Attention Lights Up New Object Representations before the Old Ones Fade Away

Paul S. Khayat, Henk Spekreijse, Pieter R. RoelfsemaJournal of Neuroscience 2006 · DOI

Abstract

We investigated how attention shifts from one object to another by recording neuronal activity in the primary visual cortex. Monkeys performed a contour-grouping task in which they had to select a target curve and ignore a distractor curve. Some trials required a shift of attention, because the target and distractor curves were switched during the course of the trial. We monitored the dynamics of this attention shift in area V1, in which neuronal responses evoked by the target curve are stronger than those evoked by the distractor. The reallocation of attention was associated with a rapid and strong enhancement of responses to the newly attended curve, followed, after approximately 60 ms, by a weaker suppression of responses to the curve from which attention was removed. We conclude that attention can be rapidly allocated to a new object before it disengages from the previously attended one.

Introduction

Classic psychological models, such as Posner’s "spotlight" theory, suggest that shifting visual attention is a strictly sequential process: the brain must first disengage from the current object, move the focus, and then engage the new target. However, behavioral and EEG studies have struggled to provide consistent timing for these phases, largely because indirect measures cannot easily separate perception from reallocation.

This study, published in The Journal of Neuroscience, addresses this debate by recording directly from the primary visual cortex (V1) in macaques. By observing the "source code" of neuronal firing rates, researchers demonstrate that the brain does not follow a simple stop-and-start sequence. Instead, it "lights up" a new representation significantly before the old one begins to fade, allowing for a brief period of attentional overlap.

Experimental Methods

The researchers trained two macaque monkeys to perform a sophisticated curve-tracing task designed to isolate the temporal dynamics of attentional shifts.

Experimental paradigm showing normal and switch trials.

In a Normal trial (Figure 1a), monkeys fixated on a central point while two curves appeared. One curve was connected to the fixation point (the target, T), while the other was not (the distractor, D). The monkey's task was to follow the target curve mentally and eventually make a saccade to the circle at its end.

In a Switch trial (Figure 1b), the "cue" at the fixation point changed mid-trial (at 400ms), effectively swapping the identities of the target and distractor. This forced the monkey to rapidly reallocate attention to the new target. During these tasks, multiunit activity was recorded from 59 sites in area V1. Because V1 neurons respond more strongly to attended curves, the researchers used a "Modulation Index" (MI) to track the precise timing of enhancement for the new target and suppression for the old distractor.

Results: The Asymmetry of Attention

The electrophysiological data revealed a striking temporal gap in how the brain handles attentional shifts. Rather than a symmetric transition, the process is characterized by a rapid "on" signal and a slower "off" signal.

  • Enhancement Latency: Neurons began signaling the new target curve approximately 144 ms after the switch.
  • Suppression Latency: The representation of the old, now-irrelevant curve only began to be suppressed 210 ms after the switch.
  • The Attentional Overlap: There is a statistically significant 66 ms lead time during which the brain is effectively attending to both objects simultaneously.
  • Magnitude Differences: The enhancement of the new target (MI = 0.41) was significantly stronger than the suppression of the old distractor (MI = 0.16).

These findings suggest that the brain prioritizes the rapid engagement of new information over the immediate disengagement from the old. The "fade away" of the previous representation appears to be a more passive or slower-acting mechanism than the active "lighting up" of the new target.

Discussion and Future Directions

This research challenges the sequential "spotlight" model and reinforces the view of area V1 as a dynamic map. Rather than just a simple feature detector, V1 acts as a blackboard where higher-order attentional signals—likely originating in the parietal or frontal cortex—are superimposed onto low-level visual representations.

While these results provide high-resolution temporal data for V1, questions remain about the universality of this overlap strategy across the visual hierarchy. Future work involving simultaneous recordings from executive areas like the prefrontal cortex could determine if the "overlap" command originates there or if it emerges as a result of processing delays within the visual cortex itself.

Summary updated

Abstract

We investigated how attention shifts from one object to another by recording neuronal activity in the primary visual cortex. Monkeys performed a contour-grouping task in which they had to select a target curve and ignore a distractor curve. Some trials required a shift of attention, because the target and distractor curves were switched during the course of the trial. We monitored the dynamics of this attention shift in area V1, in which neuronal responses evoked by the target curve are stronger than those evoked by the distractor. The reallocation of attention was associated with a rapid and strong enhancement of responses to the newly attended curve, followed, after approximately 60 ms, by a weaker suppression of responses to the curve from which attention was removed. We conclude that attention can be rapidly allocated to a new object before it disengages from the previously attended one.
BibTeX
@article{khayat2006attention,
  author = {Paul S. Khayat and Henk Spekreijse and Pieter R. Roelfsema},
  title = {{Attention Lights Up New Object Representations before the Old Ones Fade Away}},
  journal = {Journal of Neuroscience},
  year = {2006},
  doi = {10.1523/jneurosci.2784-05.2006},
}
Figure