Attention, Awareness, and the Right TPJ

Review written by Amy Ciceu (2024)

Have you ever found yourself in deep thought in a public setting only to come to your senses and arrive at the uncomfortable realization that you’re making eye contact with another person? Such is the dilemma that faces us when we lose control of our awareness. Awareness—a module of behavior that allows us to be conscious of stimuli in our environment—is fundamentally distinct from but similar to attention—the process of selectively focusing the mind on certain stimuli at the expense of others. In an insightful new study led by graduate student Andrew Wilterson of the Graziano Lab at Princeton University, researchers used stimulus-prediction tasks and MRI imaging to investigate the interrelated nature of awareness and attention in the human brain. 

A recent theory in psychology developed by Dr. Graziano, the attention schema theory (AST), provides a plausible explanation to understand the ways in which awareness and attention closely interact. The AST proposes that awareness allows the brain to govern the control of attention. In the present study, Wilterson et al. built on AST by positing that awareness serves as the linchpin of attention regulation. In other words, in order to enable the accurate and timely shifting of our attention to relevant stimuli, awareness is indispensable. However, awareness has been difficult to quantify in experimental settings due to the subjective nature of this phenomenon. Prior studies identified a plausible brain region that might mediate awareness: the right temporoparietal junction (TPJ). The right TPJ has previously been implicated in empathizing with others—a capacity termed theory of mind, which governs our ability to develop an awareness of others’ distinctive mental states. However, few studies have corroborated the role of the TPJ in regulating awareness and attention or whether awareness is necessary for precise, internally generated control of attention. 

In order to examine the role of the right TPJ in awareness and attention, the researchers conducted two versions of a target-prediction task. The target-prediction task was developed by the authors as a useful behavioral paradigm to precisely manipulate attention and awareness.  There was just one difference between the two tasks: the subjects were either aware or unaware of the implicit association between an exogenous cue and the subsequent location of the target stimulus, a tiny white circle whose location was predicted by the cue. The difference in the tasks lies in the manipulation of participants’ awareness of the cue and its relationship to the target dot, thereby allowing the experimenters to test the premises of the AST. 

Endogenous attention refers to attention directed by internal mechanisms, such as goals that motivate attention-directing behavior. For example, imagine yourself scanning the pages of a Where’s Waldo? book, searching for the titular character in his characteristic striped sweater; this action is motivated by the objective of locating the cleverly hidden Waldo. In contrast, exogenous attention is driven by the prominence of salient stimuli in a subject’s environment. Take, for instance, all of the other interesting elements contained in the pages of a Where’s Waldo? book. These eye-grabbing minutiae render the task of locating Waldo increasingly difficult, in no small part because these quirky details evoke significant attention on their own. Participants performed these tasks while in a functional magnetic resonance imaging (fMRI) machine. fMRI measures blood-oxygen-level-dependent (BOLD) signals in the brain, which can be used as a proxy for neural activity. Using this experimental setup, Wilterson et al. were able to measure BOLD activity--and inferred neural activity--in the right TPJ of participants performing each task.

The first task measured subjects’ ability to direct endogenous attention to the predicted location of a target stimulus after being presented with an exogenous, salient cue. More specifically, participants were told to fixate on a black circle at the center of the testing screen in front of them. After 1200 milliseconds, a red cue appeared at a random location on the screen for 35 milliseconds before giving way to an array of equidistant black circles. This array, considered a “visual mask”, persisted for 465 milliseconds before a target, a white dot, appeared for 80 milliseconds in a location to the right of the initial red cue (on 85% of trials, considered “predicted trials”) or to the left of the cue (on 15% of trials,  considered “unpredicted trials”). Before completing the task, participants were informed of each visible stimulus to appear in the task, including the salient red cue, and were instructed to press certain keys at the end corresponding to whether the target appeared to the right or the left of the initial red cue. This initial knowledge of all cues that would appear throughout the task allowed participants to formulate an implicit predictive association between cue and target location. The cue accurately predicted the location of the target on 85% of trials, but participants were never explicitly informed about the association between the cue and the location of the target.

The second task was identical to the first, with one key difference: when explaining the task procedure, the task administrators explained all stimuli except the cue, which was colored black rather than red to conform with the “visual mask” presented shortly thereafter. In this manner, participants’ awareness of the implicit association between the initial black cue and the location of the subsequent target was precluded. 

The authors confirmed that the precise association between cue and target was indeed implicit by asking participants post-test questions about any relationships they observed between the cue stimulus and the target stimulus. A vast majority of participants in the first task reported that they had noticed that the cue and target tended to sequentially appear in close spatial proximity to one another. Interestingly, none of the participants identified the precise association between both stimuli: that the target often appeared slightly to the right of the cue. Participants in the second task were more naive to the association between cue and target locations: they did not report consciously observing this association, let alone even spotting the black masked cue at the outset of each trial of the task. From these participant responses, the authors concluded that awareness was present in the first task and absent in the second task.

In line with previous research—such as that by Geng and Vossel (2013) suggesting a possible role for the TPJ in regulating attentional control1—the authors speculated that the right TPJ would become active relative to baseline in task one, where awareness of the cue permits the efficient command of attention to the predicted location of the target. In particular, they believed that the TPJ would become more active during unpredicted relative to predicted trials during the first task because the unanticipated location of the target would elicit more feedback from the TPJ, which is thought to maintain control over one’s awareness of the association between the stimuli in the task. Conversely, they hypothesized that the right TPJ would remain inactive in the second task during both predicted and unpredicted trials because the association between the now masked cue and the subsequent location of the target would no longer command participants’ awareness and therefore result in disrupted control of attention. 

Performance on both tasks was assessed by calculating a single reaction time score for each participant, averaged across all task trials. This score served as a proxy for the magnitude of each participant’s endogenous control of attention, or the extent to which each participant actively attended to the stimulus and implicitly constructed a relationship between the cue and target locations. 

The results for the first task revealed that the implicit association between the appearance of the red salient cue and the subsequent target was robust. During trials where target location was correctly predicted, mean reaction times were shorter than those of trials in which target stimuli did not occur to the right of the salient cue stimulus. In contrast, trials in which target locations were inaccurately predicted (i.e. the targets appeared to the left rather than the right of the cue’s original position) resulted in longer reaction times, indicating an inverse relationship between the accuracy of predicted target location and participants’ reaction times. Strikingly, fMRI revealed that right TPJ activity was significantly higher relative to baseline during unpredicted trials compared to predicted trials. In fact, during predicted trials, the right TPJ exhibited no considerable BOLD activity difference from baseline. The authors interpreted this finding as a consequence of the right TPJ’s sensitivity to the unanticipated target location during the unpredicted trials. This hypothesis concurs with previous findings that suggest that the right TPJ is sensitive to surprising stimuli that invoke exogenous attention.2 

In comparison, Wilterson et al. found that in the second task the right TPJ was inactive during both unpredicted and predicted trials. Results for the second task underscore virtually no difference between participants’ reaction times during predicted and unpredicted trials; the mean reaction time was about 460 ms for both types of trials. These results suggest that the lack of awareness of the predictive relationship between the masked cue and the target location may have occluded any differences in right TPJ activity. Therefore, the lack of activity in the right TPJ confirmed the authors’ hypothesis that an absence of awareness prevents activation of the TPJ and consequently the control of endogenous attention to the projected location of the target.

Taken together, Wilterson et al.’s study suggests that the right TPJ serves as an important arbiter of awareness and attention in the context of visual processing. Notwithstanding the ambiguity of the right TPJ’s precise role in mediating attention and awareness, the results highlight a central role of this region in governing these behavioral modules. Specifically, awareness serves as a precondition for enabling control of endogenous attention and the TPJ takes note of stimuli that transgress the implicit predictive models that we construct of the world. The next time you’re out in public operating on autopilot, make sure you start brooding over your recent dilemmas when not in the presence of other human beings. Otherwise, you could  become unaware of your surroundings and relinquish control of your attention as your eyes  maintain contact with those of a concerned passerby for an uncomfortably long time.

The original article described here was published in PNAS on June 22, 2021. Please follow this link to view the full version. 


1 J. J. Geng, S. Vossel, Re-evaluating the role of TPJ in attentional control: Contextual updating? Neurosci. Biobehav. Rev. 37, 2608–2620 (2013).

2 M. Corbetta, J. M. Kincade, G. L. Shulman, Neural systems for visual orienting and their relationships to spatial working memory. J. Cogn. Neurosci. 14, 508–523 (2002),