use terms consciousness and awareness interchangeably.
2
For a list of the abbreviations used in this study, see Appendix A.
*Bennett L. Schwartz
bennett.schwartz@fiu.edu
1
Department of Psychology, Florida International University,
Miami, FL 33199, USA
Psychonomic Bulletin & Review
https://doi.org/10.3758/s13423-020-01771-2
Author's personal copy

have used different strategies to evaluate conscious per-
ception. One of our assertions is that the lack of homoge-
neous experimental configuration and behavioral assess-
ments of conscious perception prevent us from generaliz-
ing the results of the studies that employ CFS. As we will
show, this heterogeneity results in discrepancy of evi-
dence regarding nonconscious perception of different vi-
sual categories and neural basis of nonconscious percep-
tion under the influence of CFS (it is important to note
that this heterogeneity might be one of the reasons of the
aforementioned discrepancies, and some other factors may
also contribute to these discrepancies, which are beyond
the scope of this review). Therefore, it is crucial to under-
stand how different experimental configurations and be-
havioral assessments of consciousness influence the re-
sults of the CFS studies.
The current research program that utilizes CFS to under-
stand the differences between conscious and nonconscious
perception needs to provide answers to at least the following
five questions, which are critical for the scientific understand-
ing of conscious perception:
1. What is (are) the optimal experimental configuration(s) to
obtain an optimal suppression?
2. What is (are) the optimal assessment(s) of consciousness?
3. What is the mechanism of suppression of CFS?
4. Can visual information be processed under the influence
of CFS and in absence of awareness?
5. What is the level of visual processing under the influence
of CFS?
To find an accurate response to the last three questions,
we need to first find the optimal experimental configura-
tion(s) and the optimal assessment(s) of conscious percep-
tion. Therefore, the first two questions are highly critical
for the understanding of the mechanism of suppression
and the level of nonconscious processing with CFS.
Therefore, the aim of this review is to evaluate the impact
of different experimental configurations and the assess-
ment of consciousness on the results from previous CFS
studies. To this end, we will first review different spatial
and temporal properties of target and masks, which may
impact the depth of suppression and processing of both
masks and target under CFS. Second, we will review the
methods to assess conscious perception and discuss a
model-based analysis approach based on a multidimen-
sional extension of signal detection theory, called general
recognition theory (Ashby & Soto, 2015; Soto, Zheng,
Fonseca, & Ashby, 2017). Then, we will review some of
the key findings regarding nonconscious perception with
CFS. Finally, we will review some of the studies that
explored the neural underpinnings of conscious and non-
conscious perception using CFS.
Before moving to the next section, we clarify a specific
point that is often questioned.
4
Some may rightfully point
out that “nonconscious perception”is not a proper term to
use in order to refer to the visual processing in the absence
of awareness and suggest using the term “nonconscious visual
processing”instead. Although we acknowledge the implica-
tions of the differences in terminology, we argue that any
visual processing in the absence of awareness that translates
to behavioral responses (such as adaptation or priming. See
section on Nonconscious Perception With CFS) counts as a
nonconscious perception. Based on this, if visual processing
in the absence of awareness does not translate to a behavioral
response, it does not qualify as nonconscious perception.
Regardless of this argument, we will use two terms inter-
changeably, but we will not consider any visual/neural pro-
cessing that does not translate to a behavioral response as an
instance of nonconscious perception.
Influence of experimental configuration
on CFS suppression
Since Tsuchiya and Koch (2005) first introduced CFS, many
studies used this method to evaluate nonconscious perception
at the behavioral and neural level. Some of these studies, how-
ever, did not control or manipulate masks and targets’low-
level attributes, such as spatial and temporal frequencies.
Subsequent studies, however, showed that strength of sup-
pression is tightly correlated with spatial and temporal prop-
erties of target and masks in CFS. As we will see in the fol-
lowing sections, target and maskers’low-level attributes may
explain someresults with CFS (e.g., face and facial expression
studies) and by controlling or manipulating specific attributes
of maskers and targets, some of the nonconscious effects may
disappear (for example, in case of nonconscious recognition
of manipulable objects. See the section titled Nonconscious
Perception of Manipulable Objects).
Therefore, it is possible that some CFS studies failed to
induce complete suppression because of the inability of their
experimental configurations to allow suppression of the tar-
gets. On the other hand, some settings might oversuppress the
visual processing, meaning they might suppress both con-
scious and nonconscious processing. This may be one of the
reasons of discrepancy in the CFS literature regarding differ-
ent visual categories, such as facial expression and tool recog-
nition (see the subsections titled Nonconscious Facial
Expression Recognition and Nonconscious Perception of
Manipulable Objects). Therefore, experimental configurations
with optimal temporal and spatial parameters of targets and
masks are essential parts of the CFS studies. Fortunately, in
4
We thank one of the anonymous reviewers for motivating us to clarify this
point.
Psychon Bull Rev
Author's personal copy

the past few years, a growing number of studies evaluated the
role of these attributes in CFS related suppressions, and we
review these studies in this section.
The first experimental parameter that needs to be set care-
fully is temporal frequency (TF) of the maskers (or masks). In
their original study, Tsuchiya and Koch (2005) used masks
with 10 Hz TF to render the targets invisible. The majority of
the CFS studies also followed Tsuchiya and Koch and used
masks with 10 Hz TF (e.g., Almeida, Mahon, Nakayama, &
Caramazza, 2008; Anderson, Siegel, White, & Barrett, 2012).
Some studies, however, used TFs lower or higher than 10 Hz,
ranging from close to zero Hz to 100 Hz (e.g., Geng, Zhang,
Li, Tao, & Xu, 2012). It seems that using different TFs for
maskers, in some of the studies, is arbitrary and sometimes
based on untested assumptions. For example, some studies
used maskers with higher TF than 10 Hz based on the pre-
sumption that suppressive power of maskers increases with
increase in maskers’TF (e.g., Geng et al., 2012; Xu, Zhang,
&Geng,2011). Some studies do not report maskers’TF at all,
and therefore these studies cannot be compared with other
studies along any dimensions (e.g., Costello, Jiang,
Baartman, McGlennen, & He, 2009). Because TF varies or
is not reported at all, generalization of the results across CFS
studies is questionable at best (for a discussion regarding the
generalizability of CFS studies, see Zhan, Engelen, & de
Gelder, 2018a). Therefore, one of the goals of this paper is
to stress to researchers the importance of adopting an optimal
TF for their maskers to ensure an optimal level ofsuppression.
In order to understand the influence of maskers’TF on
target visibility, we need to clarify the difference between
maskers’TF with the modulation in temporal structure of
the maskers’luminance. When masks replace each other with
a specific TF, a broad energy spectrum in the temporal dimen-
sion is produced as the result of the changes in the luminance
of the pixels within the maskers. This energy spectrum peaks
at a much lower frequency (see E. Yang & Blake, 2012). For
example, presentation of maskers with 10 Hz TF produces a
broad energy spectrum in the temporal dimension that peaks
at 5 Hz. This distinction leads to two different strategies to
manipulate and evaluate the influence of maskers’TF on tar-
get visibility. In the first strategy, maskers’TF is evaluated by
manipulating the duration of each mask (maskers’refresh
rate). For example, for presenting 10 maskers with the fre-
quency of 10 Hz, each masker will be presented for 100 ms
(for the total of 1 second), and for presenting 10 maskers with
the frequency of 20 Hz, each masker will be presented for 50
milliseconds (for the total of 500 ms). For the remaining of
this section, we refer to this strategy “refresh-based strategy.”
The second strategy involves spectrum analysis of the
maskers and filtering the energy spectrum in the temporal
dimension. In this strategy, the changes in the luminance of
the pixels within the presentation duration of the maskers are
fast Fourier transformed, and a bandpass filter (a filter that
only allows frequency within a specific range to pass) is ap-
plied to the transformed signal in the temporal dimension (see
E. Yang & Blake, 2012). Finally, by inverse Fourier transfor-
mation of the signal, a modified set of maskers are produced
from the old ones that have a specific temporal structure with-
in a very specific range. For the remaining of this section, we
refer to this strategy “spectral-based strategy.”
In the past few years, some studies have used either of the
abovementioned strategies to evaluate the influence of
maskers’TF on the visibility of the target stimuli. Studies that
adopted a spectral-based strategy showed that maskers with
TFs lower than 10 Hz are more effective in suppressing the
target stimulus than those maskers with TFs higher than 10
Hz. In the first study, E. Yang and Blake (2012,Experiment4)
showed that Mondrians with low TF (below 10 Hz) induce
similar level of suppression as Mondrians with 10 Hz TF.
Mondrians with high TF (above 10 Hz) induced weaker sup-
pression than Mondrians with 10 Hz TF. According to these
results, high-TF maskers are not as effective as maskers with
10 Hz and lower TFs to suppress the target stimuli. Further
studies not only supported E. Yang and Blake’s(2012)results,
but showed that maskers with lower TF than 10 Hz might be
more effective at suppression than 10 Hz TF maskers. For
example, Han, Lunghi, and Alais (2016) showed that filtered
noise maskers (which were different from traditional
Mondrians) with lower TF (0.375, 0.75, 1.5, and 3 Hz) in-
duced longer suppression duration than noise masker with
zero TF (static masker) and noise makers with high TF
(6.25, 12.5, and 25 Hz). High-TF noise maskers, however,
induced shorted suppression duration than did static masker.
In addition, Han, Blake, and Alais (2018, Experiment 3),
showed that Mondrians with lower TF (low-band-pass
Mondrians) have similar suppressive power (similar suppres-
sion duration) to Mondrians with 10 Hz TF (unfiltered
Mondrians). In accordance with previous results, Han et al.
(2018) showed that Mondrians with high TF (high-band-pass
Mondrians) have a lower suppressive power than low-band-
pass and unfiltered Mondrians.
Studies that adopted the refresh-based strategy supported
the results of the spectral-based studies and showed that
Mondrians with TF lower than 10 Hz have more suppressive
powers than those with TF higher than 10 Hz. For example,
Zhu, Drewes, and Melcher (2016) showed that the targets,
which were rendered invisible with 5-Hz and 7-Hz TF
Mondrians,required a higher contrast to break the suppression
than targets that were rendered invisible with Mondrians with
higher or lower TFs. Fitting skewed-normal curve to their
data, Zhu et al. (2016) showed that the break-through contrast
peaks at 6-Hz TF. Finally, Zhan et al. (2018a), showed that
Mondrians with lower TF than 10 Hz (4, 6, and 8 Hz) induce
more unseen trials (measured by yes/no detection task) than
did Mondrians with TF around 10 Hz and higher. In contrast
with the preceding studies, Kaunitz, Fracasso, Skujevskis, and
Psychon Bull Rev
Author's personal copy

Melcher (2014) showed that Mondrians with TFs higher than
10 Hz (16.6 and 28.5 Hz) are more effective in suppressing a
brief target, whereas Mondrians with TFs lower than 10 Hz (5
and 8 Hz) are not as effective as high-TF Mondrians. This
means that Mondrians with a high TF might be actually more
effective than Mondrians with low TF.
In summary, studies point to three different Mondrians’
TFs as optimal TF: TF close to zero (Han et al., 2016), TF
between 5 to 7 Hz (with 6 Hz being most effective; Zhan et al.
2018a; Zhu et al., 2016) and 28.5 Hz (Kaunitz et al., 2014).
We think that there are two potential reasons for this discrep-
ancy. First, Han et al. (2016) used noise maskers that change
smoothly, in contrast to traditional Mondrians that have sharp
updates (that change abruptly). Second, studies that presented
the target stimuli for a relatively long time, found TFs lower
than 10 Hz as optimal TF (Han et al., 2018; Han et al., 2016;
Zhan et al. 2018a; Zhu et al., 2016), whereas the study that
presented the target relatively briefly, found TFs higher than
10 Hz as optimal TF (28.5 Hz; Kaunitz et al., 2014).
Therefore, it seems the duration of target stimuli interacts with
maskers’TF, and the maskers’optimal TF depends on the
duration of the target.
Han and Alais (2018) provided supporting evidence for
this conclusion. In their study, Han and Alais adopted the
spectral-based strategy and showed that noise maskers with
low TF are more efficient and powerful at inducing suppres-
sion when the target is presented for longer duration and is
modulating with low frequency. Maskers with high TF are
more efficient and powerful at inducing suppression when
the target is presented briefly and is modulating with high
frequency. Even though maskers with high TF are more ef-
fective at suppressing a brief target, it is also possible that this
effectiveness is the results of distraction of attention by the
maskers. That is, it could be possible that maskers with high
TF do not induce enough suppression, but their rapid flicker-
ing distracts participants from attending the brief target. We
think that this possibility should be investigated in future
studies.
In addition to the maskers’(and also targets’) TF, spatial
frequency (SF) of both targets and maskers is an essential
attribute of a CFS experiment. Tsuchiya and Koch (2005,
Experiment 5) and E. Yang and Blake (2012;Experiments1
& 2) showed that Mondrians with low SF are more efficient in
suppressing targets than are Mondrians with high SF.
Moreover, E. Yang and Blake (2012) found strongest spectral
power of the Mondrians in the low SFs, which might explain
why Mondrians with low SF have more suppressive power
than do Mondrians with high SF. A follow-up experiment
showed that unfiltered and low band-pass filtered (low SF)
Mondrians are more effective in suppressing the targets than
are high band-pass filtered (high SF) Mondrians (E. Yang &
Blake, 2012, Experiment 3). Moreover, the unfiltered and
low-SF Mondrians had stronger suppression effects on the
targets with low SF than on targets with high SF. These results
indicate that low-SF properties of Mondrians are essential
parts of suppression with CFS, an important point that has
both methodological and theoretical implications.
Han et al. (2016, Experiment 3), however, found that their
noise maskers yielded longer suppression durations for the
natural image targets with high SF than those with low SF.
It is not clear why there should be a difference between tradi-
tional Mondrians and the noise maskers used by Han et al.
(2016) in suppressing targets with low versus high SF. In
order to resolve this issue, we propose a study that compares
the suppression impact of these two types on maskers on tar-
gets with different SFs, by equating the spectral power of the
two types of maskers across different SFs.
Other visual attributes also affect the depth of suppression
with CFS. For example, Han et al. (2018, Experiment 1)
showed that the internal structure of the Mondrians is an im-
portant element of suppression, and disturbing the structural
integrity of the Mondrians by phase scrambling reduces the
suppressive power of them. In addition, Han et al. (2018,
Experiment 2), showed that Mondrians with intact edges but
phase-scrambled solid areas induce longer suppression dura-
tion than do Mondrians with intact solid areas but phase-
scrambled edges. Moreover, Mondrians with intact edges in-
duce shorter suppression duration than do intact Mondrians.
Mondrians with intact solid area, however, did not induce
longer suppression duration than did phase-scrambled
Mondrians. Furthermore, reducing the structural integrity of
the edges (60% reduction) resulted in shorter suppression du-
ration, regardless of structural integrity of the solid areas.
Reducing the structural integrity of the solid areas (60% re-
duction) only resulted in shorter suppression duration, when
the structural integrity of the edges was intact or reduced
slightly (20% reduction). These results indicate that spatial
edges are essential parts of integral structure of the
Mondrians and therefore crucial elements of suppression. In
line with this, Hong and Blake (2009)showedthatremoving
the sharp edges of a bar stimulus reduced the availability of
participants to discriminate the orientation of the bar, meaning
bars without sharp edges are more susceptible to CFS suppres-
sion. Therefore, maskers with intact spatial edges are more
effective in inducing suppression than are maskers without
this property. This is important because some studies adopted
maskers that lacked spatial edges, and incomplete suppression
might explain the results of these studies (e.g., Song & Yao,
2016).
Another important visual attribute that may impact CFS-
based suppression is color. Many studies have used chromatic
Mondrians based on the presumption that colorful maskers
induced stronger suppression than do grayscale maskers
(e.g., Song & Yao, 2016). Hong and Blake (2009), however,
showed that chromatic and grayscale Mondrians did not differ
with respect to suppression of orientation of a bar stimulus.
Psychon Bull Rev
Author's personal copy

Using achromatic Mondrians to suppress a chromatic target
stimulus, on the other hand, results in bleaching through of the
color of the target. That is, bleaching means that when
suppressing a chromatic target with achromatic Mondrians,
participants reported experiencing the color of the target. In
this situation, using chromatic Mondrians suppresses the color
of the targets completely. In addition, Hong and Blake (2009)
showed that chromatic Mondrians are less effective than ach-
romatic Mondrians are in suppressing moving targets.
Therefore, it seems chromatic Mondrians are only useful
when suppressing a chromatic target, but not achromatic
one. Based on this, some studies that used chromatic
Mondrians to suppress an achromatic target may fail to induce
suppression. Hence, it is important to interpret and generalize
the results of such studies with more caution. Therefore, we
recommend using achromatic Mondrians to suppress an ach-
romatic target and using chromatic Mondrians to suppress a
chromatic target.
Motion is another visual attribute that influences the depth
of CFS related suppression. It seems traditional Mondrians are
not as effective as moving Mondrians in suppressing a moving
target (Moors, Wagemans, & de-Wit, 2014). Having said that,
there are two attributes of the moving Mondrians that need to
be evaluated when evaluating the impact of moving
Mondrians on a moving stimulus: speed and the pattern of
motion. Moors et al. (2014) showed that moving Mondrians
have more suppressive power when the speed of Mondrians
and targets are alike (3°/s in Experiment 1 and 2°/s in
Experiment 2). When the target had speed of 5°/s, however,
moving Mondrians with the same speed did not cause stronger
suppression than moving Mondrians with other speed rates.
These results indicate that matching the speed of Mondrians
and targets increases the depth of suppression. Ananyev,
Penney, and Hsieh (2017), on the other hand, found that slow
moving Mondrians (1°/s and 2°/s) are more effective than fast
moving Mondrians (3°, 5°, and 8°/s). Targets with high speed
(5°/s) broke the suppression faster and were detected more
often than targets with low speed (1°/s). When trajectory
length of slow and fast targets was controlled, Ananyev
et al. (2017, Experiment 3) found that the fast target no longer
broke the suppression faster than slow targets, but it was de-
tected more often than in the slow targets. These results indi-
cate that slow Mondrians are more effective than fast
Mondrians are, and fast targets are more immune to the
impact of suppression than slow targets are. Ananyev et al.
(2017) also examined the impact of pattern of motion in mov-
ing Mondrians on the depth of suppression and found that
matching the pattern of motion between target and moving
Mondrians results in a weaker suppression. In contrast, line-
arly moving Mondrians result in longer suppression than do
Mondrians with a rotational moving pattern. Therefore, mov-
ing Mondrians are more effective at inducingsuppression than
traditional Mondrians are, but the evidence regarding the
influence of the speed and the pattern of motion on the depth
of suppression is not conclusive.
Finally, it has been shown that spatial density of Mondrians
also affects the depth of suppression in a CFS experiment in a
way that reduction in spatial density of Mondrians weakens
the depth of suppression (Drewes, Zhu, & Melcher, 2018).
Mondrians’spatial density interacts with their TFs such that
when Mondrians’spatial density reduces, the optimal
Mondrians’TF increases. Therefore, Mondrians with low spa-
tial density have stronger suppression effect when they are
presented with high TF. Mondrians with high spatial density,
on the other hand, induce stronger suppression when they are
presented with lower TF.
There are other important attributes that to the best of our
knowledge, have not been studied yet, such as the difference
between rectangular and circular Mondrians. Evaluating these
attributes is important because there are some untested pre-
sumptions about the impact of these attributes. For example,
circular Mondrians were used to suppress face stimuli based
on the assumption that a circular component of circular
Mondrians has stronger suppressive power on face stimuli
than rectangular Mondrians do (e.g., Stein, Hebart, &
Sterzer, 2011a, Experiment 4). Therefore, future studies
should investigate the impact of these attributes on the depth
of suppression.
In summary, different visual attributes may affect the depth
of suppression with CFS. To utilize CFS, it is necessary to
control or manipulate these attributes. That is, to obtain an
optimal level of suppression with CFS, we must carefully
adopt maskers with specific levels of each of the visual attri-
butes we just reviewed. These visual attributes include tem-
poral and spatial frequencies, color, as well as the internal
structure and spatial density of the maskers. As we noted,
the optimal level of each visual attribute of the maskers de-
pends on the attributes of the targets. For example, when we
wish to suppress a moving target, we should use moving
Mondrians and when we wish to suppress a chromatic target,
we should use chromatic Mondrians. Based on the studies that
we reviewed in this section, we recommend the following
points:
1. Temporal frequency (TF): To suppress a target with a
long presentation duration, we recommend using
Mondrians with low TF. To suppress a target with a short
presentation duration, we recommend using Mondrians
with high TF.
2. Spatial frequency (SF): any study that uses CFS to
achieve suppression needs to perform a spectral power
analysis and report the results (for an example, see E.
Yang & Blake, 2012). This helps us to generalize the
results of CFS studies and facilitates replication of the
previous studies. For example, if a researcher wishes to
replicate a nonconscious effect with CFS, this researcher
Psychon Bull Rev
Author's personal copy

can adopt maskers with the same spectral power as the
previous studies. Moreover, by comparing the spectral
power of maskers across studies that evaluated noncon-
scious perception of a certain visual category (such as face
recognition), researchers can draw a clearer conclusion
regarding the opposing results.
3. Color: To suppress a chromatic target, we recommend
using chromatic Mondrians. When the target stimulus is
achromatic, we recommend using achromatic Mondrians.
As we saw, chromatic Mondrians may even be less effec-
tive to suppress some types of target stimuli such as
motion.
4. Motion: To suppress a moving target, we recommend
using moving Mondrians. When adopting moving
Mondrians, we must select moving Mondrians with spe-
cific speed and pattern of motion. Because the evidence
regarding these two components is not conclusive, and no
other study has replicated the results of Ananyev et al.
(2017) or Moors et al. (2014), we suggest that studies
attempting to suppress a moving target evaluate the depth
of suppression resulting from different speed and patterns
of motion in a preliminary study.
5. Internal structure of the maskers: We recommend using
Mondrians with intact solid areas, and especially intact
edges.
6. Spatial density: We recommend using Mondrians with
high spatial density. Moreover, Mondrians’TFs should
be adjusted based on the spatial density of them.
Hence, we suggest careful modification of the
abovementioned attributes in any CFS study. Moreover, some
researchers have matched the visual attributes of the maskers
and the targets to obtain optimal suppression. That is, they
have adopted similar visual attributes to optimally suppress
the target stimulus. For example, Moors et al. (2014)found
more effective suppression when matching the speed of
Mondrians and the targets (but see Ananyev et al., 2017).
This strategy is based on feature-selective suppression which
is a well-known feature of binocular rivalry (see Alais &
Melcher, 2007). Based on this assumption, matching the vi-
sual attributes of the masks and the targets increases the depth
of suppression (e.g., see Moors et al., 2014). For example, in
some studies, circular Mondrians (those with elliptical ovals)
were used to suppress a face stimulus based on the assumption
that similarity between ellipses and faces will induce a stron-
ger suppression (e.g., Izatt, Dubois, Faivre, & Koch, 2014;
Stein, Hebart, et al., 2011, Experiment 4; also see E. Yang,
Brascamp, Kang, & Blake, 2014).
There is, however, some evidence against this assumption.
For example, Ananyev et al. (2017) showed that matching the
pattern of motion between target and moving Mondrians
results in a weaker suppression, and E. Yang and Blake
(2012) showed that targets with high SF are less prone to the
suppression with CFS, regardless of the maskers’SF, and
maskers with low SF are more effective in inducing suppres-
sion. Therefore, matching the SF of Mondrians and mask may
not be a good strategy to obtain a strong suppression. Based
on this, we believe that to utilize the feature-selective suppres-
sion for CFS-induced suppression, future studies need to test
this assumption carefully for different visual attributes.
How to assess awareness in a CFS
experiment?
Regardless of the method used to study nonconscious percep-
tion, it is crucial to use an accurate and reliable measure of
awareness to ensure that participants are not conscious of the
target stimulus. If we cannot evaluate conscious perception
with an acceptable accuracy, we are not able to make any
conclusions about conscious and nonconscious perception.
In this section, we will discuss some of the main methods
for evaluating conscious perception and will discuss how to
obtain a reliable assessment of consciousness. The focus will
be on the differences between subjective and objective mea-
sures of conscious perception.
There are important differences between subjective and
objective measures used to evaluate conscious perception
(Persuh, 2018).When using subjective measures, researchers
ask participants to report the visibility of the target (Dehaene
&Naccache,2001;Lau,2008). That is, participants are re-
quired to indicate if they consciously perceiving the target. We
consider these measures as subjective because participants’
reports in such tasks are based on their subjective experiences.
Subjective measures’face validity is one of the main advan-
tages of these measures, as they provide a direct assessment of
subjective nature of conscious perception (Lau, 2008). There
are a variety of subjective measures, which evaluate conscious
perception directly or indirectly (by evaluating different cog-
nitive phenomena related to consciousness). When using di-
rect, subjective measures, we can ask participants to use a yes/
no detection task or a graded scale.
5
In the yes/no detection task, participants are asked to judge
the visibility of a target with two alternatives: yes (detecting or
seeing the target) and no (not detecting or not seeing the target;
see Dehaene & Naccache, 2001).
6
This task is popular
5
The difference between these yes/no detection and graded measures is based
on a controversy regarding the nature of subjective experiences: Is conscious
perception graded or dichotomous (this topic is beyond the scope of this
review; we refer our readers to some of the many excellent reviews in this
matter: Overgaard et al., 2006; Sergent & Dehaene, 2004)?
6
It is important to note that we use different terms to refer to detecting or not
detecting the target, depending on the type of sensory system we are studying.
For example, when one aims to study auditory consciousness (e.g., Dykstra,
Cariani, & Gutschalk, 2017), the two detection alternatives are hearing the
target or not hearing the target.
Psychon Bull Rev
Author's personal copy

because it is easy for participants to use, and it yields lower
variance in threshold estimation than other methods (Lesmes
et al., 2015). There are some issues with this method from
both philosophical and experimental points of view, but the
main problem is the problem of conservative criterion
(Cowey, 2010; Goldiamond, 1958; also see Peters, Ro, &
Lau, 2016): When assessing participants’awareness using
the yes/no detection task, they are more likely to use a con-
servative criterion to report the presence/absence of the target
(see below). Moreover, this method has been criticized on the
basis that it considers conscious perception as a dichotomous
phenomenon, and therefore it does not capture the subjective
nature of conscious perception exhaustively and exclusively
(Sandberg & Overgaard, 2015; Timmermans & Cleeremans,
2015). Because of this, some researchers used graded scales to
evaluate conscious perception.
A graded scale, which has been utilized in a growing num-
ber of studies in the recent years, is the Perceptual Awareness
Scale (the PAS; Ramsøy & Overgaard, 2004). When using the
PAS, participants judge the visibility of the target stimulus by
using four categories of perceptual responses: “No experi-
ence,”“brief glimpse,”“almost clear experience,”and “abso-
lutely clear experience.”The PAS has advantages that make it
an intuitive subjective measure of conscious perception: It
seems the PAS is an exhaustive measure of conscious percep-
tion, as it captures different levels of perception than complete
visibility and invisibility (Sandberg, Bibby, & Overgaard,
2013; Timmermans & Cleeremans, 2015), and a participant’s
performance in the PAS correlates better with objective per-
formance (Overgaard, Rote, Mouridsen, & Ramsøy, 2006;
Sandberg & Overgaard, 2015; Sandberg, Timmermans,
Overgaard, & Cleeremans, 2010; see also Mazzi, Bagattini,
& Savazzi, 2016; Overgaard, Fehl, Mouridsen, Bergholt, &
Cleeremans, 2008). Furthermore, at least two studies found
strong correlations between the PAS responses and the mag-
nitude of neuronal signals recorded by EEG (Tagliabue,
Mazzi, Bagattini, & Savazzi, 2016) and MEG (Andersen,
Pedersen, Sandberg, & Overgaard, 2016).
There are some drawbacks related to the PAS. For ex-
ample, the PAS effectiveness depends on the ability of
participants to report their subjective experiences
(Sandberg et al., 2010). Moreover, the borders between
different levels of the PAS are not absolute and explicit,
and this raises two issues: First, participants may get con-
fused when using the PAS, especially if they are not good
introspectionists, and second, we can probably assume that
there are more levels between the four proposed levels in
the PAS (for a solution for these two issues, see Ramsøy &
Overgaard, 2004; Sandberg et al., 2013). Finally, using the
PAS does not solve the problem of decision criterion. As
with the yes/no detection task, participants may adopt a
conservative (or even a liberal) criterion to categorize their
subjective experiences with the PAS.
Some researchers prefer to use subjective, but indirect mea-
sures of conscious perception. These measures are called in-
direct because they assess conscious perception indirectly by
evaluating related phenomena to conscious perception. One
indirect measure is confidence ratings. In this task, partici-
pants rate their confidence concerning their performance or
conscious perception (metacognitive confidence rating). This
approach is based on the views that link conscious experi-
ences with metacognition by stating that metacognitive expe-
riences are essential parts of our phenomenology (Koriat,
2007; Kunimoto, Miller, & Pashler, 2001; Metcalfe &
Schwartz, 2016; also see Lau & Rosenthal, 2011;Rosenthal,
2019). Based on this approach, if a participant is highly con-
fident, and if the confidence associates closely with accuracy,
then that individual is consciously perceiving the target (e.g.,
Kunimoto et al., 2001; also see Peters & Lau, 2015). In con-
trast, if a participant is not confident in his or her responses or
the confidence does not associate with accuracy, that individ-
ual has no conscious perception of the target. Even though
conscious perception and metacognitive experiences are tight-
ly linked, there might be cases that these two are separated
from each other (Charles, Van Opstal, Marti, & Dehaene,
2013; Jachs, Blanco, Grantham-Hill, & Soto, 2015; Reder,
1996). Consequently, metacognitive confidence ratings do
not always represent conscious perception.
In general, confidence rating suffers from a similar issue as
the PAS. As with the PAS, the different categories of confi-
dence rating do not have clear and absolute borders.
Subsequently, two participants with comparable conscious
and metacognitive experiences may respond differently to
confidence rating questions (Sandberg et al., 2010).
Moreover, the problem of conservative criterion is a main
issue when assessing conscious perception using confidence
ratings. Therefore, like the PAS and yes/no detection task,
participants may adopt a conservative criterion (see below
for a potential solution for this issue).
We now switch to the objective measures of conscious
perception. Using objective measures is another approach to
evaluate conscious perception. In this approach, participants
are asked to classify the category of the target or determine the
interval in which the target was presented using a variety of
objective tasks, such as m-alternative forced choice (mAFC),
m-interval forced choice (mIFC), or discrimination task. Here,
participants are not being asked whether they perceived a tar-
get consciously or how confident they are that they saw a
target, but where the target was or in what interval the target
occurred. If an observer cannot perform this task at above-
chance levels, we can assume that this participant is not aware
of the target. These measures are called objective because it is
believed that they provide an objective assessment of aware-
ness (See Schmidt & Vorberg, 2006). In this approach, a par-
ticipant’s performance is contrasted between a direct
(objective) task and indirect task (e.g., priming and
Psychon Bull Rev
Author's personal copy

adaptation). We can observe a nonconscious perceptual effect,
if participants are at chance level in the direct task but above
chance level in the indirect task.
Some issues cast doubt on the adequacy of the objective
measures for evaluating conscious versus nonconscious per-
ception. First, establishing that a participant’sperformanceon
an objective task is not different from chance level requires
proper statistical analysis such as power analysis (Sterzer,
Stein, Ludwig, Rothkirch, & Hesselmann, 2014;E.Yang
et al., 2014). The majority of the studies, however, use tradi-
tional p-value testing methods, which may not be sufficient to
establish the lack of awareness (accepting the null hypothesis;
see Sterzer et al., 2014;E.Yangetal.,2014). Moreover, there
are situations in which participants’performances are different
from one objective task to another. For example, a partici-
pant’s performance may be at chance level in a two-
alternative forced-choice (2AFC) color-discrimination task,
but above chance level in a 2AFC orientation-discrimination
task (Sterzer et al., 2014). Therefore, participant’sconscious
perception of the target depends on the type of the objective
task that we use to evaluate it. That is, in the above example,
we cannot assert that the participant is conscious or not. A
potential solution for this problem is to use the same dimen-
sion (feature) in both direct and indirect tasks (Sterzer et al.,
2014). For example, if researchers wish to study nonconscious
perception of orientation, they should use the 2AFC
orientation-discrimination task as the measure of awareness.
Two potential problems that we discussed already, lack of
proper statistical analysis and the different levels of perfor-
mance in the different objective tasks, are resolvable and rel-
atively easy to avoid. There are, however, more fundamental
problems that are often ignored. First, there is a common as-
sumption that objective tasks such as mAFC and mIFC are
bias free. But they are not, even though they are less suscep-
tible to the adoption of a conservative criterion than subjective
tasks (e.g., Cowey, 2010). More importantly, objective tasks
are measures of performance, not conscious experience (Lau,
2008). When considering participants’performance in an ob-
jective task as a measure of awareness, we make an assump-
tion that performance is the representation of conscious expe-
rience. But there are several situations in which there is a clear
dissociation between conscious experience (measured by sub-
jective tasks) and performance (measured by objective tasks).
The classic example of this dissociation is blindsight. People
with blindsight report that they are not able to perceive the
target stimulus, which is presented inside the blind field.
These people, however, have higher-than-chance performance
in an objective task (Weiskrantz, 1986,1996; Weiskrantz,
Barbur, & Sahraie, 1995). Therefore, if we assume that per-
formance in an objective task represents conscious perception,
then blindsight patients are not blind anymore and are able to
perceive the targets consciously and have a normal visual life
as people with normal and functional vision (see Rosenthal,
2019). Therefore, above-chance performance in an objective
task does not necessary imply conscious perception, and in
some cases, when using subjective tasks as the measure of
conscious perception, above-chance performance in an objec-
tive task actually represents nonconscious perception.
Based on this, we argue that objective measures are not
appropriate for evaluating conscious perception. Therefore,
the contrast between direct (objective) and indirect (e.g., prim-
ing and adaptation) measures as the measure of nonconscious
perception might not be an accurate measure of dissociation
between conscious and nonconscious perception. This inade-
quacy might explain some of the null results from different
blinding methods, especially CFS. That is, in some of the
previous studies, participants who had above-chance perfor-
mance in an objective task were excluded from further analy-
sis (e.g., Almeida, Mahon, & Caramazza, 2010;Almeida
et al., 2014; Barbot & Kouider, 2012;Lapateetal.,2016).
This means that the participants who were showing signs of
nonconscious perception by having above-chance perfor-
mance in an objective task, were excluded. Therefore, it is
quite possible that some studies could not find evidence for
nonconscious perception, only because they did not include
participants who were showing signs of nonconscious
perception.
Therefore, we think that subjective measures are better
measures of awareness than objective measures. There are,
however, some methodological concerns that may undermine
the efficiency of the subjective measures that need to be ad-
dressed. When using subjective measures, above-chance dis-
crimination performance (e.g., in the mAFC, adaptation or
priming task) in the trials that participants report they are not
able to see the target (miss trials) is considered as the sign of
nonconscious processing. The observed nonconscious effect
resulted from this approach may be due to a statistical artifact,
regression to mean (Shanks, 2017; for a comprehensive
discussion on the problems of this approach, see Schmidt,
2015).
When evaluating nonconscious perception, the observed
nonconscious perception might be the results of the difference
between conscious and nonconscious processes or the result
of adopting a conservative criterion in the detection task (this
might be a decision bias or a perceptual bias; see Witt, Taylor,
Sugovic, & Wixted, 2015). Obviously, the former is the aim
of studies that try to contrast these two modes of perception
(e.g., see Peters & Lau, 2015). The traditional methods, in
which the performance accuracy is contrasted against chance
level in the miss trials (we call it “accuracy-based measures”)
are not able to separate the effects of perceptual and criterion
processes. Therefore, a different method than accuracy-based
measures is required to evaluate the contrast between con-
scious and nonconscious perception. One way to evaluate
the aforementioned dissociation between conscious and non-
conscious perception is by using signal detection theory
Psychon Bull Rev
Author's personal copy

(SDT; Green & Swets, 1966; Macmillan & Creelman, 2004).
According to SDT, a perceptual effect is represented as a
change in perceptual sensitivity (d′,or d-prime), which is the
measure of ability of participants to discriminate signal from
noise. SDT also provides a measure of bias, even though this
measure represents both perceptual and decisional bias, and
SDT is not able to measure these two types of bias separately
(Witt et al., 2015). Based on this, d′can be the measure of
interest in conscious research.
An important finding in the field of visual awareness is
related to the higher sensitivity of blindsight patient G.Y. in
a 2AFC task than the yes/no detection task (Azzopardi &
Cowey, 1997), meaning G.Y.’s ability to discriminate signal
from noise in the conscious detection task is abolished
completely (yes/no detection d′= 0), but he is able to discrim-
inate signal from noise in an objective 2AFC task (discrimi-
nation d′higher than zero; also see Azzopardi & Cowey,
1998). This means that we can consider higher d′value in an
objective task as compared with a subjective task as the psy-
chophysical signature of perception in the absence of aware-
ness
7
(Balsdon & Azzopardi, 2015; Heeks & Azzopardi,
2015; Lloyd, Abrahamyan, & Harris, 2013;Yoshida&Isa,
2015; but see Peters et al., 2016), which could be a more
accurate assessment of conscious and nonconscious percep-
tion than accuracy-based measures. We can use this approach
when using confidence ratings as the measure of awareness. In
this situation, nonconscious perception is represented by
higher sensitivity in the objective task than in confidence rat-
ing sensitivity (note that metacognitive confidence rating sen-
sitivity is different from sensitivity in the detection or objec-
tive task. We will discuss metacognitive sensitivity below).
Although this new approach (which we call it “SDT-based
analysis”) is an improvement over the approach that utilizes
the accuracy-based measures, to our surprise, only a few stud-
ies have adopted this approach (e.g., Balsdon & Azzopardi,
2015; Heeks & Azzopardi, 2015; Lloyd et al., 2013;Yoshida
&Isa,2015), and to the best of our knowledge, no CFS study
has done so.
There are, however, at least two limitations regarding the
SDT-bases approach that need to be addressed. First, although
we can use this approach to investigate the dissociation be-
tween conscious and nonconscious processing, it may not be
suitable to answer some of the most important questions in the
field of consciousness research (e.g., can perceptual process-
ing happen in the absence of awareness?). Second, we can use
this approach only if the subjective and objective tasks mea-
sure two different formats of the same underlying construct
(e.g., yes/no detection and 2IFC detection tasks as the mea-
sures of conscious and nonconscious perception, respectively;
see Azzopardi & Cowey, 1997; Heeks & Azzopardi, 2015).
Therefore, one cannot compare sensitivity of two different
tasks that measure two different constructs (e.g., yes/no detec-
tion and two-choice discrimination of orientation tasks).
Therefore, a new analysis approach is needed to evaluate con-
scious and nonconscious perception. We argue that the model-
based analysis that we will describe in the next few paragraphs
will solve these issues and assess conscious and nonconscious
perception more rigorously.
Recently, Pournaghdali and colleagues introduced a novel
model-based analysis approach based on a multidimensional
extension of SDT, called general recognition theory (GRT;
Ashby & Soto, 2015; Soto et al., 2017) to evaluate perceptual
processing in the absence of awareness (Pournaghdali,
Schwartz, Hays, & Soto, 2020). By fitting GRT model to a
data collected from yes/no detection and two-choice discrim-
ination task, Pournaghdali and colleagues constructed a curve
called the sensitivity versus awareness (SvA) curve. The SvA
curve depicts conditional sensitivity of participants in a dis-
crimination task (e.g., 2AFC, priming and adaptation task) for
each level of relative likelihood of awareness (see Fig. 1;also
see Pournaghdali et al., 2020,Fig.1b). By providing an ob-
jective benchmark (criterion) based on an ideal observer, the
SvA curve divides the distribution of relative likelihood of
awareness in two areas: the area (or region) of high likelihood
of awareness and the area (or region) of low likelihood of
awareness (it is important to note that x-axis in the SvA curve
is labeled “relative likelihood of no awareness.”Based on this,
in the x-axis of an SvA curve, zero represents a high level of
awareness, and higher values in this axis represents higher
likelihood of not being aware; see Fig. 1).
By evaluating the pattern resulted from the SvA curve, we
can determine whether perceptual processing necessary for a
perceptual discrimination task is awareness-dependent (see
Fig. 1). That is, we can specify if a specific visual processing
is dependent or independent of awareness. Independence be-
tween a specific visual processing and awareness is represent-
ed in the SvA curve as a consistent level of conditional sensi-
tivity of the discrimination performance (henceforth, sensitiv-
ity) at all levels of the relative likelihood of awareness (see
Fig. 1a). The dependency of visual processing on awareness is
represented by systematic change in the sensitivity asthe func-
tion of the relative likelihood of awareness. If sensitivity drops
as the likelihood of awareness drops and reaches zero before
reaching the region of low likelihood of awareness (before
objective criterion), then this visual processing requires
awareness, and nonconscious perception of this class of visual
stimulus does not occur (see Fig. 1b). In two situations, how-
ever, dependency of visual processing on awareness implies
nonconscious perception. First, if sensitivity drops as the like-
lihood of awareness drops but does not reach zero before
objective criterion, then we will have a visual processing that
could happen in the absence of awareness (see Fig. 1c). In this
7
It is important to note that SDT assumes a mathematical relationship between
ds of detection and 2AFC (or 2IFC) tasks: 0detectionðÞ¼d02AFCðÞ*
ffiffiffi
2
p.
Psychon Bull Rev
Author's personal copy

case, the nonconscious processing is weaker than conscious
processing, but is still reliable. Finally, it is possible that the
sensitivity is above zero only in the area of low likelihood of
awareness, and it rises as the likelihood of awareness drops
(see Fig. 1d). This is a case of nonconscious perception in
which the processing of information might actually improve
in the absence of awareness (see also Newell & Shanks,
2014).
The SvA curve has some unique features that make it an
appealing approach to evaluate conscious and nonconscious
perception: First, the SvA curve considers awareness as a
nonbinary continuous phenomenon. That is, even when the
yes/no detection task is used to assess awareness, the SvA
curve converts the responses into a continuous distribution.
This potentially solves a problem that faces the yes/no
detection task. That is, even if participants report their experi-
ences with a dichotomous task, the SvA curve helps to capture
the graded nature of awareness. Second, the SvA curve is
compatible with different measures of awareness such as the
PAS or metacognitive confidence ratings (see Pournaghdali
et al., 2020, Supplementary Material). More importantly, by
utilizing the objective benchmark that is different from sub-
jective criterion, the SvA curve provides a bias-free assess-
ment of conscious and nonconscious perception. Finally, the
SvA also depicts the subjective criteria of all the participants.
This in turn will help researchers to evaluate participants’
response bias with more accuracy. Thus, we think that this
model-based analysis captures the same psychophysical ef-
fects as the previous SDT-based approach (e.g., Azzopardi
&Cowey,1997; Heeks & Azzopardi, 2015), but with a
Relative Likelihood of No Awareness
Region of High Likelihood
of Awareness
Objective Criterion Objective Criterion
Objective Criterion Objective Criterion
Region of High Likelihood
of Awareness
Region of High Likelihood
of Awareness
Region of High Likelihood
of Awareness
Region of Low Likelihood
of Awareness
Region of Low Likelihood
of Awareness
Region of Low Likelihood
of Awareness
Region of Low Likelihood
of Awareness
Sensitivity (d' or meta-d')
ab
cd
Fig. 1 Sensitivity versus awareness (SvA) curve. This graph depicts four
different SvA curves. In each graph, the x-axis represents “the relative
likelihood of no awareness,”with zero meaning that the participant is
consciously perceiving the target. As we move toward the higher values
in the x-axis, participant’s likelihood of not being aware of the target
(relative likelihood of no awareness) increases. The y-axis represents
participant’s sensitivity in the perceptual task (e.g., priming, visual
aftereffect, discrimination task). Zero sensitivity represents chance level.
The red vertical line represents the objective criterion based on the
optimal observer (see (Pournaghdali et al., 2020). This line divides the
SvA graph into two areas: the region of high likelihood of awareness (left)
and the region of low likelihood of awareness (right). The four graphs
depict different possibilities of theassociation between consciousness and
perceptual processing. aRepresentation of a condition in which there is
no association between awareness and a specific perceptual processing. b
Representation of a condition in which a specific perceptual processing
depends on awareness. In this condition, as the likelihood of awareness
drops, so does the sensitivity. However, the sensitivity in this condition
reaches zero before the objective criterion and in the area of high
likelihood of awareness. This means this specific type of perceptual
processes depends on consciousness and it does not exist in the absence
of awareness. cRepresentation of a situation in which sensitivity drops as
the likelihood of awareness drops and reaches zero after the objective
criterion and in the area of low likelihood of awareness. The higher-
than-chance sensitivity (sensitivity higher than zero) in the area of low
likelihood of awareness indicates nonconscious perception. d
Representation of a situation in which the sensitivity is at the chance
level in the area of high likelihood of awareness, but it starts improving
after the objective criterion and in the area of low likelihood of awareness.
Note that we can use different measures of awareness to evaluate
awareness with SvA curve. Moreover, the y-axis can also represent
Type 2 d′or meta-d′. For this, SvA curve represents changes in the
ability of metacognitive system to evaluate perceptual processing as the
function of awareness. (Color figure online)6
Psychon Bull Rev
Author's personal copy

greater resolution. Hence, this approach allows us to obtain a
finer understanding of the dissociation between conscious and
nonconscious perception, and the level of processing in the
absence of awareness across different perceptual paradigms
and tasks (Pournaghdali et al., 2020).
One final note is that we can evaluate sensitivity while
using metacognitive confidence ratings as the measure of con-
scious perception. For this purpose, a different measure of
sensitivity has been proposed: metacognitive (or Type 2) sen-
sitivity (Type 2 d′; for review, see Fleming & Lau, 2014).
There are, however, different versions of Type 2 d′.In one
version, Kunimoto et al. (2001) proposed a sensitivity mea-
sure called a′, which follows similar assumptions as traditional
d′(Type 1 d′). This measure, however, is not bias free (Evans
& Azzopardi, 2007) and does not measure metacognitive sen-
sitivity accurately.
Recently, another measure of metacognitive sensitivity was
proposed that does not hold the same assumptions as Type 1 d′
and is a more accurate measure of metacognition as it can
differentiate Type 2 sensitivity with more precision than other
measures: meta-d′(Maniscalco & Lau, 2012,2014;alsosee
Fleming & Lau, 2014). Therefore, if researchers wish to use
metacognitive confidence rating as the measure of conscious
perception, it is advisable to use meta-d′.We emphasize this
issue because, recently, a few studies evaluated conscious
perception using metacognitive ratings and used a′.
We can model meta-d′using the SvA curve as well. In this
situation the SvA curve represents meta-d′as the function of
likelihood of awareness. That is, instead of measuring Type 1
sensitivity, the SvA curve can assess meta-d′. We can use this
SvA curve to address some of the main questions regarding
consciousness and metacognition: What is the nature of the
association between consciousness and metacognition? Is
consciousness dependent on metacognition? Can we observe
metacognitive processes in the absence of awareness (see
Pournaghdali et al., 2020)?
In conclusion, the assessment of consciousness is a critical
part of any study in the field of consciousness, and we need to
evaluate consciousness with the best tools available and ana-
lyze the results with proper statistical techniques. In the next
two sections we focus on the behavioral and neural evidence
regarding nonconscious recognition of different visual catego-
ries with CFS, in light of the sections that we just discussed.
That is, we try to draw a clearer picture regarding the noncon-
scious perception with CFS and see if experimental configu-
ration and assessment of awareness has any impact on the
results of previous studies.
Nonconscious perception with CFS
When presenting participants with a target stimulus within
the CFS paradigm, participants report that they are not
able to see the target consciously. However, participants
show signs of nonconscious vision via a variety of
methods, including adaptation aftereffects, priming ef-
fects, breaking time of the suppressed target, and above-
chance performance in an objective discrimination task.
One way to evaluate nonconscious perception is by exam-
ining participants’performance in an objective discrimi-
nation task against chance level in the trials that partici-
pants report not seeing the target. Above-chance perfor-
mance in these trials indicates nonconscious perception of
the invisible target.
In an adaptation aftereffect study, prolonged exposure
to a stimulus (the adaptor), with a specific property (such
as left to right motion) causes participants to perceive the
subsequent target (which is neutral in regard to the di-
mension of the interest) as having a property opposite to
the adaptor (right to left motion). For example, if the
adaptor is a male face and the target is a face that can
be seen as male or female, participants report the target
face as a female face (gender aftereffect; e.g., Afraz &
Cavanagh, 2009; Ng, Boynton, & Fine, 2008). In con-
trast, in a priming study, presentation of a prime facili-
tates perception of the subsequent target, if the target is
congruent with the prime. This priming effect shows it-
self through faster reaction time and more accurate clas-
sification of the target (e.g., Bar & Biederman, 1998).
Both adaptation aftereffect and priming effect could oc-
curiftheadaptorortheprimeisvisible(consciousor
visible condition) or invisible (nonconscious or invisible
condition). Moreover, if the effect transfers between the
two eyes, and if the effect is observed despite different
target’sandprime’s (or adaptor’s) size, we can assume
that the effect is not low level. Otherwise, the effect is
high or at least midlevel.
Another way of evaluating nonconscious perception
with CFS is to use a variant of CFS called “breaking
continuous flash suppression,”or bCFS (Jiang, Costello,
&He,2007). In the bCFS paradigm, an experimenter
measures the time that a target stimulus needs to break
the suppression of the masks (detection time), such that
the participant becomes aware of the target. In this type of
study, researchers usually consider the systematic differ-
ence between the breaking times of two targets as the
nonconscious advantage of a target that overcome the
suppression faster. For example, if upright faces break
the suppression faster than inverted faces, it indicates that
face stimuli are being processed nonconsciously under the
influence of CFS (for a critical review on bCFS, see
Gayet, Van der Stigchel, & Paffen, 2014).
A main question here is the extent to which the visual
information is processed under the influence of CFS. That
is, researchers try to investigate the possibility of process-
ing of high-level visual information (e.g., face identity,
Psychon Bull Rev
Author's personal copy

facial expression, semantic information) during CFS.
There are, however, criteria for establishing high-level
visual perception in the absence of awareness. If any of
these criteria do not hold for any high-level visual cate-
gory, we question the high-level nonconscious perception
of that category. First, any study investigating high-level
nonconscious perception should rule out any involvement
of low-level features (e.g., contours, contrast, shapes) on
the observed nonconscious effect (Moors et al., 2019;
Moors, Hesselmann, Wagemans, & van Ee, 2017).
Second, a high-level nonconscious perception should be
replicated at least once, as a matter of good methodolog-
ical procedure (Moors et al., 2019). If a high-level non-
conscious perception is only observed in one study, we
should be careful in interpreting the results of that study.
Finally, positive results from bCFS studies are not suf-
ficient to conclude that there is high-level nonconscious
perception. That is, in addition to bCFS results, other
studies should provide support for high-level visual pro-
cessing in the absence of awareness (Moors et al., 2019).
This is important because, in order to establish a noncon-
scious effect with bCFS, one needs to compare the detec-
tion time resulted from bCFS with the detection time re-
sulted from a binocular control condition that mimics the
bCFS (see Jiang et al., 2007). Researchers usually consid-
er presence of the effect (difference between the detection
times of the two or more target stimuli) in the bCFS con-
dition and a null or a weaker effect in the control condi-
tion as the sign of nonconscious perception. This ap-
proach, however, does not provide a conclusive evidence
for pure nonconscious processing under the influence of
CFS, because the differences between bCFS and control
condition may arise not only from the nonconscious pro-
cesses, but also from other factors that are not related to
the nonconscious processing (Stein, Hebart, et al., 2011a;
also see Gayet et al., 2014). Therefore, establishing a pure
nonconscious perception with bCFS is difficult, and we
should always interpret bCFS results with caution.
Therefore, based on these criteria and based on our
discussion on the experimental configuration and the as-
sessment of consciousness, we review some of the impor-
tant findings regarding the nonconscious perception of
different classes of target stimuli with CFS. For each vi-
sual category, we present a table summarizing the exper-
imental configuration and the assessment of awareness,
used in each study (Tables 1through 7). In the current
section, we focus on the level of visual processing in the
absence of awareness with CFS as well. Our intention is
to explore what the evidence tell us about the nature of
nonconscious perception when employing CFS. We also
refer our readers to other reviews on CFS that we think
they are relevant (Faivre, Berthet, & Kouider, 2014;
Moors, 2019;E.Yangetal.,2014).
Nonconscious perception of orientation
Early CFS studies (Bahrami, Carmel, Walsh, Rees, &
Lavie, 2008; Kanai, Tsuchiya, & Verstraten, 2006)pro-
vided evidence for nonconscious adaptation of orientation
with CFS. However, Peremen and Lamy (2014)didnot
find a reliable nonconscious priming effect (faster and
more accurate response to the presentation of the congru-
ent target to the prime) when the target was invisible with
CFS.AccordingtoKoivistoandGrassini(2018), howev-
er, several methodological issues—such as long presenta-
tion duration of the prime, similarity of some parts of the
mask with the target, and the prime contrast—may ex-
plain the null result of Peremen and Lamy (2014).
Therefore, Koivisto and Grassini (2018)evaluatedthe
nonconscious influence of an invisible prime (right/left
oriented arrow), presented for a shorter duration than
Peremen and Lamy’s(2014) study, on the processing of
the target and found that the congruent target/prime pair
elicits a faster and more accurate response than the incon-
gruent pair, meaning CFS does not interrupt nonconscious
processing of the prime. In contrast to Peremen and Lamy
(2014), Koivisto and Grassini (2018) found evidence for
nonconscious priming of orientation with CFS. Therefore,
Koivisto and Grassini’s(2018) study and adaptation after-
effect studies of orientation provided an irrefutable sup-
port for nonconscious perception of orientation with CFS.
In addition, Song and Yao (2016) showed that partic-
ipants classify the orientation of an invisible Gabor patch
at above chance level in a nonconscious 2AFC task.
Song and Yao (2016) also showed overall improvement
in nonconscious visual perception of three visual catego-
ries (right/left orientation, face/house, and happy/sad
face) with increasing the target luminance, but it is not
clear if the nonconscious perception of orientation
changed with change in luminance, because they did
not present the luminance results for separate categories
(Song & Yao, 2016,Fig.3).Hong(2015) provided ev-
idence for preferential processing of radial orientation,
8
while suppressing the visual stimulus using CFS. In this
study, Hong (2015, Experiment 1) showed that sinusoi-
dal Gabor patch stimuli with radial orientation break the
suppression faster than those stimuli with tangential ori-
entation. This means that participants not only are able to
processes orientation nonconsciously but also have pref-
erential processing for specific orientations.
8
Radial orientation refers to a situation in which the orientation and the loca-
tion of the target are homogenous. For example, if a Gabor patch is left tilted
and the target is in the top left or bottom right side of the screen, the Gabor
patch has a radial orientation. Otherwise the orientation is tangential.
Psychon Bull Rev
Author's personal copy

To summarize, despite the differences in experimental con-
figuration and assessments of awareness between studies
discussed in this section, a majority of them showed that even
though CFS disrupts conscious perception of target, orienta-
tion is being processed in a nonconscious manner. That is,
these studies provided conclusive evidence for nonconscious
perception of orientation under the influence of CFS.
Therefore, we can conclude that orientation is being processed
under the influence of CFS (see Table 1for a summary of the
experimental configuration and the assessment of awareness
in the studies that evaluated nonconscious perception of
orientation with CFS).
Nonconscious face recognition
Early adaptation studies with CFS showed that awareness is
necessary for processing of face stimuli, and an invisible adap-
tor fails to induce the nonconscious face aftereffect with CFS
(Amihai, Deouell, & Bentin, 2011; Moradi, Koch, & Shimojo,
Table 1. Experimental configuration and the awareness task in the studies that evaluated nonconscious perception of orientation
Study Masks’
type
Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness task Is awareness assessed in the main
block?
Bahrami et al. (2008) Mondrians 30 Hz NR* Yes No Objective + CR No
Kanai et al. (2006) Mondrians 15 Hz NR Yes No Yes/No No
Peremen & Lamy
(2014)
Mondrians 10 Hz NR Yes No The PAS Yes
Koivisto & Grassini
(2018)
Mondrians 11 Hz NR Yes No Objective + The PAS Yes
Song & Yao (2016) Mondrians 33.33 Hz NR Yes No Subjective Graded
Scale
Yes
Hong (2015) Mondrians NR NR No No NA** NA
*NR: not reported. **For the bCFS studies, we did not report the awareness measures in this and the following tables, because the awareness measure in
bCFS studies have a different application than other type of nonconscious studies
Table 7 Experimental configuration and the awareness task in the CFS studies of a variety of visual categories
Study Masks’type Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness task Is awareness assessed in the
main block?
Zhan & de Gelder
(2018)
Mondrians 10 Hz NR Yes No No trial by trial
assessment of
visibility*
No
Costello et al. (2009) dynamic noise
pattern
NR NR Yes No NA NA
Y.-H. Yang & Yeh
(2011)
Mondrians 10 Hz NR No No NA NA
Zabelina et al. (2013) Mondrian like
pattern
10 Hz NR Yes No Yes/No Yes
Bahrami et al. (2010) Mondrians 30 Hz NR Yes No Subjective graded scale Yes
Hesselmann et al.
(2015)
Mondrians 10 Hz NR No No The PAS Yes
Gomes et al. (2017) Mondrians 10 Hz No** No No NA NA
Hedger et al. (2015a) Mondrian like
pattern
10 Hz NR Yes Yes Yes/No (report a
breakthrough)
Yes
Objective No
Koivisto & Rientamo
(2016)
Mondrians 11 Hz NR Yes Yes Objective Yes
Kimchi et al. (2018) Mondrians 10 Hz NR Yes No The PAS Yes
S.-Y. Lin & Yeh
(2016)
Mondrians 10 Hz NR Yes No Yes/No Yes
Raio et al. (2012) Mondrians 10 Hz NR Yes No Objective/CR Yes
Gayet et al. (2016) Mondrians 10 Hz NR No No*** NA NA
*In this study, participants debriefed at the end of the experiment,when they were asked about their perceptual experiences during the main experiment.
**This study reported the spatial frequency band of the target stimuli. ***In this study, the targets were presented inside a colorful annulus
Psychon Bull Rev
Author's personal copy

2005). In addition, Shin, Stolte, and Chong (2009)showed
that an invisible face induces gender aftereffect only if partic-
ipants direct their attention to the location of the adaptor. The
nonconscious aftereffect, however, disappears when partici-
pants direct their attention to a location other than the adap-
tor’s. Moreover, Stein and Sterzer (2011)showedthatanin-
visible adaptor induces face-shape adaptation only when the
adaptor and the target are presented to the same eye and have
the same size. According to Stein and Sterzer (2011), if we
present the target and the adaptor to different eyes or when
they are presented to the same eye but with different sizes, the
adaptation aftereffect disappears, indicating this aftereffect is
rather low level. Finally, Stein, Peelen, and Sterzer (2012)
provided evidence for nonconscious eye-gaze adaptation with
CFS, when both the target and the adaptor were presented to
the same or different eyes. This adaptation, however,
Table 3 Experimental configuration and the awareness task in the bCFS studies of face recognition
Study Masks’type Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness task Is awareness assessed in
the main block?
Jiang et al. (2007) Dynamic noise
pattern
NR NR Yes Yes NA NA
Stein, Peelen, et al. (2011b) Mondrians 10 Hz NR No No NA NA
Stein et al. (2016) Mondrians 10 Hz NR No No NA NA
Zhou et al. (2010) Dynamic noise
Patch
NR NR No No NA NA
Geng et al. (2012) Noise/Mondrians 100 Hz NR Yes Yes NA NA
Chen & Yeh, (2012) Mondrians 10 Hz NR No No NA NA
Stein, Senju, et al. (2011c) Mondrians 10 Hz NR Yes No NA NA
Yokoyama et al. (2013) Mondrians 20 Hz NR Yes No NA NA
Gobbini, Gors, Halchenko,
Rogers, et al. (2013b)
Mondrians 10 Hz NR Yes Yes NA NA
Gobbini, Gors, Halchenko,
Hughes, et al. (2013a)
Mondrians 10 Hz NR Yes Yes NA NA
Moorsetal.(2016) Mondrians 10 Hz NR No No NA NA
Hung et al. (2016) Mondrians 10 Hz NR Yes Yes
(Exp.
1–2)/
No (Exp.
3)
Subjective Graded Scale +
Objective (Exp. 3)
Yes
NA (Exp. 1–2) NA
Nakamura & Kawabata (2018) Dynamic noise
pattern
10 Hz NR Yes No NA NA
Table 2 Experimental configuration and the awareness task in the priming and aftereffect studies of nonconscious face recognition
Study Masks’type Masks’
TF
Masks’
SF
Colorful
Masks?
Colorful
Target?
Awareness task Is awareness assessed in the
main block?
Amihai et al. (2011) Mondrians 10 Hz NR Yes No Yes/No (report a
breakthrough)**
Yes
Moradi et al. (2005)MovingRandom
Dot
NR* NR No No Yes/No (report a
breakthrough)
Yes
Shin et al. (2009) Dynamic Radial
Gratings
1 Hz NR No No Yes/No (report a
breakthrough)
Yes
Stein & Sterzer
(2011)
Mondrians 10 Hz NR Yes No Yes/No (report a
breakthrough)
Yes
Stein et al. (2012) Mondrians 10 Hz NR Yes No Yes/No (report a
breakthrough)
Yes
Barbot & Kouider
(2012)
Mondrians 8 Hz NR Yes No Objective No
Izatt et al. (2014) Mondrians 10 Hz NR No No Objective + The PAS Yes
Gelbard-Sagiv et al.
(2016)
Mondrians 10 Hz NR No Yes Objective + The PAS Yes
*In this study, authors did not report the speed of the moving dots. **In such studies, the researchers asked participants to report seeing anything other
than the mask during or after the trial. Therefore, they did not use the traditional yes/no detection task
Psychon Bull Rev
Author's personal copy

disappeared when the target and the adaptor were presented
with different size. Overall, preceding studies indicate that the
nonconscious high-level adaptation of face stimuli is impaired
under the influence of CFS, as compared with face adaptation
when the adaptor is consciously visible.
Nonconscious face priming studies provide further evi-
dence for the idea that CFS impairs high-level face recognition
and indicate that nonconscious face recognition with CFS
might be due to the process of lower-level features of the face
stimuli. Barbot and Kouider (2012), for example, showed that
when an invisible face prime is presented for 60 ms, it facili-
tates perception of a subsequent face target (only when is
presented to the same eye as the prime). The invisible prime
has an inhibitory influence on the processing of the target
when it is presented for 1,000 ms. According to Barbot and
Kouider (2012), we observe the facilitatory and inhibitory
impacts of short and long invisible prime, only when both
the prime and the target are presented to the same eye. This
means that the nonconscious face priming with CFS is rather
low-level.Moreover, Izatt et al. (2014) found that face identity
priming (when prime and target have different view point) and
face repetition priming (when prime and target have same
view point) are abolished when rendering the prime invisible
for 50 ms with CFS, and concluded that CFS interferes with
face perception at a lower level of visual processing.
Moreover, Izatt et al. (2014) showed that face priming impair-
ment is associated with chance-level performance in a 2AFC
task. Gelbard-Sagiv, Faivre, Mudrik, and Koch (2016)also
showed that we can observe nonconscious face priming with
CFS, but only when participants report that they are aware of
lower-level visual features of the prime, like color
(Experiment 1) and location (Experiment 2). According to
Gelbard-Sagiv et al. (2016), in absence of low-level aware-
ness, the invisible prime is unable to induce nonconscious
Table 4 Experimental configuration and the awareness task in the priming and aftereffect studies of nonconscious facial expression recognition
Study Masks’type Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness task Is awareness assessed in the
main block?
Adams et al. (2010) Mondrians 19 Hz NR No No Objective No
E. Yang et al. (2010) Mondrian 10 Hz NR Yes No Yes/No (report a
breakthrough)
Yes
Almeida et al. (2013)Randomnoise
pattern
10 Hz NR Yes Yes Objective No
Chiesa et al. (2015)Scrambledface
images
10 Hz NR Yes Yes Objective + Yes/No
(report a break-
through)
Yes
Anderson et al. (2012) Mondrians 10 Hz NR No No Yes/no Yes
Kring et al. (2014) Mondrians 10 Hz NR No No Objective No
Lapate et al. (2014) Mondrians ~10.71
Hz
NR Yes Yes Objective No
Ye et al. (2014) Mondrians 15 Hz NR Yes Yes Yes/No (report a
breakthrough)
Yes
Doi & Shinohara (2016) Mondrians 20 Hz NR No No Yes/no No
Faivre et al. (2012) Mondrians 8 Hz NR Yes No Objective Yes*
Vetter, Badde, Phelps, &
Carrasco (2019)
Mondrians 28 NR Yes No Objective discrimination
+ the PAS
Yes
*In this study, the authors performed awareness task in a separate trial than the main trials but in the same block
Table 5 Experimental configuration and the awareness task in the bCFS studies of facial expression recognition
Study Masks’type Masks’TF Masks’SF Colorful masks? Colorful target? Awareness task Is awareness assessed
in the main block?
Sterzer et al. (2011) Mondrians 10 Hz NR No No NA NA
Tsuchiya et al., (2009) Mondrians 10 Hz NR Yes No NA NA
E. Yang et al. (2007) Mondrians 10 Hz NR No No NA NA
Stein et al. (2014) Mondrians 10 Hz NR* No No NA NA
Gray et al. (2013) Mondrians 10 Hz NR No No NA NA
Hedger et al. (2015b) Mondrians 20 Hz NR No No NA NA
Stein & Sterzer (2012) Mondrians 10 Hz NR No No NA NA
*This study used high/low and unfiltered bandpass face stimuli
Psychon Bull Rev
Author's personal copy

impact on the target. Therefore, results from priming studies
also indicate that CFS impairs nonconscious high-level face
recognition, and any observed nonconscious face priming is
rather the result of low-level features of the prime and the
target (see Table 2for a summary of the experimental
configuration and the assessment of awareness in the
priming and adaptation aftereffect studies of face recognition
with CFS).
Whereas results from adaptation and priming studies indi-
cate that face recognition is impaired by CFS, several bCFS
studies provide evidence for nonconscious face recognition un-
der the influence of CFS. Using the bCFS technique, different
studies have shown that upright faces break the suppression
faster than inverted faces do (Jiang et al., 2007; Stein, Peelen,
&Sterzer,2011b; Stein, Reeder, & Peelen, 2016; Zhou, Zhang,
Liu, Yang, & Qu, 2010), indicating that the high-level process-
ing of face stimuli may survive under the influence of CFS.
Furthermore, other bCFS studies showed that participants’
own face (Geng et al., 2012, Experiment 1), faces with direct
gaze (Chen & Yeh, 2012; Stein, Senju, Peelen, & Sterzer,
2011c; Yokoyama, Noguchi, & Kita, 2013, Experiment 1),
faces belong to familiar identities (Gobbini, Gors, Halchenko,
Rogers, et al., 2013b), faces oriented toward the observer
(Gobbini, Gors, Halchenko, Hughes, & Cipolli, 2013a), and
faces that are experienced in a congruent configuration
(Moors, Wagemans, & de-Wit, 2016) overcome the suppres-
sion faster. In addition, Hung, Nieh, and Hsieh (2016)showed
that in comparison to invisible unattractive faces, invisible at-
tractive faces break the suppression faster (Experiment 1), have
lower detection threshold (Experiment 2), and nonconsciously
direct spatial attention (Experiment 3; also see Nakamura &
Kawabata, 2018). Therefore, results from bCFS studies may
indicate the survival of face-recognition processes in the ab-
sence of awareness while employing CFS. But as we discussed
at the beginning of this section, the observed nonconscious
effects from bCFS studies may not be the result of pure non-
conscious processes, and other factors may influence these ef-
fects. Based on this, the results from bCFS studies should be
interpreted with caution.
To summarize, adaptation and priming studies showed that
CFS impairs high-level face recognition, even though results
from bCFS studies provided supports for the survival of high-
level face recognition under the influence of CFS. Because
bCFS studies are the only ones that provided support for the
nonconscious high-level face recognition with CFS, we con-
clude that CFS impairs the nonconscious high-level face rec-
ognition (see Table 3for a summary of the experimental
configuration and the assessment of awareness in the bCFS
studies of face recognition).
Nonconscious facial expression recognition
Another important visual category that may show signs of
high-level nonconscious perception under the influence of
CFS is facial expression. Several studies investigated the pos-
sibility of nonconscious facial expression recognition when
we render the face stimuli invisible with CFS (Adams, Gray,
Garner, & Graf, 2010; Almeida, Pajtas, Mahon, Nakayama, &
Caramazza, 2013; Faivre, Berthet, & Kouider, 2012;
Table 6 Experimental configuration and the awareness task in the CFS studies of manipulable objects
Study Masks’Type Masks’
TF
Masks’
SF
Colorful
Masks?
Colorful
Target?
Awareness task Is awareness assessed in
the main block?
Almeida et al.
(2010)
Random noise pattern 10 Hz NR Yes* No/with a
color filter
*
Objective No
Almeida et al.
(2008)
Random noise pattern 10 Hz NR Yes* No/with a
color filter
*
Yes/No (Exp.
1)/objective
(Exp. 2–5)
No
Almeida et al.
(2014)
Random noise pattern 10 Hz NR Yes No/with a
color filter*
Objective No
Sakuraba et al.
(2012)
Random noise pattern 10 Hz NR Yes No/with a
color filter*
Yes/no
detection
No
Hesselmann et al.
(2016)
Mondrians 10 Hz NR Yes No The PAS Yes
The PAS +
Objective
No
Hesselmann et al.
(2018)
Mondrians (Exp. 1–5)/random
noise pattern (Exp. 7)
10 Hz NR Yes No The PAS Yes
Objective +
The PAS
No
Rothkirch &
Hesselmann
(2018)
Random noise pattern 10 Hz NR No No Objective +
The PAS
No
*These studies used anaglyph glasses which means target and masks needed to be presented with colorful filters
Psychon Bull Rev
Author's personal copy

Tsuchiya, Moradi, Felsen, Yamazaki, & Adolphs, 2009;E.
Yang, Hong, & Blake, 2010). Evidence regarding noncon-
scious facial expression recognition with CFS, however, is
not conclusive. Whereas adaptation aftereffect studies provid-
ed conflicting evidence for nonconscious facial expression
recognition, priming studies showed that an invisible emo-
tional face prime can influence how we perceive and evaluate
a subsequent target. Moreover, even though bCFS studies
showed that fearful faces break the suppression faster, we
cannot rule out the possibility of influence of low-level effects
on the privilege of fearful faces in bCFS.
Adams et al. (2010) showed that invisible face stimuli with
happy, fearful, and angry emotions can induce significant high-
level facial expression aftereffect (FEA), even though the mag-
nitude of FEA was significantly lower than the visible condi-
tion. Yang et al., (2010), on the other hand, found no evidence
of nonconscious FEA with CFS (also see Adams et al., 2011).
Accordingtotheprimingstudies, an invisible emotional face
(prime) can influence how we perceive and evaluate the target
stimulus belong to different visual categories, such as unfamil-
iar alphabets (Almeida et al., 2013; Chiesa, Liuzza, Acciarino,
& Aglioti, 2015; but see Faivre et al., 2012, Experiments 5a–b),
neutral faces (Anderson et al., 2012; Kring, Siegel, & Barrett,
2014; Lapate, Rokers, Li, & Davidson, 2014; but see Faivre
et al., 2012, Experiment 5b) and faces with a congruent emotion
to the prime (Ye, He, Hu, Yu, & Wang, 2014). An invisible
emotional face also has the ability to influence numerosity es-
timation (Doi & Shinohara, 2016) and induce reliable skin con-
ductance response (only when the face is fearful; Lapate et al.,
2014;seeTable4for a summary of the experimental
configuration and the assessment of awareness in the priming
and adaptation aftereffect studies that evaluated nonconscious
facial expression recognition with CFS).
Moreover, results from bCFS studies indicate that faces with
fearful expression overcome the suppression faster than neutral
faces or faces with other emotional expressions (Sterzer et al.,
2011;Tsuchiyaetal.,2009;E.Yangetal.,2007), meaning that
emotional content of face stimuli is being processed under the
influence of CFS. According to Stein, Seymour, Hebart, and
Sterzer (2014), the advantage of fearful faces to break the sup-
pression relies on high spatial frequency of fearful faces, meaning
that the parvocellular visual pathway is processing the noncon-
scious information. Although fearful faces overcome the sup-
pression faster than faces with other emotional expressions, it
seems this privilege, to some extent, may not be exclusively
due to the emotional content of faces but also because of low-
level features (Gray, Adams, Hedger, Newton, & Garner, 2013;
Hedger, Adams, & Garner, 2015b). In contrast, Stein and Sterzer
(2012) showed that schematic positive (happy) faces break the
suppression faster than schematic negative (threatening) faces,
but it seems the configuration of faces in their study was respon-
sible for the breaking time difference between positive and neg-
ative faces (see Table 5for a summary of the experimental
configuration and the assessment of awareness in the bCFS
studies of facial expression recognition).
Emotional faces can also influence eye movements under
the influence of CFS. Recently, Vetter et al. (2019)usedCFS
to suppress emotional faces (fearful, angry and neutral) from
awareness while recording participant’eyemovementinre-
sponse to the face stimuli belong to the abovementioned emo-
tional categories. The results of this study indicated that, com-
pared with the invisible neutral faces, participants direct their
gaze toward the invisible fearful faces and away from the
invisible angry faces. According to these results, under the
influence of CFS, suppressed emotional faces can impact
eye movements (Vetter et al., 2019).
Evidence for nonconscious high-level recognition of facial
expression under the influence of CFS is not conclusive.
Although adaptation, priming, and bCFS studies showed that
the perception of facial expression survives under the influ-
ence of CFS, low-level features of face stimuli might be re-
sponsible for observed nonconscious perception of facial ex-
pression. That is, the observed nonconscious recognition of
facial expression might be because of the processes of lower-
level features but not the emotional content of the emotional
faces. Therefore, it is crucial to evaluate the extent to which
low-level features and emotional content of face stimuli sep-
arately influence nonconscious recognition of facial expres-
sion. Based on this, we cannot conclude that high-level facial
expression recognition survives under the influence of CFS.
Nonconscious perception of manipulable
objects
Manipulable objects is another important visual category that
has been studied extensively with CFS. Evaluating noncon-
scious recognition of manipulable objects is important be-
cause this evaluation provides a mean to evaluate prediction
of the two-stream theory of visual perception by Milner and
Goodale (2008). According to the two-stream theory of visual
perception, visual consciousness is related to neural activities
in the ventral pathway (which includes important visual areas,
such as V4, the occipital face area, and different parts of tem-
poral lobe, such as the inferotemporal cortex [IT] and superior
temporal sulcus [STS]), and activities in the dorsal pathway
(which includes areas such as V5, V7, and different parts of
parietal lobe, such as intraparietal sulcus) may not be neces-
sary for conscious vision, but may support nonconscious vi-
sion. In accordance with this theory, Z. Lin and He (2009)
proposed that CFS disrupts activities in the ventral pathway,
but not the dorsal pathway. Therefore, presence of high-level
perception of manipulable objects, which are being processed
in the dorsal stream, and absence of high-level perception of
nonmanipulable objects under the influence of CFS is a strong
evidence for the two-stream theory of visual perception and
Psychon Bull Rev
Author's personal copy

the aforementioned proposal by Z. Lin and He (2009). Some
perceptual priming studies have tried to evaluate this proposal
by comparing priming effect of manipulable (tools) versus
nonmanipulable objects.
Some studies using CFS showed that suppressed tool
primes have facilitatory effect on the perception of visible
target from the same category (category priming), whereas
suppressed nonmanipulable objects did not influence subse-
quent perception of visible target from the same category
(Almeida et al., 2010; Almeida et al., 2008). Subsequent stud-
ies, however, related this priming effect to the elongation of
tools and showed elongated objects other than tools (like a fish
or vegetables
9
), have similar category priming effect (Almeida
et al., 2014; Sakuraba, Sakai, Yamanaka, Yokosawa, &
Hirayama, 2012; for review, see Hebart & Hesselmann,
2012; Ludwig & Hesselmann, 2015). Thus, these studies sug-
gest that tool priming is a low-level effect about length and
orientation rather than a high-level effect about tool
categorization.
Recently, however, both results have been challenged by a
series of studies that showed neither tools nor other elongated
objects induce category priming (Hesselmann, Darcy,
Ludwig, & Sterzer, 2016; Hesselmann, Darcy, Rothkirch, &
Sterzer, 2018; Rothkirch & Hesselmann, 2018). There are at
least two differences between earlier and more recent studies,
which may explain the results. First, the tool targets were all
elongated in the earlier studies (Almeida et al., 2010;Almeida
et al., 2008; Almeida et al., 2014; Sakuraba et al., 2012),
whereas more recent studies (Hesselmann et al., 2016,2018;
Rothkirch & Hesselmann, 2018) used both elongated and
nonelongated tool targets. Second, earlier studies (Almeida
et al., 2010; Almeida et al., 2008; Almeida et al., 2014;
Sakuraba et al., 2012)usedred–green anaglyph glasses,
whereas the most recent studies used a mirror stereoscope to
induce the rivalry (Hesselmann et al., 2016; Hesselmann et al.,
2018; but see Rothkirch & Hesselmann, 2018). The suppres-
sion may not be complete with anaglyph glasses, and some
parts of the “to be suppressed”image might bleed through the
glass. Therefore, it is possible that some cross talk exists be-
tween the two eyes when we use these glasses (Carmel et al.,
2010; but see Hesselmann et al., 2018,Experiment7).Finally,
earlier studies used random noise patterns to induce suppres-
sion (Almeida et al., 2010; Almeida et al., 2008; Almeida
et al., 2014; Sakuraba et al., 2012), as compared with the
recent studies that used Mondrians to induce suppression
(but see Rothkirch & Hesselmann, 2018). This means that
studies that used random noise patterns might have induced
an incomplete suppression, and a small part of the target might
have been visible in these studies. This suggests that some of
the effects in the earlier studies are because of incomplete
suppression of the target. Therefore, conscious perception of
the tool targets has influenced the results of the earlier studies.
Therefore, it seems that CFS interferes with category prim-
ing of manipulable objects, even though elongation shape of
tools may induce nonconscious shape priming under the in-
fluence of CFS (Hesselmann et al., 2016;Hesselmannetal.,
2018; Rothkirch & Hesselmann, 2018). Based on this, CFS
studies of manipulable objects did not provide support for the
two-stream theory of visual perception and indicated that cat-
egorical representation of this class of visual objects does not
survive under the influence of CFS. That is, the results of these
studies indicate that nonconscious perception of manipulable
object does not occur with CFS (see Table 6for a summary of
the experimental configuration and the assessment of
awareness in the studies that evaluated nonconscious
perception of manipulable objects with CFS).
Nonconscious recognition of other visual
categories
Finally, there are handful of other visual categories for which
research has examined if there is nonconscious perception
under the influence of CFS. Some other studies provided ev-
idence for nonconscious visual abilities such as body posture
recognition (Zhan & de Gelder, 2018), word recognition
(Costello et al., 2009; Jiang et al., 2007; Y.-H. Yang & Yeh,
2011; Zabelina et al., 2013), numeric estimation (Bahrami
et al., 2010; but see Hesselmann, Darcy, Sterzer, & Knops,
2015), recognition of threatening species (Gomes, Silva,
Silva, & Soares, 2017; but see Hedger, Adams, & Garner,
2015a), discrimination of animals from nonanimal stimuli
(Koivisto & Rientamo, 2016), and perceptual grouping
(Kimchi, Devyatko, & Sabary, 2018; S.-Y. Lin & Yeh,
2016). Furthermore, research showed that fear learning can
happen under the influence of CFS (Raio, Carmel, Carrasco,
&Phelps,2012) and also a previously learned target could
break the suppression faster than an unlearned target (Gayet,
Paffen, Belopolsky, Theeuwes, & Van der Stigchel, 2016;see
Table 7for a summary of the experimental configuration and
the assessment of awareness in the studies that evaluated
nonconscious perception of visual categories discussed in
this section).
Is there any conclusive evidence for high-level
nonconscious perception with CFS?
The behavioral studies that we reviewed in this section
failed to provide conclusive evidence for high-level non-
conscious perception under the influence of CFS (see
9
Note that both manipulable and nonmanipulable objects can be elongated or
nonelongated. An example of an elongated manipulable object is a screwdriver
because its length is longer than its width. An example of a nonelongated
manipulable object is a doorknob because the length is not significantly longer
than the width.
Psychon Bull Rev
Author's personal copy

Tables 8and 9), even though the results of behavioral stud-
ies indicate that nonconscious perception of low-level and
midlevel visual information (such as orientation and shape)
may survive under the influence of CFS. Although some
studies provided evidence for high-level nonconscious per-
ception with CFS, such as bCFS and priming studies of
facial expression recognition, they failed to rule out the
impact of low-level visual attributes in the observed non-
conscious effects. Therefore, the general body of the behav-
ioral literature of CFS did not provide support for high-level
nonconscious perception with CFS.
Moreover, the heterogeneity in the experimental configu-
ration and the incongruent assessment of awareness across
different studies seem to play a major role in the inconsistent
results regarding nonconscious high-level recognition of dif-
ferent categories with CFS. Therefore, it is crucial for future
studies to take these issues into consideration. Indeed, such
future studies could draw a better picture regarding level of
visual processing in the absence of awareness under the influ-
ence of CFS.
The neural correlates of conscious
and nonconscious perception with CFS
One current aim of science of consciousness is to find the
neural activities that are associated with conscious experi-
ences (the NCCs; Chalmers, 2000; Crick & Koch, 2003),
especially those neural activities that are substrates of con-
scious perception (the neural substrates of consciousness;
Aru, Bachmann, Singer, & Melloni, 2012; de Graaf, Hsieh,
&Sack,2012; for recent reviews on this topic, see Boly et al.,
2017; Koch, Massimini, Boly, & Tononi, 2016; Odegaard,
Knight, & Lau, 2017; Tsuchiya, Wilke, Frässle, & Lamme,
2015). The differences between conscious and nonconscious
perception in term of neural mechanisms is critical to
understanding these neurocomputational mechanisms. That
is, it is critical to isolate the neurocomputational mechanisms
that are directly related to conscious perception from those
that are related to the visual processing in the absence of
awareness. Many studies have taken advantage of CFS prop-
erties in order to evaluate neural basis of conscious and non-
conscious perception.
To evaluate the neural basis of nonconscious perception,
we must use three criteria that are similar to those we used in
the previous section, which we review again here. First, any
study investigating the neural basis of a nonconscious high-
level recognition of a visual category (such as face or facial
expressionrecognition) must rule out any involvement of low-
level features (e.g., contours, contrast, shapes) on the observed
nonconscious effect and the neural basis of this effect. If not,
we cannot conclude that the observed neural activity is
representing the high-level representation of the target.
Second, the association between a neural activity and noncon-
scious perception of a specific visual category should be rep-
licated at least once. In the absence of replication, we should
be careful in interpreting the observed association between
that neural activity and the nonconscious perception. Finally,
positive results from bCFS studies is not enough to establish
the neural basis of nonconscious perception (Moors et al.,
2019).
By taking into account these three criteria, we review some
of the CFS studies that evaluated the neural basis of noncon-
scious perception. First, we start with the primary visual cor-
tex and then continue by discussing evidence regarding the
role of the dorsal versus ventral streams in nonconscious rec-
ognition of tools and body expression. Then, we return to
discuss the neural basis of nonconscious perceptionof specific
important visual categories with CFS. We discuss the neural
correlates of nonconscious face and facial expression recogni-
tion. Similar to the previous section, for each visual category,
we present a table summarizing the experimental configura-
tion and the assessment of awareness, used in each study.
Table 8 Summary of the evidence for the nonconscious perception of
different categories
Visual Category Adaptation aftereffect Priming bCFS mAFC
Orientation ✓✓✓✓
Face × × ✓–
Facial expression ? ✓?–
Manipulable Objects –×––
Elongated objects –?––
Note. This table summarizes the evidence for the nonconscious percep-
tion of the main visual categories that we discussed in section
“Nonconscious perception with CFS.”“✓“indicates the evidence for
the existence of the nonconscious perception. “×”indicates the evidence
against the existence of the nonconscious perception. “?”indicates mixed
evidence for the existence of the nonconscious perception. “–” indicates
lack of evidence for or against of the nonconscious perception
Table 9 Summary of the evidence for nonconscious perception of
different visual categories
Visual category Evidence for nonconscious perception
Body posture ✓
Word ✓
Animals vs. nonanimal ✓
Perceptual grouping ✓
Fear learning ✓
Numeric estimation ?
Threatening species ?
Note. “✓“indicates the evidence for the existence of the nonconscious
perception. “?”indicates mixed evidence for the existence of the noncon-
scious perception
Psychon Bull Rev
Author's personal copy

The primary visual cortex (V1)
V1 has been a controversial region regarding its association
with conscious perception. Some people without V1 show a
clear dissociation between conscious and nonconscious per-
ception. That is, people with this condition report that they
cannot consciously see target stimuli when they are presented
inside the blind field (scotoma), but they can classify different
attributes of the target above chance level in a nonconscious
task (blindsight; Weiskrantz, 1986,1996). Inhibiting V1 ac-
tivity with transcranial magnetic stimulation (TMS) results in
a similar pattern of perception to blindsight (TMS-induced
blindsight; Allen, Sumner, & Chambers, 2014;Boyer,
Harrison, & Ro, 2005; Ro, Shelton, Lee, & Chang, 2004).
Lack of conscious perception in both blindsight and TMS-
induced blindsight implies that V1 is essential for conscious
perception (but see Lloyd et al., 2013; Mazzi, Mancini, &
Savazzi, 2014). However, evidence from neuroimaging and
single neuron studies of binocular rivalry, have been mixed
with respect to the role of V1 in conscious perception.
Neuroimaging studies in human participants showed that ac-
tivities in V1 closely correlate with conscious vision (e.g.,
Polonsky, Blake, Braun, & Heeger, 2000), whereas single
neuron studies in nonhuman primates found no supporting
evidence for this association (e.g., Leopold & Logothetis,
1996; for review, see Tong, 2003).
CFS has been employed to evaluate the role of V1 in con-
scious perception. In a CFS study, Watanabe et al. (2011)
manipulated attention and awareness independently and stud-
ied the impact of these two factors on blood oxygenation
level-dependent (BOLD) signals from V1. Important to this
study was the independent manipulation of the attention from
visibility. For this, Watanabe et al. (2011) asked participants
to report the visibility of the target in the “attention to target
block”or to report the detection of a nontarget letter at the
center of the fixation in the “attention to nontarget block.”By
using this design, Watanabe et al. (2011)showedthatatten-
tion, but not awareness, is associated with change in BOLD
signals from V1. This study, however, was criticized for being
statistically underpowered (Yuval-Greenberg & Heeger,
2013; also see Sterzer et al., 2014). Yuval-Greenberg and
Heeger (2013), with a similar design but more trials per par-
ticipant, tested the association between conscious perception
and V1 BOLD signal and showed that there is a correlation
between visibility of target under CFS and the magnitude of
BOLD signal from V1.
V1 also processes invisible targets and may be involved in
nonconscious perception. For example, Yamashiro et al.
(2014) showed robust V1 retinotopic response to the onset
of a target stimulus, even when the target was invisible.
According to this study, the magnitude of V1 responses were
not associated with individual differences in suppression du-
ration. The magnitude of retinotopic responses in V3 and V4v,
on the other hand, were associated with suppression duration,
meaning weaker response in V3 and V4v, but not in V1 and
V2, were associated with longer suppression duration.
Regardless of interindividual differences, Yamashiro et al.’s
(2014) results indicate that V1 is important for nonconscious
perception during CFS. Finally, Bahrami, Lavie, and Rees
(2007) showed that V1 is responsive to an invisible target,
but the intensity of V1 response was reduced when partici-
pants had to do a high-load attentional task. Therefore, where-
as V1’s role in conscious perception is not conclusive, it
seems V1 is responsible for nonconscious processing of visual
information (see Table 10 for a summary of the experimental
configuration and the assessment of awareness in the CFS
studies of V1).
Processing of objects in dorsal versus ventral
visual stream
Researchers have investigated the possibility that the CFS-
suppressed target may be represented in higher visual areas.
If these areas are involved, this would provide further evi-
dence for the neural basis of nonconscious perception with
CFS and resolve some of the inconsistent evidence from
behavioral studies as well. In a seminal study in this area,
Fang and He (2005) measured BOLD signals in response to
visible and invisible tools and faces in the regions of interest
(ROIs) in the dorsal (V3A, V7, and intraparietal sulcus [IPS])
and ventral (lateral occipital cortex [LO] and the anterior fu-
siform gyrus) pathways. In response to the invisible tools,
dorsal ROIs showed robust BOLD activity. The same re-
sponse was absent when the invisible target was a face.
Ventral ROIs, however, were only responsive to visible tool
and face objects, and were unresponsive to the invisible tar-
gets belonging to both categories.
The results of Fang and He (2005)supportsthetwo-stream
theory of visual perception (Milner & Goodale, 2008), in
which activities in dorsal visual areas are associated with non-
conscious perception and activities in ventral visual areas are
associated with conscious perception. Therefore, Z. Lin and
He (2009) proposed that CFS selectively impairs neural activ-
ities in the ventral stream,but not the dorsal stream. According
to this proposition, dorsal areas might be responsible for non-
conscious perception of different object categories under the
influence of CFS.
In accordance to the two-stream theory of visual perception
(Milner & Goodale, 2008), the majority of the following stud-
ies also pointed to the association between conscious percep-
tion and activities in the ventral ROIs. According to these
studies, a visible target induces stronger signal in the ventral
ROIs than an invisible target (Hesselmann, Hebart, & Malach,
2011; Hesselmann & Malach, 2011; Ludwig, Kathmann,
Sterzer, & Hesselmann, 2015; Zhan, Goebel, & de Gelder,
Psychon Bull Rev
Author's personal copy

2018b). The results regarding the neural correlates of noncon-
scious perception with CFS, however, are mixed, and did not
provide supporting evidence for this theory. Behavioral prim-
ing studies, for example, provided no evidence for this ac-
count, and they suggested that areas in the dorsal regions do
not remain active in the absence of conscious perception.
Further evidence from other neuroimaging studies, did
not support this account as well and failed to replicate Fang
and He’s(2005) results. For example, Hesselmann and
Malach (2011) showed that, in response to the invisible tar-
get, BOLD signals from ventral (LO and posterior fusiform
gyrus) and dorsal (V3A, V7 and IPS) ROIs attenuated
completely. Hesselmann and Malach (2011), by using a
more fine-grain analysis method called multivariate pattern
analysis (MVPA; Kamitani & Tong, 2005), decoded re-
sponses to the invisible target only in area LO. Moreover,
Hesselmann et al. (2011), by matching performance between
visible and invisible trials, found BOLD results similar to
Hesselmann and Malach’s(2011) study. The MVPA analy-
sis by Hesselmann et al. (2011), however, found decodable
activity to the invisible tool and face only in the posterior
fusiform gyrus. Presence of neural responses in the ventral,
but not dorsal ROIs in the absence of awareness in these
studies clearly contradicts the predictions of the two-stream
theory of visual perception.
The results ofmore recent studies also did not support Fang
and He’s(2005)results.Ludwigetal.(2015), for example,
showed similar stream-invariant BOLD signal reduction in
response to invisible tools and faces and showed nondifferent
classification of tools and faces with MVPA, even though by
using MVPA, they could decode elongated from non-
elongated tools in the fusiform face area (FFA) and right
V3A and V7, but not left V3A and V7. Ludwig et al.’s
(2015) results imply that shape, but not category of the invis-
ible objects, might be processed in the dorsal ROIs under the
influence of CFS. Moreover, by using MVPA, Fogelson,
Kohler, Miller, Granger, and Tse (2014) showed category-
specific classification in the middle occipital gyrus, the middle
occipital and lunate sulci, and the lingual gyrus (all belong to
the ventral stream in the occipital cortex). These results again
contradict the predictions of the two-stream theory of visual
perception. On the other hand, Ludwig, Sterzer, Kathmann,
and Hesselmann (2016), using MVPA, showed higher than
chance category classification in the ventral and dorsal ROIs,
when target was presented without the masks or with low-
contrast masks (visible condition). The classification accura-
cy, however, fell to the chance level when the target was
masked by high-contrast masks (invisible condition) in both
ventral and dorsal ROIs. Finally, Tettamanti, Conca, Falini,
and Perani (2017) compared responses to manipulable and
nonmanipulable objects using BOLD and MVPA and found
BOLD response to an invisible manipulable object in the dor-
sal (the ventral premotor area, and inferior and superior pari-
etal cortex) and the ventral (the lateral middle temporal gyrus)
ROIs. Nonmanipulable objects, on the other hand, caused
activity in the bilateral fusiform gyrus. Using MVPA,
Tettamanti et al. (2017) decoded manipulable and
nonmanipulable objects from the ventral premotor cortex, in-
ferior parietal cortex, lateral middle temporal gyrus, and right
fusiform gyrus. These results, therefore, did not provide sup-
port for the two-stream theory of visual perception.
In summary, studies following up to Fang and He’s(2005)
study failed to provide evidence for the proposition that neural
activities in the dorsal visual stream survive under the influ-
ence of CFS (see Table 11). According to these studies, invis-
ible tool objects do not induce distinguishable neural activity
in the dorsal areas from those in the ventral areas. Based on
these studies, the invisible objects belong to this visual cate-
gory may induce neural activity in both ventral and dorsal
areas. Moreover, it seems that the induced neural activities
in the dorsal and ventral areas are the results of representation
of lower level visual attributes such as elongation and not the
results of representation of high-level category information
about the manipulable objects. Therefore, the neuroimaging
and behavioral studies of manipulable objects did not provide
support for high-level processing of this visual category under
Table 10 Experimental configuration and the awareness task in the CFS studies of V1
Study Masks’type Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness
task
Is awareness assessed in
the main block?
Conscious
perception
Nonconscious
perception
Watanabe et al.
(2011)
Mondrians** 10 Hz NR No No Yes/No
detection
Yes ×–
Yuval-Greenberg &
Heeger (2013)
Mondrians** 12 Hz NR* No No Yes/No
detection
No*** ✓–
Yamashiro et al.
(2014)
Mondrians 7 Hz NR Yes No Yes/No
detection
Yes –✓
Bahrami et al.
(2007)
Mondrian
like
pattern
10 Hz NR No No Objective Yes –✓
*In this study, the spatial frequencies of the different component of the mask were reported. **In these studies, masks (Mondrians) were composed of
moving grating patches. ***In this study, the subjective responses were collected in a pretest run
Psychon Bull Rev
Author's personal copy

the influence of CFS. Hence, CFS studies do not support the
predictions of the two-stream theory of visual perception.
In contrast with the studies that used tool categories to
investigate the association between neural activities in the
dorsal visual pathway and nonconscious perception, Zhan
et al. (2018b) used images of upright and inverted body pos-
tures with fearful expression to investigate the dissociation
between dorsal and ventral pathways and their role in con-
scious and nonconscious perception. Because the upright
and inverted body postures are not different in shape and
elongation, any difference between neural responses to these
two categories should be higher level and not reducible to the
lower-level features (also see Zhan & de Gelder, 2018). Based
on this, Zhan et al. (2018b) found that dorsal ROIs (posterior
and middle IPS) show higher BOLD activation to the upright
than to the inverted body postures, and this activity was inde-
pendent of awareness. They also found higher BOLD activa-
tion in ventral ROIs (fusiform body area) in response to the
upright body postures, even when the body postures were in-
visible. Thus, the general pattern of results from Zhan et al.’s
(2018b) study indicated that the BOLD signals in dorsal stream
was independent of awareness. Because this study has not been
replicated, we warn readers to be cautious on making a firm
conclusion regarding its results. In summary, it seems the dorsal
visual stream does not process invisible image of tools, though
it might be responsible for nonconscious processes of body
posture (see Table 12 for a summary of the experimental
configuration and the assessment of awareness in the CFS
studies of dorsal vs. ventral visual pathway).
Neural correlates of conscious
and nonconscious face and facial expression
recognition
Because of the significance of face and facial expression rec-
ognition in humans’cognitive life, face recognition and facial
expression recognition are two of the best-studied visual
categories with CFS, and several studies investigated noncon-
scious face and facial expression recognition with CFS. Other
studies tried to investigate the neural basis of conscious and
nonconscious perception of these two categories while sup-
pressing the face stimuli with CFS.
Jiang and He (2006) were the first to evaluate the neural
differences between conscious and nonconscious perception
of face and facial expression (fearful vs. neutral) recognition
withCFS.AccordingtoJiangandHe(2006), face processing
areas (FFA and STS) and amygdala responded significantly to
visible face stimuli, with stronger response to the fearful faces
than the neutral faces in FFA and amygdala but not in STS. In
the invisible condition, FFA responded significantly to both
neutral and fearful faces with no difference between two emo-
tions. STS, on the other hand, only responded to the invisible
fearful, but not neutral faces. Finally, the amygdala responded
significantly to both fearful and neutral face stimuli, with stron-
ger response to the fearful face than the neutral faces. Based on
Jiang and He’s results, we suspect that FFA response is associ-
ated with nonconscious face recognition (because of similar
response to neutral and fearful faces), and amygdala and STS
activities are associated with nonconscious facial expression
recognition (for similar results with a different sample, see
Vizueta, Patrick, Jiang, Thomas, & He, 2012).
Moreover, Sterzer, Haynes, and Rees (2008) used univar-
iate (BOLD) and multivariate (MVPA) analysis to explore the
neural basis of nonconscious processing of faces and houses.
Whereas univariate analysis did not provide evidence for non-
conscious processing of faces and houses in the FFA and
parahippocampal place area (PPA) respectively, by using
MVPA, Sterzer et al. (2008) decoded visible and invisible
faces and houses from FFA and PPA above chance level,
with higher accuracy in decoding in the visible condition.
Based on the results from Jiang and He (2006)andSterzer
et al. (2008), we conclude that FFA might be responsible for
nonconscious face recognition, and STS and the amygdala
might be responsible for nonconscious facial expression
recognition. Two EEG/MEG studies from the same labs also
Table 11 Summary of the studies that evaluated the involvement of dorsal and ventral ROIs in nonconscious recognition of manipulable objects
Study Ventral/BOLD Ventral/MVPA Dorsal/BOLD Dorsal/MVPA
Fang & He (2005)× –✓–
Hesselmann & Malach (2011)× ✓××
Hesselmann et al. (2011)× ✓××
Ludwig et al. (2015)× × × ×
Fogelson et al. (2014)–✓–×
Ludwig et al. (2016)–×–×
Tettamanti et al. (2017)✓✓ ✓✓
“✓“indicates the evidence for the involevement of a dorsal or ventral ROI in noncosncious recognition of manipulable objects. “×”indicates the
evidence against the involevement of a dorsal or ventral ROI in noncosncious recognition of manipulable objects. “–” indicates lack of evidence for or
against the involevement of a dorsal or ventral ROI in noncosncious recognition of manipulable objects
Psychon Bull Rev
Author's personal copy

support this conclusion. In the first study, Jiang et al. (2009)
observed reliable face related negative deflection around
170 ms after the target’s onset (N170 component) in the bilat-
eral temporal recording sites, in response to the invisible fear-
ful, but not neutral and scrambled faces. According to Jiang
et al. (2009), STS is a generator of this component, even
though defining the exact origin of an EEG signal may not
be possible (see Luck, 2014). In another study, Sterzer,
Jalkanen, and Rees (2009) found a stronger M170 (N170
equivalent in MEG) component in the lateral occipitotemporal
sites in response to the invisible face than to the invisible
house. Hence, these two EEG/MEG studies provided addi-
tional support for Jiang and He’s(2006) and Sterzer et al.’s
(2008) results. Consequently, we consider these early studies
as the evidence for the involvement of the FFA in noncon-
scious face recognition and the involvement of the STS and
amygdala in nonconscious facial expression recognition.
Other neuroimaging studies from other labs also supported
Jiang and He’s(2006) and Sterzer et al.’s(2008) results. For
example, Troiani, Price, and Schultz (2014) found greater task-
irrelevant activation in the left amygdala and left FFA in
response to fearful faces than to houses when they masked
the targets with checkboards instead of Mondrians or noise
maskers. Lapate et al. (2016) showed that the right amygdala’s
response to the fearful faces is independent of visual awareness,
and also that its response to the invisible fearful faces correlates
with likability of the visible neutral faces, presented after the
invisible fearful faces. This means activity of the right
amygdala might be responsible for nonconscious priming of
emotional faces. Another EEG study also supported the
earlier results from neuroimaging and EEG/MEG results. In
this study, Suzuki and Noguchi (2013) found a larger N170
in response to an invisible upright face than to an invisible
inverted face in the occipital, temporal, and parietal electrodes.
Interestingly, this effect was inverse of the effect in the aware
condition (larger N170 in response to inverted than to upright
faces) and was the result of preserved N170 response to the
invisible upright face. In accordance with earlier studies, the
results from Suzuki and Noguchi’s(2013) study indicate that
face processing areas are responsive to invisible face stimuli
under the influence of CFS.
In contrast to these results, some other studies found no
evidence for the association between the nonconscious face
recognition and FFA activity, on one hand, and the
nonconscious facial expression recognition and STS and
amygdala activity, on the other hand. For example, Tsuchiya
et al. (2009, Experiment 3) showed that subject S.M., who
suffers from complete bilateral amygdala lesion, could detect
fearful faces faster than happy faces, similar to the same fear
advantage in the control participants. According to these results,
the amygdala may be unnecessary for nonconscious facial ex-
pression (fearful) recognition. Moreover, by using Mondrians
as the maskers, Troiani and Schultz (2013) failed to replicate
their previous results (Troiani et al., 2014).BasedonTroiani
and Schultz (2013), fearful face versus house stimuli do not
trigger distinguishable responses in the amygdala, FFA, and
PPA, even though the amygdala can discern stimulus-present
from stimulus-absent trials. Therefore, the results of these stud-
ies, in contrast to earlier studies, did not provide support for the
proposition that face processing areas are responsive to invisi-
ble face stimuli under the influence of CFS.
Some EEG/MEG results also refute the involvement of face
processing areas in nonconscious recognition of face and facial
expression under the influence of CFS. For example,
Schlossmacher, Junghöfer, Straube, and Bruchmann (2017)
showed that there is no difference in the magnitude of the
Table 12 Experimental configuration and the awareness task in the CFS studies of dorsal versus ventral visual pathway
Study Masks’
type
Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness task Is awareness assessed in the
main block?
Fang & He (2005) Dynamic
noise
10 Hz NR Yes* Yes* Objective No
Hesselmann & Malach
(2011)
Mondrians 10 Hz NR No No Subjective graded scale Yes
Hesselmann et al.
(2011)
Mondrians 10 Hz NR No No Subjective graded scale Yes
Ludwig et al. (2015) Mondrians 10 Hz NR No No Objective + The PAS Yes
Fogelson et al. (2014) Mondrians 10 Hz NR Yes Yes Yes/No + Objective Yes
Ludwig et al. (2016) Mondrians 10 Hz NR No No Objective + The PAS Yes
Tettamanti et al.
(2017)
Mondrians 10 Hz NR Yes* Yes The PAS Yes
Zhan et al. (2018) Mondrians 10 Hz NR Yes No No trial by trial assessment of
visibility**
No
Yes/No No
*These studies used anaglyph glasses, which means target and masks needed to be presented with colorful filters. **In this study, participants debriefed
at the end of each experimental run, when they were asked about their perceptual experiences during the run
Psychon Bull Rev
Author's personal copy

N170 (in the posterior sites) and late positive potential (LPP; in
the centroparietal sites) in response to fearful, happy, and
neutral faces. In addition, Sakuraba, Kobayashi, Sakai, and
Yokosawa (2013) found no difference between M170 re-
sponses in the invisible face, tools, and no-stimulus trials.
Sakuraba et al. (2013) found that invisible face stimuli suppress
alpha band EEG responses in the occipital sites between 0 and
200, and again between 500 and 1,000 ms after stimulus pre-
sentation. Finally, according to Geng et al. (2012), noncon-
scious self-face versus famous-other face has no impact on
N170 amplitude. In contrast, Geng et al. (2012) found reduced
vertex positive potential (VPP) in the central and frontocentral
sitesinresponsetoself-facethantofamous-otherface.Insum-
mary, the studies that we reviewed in the past two paragraphs
provided no indication for the involvement of face processing
areas in face and facial expression recognition under the influ-
ence of CFS (see Tables 13 and 14).
Gaze direction is an important factor in face recognition. As
we mentioned in the subsection titled Nonconscious Face
Recognition, direct gaze breaks the suppression faster than does
averted gaze (Chen & Yeh, 2012; Stein, Senju, et al., 2011;
Yokoyama et al., 2013). At least two studies evaluated neural
basis of gaze direction (direct vs. averted gaze) with CFS. In an
fMRI study, Madipakkam, Rothkirch, Guggenmos, Heinz, and
Sterzer (2015) found significantly stronger response to the vis-
ible faces with direct gaze in the FFA and IPS and close to
significant stronger response to the visible faces with direct
gaze in area STS, but not in the amygdala. In the invisible
condition, Madipakkam et al. (2015) found significant stronger
amygdala and STS and close to significant stronger FFA re-
sponse to the face with averted gaze. These results are in accor-
dance with Jiang and He’s(2006) and Sterzer et al.’s(2009)
results. In contrast, in an event-related potential (ERP) study,
Yokoyama et al. (2013) found significant larger negative visual
evoked potential deflection (around 200 ms after stimulus
presentation) to the faces with direct gaze in the frontal and
parietal sites, but not in the occipitotemporal areas.
Yokoyama et al.’s(2013) results indicate that temporal areas
(STS and FFA) are not responsible for nonconscious recogni-
tion of gaze. To summarize, the evidence regarding the neural
basis of gaze direction recognition under the influence of CFS is
not conclusive. Whereas the neuroimaging results from
Madipakkam et al. (2015) provided support for the involvement
of face processing areas (the FFA, STS, and amygdala), ERP
results from Yokoyama et al.’s(2013) study did not support
these results, even though a direct comparison between ERP
and neuroimaging results may not be possible.
In conclusion, according to the majority of the studies, FFA
might be responsible for nonconscious face recognition, and
the amygdala and STS might be responsible for nonconscious
facial expression recognition, although there are studies that
do not support these associations. One reason behind this dis-
parity might be related to the maskers’attributes used in the
studies. For example, Troiani and colleagues (Troiani &
Schultz, 2013; Troiani et al., 2014) found different results
when they used different maskers. That is, when they used
checkerboards as the maskers, they found neural responses
in the left amygdala and left FFA to fearful faces, but when
they used Mondrians to achieve the suppression, the neural
responses to fearful faces disappeared. Use of different types
of masks and incomplete suppression of face stimuli with
checkerboards in Troiani et al.’s(2014) study is the reason
for the discordance between these two studies. Moreover,
studies we reviewed in this section, presented their maskers
with different TFs. This may also explain some of the incon-
sistent results. Therefore, it is possible that some of the studies
that found evidence for the aforementioned associations failed
to suppress the face stimuli completely.
Table 14 Neural basis of nonconscious facial expression recognition
with CFS
Study FFA STS Amygdala N170/
M170
Jiang & He (2006)–✓✓ –
Troiani et al. (2014)✓–✓–
Lapate et al. (2016)––✓–
Jiang et al. (2009)––– ✓
Tsuchiya et al. (2009 ––×–
Troiani & Schultz (2013)× –×–
Schlossmacher et al. (2017)––– ×
FFA = fusiform face area; STS = superior temporal sulcus
“✓“indicates the evidence for the involevement of an area or ERP com-
ponent in noncosncious facial expression recognition. “×”indicates the
evidence against the involevement of an area or ERP component in
noncosncious facial expression recognition. “–” indicates lack of evi-
dence for or against the involevement of an area or ERP component in
noncosncious facial expression recognition
Table 13 Neural basis of nonconscious face recognition with CFS
Study FFA STS Amygdala N170/
M170
Jiang & He (2006)✓–– –
Sterzer et al. (2008)✓–– –
Sterzer et al. (2009)––– ✓
Suzuki & Noguchi (2013)––– ✓
Sakuraba et al. (2013)––– ×
Geng et al. (2012)––– ×
FFA = fusiform face area; STS = superior temporal sulcus
“✓“indicates the evidence for the involevement of an area or ERP com-
ponent in noncosncious face recognition. “×”indicates the evidence
against the involevement of an area or ERP component in noncosncious
face recognition. “–” indicates lack of evidence for or against the
involevement of an area or ERP component in noncosncious face
recognition
Psychon Bull Rev
Author's personal copy

Furthermore, studies that found evidence for the involvement
of face processing areas in nonconscious recognition of face and
facial expression under the influence of CFS do not rule out the
response of the ROIs to lower-level features of face stimuli. For
example, it is not clear if amygdala responses to an invisible
fearful face (Jiang & He, 2006; Lapate et al., 2016;Troiani
et al., 2014) is because of the emotional content or because of
the lower-level attributes of the faces. By taking the evidence
from behavioral studies that we reviewed in the sections titled
Nonconscious Face Recognition and Nonconscious Facial
Expression Recognition to account, we can assume that the
lower-level features of the face stimuli have substantive influence
on the neural responses in the FFA, the STS, and the amygdala.
We believe that future studies must investigate the extent of the
influence of high-level and lower-level attributes of an invisible
face stimulus on the neural responses in the aforementioned
ROIs. That is, future studies should investigate whether the neu-
ral responses in the three main regions are caused by high-level
visual information of the faces or by the lower-level features (see
Table 15 for a summary of the experimental configuration and
the assessment of awareness in the studies that evaluated
nonconscious perception of visual categories discussed in this
section).
Some authors also directly investigated the neural bases of
conscious face perception using CFS. In a recent study,
Baroni et al. (2017), utilized CFS to investigate the neural
basis of conscious face recognition. In this study, Baroni
et al. (2017) collected electrocorticographic (ECoG) responses
from five participants with epilepsy while manipulating the
visibility of the face targets by presenting the face stimuli with
different contrast. Baroni et al. used classifiers to decode per-
ceptual state (conscious perception vs. no perception) and
contrast of the target stimulus (high to low contrast).
According to Baroni et al., perceptual state of participants
was decodable from electrode sites in the face processing
areas in the fusiform gyrus, the lateral-temporal/inferior-pari-
etal cortex (based on authors’claim this area might correspond
to the STS). This study provided additional support for the
view that activities in the higher-order face areas (the FFA
and the STS) are responsible for conscious face recognition
(e.g., Tong, Nakayama, Vaughan, & Kanwisher, 1998).
What is the level visual processing
in the absence of awareness
under the influence of CFS?
The neuroimaging (e.g., fMRI) and electrophysiology (e.g.,
ERP) studies of CFS tried to shed more light on the question
that titles this section. To review, there is compelling evidence
that V1 is processing visual information in the absence of
awareness, even though attention may influence V1 activity
in the absence of awareness. Beyond V1, recent CFS studies
found no evidence supporting the Z. Lin and He (2009)prop-
osition that CFS selectively impairs neural activities in the
ventral stream, but not the dorsal. That is, evidence from
CFS studies challenges a prediction of the two-stream theory
of conscious perception (Milner & Goodale, 2008). In addi-
tion, the results of some CFS studies indicate that the FFA is
responsible for nonconscious face recognition, and the STS
and amygdala are responsible for nonconscious recognition of
facial expression, even though there are evidence against both
of these conclusions. Moreover, neuroimaging and electro-
physiology studies that provided support of involvement of
the FFA in nonconscious face recognition and the STS and
amygdala in nonconscious facial expression recognition failed
to exclude the influence of lower-level features of the face
stimuli on the observed neural responses. Therefore, these
studies did not provide conclusive evidence for high-level
visual processing under the influence of CFS.
Moreover, the brain recording studies of CFS suffer from the
heterogeneity in the experimental configuration and the incon-
gruent assessment of awareness. That is, this heterogeneity may
be a reason for some of the discrepancies in the previous studies.
Therefore, we can conclude that there is not enough evidence to
support the survival of high-level nonconscious visual processing
under the influence of CFS, even though evidence support sur-
vival of low-level nonconscious visual processing with CFS.
Concluding remarks
The aim of this review was to evaluate the ability of the cur-
rent literature on CFS to answer some of the crucial questions
in the science of consciousness. For this, we discussed how
different experimental parameters such as maskers’temporal
and spatial frequencies affect the depth of CFS-related sup-
pression. We also reviewed the optimal assessment methods
to evaluate awareness. Finally, we reviewed the CFS literature
regarding the dissociation between conscious and noncon-
scious perception at both the behavioral and neural levels.
By reviewing these issues, we hope to provide answer to the
questions that we will discuss in the following paragraphs.
Is there residual high-level perception of different visual
categories under the influence of CFS? Behavioral and neural
studies of higher-level visual categories with CFS failed to pro-
vide substantive evidence for nonconscious high-level percep-
tion with CFS. In some cases, in which the nonconscious effects
were observed (e.g., nonconscious recognition of manipulable
objects and facial expression), the observed effect was because
of the impact of lower-level features of the target stimuli. That
is, the studies that found the nonconscious effect for a high-
level visual category failed to rule out the influence of lower-
level features on the observed effect. In addition to this, other
criteria for establishing high-level nonconscious perception
with CFS (presence of replication and positive results from
Psychon Bull Rev
Author's personal copy

studies other than bCFS studies) has not been met by the pre-
vious literature. Therefore, we conclude that high-level noncon-
scious perception is abolished under the influence of CFS.
Based on this, we assert that the lower-level visual information
is being processed in lower (e.g., V1) and higher levels (e.g.,
FFA and STS) of visual hierarchy with CFS, but CFS deci-
mates high-level visual information.
One can question this conclusion and claim that neuroimag-
ing and electrophysiology studies support the high-level pro-
cessing of the targets in the absence of awareness with CFS, as
we saw in different fMRI and EEG studies. There are at least
three problem with this claim: First, neuroimaging and electro-
physiology studies did not provide conclusive evidence regard-
ing high-level visual perception under the influence of CFS.
Some studies found supporting evidence for the association
between activity in certain ROIs and nonconscious perception
of a specific visual category, and some other studies found no
supporting evidence for those associations. Moreover, some of
the observed associations might be because of the response of
the ROIs to low-level or midlevel features of target stimuli. For
example, in case of manipulable-object recognition, the elonga-
tion of the object may explain the activity in the area LO.
Therefore, we conclude that it is premature to conclude that
there is high-level processing of unseen stimuli in CFS.
Furthermore, because some of the neuroimaging and elec-
trophysiology studies did not evaluate conscious and noncon-
scious perception accurately, we cannot be sure that the ob-
served activities under the influence of CFS represents con-
scious or nonconscious processing. It is possible that some of
the observed activities were not strong enough to cause any
conscious and nonconscious perception. Therefore, in absence
of rigorous evaluation of conscious and nonconscious percep-
tion, neuroimaging and electrophysiology studies cannot pro-
vided conclusive evidence regarding the level of visual pro-
cessing under the influence of CFS. Therefore, we consider
the lack of conclusive evidence as evidence for the impair-
ment of nonconscious perception of high-level visual infor-
mation under the influence of CFS (also, see Moors, 2019).
Table 15 Experimental configuration and theawareness task in the CFS studies that evaluated theneural basis of conscious and nonconscious face and
facial expression recognition
Study Masks’type Masks’
TF
Masks’
SF
Colorful
masks?
Colorful
target?
Awareness task Is awareness assessed in the
main block?
Jiang & He (2006) Mondrians 10 Hz NR No* No No trial by trial assessment of
visibility**
No
Objective No
Sterzer et al. (2008)Mondrianlike
pattern
30 Hz NR Yes No CR Yes
Vizueta et al.
(2012)
Mondrians NR NR Yes* Yes Objective No
Jiang et al. (2009) Mondrians 20 Hz NR No No Objective No
Sterzer et al. (2009) Mondrians 30 Hz NR Yes No Objective Yes
Troiani et al. (2014) Checkerboard NR NR Yes* Yes Yes/No (report a
breakthrough)
Yes
Troiani & Schultz
(2013)
Mondrians 10 Hz NR Yes No Yes/No Yes
Lapate et al. (2016) Mondrians 10 Hz NR Yes Yes Yes/No (report a
breakthrough)
Yes
Objective No
Suzuki & Noguchi
(2013)
Mondrians 20 Hz NR Yes No Yes/No Main
Objective Control
Tsuchiya et al.
(2009)
Mondrians 10 Hz NR Yes No NA NA
Schlossmacher
et al. (2017)
Mondrians 20 Hz NR Yes No Yes/No (report a
breakthrough)
Main
Sakuraba et al.
(2013
Dynamic
Random Noise
10 Hz NR Yes* Yes* No Awareness Measure NA
Madipakkam et al.
(2015)
Mondrians 10 Hz NR No No CR Main
Yokoyama et al.
(2013)
Mondrians 20 Hz NR Yes No Yes/No Main
Baroni et al. (2017) Mondrians 15 Hz NR Yes No The PAS Main
*These studies used anaglyph glasses which means target and masks needed to be presented with colorful filters. **In this study, participants were asked
about their perceptual experiences during the run, after each scanning run
Psychon Bull Rev
Author's personal copy

The puzzle of the mechanism of suppression with CFS is
also a crucial question that should be discussed. As of now, we
found no studies that examined the mechanism of suppression
with CFS, even though we have some indirect evidence re-
garding this mechanism. Therefore, future studies must eval-
uate the mechanism of suppression with CFS. We believe that
future studies should accurately set and report the experimen-
tal configurations, which could have a tremendous impact on
the depth of suppression with CFS.
Finally, we wonder if results from CFS studies support major
theories of conscious perception. The only theories that have
been extensively examined using CFS is the two-stream theory
of consciousness, and both behavioral and neural studies of CFS
are not consistent with this theory. Other theories have not been
tested extensively with CFS, and we have no direct evidence
from CFS studies to support or disprove these theories. The trend
of results from CFS literature, however, is in line with predictions
of some of the major theories. That is, the disruption of high-level
information processing and survival of low-level information
processing under the influence of CFS is in line with some of
the major theories of consciousness, including the global neural
workspace theory (GNW; Dehaene & Naccache, 2001), the
recurrent/feedback model of visual awareness (Lamme, 2006),
the information integration theory of consciousness (IIT; Tononi,
2012), and the higher-order theory of consciousness (HOT; Lau
&Rosenthal,2011). CFS studies, however, provided no evi-
dence in favor of a specific theory.
Acknowledgements We would like to thank Drs. Fabian Soto, Bethany
Reeb-Sutherland, Jorge Riera Diaz, and Sang Wook Hong for their com-
ments on the earlier version of this manuscript.
Appendix A
Table 16 Important terms and their abbreviations
Abbreviation Term Abbreviation Term
bCFS Breaking continuous flash suppression BOLD Blood oxygenation level-dependent
CFS Continuous flash suppression FEA Facial expression aftereffect
FFA Fusiform face area GRT General recognition theory
IPS Intraparietal sulcus IT Inferotemporal cortex
LGN Lateral geniculate nucleus LO Lateral occipital cortex
LPP Late positive potential mAFC M-alternative forced choice
mIFC M-interval forced choice MVPA Multivariate pattern analysis
NCCs Neural correlates of consciousness PAS Perceptual Awareness Scale
PPA Parahippocampal place area ROIs Regions of Interest
SDT Signal detection theory SF Spatial frequency
STS Superior temporal sulcus SvA Curve Sensitivity versus awareness curve
TF Temporal frequency TMS Transcranial magnetic stimulation
V1 Primary visual cortex VPP Vertex positive potential
Psychon Bull Rev
Author's personal copy

References
Adams, W. J., Gray, K. L. H., Garner, M., & Graf, E. W. (2010). High-
level face adaptation without awareness. Psychological Science,
21(2), 205–210. https://doi.org/10.1177/0956797609359508
Adams, W. J., Gray, K. L. H., Garner, M., & Graf, E. W. (2011). On the
“special”status of emotional faces—Comment on Yang, Hong, and
Blake (2010). JournalofVision,11(3), 10. https://doi.org/10.1167/
11.3.10
Afraz, A., & Cavanagh, P. (2009). The gender-specific face aftereffect is
based in retinotopic not spatiotopic coordinates across several natu-
ral image transformations. Journal of Vision,9(10), 10–10. https://
doi.org/10.1167/9.10.10
Alais, D., & Melcher, D. (2007). Strength and coherence of binocular
rivalry depends on shared stimulus complexity. Vision Research,
47(2), 269–279. https://doi.org/10.1016/j.visres.2006.09.003
Allen, C. P. G., Sumner, P., & Chambers, C. D. (2014). The timing and
neuroanatomy of conscious vision as revealed by TMS-induced
blindsight. Journal of Cognitive Neuroscience,26(7), 1507–1518.
https://doi.org/10.1162/jocn_a_00557
Almeida, J., Mahon, B. Z., Nakayama, K., & Caramazza, A. (2008).
Unconscious processing dissociates along categorical lines.
Proceedings of the National Academy of Sciences of the United
States of America,105(39), 15214–15218. https://doi.org/10.1073/
pnas.0805867105
Almeida, J., Mahon, B. Z., & Caramazza, A. (2010). The role of the
dorsal visual processing stream in tool identification.
Psychological Science,21(6), 772–778. https://doi.org/10.1177/
0956797610371343
Almeida, J., Pajtas, P. E., Mahon, B. Z., Nakayama, K., & Caramazza, A.
(2013). Affect of the unconscious: Visually suppressed angry faces
modulate our decisions. Cognitive, Affective, & Behavioral
Neuroscience,13(1), 94–101. https://doi.org/10.3758/s13415-012-
0133-7
Almeida, J., Mahon, B. Z., Zapater-Raberov, V., Dziuba, A., Cabaço, T.,
Marques, J. F., & Caramazza, A. (2014). Grasping with the eyes:
The role of elongation in visual recognition of manipulable objects.
Cognitive, Affective, & Behavioral Neuroscience,14(1), 319–335.
https://doi.org/10.3758/s13415-013-0208-0
Amihai, I., Deouell, L., & Bentin, S. (2011). Conscious awareness is
necessary for processing race and gender information from faces.
Consciousness and Cognition,20(2), 269–279. https://doi.org/10.
1016/j.concog.2010.08.004
Ananyev, E., Penney, T. B., & Hsieh, P.-J. B. (2017). Separate require-
ments for detection and perceptual stability of motion in interocular
suppression. Scientific Reports,7(1), 7230. https://doi.org/10.1038/
s41598-017-07805-5
Andersen, L. M., Pedersen, M. N., Sandberg, K., & Overgaard, M.
(2016). Occipital MEG activity in the early time range (<300 ms)
predicts graded changes in perceptual consciousness. Cerebral
Cortex,26(6), 2677–2688. https://doi.org/10.1093/cercor/bhv108
Anderson, E., Siegel, E., White, D., & Barrett, L. F. (2012). Out of sight
but not out of mind: Unseen affective faces influence evaluations
and social impressions. Emotion,12(6), 1210–1221. https://doi.org/
10.1037/a0027514
Aru, J., Bachmann, T., Singer, W., & Melloni, L. (2012). Distilling the
neural correlates of consciousness. Neuroscience and Biobehavioral
Reviews,36(2), 737–746. https://doi.org/10.1016/j.neubiorev.2011.
12.003
Ashby, F. G., & Soto, F. A. (2015). Multidimensional signal detection
theory. In J. R. Busemeyer, Z. Wang, J. T. Townsend, & A. Eidels
(Eds.), Oxford Library of Psychology: The Oxford handbook of
computational and mathematical psychology (pp. 13–34). New
York, NY: Oxford University Press.
Azzopardi, P., & Cowey, A. (1997). Is blindsight like normal, near-
threshold vision? Proceedings of the National Academy of
Sciences,94(25), 14190–14194.
Azzopardi, P., & Cowey, A. (1998). Blindsight and visual awareness.
Consciousness and Cognition,7(3), 292–311. https://doi.org/10.
1006/ccog.1998.0358
Bachmann, T., & Francis, G. (2013). Visual masking: Studying percep-
tion, attention, and consciousness. San Diego, CA: Elsevier
Academic Press.
Bahrami, B., Lavie, N., & Rees, G. (2007). Attentional load modulates
responses of human primary visual cortex to invisible stimuli.
Current Biology,17(6), 509–513. https://doi.org/10.1016/j.cub.
2007.01.070
Bahrami, B., Carmel, D., Walsh, V., Rees, G., & Lavie, N. (2008).
Unconscious orientation processing depends on perceptual load.
Journal of Vision,8(3). https://doi.org/10.1167/8.3.12
Bahrami, B., Vetter, P., Spolaore, E., Pagano, S., Butterworth, B., &
Rees, G. (2010). Unconscious numerical priming despite interocular
suppression. Psychological Science,21(2), 224–233. https://doi.org/
10.1177/0956797609360664
Balsdon, T., & Azzopardi, P. (2015). Absolute and relative blindsight.
Consciousness and Cognition,32,79–91. https://doi.org/10.1016/j.
concog.2014.09.010
Bar, M., & Biederman, I. (1998). Subliminal visual priming.
Psychological Science,9(6), 464–468. https://doi.org/10.1111/
1467-9280.00086
Barbot, A., & Kouider, S. (2012). Longer is not better: Nonconscious
overstimulation reverses priming influences under interocular sup-
pression. Attention, Perception, & Psychophysics,74(1), 174–184.
https://doi.org/10.3758/s13414-011-0226-3
Baroni, F., van Kempen, J., Kawasaki, H., Kovach, C. K., Oya, H.,
Howard, M. A.,…Tsuchiya, N. (2017). Intracranial markers of con-
scious face perception in humans. NeuroImage,162,322–343.
https://doi.org/10.1016/j.neuroimage.2017.08.074
Boly, M., Massimini, M., Tsuchiya, N., Postle, B. R., Koch, C., &
Tononi, G. (2017). Are the neural correlates of consciousness in
the front or in the back of the cerebral cortex? Clinical and neuro-
imaging evidence. The Journal of Neuroscience,37(40), 9603–
9613. https://doi.org/10.1523/JNEUROSCI.3218-16.2017
Boyer, J. L., Harrison, S., & Ro, T. (2005). Unconscious processing of
orientation and color without primary visual cortex. Proceedings of
the National Academy of Sciences of the United States of America,
102(46), 16875–16879. https://doi.org/10.1073/pnas.0505332102
Carmel, D., Arcaro, M., Kastner, S., & Hasson, U. (2010). How to create
and use binocular rivalry. Journal of Visualized Experiments,45.
https://doi.org/10.3791/2030
Chalmers, D. J. (2000). What is a neural correlate of consciousness? In T.
Metzinger (Ed.), Neural correlates of consciousness: Empirical and
conceptual questions.https://doi.org/10.1093/acprof:oso/
9780195311105.001.0001
Charles, L., Van Opstal, F., Marti, S., & Dehaene, S. (2013). Distinct
brain mechanisms for conscious versus subliminal error detection.
NeuroImage,73,80–94. https://doi.org/10.1016/j.neuroimage.
2013.01.054
Chen, Y.-C., & Yeh, S.-L. (2012). Look into my eyes and I will see you:
Unconscious processing of human gaze. Consciousness and
Cognition,21(4), 1703–1710. https://doi.org/10.1016/j.concog.
2012.10.001
Chiesa, P. A., Liuzza, M. T., Acciarino, A., & Aglioti, S. M. (2015).
Subliminal perception of others’physical pain and pleasure.
Experimental Brain Research,233(8), 2373–2382. https://doi.org/
10.1007/s00221-015-4307-8
Costello, P., Jiang, Y., Baartman, B., McGlennen, K., & He, S. (2009).
Semantic and subword priming during binocular suppression.
Consciousness and Cognition,18(2), 375–382. https://doi.org/10.
1016/j.concog.2009.02.003
Psychon Bull Rev
Author's personal copy

Cowey, A. (2010). The blindsight saga. Experimental Brain Research,
200(1), 3–24. https://doi.org/10.1007/s00221-009-1914-2
Crick, F., & Koch, C. (2003). A framework for consciousness. Nature
Neuroscience,6(2), 119–126. https://doi.org/10.1038/nn0203-119
de Graaf, T. A., Hsieh, P.-J., & Sack, A. T. (2012). The “correlates”in
neural correlates of consciousness. Neuroscience and Biobehavioral
Reviews,36(1), 191–197. https://doi.org/10.1016/j.neubiorev.2011.
05.012
Dehaene, S., & Naccache, L. (2001). Towards a cognitive neuroscience
of consciousness: Basic evidence and a workspace framework.
Cognition,79(1/2), 1–37.
Doi, H., & Shinohara, K. (2016). Emotional faces influence numerosity
estimation without awareness. Cognitive Processing,17(4), 389–
397. https://doi.org/10.1007/s10339-016-0774-5
Drewes, J., Zhu, W., & Melcher, D. (2018). The edge of awareness: Mask
spatial density, but not color, determines optimal temporal frequen-
cy for continuous flash suppression. Journal of Vision,18(1), 12.
https://doi.org/10.1167/18.1.12
Dykstra, A. R., Cariani, P. A., & Gutschalk, A. (2017). A roadmap for the
study of conscious audition and its neural basis. Philosophical
Transactions of the Royal Society of London: Series B, Biological
Sciences,372(1714). https://doi.org/10.1098/rstb.2016.0103
Evans, S., & Azzopardi, P. (2007). Evaluation of a “bias-free”measure of
awareness. Spatial Vision,20(1/2), 61–77.
Faivre, N., Berthet, V., & Kouider, S. (2012). Nonconscious influences
from emotional faces: A comparison of visual crowding, masking,
and continuous flash suppression. Frontiers in Psychology,3,129.
https://doi.org/10.3389/fpsyg.2012.00129
Faivre, N., Berthet, V., & Kouider, S. (2014). Sustained invisibility
through crowding and continuous flash suppression: A comparative
review. Frontiers in Psychology,5.https://doi.org/10.3389/fpsyg.
2014.00475
Fang, F., & He, S. (2005). Cortical responses to invisible objects in the
human dorsal and ventral pathways. Nature Neuroscience,8(10),
1380–1385. https://doi.org/10.1038/nn1537
Fleming, S. M., & Lau, H. C. (2014). How to measure metacognition.
Frontiers in Human Neuroscience,8.https://doi.org/10.3389/
fnhum.2014.00443
Fogelson, S. V., Kohler, P. J., Miller, K. J., Granger, R., & Tse, P. U.
(2014). Unconscious neural processing differs with method used to
render stimuli invisible. Frontiers in Psychology,5,601.https://doi.
org/10.3389/fpsyg.2014.00601
Gayet, S., Van der Stigchel, S., & Paffen, C. L. E. (2014). Breaking
continuous flash suppression: Competing for consciousness on the
pre-semantic battlefield. Frontiers in Psychology,5,460.https://doi.
org/10.3389/fpsyg.2014.00460
Gayet, S., Paffen, C. L. E., Belopolsky, A. V., Theeuwes, J., & Van der
Stigchel, S. (2016). Visual input signaling threat gains preferential
access to awareness in a breaking continuous flash suppression par-
adigm. Cognition,149,77–83. https://doi.org/10.1016/j.cognition.
2016.01.009
Gelbard-Sagiv, H., Faivre, N., Mudrik, L., & Koch, C. (2016). Low-level
awareness accompanies “unconscious”high-level processing during
continuous flash suppression. Journal of Vision,16(1), 3. https://doi.
org/10.1167/16.1.3
Geng, H., Zhang, S., Li, Q., Tao, R., & Xu, S. (2012). Dissociations of
subliminal and supraliminal self-face from other-face processing:
Behavioral and ERP evidence. Neuropsychologia,50(12), 2933–
2942. https://doi.org/10.1016/j.neuropsychologia.2012.07.040
Gobbini, M. I., Gors, J. D., Halchenko, Y. O., Hughes, H. C., & Cipolli,
C. (2013a). Processing of invisible social cues. Consciousness and
Cognition,22(3), 765–770. https://doi.org/10.1016/j.concog.2013.
05.002
Gobbini, M. I., Gors, J. D., Halchenko, Y. O., Rogers, C., Guntupalli, J.
S., Hughes, H., & Cipolli, C. (2013b). Prioritized detection of
personally familiar faces. PLOS ONE,8(6), e66620. https://doi.
org/10.1371/journal.pone.0066620
Goldiamond, I. (1958). Indicators ofperception: I. Subliminal perception,
subception, unconscious perception: An analysis in terms of psycho-
physical indicator methodology. Psychological Bulletin,55(6),
373–411. https://doi.org/10.1037/h0046992
Gomes, N., Silva, S., Silva, C. F., & Soares, S. C. (2017). Beware the
serpent: The advantage of ecologically-relevant stimuli in accessing
visual awareness. Evolution and Human Behavior,38(2), 227–234.
https://doi.org/10.1016/j.evolhumbehav.2016.10.004
Gray, K. L. H., Adams, W. J., Hedger, N., Newton, K. E., & Garner, M.
(2013). Faces and awareness: Low-level, not emotional factors de-
termine perceptual dominance. Emotion,13(3), 537–544. https://
doi.org/10.1037/a0031403
Green, D. M., & Swets, J. A. (1966). Signal detection theory and
psychophysics. Oxford, England: John Wiley.
Han, S., & Alais, D. (2018). Strength of continuous flash suppression is
optimal when target and masker modulation rates are matched.
Journal of Vision,18(3), 3. https://doi.org/10.1167/18.3.3
Han, S., Lunghi, C., & Alais, D. (2016). The temporal frequency tuning
of continuous flash suppression reveals peak suppression at very
low frequencies. Scientific Reports,6, 35723. https://doi.org/10.
1038/srep35723
Han, S., Blake, R., & Alais, D. (2018). Slow and steady, not fast and
furious: Slow temporal modulation strengthens continuous flash
suppression. Consciousness and Cognition,58,10–19. https://doi.
org/10.1016/j.concog.2017.12.007
Hebart, M. N., & Hesselmann, G. (2012). What Visual Information Is
Processed in the Human Dorsal Stream? Journal of Neuroscience,
32(24), 8107–8109. https://doi.org/10.1523/JNEUROSCI.1462-12.
2012
Hedger, N.,Adams, W.J., & Garner, M. (2015a).Autonomic arousal and
attentional orienting to visual threat are predicted by awareness.
Journal of Experimental Psychology: Human Perception and
Performance,41(3), 798–806. https://doi.org/10.1037/xhp0000051
Hedger, N., Adams, W. J., & Garner, M. (2015b). Fearful faces have a
sensory advantage in the competition for awareness. Journal of
Experimental Psychology: Human Perception and Performance,
41(6), 1748–1757. https://doi.org/10.1037/xhp0000127
Heeks, F., & Azzopardi, P. (2015). Thresholds for detection and aware-
ness of masked facial stimuli. Consciousness and Cognition,32,68–
78. https://doi.org/10.1016/j.concog.2014.09.009
Hesselmann, G., & Malach, R. (2011). The link between fMRI-BOLD
activation and perceptual awareness is “stream-invariant”in the hu-
man visual system. Cerebral Cortex,21(12), 2829–2837. https://
doi.org/10.1093/cercor/bhr085
Hesselmann, G., Hebart, M., & Malach, R. (2011). Differential BOLD
activity associated with subjective and objective reports during
“blindsight”in normal observers. The Journal of Neuroscience,
31(36), 12936–12944. https://doi.org/10.1523/JNEUROSCI.1556-
11.2011
Hesselmann, G., Darcy, N., Sterzer, P., & Knops, A. (2015). Exploring
the boundary conditions of unconscious numerical priming effects
with continuous flash suppression. Consciousness and Cognition,
31,60–72. https://doi.org/10.1016/j.concog.2014.10.009
Hesselmann, G., Darcy, N., Ludwig, K., & Sterzer, P. (2016). Priming
in a shape task but not in a category task under continuous flash
suppression. Journal of Vision,16(3), 17. https://doi.org/10.1167/
16.3.17
Hesselmann, G., Darcy, N., Rothkirch, M., & Sterzer, P. (2018).
Investigating masked priming along the “vision-for-perception”
and “vision-for-action”dimensions of unconscious processing.
Journal of Experimental Psychology: General https://doi.org/10.
1037/xge0000420
Hong, S. W. (2015). Radial bias for orientation and direction of motion
modulates access to visual awareness during continuous flash
Psychon Bull Rev
Author's personal copy

suppression. Journal of Vision,15(1), 15.1.3. https://doi.org/10.
1167/15.1.3
Hong, S. W., & Blake, R. (2009). Interocular suppression differentially
affects achromatic and chromatic mechanisms. Attention,
Perception, & Psychophysics,71(2), 403–411. https://doi.org/10.
3758/APP.71.2.403
Hung, S.-M., Nieh, C.-H., & Hsieh, P.-J. (2016). Unconsciousprocessing
of facial attractiveness: Invisible attractive faces orient visual atten-
tion. Scientific Reports,6, 37117. https://doi.org/10.1038/srep37117
Izatt, G., Dubois, J., Faivre, N., & Koch, C. (2014). A direct comparison
of unconscious face processing under masking and interocular sup-
pression. Frontiers in Psychology,5.https://doi.org/10.3389/fpsyg.
2014.00659
Jachs, B., Blanco, M. J., Grantham-Hill, S., & Soto, D. (2015). On the
independence of visual awareness and metacognition: A signal de-
tection theoretic analysis. Journal of Experimental Psychology:
Human Perception and Performance,41(2), 269–276. https://doi.
org/10.1037/xhp0000026
Jiang, Y., & He, S. (2006). Cortical responses to invisible faces:
Dissociating subsystems for facial-information processing.
Current Biology,16(20), 2023–2029. https://doi.org/10.1016/j.
cub.2006.08.084
Jiang, Y., Costello, P., & He, S. (2007). Processing of invisible stimuli:
Advantage of upright faces and recognizable words in overcoming
interocular suppression. Psychological Science,18(4), 349–355.
https://doi.org/10.1111/j.1467-9280.2007.01902.x
Jiang, Y., Shannon, R. W., Vizueta, N., Bernat, E. M., Patrick, C. J., &
He, S. (2009). Dynamics of processing invisible faces in the brain:
Automatic neural encoding of facial expression information.
NeuroImage,44(3), 1171–1177. https://doi.org/10.1016/j.
neuroimage.2008.09.038
Kamitani, Y., & Tong, F. (2005). Decoding the visual and subjective
contents of the human brain. Nature Neuroscience,8(5), 679–685.
https://doi.org/10.1038/nn1444
Kanai, R., Tsuchiya, N., & Verstraten, F. A. J. (2006). The scope and
limits of top-down attention in unconscious visual processing.
Current Biology,16(23), 2332–2336. https://doi.org/10.1016/j.cub.
2006.10.001
Kaunitz, L. N., Fracasso, A., Skujevskis, M., & Melcher, D. (2014).
Waves of visibility: Probing the depth of inter-ocular suppression
with transient and sustained targets.Frontiers in Psychology,5,804.
https://doi.org/10.3389/fpsyg.2014.00804
Kimchi, R., Devyatko, D., & Sabary, S. (2018). Can perceptual grouping
unfold in the absence of awareness? Comparing grouping during
continuous flash suppression and sandwich masking.
Consciousness and Cognition,60,37–51. https://doi.org/10.1016/
j.concog.2018.02.009
Koch, C., Massimini, M., Boly, M., & Tononi, G. (2016). Neural corre-
lates of consciousness: Progress and problems. Nature Reviews
Neuroscience,17(5), 307–321. https://doi.org/10.1038/nrn.2016.22
Koivisto, M., & Grassini, S. (2018). Unconscious response priming dur-
ing continuous flash suppression. PLOS ONE,13(2). https://doi.org/
10.1371/journal.pone.0192201
Koivisto, M., & Rientamo, E. (2016). Unconscious vision spots the ani-
mal but not the dog: Masked priming of natural scenes.
Consciousness and Cognition,41,10–23. https://doi.org/10.1016/
j.concog.2016.01.008
Koriat, A. (2007). Metacognition and consciousness. In P. D. Zelazo, M.
Moscovitch, & E. Thompson (Eds.), The Cambridge handbook of
consciousness (pp. 289–325). https://doi.org/10.1017/
CBO9780511816789.012
Kring, A. M., Siegel, E. H., & Barrett, L. F. (2014). Unseen Affective
Faces Influence Person Perception Judgments in Schizophrenia.
Clinical Psychological Science,2(4), 443–454. https://doi.org/10.
1177/2167702614536161
Kunimoto, C., Miller, J., & Pashler, H. (2001). Confidence and accuracy
of near-threshold discrimination responses. Consciousness and
Cognition,10(3), 294–340. https://doi.org/10.1006/ccog.2000.0494
Lamme, V. A. F. (2006). Towards a true neural stance on consciousness.
Trends in Cognitive Sciences,10(11), 494–501. https://doi.org/10.
1016/j.tics.2006.09.001
Lapate, R. C., Rokers, B., Li, T., & Davidson, R. J. (2014). Nonconscious
emotional activation colors first impressions: A regulatory role for
conscious awareness. Psychological Science,25(2), 349–357.
https://doi.org/10.1177/0956797613503175
Lapate, R. C., Rokers, B., Tromp, D. P. M., Orfali, N. S., Oler, J. A.,
Doran, S. T., …Davidson, R. J. (2016). Awareness of emotional
stimuli determines the behavioral consequences of amygdala activa-
tion and amygdala-prefrontal connectivity. Scientific Reports,6,
25826. https://doi.org/10.1038/srep25826
Lau, H., & Rosenthal, D. (2011). Empirical support for higher-order
theories of conscious awareness. Trends in Cognitive Sciences,
15(8), 365–373. https://doi.org/10.1016/j.tics.2011.05.009
Lau, H. C. (2008). Are we studying consciousness yet? In Frontiers of
Consciousness: Chichele Lectures. Oxford, NY: Oxford
University Press.
Leopold, D. A., & Logothetis, N. K. (1996). Activity changes in early
visual cortex reflect monkeys’percepts during binocular rivalry.
Nature,379(6565), 549–553. https://doi.org/10.1038/379549a0
Lesmes, L. A., Lu, Z.-L., Baek, J., Tran, N., Dosher, B. A., & Albright,
T. D. (2015). Developing Bayesian adaptive methods for estimat-
ing sensitivity thresholds (d′) in yes-no and forced-choice tasks.
Frontiers in Psychology,6.https://doi.org/10.3389/fpsyg.2015.
01070
Lin, S.-Y., & Yeh, S.-L. (2016). Interocular grouping without awareness.
Vision Research,121,23–30. https://doi.org/10.1016/j.visres.2016.
01.004
Lin, Z., & He, S. (2009). Seeing the invisible: The scope and limits of
unconscious processing in binocular rivalry. Progress in
Neurobiology,87(4), 195–211. https://doi.org/10.1016/j.
pneurobio.2008.09.002
Lloyd, D. A., Abrahamyan, A., & Harris, J. A. (2013). Brain-stimulation
induced blindsight: Unconscious vision or response bias? PLOS
ONE,8(12), e82828. https://doi.org/10.1371/journal.pone.0082828
Luck, S. J. (2014). An introduction to the event-related potential
technique (2nd). Cambridge, MA: MIT Press.
Ludwig, K., & Hesselmann, G. (2015). Weighing the evidence for a
dorsal processing bias under continuous flash suppression.
Consciousness and Cognition,35, 251–259. https://doi.org/10.
1016/j.concog.2014.12.010
Ludwig,K.,Kathmann,N.,Sterzer,P.,&Hesselmann,G.(2015).
Investigating category- and shape-selective neural processing in
ventral and dorsal visual stream under interocular suppression.
Human Brain Mapping,36(1), 137–149. https://doi.org/10.1002/
hbm.22618
Ludwig,K.,Sterzer,P.,Kathmann,N.,&Hesselmann,G.(2016).
Differential modulation of visual object processing in dorsal and
ventral stream by stimulus visibility. Cortex,83,113–123. https://
doi.org/10.1016/j.cortex.2016.07.002
Mack, A. (2003). Inattentional blindness: Looking without seeing.
Current Directions in Psychological Science,12(5), 180–184.
https://doi.org/10.1111/1467-8721.01256
Macmillan, N. A., & Creelman, C. D. (2004). Detection theory: A user’s
guide. Mahwah, NJ: Erlbaum.
Madipakkam, A. R., Rothkirch, M., Guggenmos, M., Heinz, A., &
Sterzer, P. (2015). Gaze direction modulates the relation between
neural responses to faces and visual awareness. The Journal of
Neuroscience,35(39), 13287–13299. https://doi.org/10.1523/
JNEUROSCI.0815-15.2015
Maniscalco, B., & Lau, H. (2012). A signal detection theoretic approach
for estimating metacognitive sensitivity from confidence ratings.
Psychon Bull Rev
Author's personal copy

Consciousness and Cognition,21(1), 422–430. https://doi.org/10.
1016/j.concog.2011.09.021
Maniscalco, B., & Lau, H. (2014). Signal detection theory analysis of
Type 1 and Type 2 Data: Meta-d′, response-specific meta-d′,and the
unequal variance SDT model. In S. M. Fleming & C. D. Frith (Eds.),
The cognitive neuroscience of metacognition (pp. 25–66). New
York, NY: Springer-Verlag. https://doi.org/10.1007/978-3-642-
45190-4_3
Martens, S., & Wyble, B. (2010). The attentional blink: Past, present, and
future of a blind spot in perceptual awareness. Neuroscience and
Biobehavioral Reviews,34(6), 947–957. https://doi.org/10.1016/j.
neubiorev.2009.12.005
Mazzi, C., Mancini, F., & Savazzi, S. (2014). Can IPS reach visual
awareness without V1? Evidence from TMS in healthy subjects
and hemianopic patients. Neuropsychologia,64,134–144. https://
doi.org/10.1016/j.neuropsychologia.2014.09.026
Mazzi, C., Bagattini, C., & Savazzi, S. (2016). Blind-sight vs. degraded-
sight: Different measures tell a different story. Frontiers in
Psychology,7.https://doi.org/10.3389/fpsyg.2016.00901
Metcalfe, J., & Schwartz, B. L. (2016). The ghost in the machine. In J.
Dunlosky & S. K. Tauber (Eds.), The Oxford handbook of
metamemory. Retrieved from http://www.oxfordhandbooks.com/
view/10.1093/oxfordhb/9780199336746.001.0001/oxfordhb-
9780199336746-e-19
Milner, A. D., & Goodale, M. A. (2008). Two visual systems re-viewed.
Neuropsychologia,46(3), 774–785. https://doi.org/10.1016/j.
neuropsychologia.2007.10.005
Moors, P. (2019). What’s up with high-level processing during continu-
ous flash suppression?In G. Hesselmann (Ed.), Transitions between
consciousness and unconsciousness. New York, NY: Routledge.
Moors, P., Wagemans, J., & de-Wit, L. (2014). Moving stimuli are less
effectively masked using traditional continuous flash suppression
(CFS) compared with a moving Mondrian mask (MMM): A test
case for feature-selective suppression and retinotopic adaptation.
PLOS ONE,9(5), e98298. https://doi.org/10.1371/journal.pone.
0098298
Moors, P., Wagemans, J., & de-Wit, L. (2016). Faces in commonly ex-
perienced configurations enter awareness faster due to their curva-
ture relative to fixation. PeerJ,4,e1565.https://doi.org/10.7717/
peerj.1565
Moors, P., Hesselmann, G., Wagemans, J., & van Ee, R. (2017).
Continuous flash suppression: Stimulus fractionation rather than
integration. Trends in Cognitive Sciences,21(10), 719–721.
https://doi.org/10.1016/j.tics.2017.06.005
Moors, P., Gayet, S., Hedger, N., Stein, T., Sterzer, P., van Ee, R., …
Hesselmann, G. (2019). Three criteria for evaluating high-level
processing in continuous flash suppression. Trends in Cognitive
Sciences,23(4), 267–269. https://doi.org/10.1016/j.tics.2019.01.
008
Moradi, F., Koch, C., & Shimojo, S. (2005). Face adaptation depends on
seeing the face. Neuron,45(1), 169–175. https://doi.org/10.1016/j.
neuron.2004.12.018
Nakamura, K., & Kawabata, H. (2018). Preferential access to awareness
of attractive faces in a breaking continuous flash suppression para-
digm. Consciousness and Cognition,65,71–82. https://doi.org/10.
1016/j.concog.2018.07.010
Newell, B. R., & Shanks, D. R. (2014). Unconscious influences on deci-
sion making: A critical review. The Behavioral and Brain Sciences,
37(1), 1–19. https://doi.org/10.1017/S0140525X12003214
Ng, M., Boynton, G. M., & Fine, I. (2008). Face adaptation does not
improve performance on search or discrimination tasks. Journal of
Vision,8(1), 1.1-20. https://doi.org/10.1167/8.1.1
Odegaard, B., Knight, R. T., & Lau, H. (2017). Should a few null findings
falsify prefrontal theories of conscious perception? The Journal of
Neuroscience,37(40), 9593–9602. https://doi.org/10.1523/
JNEUROSCI.3217-16.2017
Overgaard, M., Rote, J., Mouridsen, K., & Ramsøy, T. Z. (2006). Is
conscious perception gradual or dichotomous? A comparison of
report methodologies during a visual task. Consciousness and
Cognition,15(4), 700–708. https://doi.org/10.1016/j.concog.2006.
04.002
Overgaard, M., Fehl, K., Mouridsen, K., Bergholt, B., & Cleeremans, A.
(2008). Seeing without seeing? Degraded conscious vision in a
blindsight patient. PLOS ONE,3(8). https://doi.org/10.1371/
journal.pone.0003028
Peremen, Z., & Lamy, D. (2014). Comparing unconscious processing
during continuous flash suppression and meta-contrast masking just
under the limen of consciousness. Frontiers in Psychology,5.
https://doi.org/10.3389/fpsyg.2014.00969
Persuh, M. (2018). Measuring perceptual consciousness. Frontiers in
Psychology,8.https://doi.org/10.3389/fpsyg.2017.02320
Peters, M. A. K., & Lau, H. (2015). Human observers have optimal
introspective access to perceptual processes even for visually
masked stimuli. ELife,4, e09651. https://doi.org/10.7554/eLife.
09651
Peters, M. A. K., Ro, T., & Lau, H. (2016). Who’s afraid of response
bias? Neuroscience of Consciousness,2016(1). https://doi.org/10.
1093/nc/niw001
Polonsky, A., Blake, R., Braun, J., & Heeger, D. J. (2000). Neuronal
activity in human primary visual cortex correlates with perception
during binocular rivalry. Nature Neuroscience,3(11), 1153. https://
doi.org/10.1038/80676
Pournaghdali, A., Schwartz, B. L., Hays, J. S., & Soto, F. (2020).
Sensitivity vs. awareness curve: A novel model-based analysis to
uncover the processes underlying nonconscious perception.
Manuscript under revision. 10.31234/osf.io/akem2
Raio,C.M.,Carmel,D.,Carrasco,M.,&Phelps,E.A.(2012).
Nonconscious fear is quickly acquired but swiftly forgotten.
Current Biology,22(12), R477–R479. https://doi.org/10.1016/j.
cub.2012.04.023
Ramsøy, T. Z., & Overgaard, M. (2004). Introspection and subliminal
perception. Phenomenology and the Cognitive Sciences,3(1), 1–23.
https://doi.org/10.1023/B:PHEN.0000041900.30172.e8
Reder, L. M. (1996). Implicit memory and metacognition. New York,
NY: Psychology Press.
Ro, T., Shelton, D., Lee, O. L., & Chang, E. (2004). Extrageniculate
mediation of unconscious vision in transcranial magnetic
stimulation-induced blindsight. Proceedings of the National
Academy of Sciences of the United States of America,101(26),
9933–9935. https://doi.org/10.1073/pnas.0403061101
Rosenthal, D. (2019). Consciousness and confidence.
Neuropsychologia,128, 255–265. https://doi.org/10.1016/j.
neuropsychologia.2018.01.018
Rothkirch, M., & Hesselmann, G. (2018). No evidence for dorsal-stream-
based priming under continuous flash suppression. Consciousness
and Cognition,64,84–94. https://doi.org/10.1016/j.concog.2018.
05.011
Sakuraba, S., Sakai, S., Yamanaka, M., Yokosawa, K., & Hirayama, K.
(2012). Does the human dorsal stream really process a category for
tools? The Journal of Neuroscience,32(11), 3949–3953. https://doi.
org/10.1523/JNEUROSCI.3973-11.2012
Sakuraba, S., Kobayashi, H., Sakai, S., & Yokosawa, K. (2013). Alpha-
band rhythm modulation under the condition of subliminal face
presentation: MEG study. Conference Proceedings: 35th Annual
International Conference of the IEEE Engineering in Medicine
and Biology Society,2013, 6909–6912. https://doi.org/10.1109/
EMBC.2013.6611146
Sandberg, K., & Overgaard, M. (2015). Using the Perceptual Awareness
Scale (PAS). In M. Overgaard (Ed.), Behavioral methods in con-
sciousness research. New York, NY: Oxford University Press.
https://doi.org/10.1093/acprof:oso/9780199688890.003.0011
Psychon Bull Rev
Author's personal copy

Sandberg,K.,Timmermans,B.,Overgaard,M.,&Cleeremans,A.
(2010). Measuring consciousness: Is one measure better than the
other? Consciousness and Cognition,19(4), 1069–1078. https://
doi.org/10.1016/j.concog.2009.12.013
Sandberg, K., Bibby, B. M., & Overgaard, M. (2013). Measuring and
testing awareness of emotional face expressions. Consciousness and
Cognition,22(3), 806–809. https://doi.org/10.1016/j.concog.2013.
04.015
Schlossmacher, I., Junghöfer, M., Straube, T., & Bruchmann, M. (2017).
No differential effects to facial expressions under continuous flash
suppression: An event-related potentials study. NeuroImage,163,
276–285. https://doi.org/10.1016/j.neuroimage.2017.09.034
Schmidt, T., & Vorberg, D. (2006). Criteria for unconscious cognition:
three types of dissociation. Perception & Psychophysics, 68(3),
489–504
Schmidt, T. (2015). Invisible Stimuli, Implicit Thresholds: Why
Invisibility Judgments Cannot be Interpreted in Isolation.
Advances in Cognitive Psychology,11(2), 31–41. https://doi.org/
10.5709/acp-0169-3
Sergent, C., & Dehaene, S. (2004). Is consciousness a gradual phenom-
enon? Evidence for an all-or-none bifurcation during the attentional
blink. Psychological Science,15(11), 720–728. https://doi.org/10.
1111/j.0956-7976.2004.00748.x
Shanks, D. R. (2017). Regressive research: The pitfalls of post hoc data
selection in the study of unconscious mental processes.
Psychonomic Bulletin & Review,24(3), 752–775. https://doi.org/
10.3758/s13423-016-1170-y
Shin, K., Stolte, M., & Chong, S. C. (2009). The effect of spatial attention
on invisible stimuli. Attention, Perception, & Psychophysics,71(7),
1507–1513. https://doi.org/10.3758/APP.71.7.1507
Song, C., & Yao, H. (2016). Unconscious processing of invisible visual
stimuli. Scientific Reports,6, 38917. https://doi.org/10.1038/
srep38917
Soto, F. A., Zheng, E., Fonseca, J., & Ashby, F. G. (2017). Testing
separability and independence of perceptual dimensions with gen-
eral recognition theory: A tutorial and new R package (grtools).
Frontiers in Psychology,8, 696. https://doi.org/10.3389/fpsyg.
2017.00696
Stein, T., & Sterzer, P. (2011). High-level face shape adaptation depends
on visual awareness: Evidence from continuous flash suppression.
Journal of Vision,11(8), 5. https://doi.org/10.1167/11.8.5
Stein, T., & Sterzer, P. (2012). Not just another face in the crowd:
Detecting emotional schematic faces during continuous flash sup-
pression. Emotion,12(5), 988–996. https://doi.org/10.1037/
a0026944
Stein, T., Hebart, M. N., & Sterzer, P. (2011a). Breaking continuous flash
suppression: A new measure of unconscious processing during
interocular suppression? Frontiers in Human Neuroscience,5,
167. https://doi.org/10.3389/fnhum.2011.00167
Stein, T., Peelen, M. V., & Sterzer, P. (2011b). Adults’awareness of faces
follows newborns’looking preferences. PLOS ONE,6(12), e29361.
https://doi.org/10.1371/journal.pone.0029361
Stein, T., Senju, A., Peelen, M. V., & Sterzer, P. (2011c). Eye contact
facilitates awareness of faces during interocular suppression.
Cognition,119(2), 307–311. https://doi.org/10.1016/j.cognition.
2011.01.008
Stein, T., Peelen, M. V., & Sterzer, P. (2012). Eye gaze adaptation under
interocular suppression. Journal of Vision,12(7), 1. https://doi.org/
10.1167/12.7.1
Stein, T., Seymour, K., Hebart, M. N., & Sterzer, P. (2014). Rapid fear
detection relies on high spatial frequencies. Psychological Science,
25(2), 566–574. https://doi.org/10.1177/0956797613512509
Stein, T., Reeder, R. R., & Peelen, M. V. (2016). Privileged access to
awareness for faces and objects of expertise. Journal of
Experimental Psychology: Human Perception and Performance,
42(6), 788–798. https://doi.org/10.1037/xhp0000188
Sterzer, P., Haynes, J.-D., & Rees, G. (2008). Fine-scale activity patterns
in high-level visual areas encode the category of invisible objects.
Journal of Vision,8(15), 10.1-12. https://doi.org/10.1167/8.15.10
Sterzer, P., Jalkanen, L., & Rees, G. (2009). Electromagnetic responses to
invisible face stimuli during binocular suppression. NeuroImage,
46(3), 803–808. https://doi.org/10.1016/j.neuroimage.2009.02.046
Sterzer, P., Hilgenfeldt, T., Freudenberg, P., Bermpohl, F., & Adli, M.
(2011). Access of emotional information to visual awareness in pa-
tients with major depressive disorder. Psychological Medicine,
41(8), 1615–1624. https://doi.org/10.1017/S0033291710002540
Sterzer, P., Stein, T., Ludwig, K., Rothkirch, M., & Hesselmann, G.
(2014). Neural processing of visual information under interocular
suppression: A critical review. Frontiers in Psychology,5, 453.
https://doi.org/10.3389/fpsyg.2014.00453
Suzuki, M., & Noguchi, Y. (2013). Reversal of the face-inversion effect
in N170 under unconscious visual processing. Neuropsychologia,
51(3), 400–409. https://doi.org/10.1016/j.neuropsychologia.2012.
11.021
Tagliabue, C. F., Mazzi, C., Bagattini, C., & Savazzi, S. (2016). Early
local activity in temporal areas reflects graded content of visual
perception. Frontiers in Psychology,7.https://doi.org/10.3389/
fpsyg.2016.00572
Tettamanti, M., Conca, F., Falini, A., & Perani, D. (2017). Unaware
processing of tools in the neural system for object-directed action
representation. The Journal of Neuroscience,37(44), 10712–10724.
https://doi.org/10.1523/JNEUROSCI.1061-17.2017
Timmermans, B., & Cleeremans, A. (2015). How can we measure aware-
ness? An overview of current methods. In M. Overgaard (Ed.),
Behavioral methods in consciousness research. Retrieved from
http://www.oxfordscholarship.com/view/10.1093/acprof:oso/
9780199688890.001.0001/acprof-9780199688890-chapter-3
Tong, F. (2003). Primary visual cortex and visual awareness. Nature
Reviews. Neuroscience,4(3), 219–229. https://doi.org/10.1038/
nrn1055
Tong, F., Nakayama, K., Vaughan, J. T., & Kanwisher, N. (1998).
Binocular rivalry and visual awareness in human extrastriate cortex.
Neuron,21(4), 753–759.
Tong, F., Meng, M., & Blake, R. (2006). Neural bases of binocular
rivalry. Trends in Cognitive Sciences,10(11), 502–511. https://doi.
org/10.1016/j.tics.2006.09.003
Tononi, G. (2012). Integrated information theory of consciousness: An
updated account. Archives Italiennes De Biologie,150(2/3), 56–90.
https://doi.org/10.4449/aib.v149i5.1388
Troiani, V., & Schultz, R. T. (2013). Amygdala, pulvinar, and inferior
parietal cortex contribute to early processing offaces without aware-
ness. Frontiers in Human Neuroscience,7,241.https://doi.org/10.
3389/fnhum.2013.00241
Troiani, V., Price, E. T., & Schultz, R. T. (2014). Unseen fearful faces
promote amygdala guidance of attention. Social Cognitive and
Affective Neuroscience,9(2), 133–140. https://doi.org/10.1093/
scan/nss116
Tsuchiya, N., & Koch, C. (2005). Continuous flash suppression reduces
negative afterimages. Nature Neuroscience,8(8), 1096–1101.
https://doi.org/10.1038/nn1500
Tsuchiya, N., Koch, C., Gilroy, L. A., & Blake, R. (2006). Depth of
interocular suppression associated with continuous flash suppres-
sion, flash suppression, and binocular rivalry. Journal of Vision,
6(10), 1068–1078. https://doi.org/10.1167/6.10.6
Tsuchiya, N., Moradi, F., Felsen, C., Yamazaki, M., & Adolphs, R.
(2009). Intact rapid detection of fearful faces in the absence of the
amygdala. Nature Neuroscience,12(10), 1224–1225. https://doi.
org/10.1038/nn.2380
Tsuchiya, N., Wilke, M., Frässle, S., & Lamme, V. A. F. (2015). No-
report paradigms: Extracting the true neural correlates of conscious-
ness. Trends in Cognitive Sciences,19(12), 757–770. https://doi.org/
10.1016/j.tics.2015.10.002
Psychon Bull Rev
Author's personal copy

Vetter, P., Badde, S., Phelps, E. A., & Carrasco, M. (2019). Emotional
faces guide the eyes in the absence of awareness. ELife,8.https://
doi.org/10.7554/eLife.43467
Vizueta, N., Patrick, C. J., Jiang, Y., Thomas, K. M., & He, S. (2012).
Dispositional fear, negative affectivity, and neuroimaging response
to visually suppressed emotional faces. NeuroImage,59(1), 761–
771. https://doi.org/10.1016/j.neuroimage.2011.07.015
Watanabe, M., Cheng, K., Murayama, Y., Ueno, K., Asamizuya, T.,
Tanaka, K., & Logothetis, N. (2011). Attention but not awareness
modulates the BOLD signal in the human V1 during binocular sup-
pression. Science,334(6057), 829–831. https://doi.org/10.1126/
science.1203161
Weiskrantz, L. (1986). Blindsight: Acase study and implications. Oxford,
England: Oxford University Press.
Weiskrantz, L. (1996). Blindsight revisited. Current Opinion in
Neurobiology,6(2), 215–220.
Weiskrantz, L., Barbur, J. L., & Sahraie, A. (1995). Parameters affecting
conscious versus unconscious visual discrimination with damage to
the visual cortex (V1). Proceedings of the National Academy of
Sciences of the United States of America,92(13), 6122–6126.
Witt, J. K., Taylor, J. E. T., Sugovic, M., & Wixted, J. T. (2015). Signal
detection measures cannot distinguish perceptual biases from re-
sponse biases. Perception,44(3), 289–300. https://doi.org/10.
1068/p7908
Xu, S., Zhang, S., & Geng, H. (2011). Gaze-induced joint attention per-
sists under high perceptual load and does not depend on awareness.
Vision Research,51(18), 2048–2056. https://doi.org/10.1016/j.
visres.2011.07.023
Yamashiro, H., Yamamoto, H., Mano, H., Umeda, M., Higuchi, T., &
Saiki, J. (2014). Activity in early visual areas predicts interindividual
differences in binocular rivalry dynamics. Journal of
Neurophysiology,111(6), 1190–1202. https://doi.org/10.1152/jn.
00509.2013
Yang, E., & Blake, R. (2012). Deconstructing continuous flash suppres-
sion. Journal of Vision,12(3), 8. https://doi.org/10.1167/12.3.8
Yang, E., Zald, D. H., & Blake, R. (2007). Fearful expressions gain
preferential access to awareness during continuous flash suppres-
sion. Emotion,7(4), 882–886. https://doi.org/10.1037/1528-3542.
7.4.882
Yang, E., Hong, S.-W., & Blake, R. (2010). Adaptation aftereffects to
facial expressions suppressed from visual awareness. Journal of
Vision,10(12), 24. https://doi.org/10.1167/10.12.24
Yang, E., Brascamp, J., Kang, M.-S., & Blake, R. (2014). On the use of
continuous flash suppression for the study of visual processing out-
side of awareness. Frontiers in Psychology,5,724.https://doi.org/
10.3389/fpsyg.2014.00724
Yang, Y.-H., & Yeh, S.-L. (2011). Accessing the meaning of invisible
words. Consciousness and Cognition,20(2), 223–233. https://doi.
org/10.1016/j.concog.2010.07.005
Ye, X., He, S., Hu, Y., Yu, Y. Q., & Wang, K. (2014). Interference
between conscious and unconscious facial expression information.
PLOS ONE,9(8), e105156. https://doi.org/10.1371/journal.pone.
0105156
Yokoyama, T., Noguchi, Y., & Kita, S. (2013). Unconscious processing
of direct gaze: Evidence from an ERP study. Neuropsychologia,
51(7), 1161–1168. https://doi.org/10.1016/j.neuropsychologia.
2013.04.002
Yoshida, M., & Isa, T. (2015). Signal detection analysis of blindsight in
monkeys. Scientific Reports,5, 10755. https://doi.org/10.1038/
srep10755
Yuval-Greenberg, S., & Heeger, D. J. (2013). Continuous flash suppres-
sion modulates cortical activity in early visual cortex. The Journal of
Neuroscience,33(23), 9635–9643. https://doi.org/10.1523/
JNEUROSCI.4612-12.2013
Zabelina, D. L., Guzman-Martinez, E., Ortega, L., Grabowecky, M.,
Suzuki, S., & Beeman, M. (2013). Suppressed semantic informa-
tion accelerates analytic problem solving. Psychonomic Bulletin &
Review,20(3), 581–585. https://doi.org/10.3758/s13423-012-
0364-1
Zhan, M., & de Gelder, B. (2018). Unconscious fearful body perception
enhances discrimination of conscious anger expressions under con-
tinuous flash suppression. Neuropsychologia,128,325–331. https://
doi.org/10.1016/j.neuropsychologia.2018.04.019
Zhan, M., Engelen, T., & de Gelder, B. (2018a). Influence of continuous
flash suppression mask frequency on stimulus visibility.
Neuropsychologia,128,65–72. https://doi.org/10.1016/j.
neuropsychologia.2018.05.012
Zhan, M., Goebel, R., & de Gelder, B. (2018b). Ventral and Dorsal
pathways relate differently to visual awareness of body postures
under continuous flash suppression. ENeuro,5(1). https://doi.org/
10.1523/ENEURO.0285-17.2017
Zhou, G., Zhang, L., Liu, J., Yang, J., & Qu,Z. (2010). Specificity of face
processing without awareness. Consciousness and Cognition,19(1),
408–412. https://doi.org/10.1016/j.concog.2009.12.009
Zhu, W., Drewes, J., & Melcher, D. (2016). Time for awareness: The
influence of temporal properties of the mask on continuous flash
suppression effectiveness. PLOS ONE,11(7). https://doi.org/10.
1371/journal.pone.0159206
Publisher’snoteSpringer Nature remains neutral with regard to jurisdic-
tional claims in published maps and institutional affiliations.
Psychon Bull Rev
Author's personal copy