Welcome to Daniel Weissman's
Attention and Cognitive Control Laboratory


Interactions between attention and working memory

The degree to which irrelevant stimuli capture attention is modulated by whether the features of such stimuli (e.g., colors, conceptual categories, etc.) match the current contents of working memory. For example, a flashing red light may attract attention more when we are thinking of a bull fighter (who carries a red flag) than when we are thinking of a tank (which is usually green). In a series of behavioral studies, we are investigating more fully how interactions between attention and working memory underlie such effects. In our first study (Moore et al., 2010), we found that such effects can be 2-3 times larger when people fill the contents of working memory with multiple representations (e.g., red, green, & blue) than when they maintain only a single representation (e.g., red). Two additional studies have revealed that such effects can be prevented under certain conditions (Moore et al., submitted) by taking advantage of the fact that only a single representation can occupy a limited-capacity focus of attention in working memory (Moore et al., in preparation).

Role of the anterior cingulate cortex in cognitive control

Using fMRI, we are currently investigating regional specialization in the anterior cingulate cortex for processes that increase attention to relevant stimuli and processes that detect the presence of competing irrelevant stimuli. Our findings thus far have revealed such specialization in dorsal and rostral subregions of the cognitive division of the anterior cingulate cortex. Moreover, activity in dorsal subregions, which appears to index processes that increase attention to relevant stimuli, correlates with behavioral measures of orienting attention. More generally, we are also investigating how the detection of competing irrelevant stimuli by the anterior cingulate cortex leads to strategic adjustments of upcoming behavior.

Attention and cognitive control

The ability to voluntarily control the focus of our attention is a crucial component of cognitive control. In our laboratory, we use a combination of behavioral, fMRI, and EEG methodologies to study the psychological and neural processes underlying such control. Currently, we have several ongoing projects.

Hijacking of cognitive control systems by irrelevant stimuli

The inability to maintain attention to a particular task produces serious deficits in numerous clinical syndromes, including ADHD, drug addiction, and Alzheimer's Disease. Using fMRI, we are currently investigating how activity in prefrontal, parietal, and sensory regions is affected by the presence of distracting task cues that signal subjects to perform a task that is not currently relevant. Our findings to date have revealed that such stimuli not only increase activity in fronto-parietal control circuitry, but also lead subjects to prepare for the irrelevant task, as evidenced by altered patterns of pre-target biasing activity in the sensory cortices. Moreover, these patterns are further modulated by the timing with which the distracting task cues are presented, being more pronounced when the cues are presented earlier rather than later (Moore et al., 2009).

The neural bases of momentary lapses in attention

We all zone out occassionally, which leads us to respond more slowly than usual. By investigating trial-by-trial relationships between brain activity and response time in humans, we determined that attentional lapses begin with reduced prestimulus activity in anterior cingulate and right prefrontal regions involved in controlling attention. Less efficient stimulus processing during attentional lapses was also characterized by less deactivation of a 'default-mode' network, reduced stimulus-evoked sensory activity, and increased activity in widespread regions of frontal and parietal cortex. Finally, consistent with a mechanism for recovering from attentional lapses, increased stimulus-evoked activity in the right inferior frontal gyrus and the right temporal-parietal junction predicted better performance on the next trial. Future studies are examining the effects of lapses in attention on our processing of irrelevant distracting stimuli.

Figure 2. Relatively small amounts of prestimulus activity in frontal control regions predict relatively slow response times.
(a) A statistical map indicating voxels in the anterior cingulate cortex (ACC), right middle frontal gyrus (MFG) and right inferior frontal gyrus (IFG), where slower RTs were associated with reduced activity one time point (that is, 1.25–2.5 s) after stimulus onset. These activations are overlaid on several slices of the MNI-normalized brain. In this and subsequent figures, coordinates in parentheses refer to the center of mass in Talairach space. (b) The relationship between longer RTs and target-related activity (in units of percent change in fMRI signal per second of increased RT above the mean RT) across time in the ACC. Notice that fMRI activity one time point before stimulus onset decreases as RT increases. Furthermore, this negative relationship between longer RTs and target-related activity reaches its minimum value one time point after stimulus onset (arrow). This gradual decrease is highly consistent with a momentary reduction of neural activity that occurs during the intertrial interval before a stimulus is presented, and achieves its peak hemodynamic effect 3.75–6 s later. (c) The average fMRI response to target stimuli (in units of percent change from baseline) across time in the ACC.

Voluntary Task Switching

Multitasking is a way of life in modern society, but we are only beginning to understand the factors that govern when we choose to perform one task as opposed to another. In a series of behavioral and ERP studies, Joe Orr and Prof. Weissman are investigating whether various forms of distraction influence which task a person will subsequently choose to perform. They are also investigating the degree to which errors influence task choice (e.g., "if at first you don't succeed, try, try again"). The first paper on these studies was recently submitted and the second is underway!

Figure 2.  Double dissociation for cognitive control in the ACC cd .  (a) Saggital slice indicating our rostral ACC cd subregion (green) and our dorsal ACC cd subregion (red) on the MNI-normalized brain.  (b) The average fMRI signal across time (in units of percent signal change from baseline) for the various cue and target stimuli in the rostral ACC cd .  There was a significant effect of target congruency (i.e., greater peak activity for incongruent target-distracter pairs than for congruent target-distracter pairs; shown with a dashed circle), but not of cue congruency (i.e., greater peak activity for incongruent cues than for congruent cues).    (c) The average fMRI signal across time in the dorsal ACC cd for the various cue and target stimuli.  There were significant effects of both cue congruency and target congruency.  (d) Activity specific to cue congruency and target congruency in the rostral ACC cd and in the dorsal ACC cd .  In the rostral ACC cd, we observed significantly greater activity specific to target congruency than to cue congruency, while in the dorsal ACC cd we observed the exactly the opposite effect.  Dashed circles in b and c indicate significant differences in peak activity ( P < 0.05).  In d, a single asterisk denotes P < 0.05 while two asterisks denote P < 0.005.  Error bars represent S.E.M.