Attention

Early selection models (Broadbent, 1958) propose that a filter selects inputs for further processing based on physical characteristics — before semantic analysis has occurred. This explains why we retain little of the unattended message in dichotic listening (it is blocked before meaning is extracted). Late selection models (Deutsch & Deutsch, 1963) propose that all inputs are analysed to the level of meaning before selection — the bottleneck is at response selection, not perceptual analysis. The evidence from cocktail party effects (one's name in the unattended channel) and priming studies from unattended channels favours some form of intermediate or attenuation model (Treisman), in which unattended inputs receive attenuated rather than absent processing, with highly relevant content still capable of breaking through.

Inattentional blindness is the failure to perceive an unexpected but fully visible stimulus when attention is engaged elsewhere. In Simons and Chabris's (1999) gorilla experiment, ~50% of participants watching a basketball passing task failed to notice a person in a gorilla costume. The phenomenon demonstrates that perception is not automatic or comprehensive — conscious visual awareness depends critically on the allocation of attention to a stimulus. Even large, salient events can be missed if they fall outside the current focus of attention. Inattentional blindness has practical implications for safety (surgeons missing anomalies, air traffic controllers missing aircraft) and for understanding cognitive limitations in multitasking.

Top-down (endogenous, goal-directed) attention is deliberately deployed toward stimuli that are relevant to current goals and expectations — you deliberately read this sentence because your goal is to understand attention. It is controlled by the frontal eye fields (FEF) and intraparietal sulcus (IPS) — the dorsal attention network — and is voluntary and slower to deploy. Bottom-up (exogenous, stimulus-driven) attention is automatically captured by salient stimuli — a sudden flash of light, a loud noise, or one's own name — regardless of current goals. It is controlled by the temporoparietal junction (TPJ) and inferior frontal gyrus — the ventral attention network — and operates rapidly and automatically. In practice, most attentional behaviour reflects the interaction of both systems.

The attentional blink is the temporary impairment in detecting a second target (T2) appearing within about 200–500 ms after a first target (T1) in a rapid serial visual presentation (RSVP) stream. First described by Raymond et al. (1992), it suggests that consciously identifying a target requires cognitive resources that take time to replenish. T2 items presented at a lag of ~2–4 items after T1 are frequently missed — they appear to enter the visual processing stream but fail to be consolidated into conscious awareness. The attentional blink provides a tool for studying the temporal dynamics of conscious access and the bottleneck in attentional processing, which correlates with global workspace activity in frontal and parietal regions.