Color modulates the timing of conscious perception via bottom-up and top-down attention

Color can direct visual attention to specific locations through bottom-up and top-down 1 mechanisms. Using Continuous Flash Suppression (CFS) as way to investigate the factors that 2 gate access to consciousness, the current study investigated whether color also directly affected 3 the timing of conscious perception. Low or high spatial frequency (SF) gratings with different 4 orientations were shown as targets to the non-dominant eye of human participants. CFS patterns 5 were presented at a rate of 10Hz to the dominant eye to delay conscious perception of the targets, 6 and participants had the task to report the target’s orientation as soon as they could see it. With 7 low-SF targets, two types of color-based effects became evident. First, when the targets and the 8 CFS patterns had different colors, the targets entered consciousness faster than in trials where the 9 targets and CFS patterns had the same color. Second, when participants searched for a specific 10 target color, targets that matched these search settings entered consciousness faster compared 11 to conditions where the target color was irrelevant and could vary from trial to trial. Thus, the 12 current study demonstrates that color is a central feature of human perception and leads to faster 13 conscious perception of visual stimuli through bottom-up and top-down attentional mechanisms. 14


Introduction
1.1. Background 18 Color vision allows to distinguish different wavelengths within the visible light 19 spectrum and is central to the human visual system. Color vision has probably evolved 20 because the ability to discriminate colors provided critical survival benefits in foraging 21 [1] and social [2] situations. Color also is a central feature for guiding selective visual 22 attention [7,8]. Local color contrasts increase the bottom-up saliency of an object [3,5], 23 which in turn directs spatial attention, resulting in prioritized neurocognitive processing 24 of the attended information [4]. In everyday environments, local color contrasts are 25 usually associated with interesting and often relevant objects which can grab attention 26 in a bottom-up manner, such as red light or a stop sign at an intersection [4,5]. However, 27 color can also be used for directing attention in a top-down way, such as when looking for 28 friends in a crowd of people and expecting them to wear clothes of a particular color [7,8]. 29 While it is well established that color is a powerful feature for guiding selective visual 30 attention, it is not entirely clear through which mechanisms color affects how quickly 31 an object is consciously perceived [9]. A traditional view of the relationship between 32 attention and consciousness posits that attention acts as a 'gatekeeper for consciousness,' 33 i.e., that the selection of a stimulus by attention precedes conscious perception. However, 34 attention and consciousness are distinct processes that are not always aligned [11], which 35 is why top-down attention does not necessarily entail conscious processing [9,10]. 36 To investigate the factors involved in conscious perception and to understand which 37 types of processing occur unconsciously, a range of perceptual suppression techniques 38 has been developed to render stimuli temporarily invisible [12][13][14]. Continuous Flash 39 version of BR in which one is presented with a high-contrast pattern stimulus that is 48 updated over time, often at a frequency of 10 Hz [17,20]. The other eye is presented with 49 a relatively static and usually smaller target stimulus. The steep difference in stimulus 50 strength between these two eyes' inputs ensures that the CFS pattern is always perceived 51 at the beginning of each experimental trial [19], and it usually takes several seconds 52 until the initially invisible target stimulus breaks through the suppression and enters 53 consciousness [9,16]. 54 1.2. The present study 55 The current study used CFS to investigate whether colors and color contrasts 56 can accelerate conscious perception of a relevant target stimulus. The present study 57 used a "breaking CFS" procedure, in which the breakthrough time measures the time it 58 takes for an initially invisible stimulus to break CFS and become consciously accessible. 59 Specifically, the present study investigated how breakthrough time is influenced by 60 the color contrast between the relevant target stimuli and the CFS pattern on the one 61 hand, and the participants' attentional tuning for a specific color on the other hand. To 62 that end, the current study used oriented gratings as targets and asked participants 63 to report the orientation as soon as they could consciously perceive it. The relevant 64 targets were rendered in of two possible colors (cyan or yellow). Typical CFS pattern 65 stimuli were used to suppress conscious perception of the targets. The (bottom-up) color 66 contrast between the targets and the CFS patterns was manipulated by either using the 67 same color for the target and the CFS pattern (resulting in a low color contrast), or the 68 different color for the target and the CFS pattern (resulting in a high color contrast). To 69 investigate whether color can also have top-down effects on conscious perception [9], 70 the present study also manipulated the participants' task. In one block of experimental 71 trials, participants performed a singleton search task, where color was irrelevant. Here, 72 only one grating was presented in each trial, and the color of this target grating could 73 change randomly from trial to trial. The time it took for the targets to break suppression 74 was compared to another block of experimental trials, in which participants performed a 75 color search task. In the color search task, participants were always presented with two 76 gratings of different colors, and needed to report the grating with a specific task-relevant 77 color.

78
The hypotheses for the different conditions in the present study were the following.

79
First, suppose color leads to faster conscious perception via bottom-up feature salience.

80
In that case, the breakthrough time should be shorter in trials in which the color of 81 the target and the CFS pattern color were different (i.e., where the color contrast was  To verify whether any color-related effects are indeed related to faster conscious 98 perception of a stimulus rather than merely attentional processing after the target entered 99 consciousness, the present study also manipulated the effectiveness of the CFS method. 100 Previous research demonstrated that CFS could strongly and robustly suppress gratings 101 with a relatively low spatial frequency (SF) [22,23]. In contrast, gratings with a higher 102 spatial frequency usually remain relatively unperturbed by CFS. With the low-spatial 103 frequency (LSF) gratings, CFS can thus be used to test whether factors related to color 104 can reliably accelerate or delay conscious perception of a stimulus. Conversely, the 105 perception of high spatial frequency (HSF) gratings is barely affected by CFS. Therefore,      minimal, whereas strong suppression effects were expected for the low SF of 2 cpd [22].

154
The maximum contrast of the targets was set to 30% to ensure robust suppression effects.

155
Conscious perception of the targets was delayed by presenting a high-contrast 156 CFS pattern to the participants' dominant eye, which was updated at a rate of 10 Hz.  Trial procedure of the present study. The relevant target stimuli were left-or rightoriented Gabor gratings presented above or below the fixation point. The targets were delivered to the non-dominant eye using a mirror stereoscope. Conscious perception of the targets was interocularly suppressed by presenting a random sequence of CFS patterns with a 10Hz update frequency to the dominant eye. The targets had either the same or different color as the CFS patterns. The color of the CFS patterns varied randomly between cyan and yellow from trial to trial. In each trial, the participants' task was to report the target's orientation using key presses as soon as they could perceive it. As soon as the participants responded, the trial ended. The target grating's contrast was ramped up in the first phase of the trial, while the contrast of the CFS patterns decreased over time in the second phase of the trial. The trials had a maximum duration of 10 seconds but ended as soon as the participants gave a response. Left: in singleton search condition, color was irrelevant, and only one grating was shown in each trial, which varied randomly between cyan and yellow from one trial to the next. Right: In color search, participants were explicitly instructed to report the orientation of gratings with a specific color. This target color remained constant throughout the whole experimental block. Two gratings were presented in each trial of the color search block, but only one of these gratings had the relevant color.

194
The participants' task required to press the left "control" key using the left index

232
The full dataset encompassed 30,720 trials (768 trials for each of the 40 participants).  Figure 2. The statistical analysis was conducted using R [25].

Interaction effects 257
The analysis revealed a significant interaction of Target SF × Target/CFS color, F(1,  suppressed from conscious perception, which was not the case for high-SF targets [22,23].

286
Color accelerated conscious perception of the low-SF targets through two mecha- confirms that color has a substantial top-down effect on conscious perception [9]: When 298 a specific color is instantiated as an attentional template, a grating that matches this 299 template breaks into consciousness more quickly.   [27]. represent a good signal-to-noise ratio in terms of neural processing of the sensory signals.

337
In contrast, with low contrast, the resulting signal-to-noise ratio would be poor.

338
The current study draws a parallel between the mechanisms responsible for (bottom-339 up) attentional selection, where color contrasts also play a central role [3,4], and the   Interestingly, previous research using the CFS paradigm also showed that expectations 452 about the appearance of more "high-level" stimulus categories could also shorten the 453 time that stimuli require to break into consciousness [31,32]. In these experiments,