We provide evidence that the locus coeruleus-norephinephrine (LC-NE) system is the neurophysiological basis of the attentional blink. The attentional blink refers to decreased accuracy for reporting the second of two targets in a rapid serial visual presentation of distractors. The LC-NE account of the attentional blink posits that targets elicit a facilitative LC-NE system response that is available for the first target but subsequently unavailable to the second, due to the autoinhibitory nature of the LC-NE system. We propose a modification of the LC-NE account, suggesting that the LC-NE system response is elicited by interference between mutually exclusive responses demanded by temporally proximal targets and distractors. We increased the interference between the first target and the following distractor by reducing the time between them. For identifying the second target this high-interference condition yielded a benefit up to 200 msec after onset of the first, followed by a decrease in accuracy. Consistent with our modification of the LC-NE account, this result suggests a temporarily enhanced LC-NE system response to increased target-distractor interference.
When two targets are inserted into a rapid serial visual presentation (RSVP) of distractors, there is decreased accuracy for identification of the second target (T2) if it occurs approximately 100-600 msec after the first target (T1). This phenomenon is referred to as the attentional blink (AB; Raymond, Shapiro, & Arnell, 1992). The AB is considered an important phenomenon because of its relation to the process of consolidating an iconic memory into working memory, inferred from evidence that perceptual and semantic information is extracted from T2 during the AB even when T2 cannot be identified (e.g., Vogel, Luck, & Shapiro, 1998). This finding suggests that the AB deficit is particular to an attentional process that facilitates successful encoding of a stimulus into working memory for later report.
Figure 1 illustrates our variation of the AB task. The temporal profile of the AB is described using accuracy scores for the identification of T2 at varying lags. When T2 follows T1 consecutively, it is said to follow at lag 1. If a single distractor occurs between T1 and T2, T2 follows at lag 2, and so on. Identification of T2 is typically preserved at lag 1 and then most impaired at lags 2 and 3, with performance gradually recovering from lags 4-6, when the deficit typically ends. The unimpaired accuracy at lag 1 is termed lag 1 sparing. Most AB research uses a presentation rate of about 10 items/sec, so lag 1 typically corresponds to a stimulus onset asynchrony (SOA) of 100 msec after the onset of T1, and successive lags occur at 100-msec intervals.
Nieuwenhuis, Gilzenrat, Holmes, and Cohen (2005) described how the AB phenomenon maps onto the behavior of a neuromodulatory system theorized to play a role in attention. The locus coeruleus-norepinephrine (LC-NE) account of the AB is based on evidence that the LC responds to the occurrence of a motivationally salient event by releasing NE into the forebrain. Computational analyses of the effects of NE in the forebrain suggest that the presence of NE will increase the responsivity of target neurons, enhancing signal detection and stabilizing a neural representation in the face of noise or interference (Cohen, Aston-Jones, & Gilzenrat, 2004). Critical to accounts of the AB, the LC is autoinhibitory, so that a large release of NE will be followed by a short period of reduced NE release. This event sequence means that a short period of benefit will be followed by a period of deficit. Under the LC-NE account of the AB, NE is released by the LC upon target detection, facilitating target processing. Sparing observed at short SOAs following T1 (typically represented by lag 1 sparing at 100 msec) represents the continuing presence of NE before it dissipates, and the AB deficit corresponds to the period after T1 when the initial recruitment of NE has dissipated and further release is inhibited. …