Priority access to awareness? A new 'CFS tracking' paradigm reveals no priority for fearful faces or a face inversion effect

Abstract

When the eyes view separate and incompatible images, the brain suppresses one image-removing it from visual awareness. A popular paradigm for doing this is continuous flash suppression (CFS). One eye views a static 'target', the other is presented with a complex dynamic stimulus which very effectively suppresses the target. Measuring the time needed for the suppressed target to break suppression as it slowly increases in contrast (bCFS) has been widely used to investigate unconscious processing and the results have generated controversy regarding the scope of visual processing without awareness. In particular, upright faces and fearful faces have been claimed to have priority access to awareness. Here, we address this controversy with a new 'CFS tracking' paradigm (tCFS) in which the suppressed monocular target steadily increases in contrast until breaking into awareness (as in bCFS) after which it decreases until it becomes suppressed again (reCFS), with this cycle continuing for many reversals. Unlike bCFS, tCFS provides measures of breakthrough thresholds as well as suppression thresholds, and the difference between breakthrough and suppression thresholds defines the important metric of 'suppression depth'. The suppression depth results over two experiments are consistent in showing no face inversion effects (i.e. no priority for upright faces relative to inverted) and no effect of emotion (no priority for fearful faces relative to happy or neutral). Given this consistent non-selectivity, we conclude that CFS elicits a strong suppression in early visual cortex at a level preceding face processing.

Keywords: CFS; awareness; early suppression; face inversion; face priority; non-selective suppression.

Figure 1

Top: In the tCFS paradigm, the target continually changes contrast and participants indicate when their perceptual state changes with a mouse click, which then reverses the direction of contrast change. The target starts visible at high contrast and declines until suppression is indicated, then it rises in contrast until breakthrough into awareness, and then it declines again, and so on. This provides a quick and efficient measure of bCFS threshold, but also of reCFS threshold and thus suppression depth (the difference between bCFS and reCFS thresholds). Middle: Both experiments used Mondrian masks made of achromatic circles with relatively low RMS contrast (to allow targets to breakthrough before reaching maximum contrast). In Experiment 2, the Mondrian mask was compared with a mask made of circular face images. All target contrast change was linear on a log scale. Bottom: All targets were faces expressing one of three emotions and were presented either upright or inverted.

Figure 2

(a) Group mean contrast thresholds for breakthrough (bCFS: dark colours) and suppression (reCFS: light colours) with error bars showing ±1 standard error of the mean (SEM). Apart from an obvious dominance/suppression main effect, the main effects for facial expression and face orientation were not significant. ‘Up’ refers to the upright face condition, ‘Inv’ to the inverted condition. The second y-axis shows contrast in decibels converted back to linear values and shows bCFS thresholds are consistently higher than reCFS thresholds by approximately a factor of five. (b) Group mean suppression depths (±1 SEM), calculated from the thresholds in panel a. Suppression depth is defined as bCFS threshold minus reCFS threshold (in decibels). Mean suppression depth overall was 15.47 dB, several times greater than suppression depths typically obtained in binocular rivalry.

Figure 3

(a) Group mean contrast thresholds for breakthrough (bCFS: dark colours) and suppression (reCFS: light colours) with error bars showing ±1 standard error of the mean (SEM). The second y-axis shows contrast in decibels converted back to linear values. (b) Group mean suppression depths (±1 SEM) in decibels, calculated from the thresholds in panel a (suppression depth = bCFS threshold minus reCFS threshold). The face targets were more strongly suppressed by the face masker than the Mondrian masker, with no effect of the target face’s orientation or target face emotion.