‘‘Change of Mind’’ within and between nonconscious (masked) and conscious (unmasked) visual processing

https://doi.org/10.1016/j.concog.2007.08.001Get rights and content

Abstract

We examined the updating of decisions made during visuo-motor processing when two sequentially presented stimuli, Prime1 and Prime2, primed discriminative responses to a following probe. In Experiment 1, the visibility of the two primes was suppressed or left intact by varying the stimulus onset asynchrony (SOA) of the stimuli immediately following them. In Experiment 2, the visibility of Prime2 was suppressed or left intact by varying its spatial separation from the following probe. We found (1) that Prime2 dominated the effects of Prime1; (2) that Prime2’s updating was stronger when the Prime2–probe SOA was 200 as compared to 53 ms; and (3) that Prime2’s updating was weaker when the Prime2–probe spatial separation was 58 as compared to 0 minarc. We conclude that these effects are due to an interaction of spatial attention and the state (conscious vs. nonconscious) of processing of Prime2.

Introduction

Priming effects produced by one visual stimulus on the processing of a subsequent stimulus are a common way of studying visual cognition (Bar and Biederman, 1998, Greenwald et al., 1995, Kinoshita and Lupker, 2003). Of particular relevance is a variety of masked priming in which the second, probe stimulus acts as a metacontrast mask to suppress the visibility of the first, priming stimuli. Metacontrast masking prevails when the visibility of a brief stimulus is suppressed by a spatially surrounding and brief second stimulus and is optimal when the onset asynchrony (SOA) between the two stimuli ranges between 30 and 100 ms (Breitmeyer & Ogmen, 2006). This type of masked priming technique, introduced by Neumann and colleagues (Ansorge and Neumann, 2005, Klotz and Neumann, 1999, Klotz and Wolff, 1995, Neumann and Klotz, 1994, Scharlau and Neumann, 2003) to investigate aspects of direct (i.e., nonconscious) parameter specification (DPS) of discriminative motor responses to the probe stimulus, has also been used subsequently to study the types and levels of nonconscious visual processing (Breitmeyer et al., 2004a, Breitmeyer et al., 2005, Breitmeyer et al., 2004b, Ogmen et al., 2003). In these priming studies, one of two probes, e.g., a square or a diamond, is preceded by one of two primes, e.g., again a square or a diamond, and the observer is asked to respond as rapidly and accurately as possible by pressing one of two response keys as to which probe was presented. According to the theory of DPS the nonconsciously processed features of the masked stimuli can prime discriminative responses to a following probe stimulus. The priming stimuli can be either congruent (e.g., a square prime) or incongruent (e.g., a diamond prime) to the consciously perceived following probe (e.g., a square). In case of congruence, the discriminative reaction time (RT) to the probe is fast, since the appropriate response has been correctly primed; in case of incongruence, the RT to the probe is slower, since now the priming of the incorrect response must be overridden before generating the correct alternative response to the probe.

While some studies have compared effects of visible, consciously processed primes to those of invisible, nonconsciously processed primes (Scharlau & Neumann, 2003), little is known of how the interaction either of two visible primes, of two invisible primes, or of one visible and one invisible prime determines the response to the probe stimulus. In the following experiments we use two sequentially presented primes, Prime1 and Prime2, followed in turn by a probe stimulus to explore the above interactive effects of the two primes on responses to the probe. The two primes and the probe are designed so that (a) Prime2 can, under appropriately chosen conditions, serve as a metacontrast mask of Prime1 and (b) the probe in turn can, again under appropriately chosen conditions, serve as a metacontrast mask of Prime2.

Specifically, we compare the effects of the two successive primes on choice RT to the probe when either a masked, invisible Prime1 or an unmasked, visible Prime1 is followed either by an invisible, masked Prime2 or else by a visible, unmasked Prime2. Based on the prior work on masked priming, we reasonably expect the fastest probe RTs when both Prime1 and Prime2 are congruent with the probe and slower RTs when one of them is incongruent. However, several additional questions emerge. Question 1 asks how the effects of the two primes relate to each other. Several possible outcomes exist, and several specific hypotheses can express the relationship.

H1: The two primes contribute equally to the total priming effect, and thus equal weighting is given to Prime2’s updating and Prime1’s predating the response choice to the probe. Here one would expect the effects of the two primes to additively average. This will be our default hypothesis. According to it, the estimate of the expected priming effect of either Prime1 or Prime2 should be the average of the priming effects obtained for Prime1 and Prime2.

H2: Prime1 enjoys a strong “priority effect” while Prime2 enjoys a weak “recency effect”. Here Prime1’s relation to the probe (congruency or incongruency) should dominate the priming effect, thus limiting Prime2’s ability to update probe-response decisions. This is not an unreasonable hypothesis since prior work, using a single prime (Schmidt, 2002, Vorberg et al., 2003), has shown that the priming effect, defined here as the difference between RTs to an incongruent prime and RTs to a congruent prime, increases as the SOA between the prime and the probe increases. Since in the experiments to be reported here the SOA between Prime1 and the probe is always longer than that between Prime2 and the probe, one might expect Prime1 to yield the larger priming effect. Specifically, here the effect of Prime1 should be larger than the average of the Prime1 and Prime2 effects, and the effect of Prime2 should correspondingly be smaller than the average.

H3: In contrast, Prime1 enjoys a relatively weak priority effect while Prime2 enjoys a strong recency effect. Consider, for instance, the vision-for-action system (Milner & Goodale, 1995). Because it must be able to adjust to the rapidly changing stimulus-response contingencies in the visual environment, its control of motor activity, informed by very brief “on-line” visual processing, must also be rapidly updatable by novel visual input. The information provided by Prime2 therefore would dominate the priming effect on the probe so that it is able to update probe-response decisions. Here, in contrast to the prediction of H2, the effect of Prime1 should be smaller than the average of the Prime1 and Prime2 effects, and the effect of Prime2 should correspondingly be larger than the average.

Question 2 is if and to what extent the pre- or updating effects of a masked, invisible prime are different from its effects of an unmasked and visible prime. In other words, regardless of whether Prime1’s effect dominates Prime2’s or vice versa, does such dominance depend on prime visibility? Such dependencies would shed light on differences between nonconscious and conscious visual processing of primes.

Section snippets

Experiment 1: Varying prime visibility by varying SOA

In this experiment we vary the visibility1 of a prime by

Experiment 2: Varying prime visibility by varying spatial separation

In the prior experiment, the visibility of a prime was manipulated by varying the SOA separating it from its immediately following stimulus. Hence, the effects of prime visibility were confounded with SOA. In the current experiment we set the Prime1–Prime2 SOA and the Prime2–probe SOA to the same value of 53 ms. Given the stimuli used in the prior experiment, both Prime1 and Prime2 would be rendered nearly invisible. However, it is known that the strength of metacontrast suppression decreases as

General discussion

Several results of the two experiments are noteworthy. First, they replicate numerous prior findings showing that RTs to a probe stimulus are higher when a prime, regardless of whether it is invisible or (at least partly) visible (Schmidt, 2002, Vorberg et al., 2003), is incongruent than when it is congruent. Moreover, similar to previous findings obtained with successively masked primes (Jaśkowski, Skalska, & Verleger, 2002), the present findings also show that when more than one prime are

Acknowledgments

I thank Rashmi Sundararajan for assistance in the data collection. This research was supported by NSF Grant BCS-0114533.

References (25)

  • B.G. Breitmeyer et al.

    Unconscious priming by color and form: Different processes and levels

    Consciousness and Cognition

    (2004)
  • H. Ogmen et al.

    The what and where in visual masking

    Vision Research

    (2003)
  • U. Ansorge et al.

    Intentions determine the effect of nonconsciously registered visual information: Evidence for direct parameter specification in the metacontrast dissociation

    Journal of Experimental Psychology: Human Perception and Performance

    (2005)
  • M. Bar et al.

    Subliminal visual priming

    Psychological Science

    (1998)
  • B.G. Breitmeyer et al.

    Visual masking: Time slices through conscious and unconscious vision

    (2006)
  • B.G. Breitmeyer et al.

    Unconscious and conscious priming by forms and their parts

    Visual Cognition

    (2005)
  • B.G. Breitmeyer et al.

    Unconscious priming with chromatic stimuli occurs at stimulus- not percept-dependent levels of visual processing

    Psychological Science

    (2004)
  • M. Eimer et al.

    Effects of masked stimuli on motor activation: Behavioral and electrophysiological evidence

    Journal of Experimental Psychology: Human Perception and Performance

    (1998)
  • M. Eimer et al.

    Links between conscious awareness and response inhibition: Evidence from masked priming

    Psychonomic Bulletin &Review

    (2002)
  • C.W. Erisksen et al.

    Visual attention within and around the field of focal attention: A zoom lens model

    Perception & Psychophysics

    (1986)
  • A.G. Greenwald et al.

    Activation by marginally perceptible (“subliminal”) stimuli: Dissociation of unconscious from conscious cognition

    Journal of Experimental Psychology: General

    (1995)
  • P. Jaśkowski

    The effect of nonmasking distractors on the priming of motor responses

    Journal of Experimental Psychology: Human Perception and Performance

    (2007)
  • Cited by (4)

    • Individually different weighting of multiple processes underlies effects of metacontrast masking

      2016, Consciousness and Cognition
      Citation Excerpt :

      Metacontrast masking occurs when a masking stimulus follows a target stimulus and the contours of the mask are in close contiguity to the contours of the target stimulus. A broad range of stimuli with specific geometric configurations has been employed to produce metacontrast masking including flanking bars (e.g., Ramachandran & Cobb, 1995), circular shape displays (e.g., Cohen, van Gaal, Ridderinkhof, & Lamme, 2009; Fehrer & Raab, 1962; Schmidt, 2002), squares and diamonds (e.g., Ansorge, Becker, & Breitmeyer, 2009; Ansorge, Breitmeyer, & Becker, 2007; Breitmeyer & Hanif, 2008; Breitmeyer, Ogmen, & Chen, 2004; Lau & Passingham, 2007; Mattler, 2003; Neumann & Klotz, 1994; Tapia, Breitmeyer, & Shooner, 2010) or arrow stimuli (e.g., Kiesel et al., 2006; Mattler & Palmer, 2012; Tapia & Breitmeyer, 2011; Van Gaal, Scholte, Lamme, Fahrenfort, & Ridderinkhof, 2011; Vorberg et al., 2003; Wenke, Fleming, & Haggard, 2010). Across all displays, metacontrast is characterized by the phenomenon that the visibility of the target is a function of stimulus onset asynchrony (SOA) between target and mask.

    • Visual Masking: Studying Perception, Attention, and Consciousness

      2013, Visual Masking: Studying Perception, Attention, and Consciousness
    View full text