Loads of unconscious processing: The role of perceptual load in processing unattended stimuli during inattentional blindness

Inattentional blindness describes the failure to detect an unexpected but clearly visible object when our attention is engaged elsewhere. While the factors that determine the occurrence of inattentional blindness are already well understood, there is still a lot to learn about whether and how we process unexpected objects that go unnoticed. Only recently it was shown that although not consciously aware, characteristics of these stimuli can interfere with a primary task: Classification of to-be-attended stimuli was slower when the content of the task-irrelevant, undetected stimulus contradicted that of the attended, to-be-judged stimuli. According to Lavie’s perceptual load model, irrelevant stimuli are likely to reach awareness under conditions of low perceptual load, while they remain undetected under high load, as attentional resources are restricted to the content of focused attention. In the present study, we investigated the applicability of Lavie’s predictions for the processing of stimuli that remain unconscious due to inattentional blindness. In two experiments, we replicated that unconsciously processed stimuli can interfere with intended responses. Also, our manipulation of perceptual load did have an effect on primary task performance. However, against our hypothesis, these effects did not interact with each other. Thus, our results suggest that high perceptual load cannot prevent task-irrelevant stimuli that remain undetected from being processed to an extent that enables them to affect performance in a primary task. Surprisingly often we are unaware of clearly visible stimuli in our direct view simply because they appear unexpectedly and outside our attentional focus, a phenomenon called inattentional blindness (Mack & Rock, 1998). Inattentional blindness paradigms are thought to be especially rigorous measures of attentional capture as critical stimuli appear completely unexpectedly, and thus, voluntary attention cannot be directed towards its detection (e.g., New & German, 2015). Therefore, all factors influencing the processing of the critical stimulus and its crossing of the threshold of awareness rely on involuntary distribution of attention. By now, a lot of research has been conducted to unravel the factors that determine whether or not inattentional blindness occurs — namely, whether or not an unexpected object crosses the threshold of awareness (i.e., was reported by the participants; Kreitz, Furley, Memmert, & Simons, 2016; Mack, Pappas, Silverman, & Gay, 2002; Most et al., 2001; Simons & Jensen, 2009). This binary view on awareness is a general pattern in inattentional blindness research, while much less is known about the fate of those stimuli that remain undetected due to inattentional blindness. One theoretical framework that distinguishes different types of processing and provides distinguishable predictions for those types is the global neuronal workspace hypothesis (Dehaene, Changeux, Naccache, Sackur, & Sergent, 2006). Among other things, this framework predicts substantial processing of stimuli that have sufficient bottom-up strength but remain undetected due to a lack of attentional amplification (preconscious processingFootnote1), as is the case in inattentional blindness paradigms. And, indeed, the existing body of research on that topic indicates that there can be perceptual processing of unexpected objects even if they are not consciously noticed and therefore cannot be reported. In their seminal book, Mack and Rock (1998) report experiments in which they investigated whether features of an undetected object could prime following responses. They found evidence for perceptual processing in the absence of awareness; in a stem completion task that immediately followed an inattentional blindness task, participants were significantly more likely to complete the stem in accordance with the word that was unexpectedly presented in the inattentional blindness task. Other research has shown that behavioural responses can be influenced by grouping processes that occur completely outside of our awareness (Lamy, Segal, & Ruderman, 2006; Moore & Egeth, 1997; Wood & Simons, 2018) and that a large illusion-inducing rectangle, although never consciously noticed, still affected the reported position of a target (Lathrop, Bridgeman, & Tseng, 2011). Jiang and Leung (2005) provided evidence that latent learning of repeated information occurred without attention and facilitated performance in a later performed task. Also, imaging methods were able to demonstrate substantial neuronal marks of stimuli that remain beneath the threshold of awareness due to inattentional blindness (Pitts, Martínez, & Hillyard, 2012): Amplitudes of early event-related potentials (ERPs) that were recorded over the occipital pole differentiated between visually presented shapes and random arrays regardless of whether or not participants were aware of the shapes. Finally, there are even hints in inattentional blindness research indicating that the processing of objects that remain undetected goes beyond mere perceptual processing: A recent study provides direct evidence for the conclusion that the semantics of the undetected stimuli can be processed preconsciously; in two experiments, reaction times for target stimuli in a primary task were significantly slower when the semantic content of an undetected stimulus contradicted that of the attended, to-be-judged stimulus (Schnuerch, Kreitz, Gibbons, & Memmert. 2016; for indirect evidence see also Calvillo & Jackson, 2014; Koivisto & Revonsuo, 2009). Summed up, to this point evidence suggests that unattended and unnoticed stimuli during inattentional blindness can be processed to some degree. It remains unclear however, which factors determine the occurrence and strength of such preconscious processing. One influencing factor that has already been shown to determine whether or not inattentional blindness occurs, and that could, therefore, play a role in preconscious processing of unnoticed objects as well, is the perceptual load of the primary task. The perceptual load model (e.g., Lavie, 1995) suggests that the processing of task-irrelevant stimuli depends on the perceptual load of the primary task; low perceptual load of the attended information leaves free attentional capacities that can spill over to task-irrelevant stimuli. In contrast, high load captured more attentional resources and prevents the processing of irrelevant stimuli. This model was tested and supported by a variety of empirical data, mainly demonstrating an influence on reaction times (e.g., Furley, Memmert, & Schmid, 2013; Lavie, 1995, 2005; Lavie & Fox, 2000; Lavie, Hirst, de Fockert, & Viding, 2004). Some studies already investigated the influence of perceptual load on whether or not an unexpected object crosses the threshold of awareness (i.e., was reported by the participants). And indeed, this line of research suggests that unexpected and unattended stimuli are more likely to be detected when the perceptual load of the primary task is low (Calvillo & Jackson, 2014; Cartwright-Finch & Lavie, 2007; but see Koivisto & Revonsuo, 2009). In a more applied setting, high perceptual load also significantly increased inattentional blindness and reaction times to hazards in a driving simulator (Murphy & Greene, 2016, 2017). The question that remains is whether perceptual load also moderates the processing of stimuli that do not receive voluntary attention and thus go unnoticed in an inattentional blindness setting. As argued above, previous theorizing and empirical data suggest that the processing of unexpected objects is a continuous process that oversteps the binary threshold of awareness at some point, but can also be measured beneath that threshold. We therefore hypothesize that those factors that influence whether or not a stimulus crosses the threshold of awareness also determine the degree of preconscious processing, and that one of these factors might be the perceptual load of the primary task. To investigate this question, we adapted a recently established inattentional blindness paradigm that uncovered preconscious processing of unattended and undetected stimuli through an interference effect measured by reaction times (Schnuerch et al., 2016). We systematically varied perceptual load of the primary task to test the assumption that predictions of the perceptual load model also hold true for stimuli that remain unconscious due to inattentional blindness. The present study will not only expand knowledge on inattentional blindness but will also provide evidence for the generalizability of the perceptual load model to the processing of stimuli that remain unconscious.