Introduction: Selective attention is fundamental to perception and cognition, allowing brain to prioritize relevant information while suppressing distractions. While neural oscillations are widely implicated in this process, their specific causal roles and underlying mechanisms remain a key area of investigation. A central debate concerns whether oscillations, particularly in the alpha band, primarily mediate attentional selection through target enhancement or distractor suppression. Furthermore, it is unclear if the oscillatory mechanisms governing spatial attention are distinct from those controlling non-spatial, feature-based attention. This review synthesizes recent human electrophysiological and brain stimulation studies to clarify distinct, causal roles of parietal and frontal cortical oscillations in implementing different forms of attentional gating.
Methods: A systematic literature search was conducted in PubMed, Web of Science, and Scopus databases for articles published between January 2019 and December 2024 using keywords: Spatial Attention, Alpha Oscillations, Corticocortical Feedback, Neurostimulation, and Causal Inference. Inclusion criteria were restricted to Q1, original English-language research articles, excluding reviews and case studies, which resulted in a final selection of four papers. The synthesized studies utilized complementary methodologies in human participants to establish causal links. These approaches typically combined non-invasive brain stimulation (tACS) with high-resolution electrophysiology (EEG, ECoG) during auditory and visual tasks to dissociate different forms of attention, integrating behavioral performance with computational modeling.
Results: The synthesized findings reveal a clear functional and anatomical dissociation in the oscillatory control of attention.
First, studies consistently demonstrate that parietal alpha-band oscillations causally implement spatial gating. Exogenously boosting alpha activity in the right intraparietal sulcus (IPS) with 10 Hz tACS selectively impaired auditory spatial attention to the contralateral (left) side, supporting a suppressive function for alpha rhythms. This causal finding is mechanistically refined by invasive electrocorticography (ECoG) evidence in visual attention, which shows that alpha-band activity is spatially, temporally, and functionally dissociated: alpha power
increases (synchronization) early in ipsilateral, unattended cortical regions to suppress distractors, whereas it decreases (desynchronization) later in contralateral, attended regions to facilitate targets. This suppressive role at unattended locations appears driven by local neural activity modulating alpha phase. In contrast, facilitation at attended locations is associated with increased feedforward connectivity, where alpha phase in lower-order visual areas modulates neural firing in higher-order areas.
Second, the control of non-spatial, feature-based attention is causally linked to slow oscillations in prefrontal cortex, not parietal alpha. Parietal alpha stimulation did not affect non-spatial auditory attention. Instead, multiple experiments showed that bilateral inferior frontal junction (IFJ) is selectively engaged during non-spatial attention. Critically, phase-locked tACS timed to align with predicted high-excitability states in the IFJ causally enhanced performance on a feature-based attention task. Drift-diffusion modeling revealed this improvement was specifically due to an increased rate of sensory evidence accumulation for the attended feature.
Finally, the effects reported across all studies were highly specific. The causal influence of parietal alpha was frequency-specific (present at 10 Hz but not 6 Hz) and transient, occurring only during stimulation. Similarly, the effect of frontal stimulation was site-specific (present for IFJ but not a control site) and phase-specific, with performance varying sinusoidally with stimulation phase.
Conclusion: Collectively, these studies provide strong, causal evidence that distinct oscillatory mechanisms in different cortical networks govern attentional selection. The primary role of parietal alpha oscillations in spatial attention appears to be the functional suppression of task-irrelevant locations, implementing a "gating by inhibition" mechanism that facilitates target selection by reducing distractor interference. In contrast, non-spatial, feature-based attention is causally controlled by excitability state of the prefrontal cortex, likely governed by slow rhythmic activity, which appears to directly enhance the gain on sensory evidence for an attended feature.
These findings carry significant implications, establishing a double dissociation for the oscillatory control of "where" versus "what" attention. This framework provides a robust, mechanistic basis for understanding how attentional control is flexibly deployed and may inform pathophysiology of clinical disorders marked by attentional deficits. Future research should investigate how these parietal and frontal networks interact during complex tasks requiring integration of spatial and feature-based cues. Furthermore, exploring the translational potential of precisely-timed, phase-locked brain stimulation to remediate specific attentional impairments represents a promising avenue for clinical neuroscience.