Pupil dilation: a fingerprint of temporal selection during the "attentional blink".

Ariel Zylberberg,Mariano Sigman,Manuel Oliva

doi:10.3389/fpsyg.2012.00316

Abstract

Pupil dilation indexes cognitive events of behavioral relevance, like the storage of information to memory and the deployment of attention. Yet, given the slow temporal response of the pupil dilation, it is not known from previous studies whether the pupil can index cognitive events in the short time scale of ∼100 ms. Here we measured the size of the pupil in the Attentional Blink (AB) experiment, a classic demonstration of attentional limitations in processing rapidly presented stimuli. In the AB, two targets embedded in a sequence have to be reported and the second stimulus is often missed if presented between 200 and 500 ms after the first. We show that pupil dilation can be used as a marker of cognitive processing in AB, revealing both the timing and amount of cognitive processing. Specifically, we found that in the time range where the AB is known to occur: (i) the pupil dilation was delayed, mimicking the pattern of response times in the Psychological Refractory Period (PRP) paradigm, (ii) the amplitude of the pupil was reduced relative to that of larger lags, even for correctly identified targets, and (iii) the amplitude of the pupil was smaller for missed than for correctly reported targets. These results support two-stage theories of the Attentional Blink where a second processing stage is delayed inside the interference regime, and indicate that the pupil dilation can be used as a marker of cognitive processing in the time scale of ∼100 ms. Furthermore, given the known relation between the pupil dilation and the activity of the locus coeruleus, our results also support theories that link the serial stage to the action of a specific neuromodulator, norepinephrine.

Highlights

When two masked stimuli are presented in close succession, identification of the first one (T1) hinders the detection of the second (T2) if both are presented within 200–500 ms (Raymond et al, 1992)
The Psychological Refractory Period (PRP) and the Attentional Blink (AB) are related phenomena which can even be obtained within the same paradigm (Wong, 2002)
We observed a clear “lag-1 sparing” (Martin and Shapiro, 2008), whereby accuracy is higher at lag-1 than at lags 2 and 3 (Figure 1B). These results indicate that our paradigm produced a classic AB effect, which allowed us to investigate how the pupil varies with performance and lag

Summary

Introduction

When two masked stimuli are presented in close succession, identification of the first one (T1) hinders the detection of the second (T2) if both are presented within 200–500 ms (Raymond et al, 1992). This observation, referred as the Attentional Blink (AB), has been described in terms of two-stage theories of conscious access (Chun and Potter, 1995), where an initial effortless feedforward propagation of stimulus information is followed by a second wave of stimulus amplification incorporating contextual and task-related information (Dehaene et al, 2006).

Methods

Results

Conclusion