Definition
Attentional control is the capacity to direct attention according to internal priorities — task goals, current intentions, abstract relevance — rather than passively following stimulus salience. The construct corresponds to what Posner termed the executive attention network (anchored in anterior cingulate and lateral prefrontal cortex) and what subsequent literature describes as the dorsal frontoparietal control network operating in opposition to the ventral attention network (which is more reflexively engaged by salient stimuli).
Operationally, attentional control comprises several dissociable components: attentional engagement (allocating resources to a target), attentional disengagement (releasing from a current target so attention can move), resistance to distraction (suppressing competing inputs), and flexible reallocation (shifting among targets in response to changing demands). These components dissociate in patients with various lesion profiles and in conditions such as ADHD (where disengagement and resistance to distraction are most affected) and autism spectrum (where reallocation is selectively impaired in some paradigms).
The neural mechanism centers on biased competition: posterior sensory cortices represent multiple inputs in parallel with mutual inhibition, and frontoparietal control signals bias the competition in favor of the goal-relevant input. The control signal is sustained as long as the goal is active; when control fails, sensory cortices revert to salience-driven competition in which the most physically distinctive or affectively-charged input wins by default.
Context
Within Prismatic's threat model, attentional control is the gating mechanism through which all higher epistemic processing must be supplied with input. If attention is captured by stimuli that the manipulator has engineered to be salient (high-affect, novel, conflict-evoking, fear-inducing), then whatever evidence would be needed to detect the manipulation is never selected for processing in the first place — the failure occurs before any deliberation could engage with it.
Many kompendium techniques operate at this level. Distraction techniques (EA-1.3 family) generate competing salient stimuli that displace task-relevant attention; affect-loaded inputs (EA-2.x family) exploit the priority that affectively-charged content has in the salience computation; information saturation (EA-1.1 family) overwhelms control capacity until salience-driven selection becomes the only available mode. Each operates upstream of the cognitive functions that would, in their absence, detect the corruption.
The Decision Engine treats attention-state proxies as explicit calibration inputs: extended sessions, high-density information environments, multitasking contexts, and time-pressure conditions all signal compromised attentional control and produce confidence downweighting in resulting decisions. Platform UX design follows control-supportive principles: progressive disclosure rather than information dumps, explicit foregrounding of decision-critical elements, removal of decorative salience competitors, and pacing that protects control capacity.
Interactive: Biased Competition
Posterior sensory cortices represent goal-relevant and salience-driven inputs in parallel with mutual inhibition. Frontoparietal control signals bias the competition. When control is high, the goal wins; when control collapses, the most physically distinctive or affectively-charged input wins by default. Adjust the sliders or click a kompendium attack preset.
Top-down bias toward the goal-relevant input
Physical/affective intensity of the competing input
See Also
- Dorsolateral PFC — Control signal source
- Anterior Cingulate — Control recruitment
- Cognitive Load — Saturates control
- Inhibitory Control — Sister executive function
- Sustained Attention — Time-extended control
- System 1 / System 2 — System 2 requires controlled attention
- Kompendium — EA technique taxonomy
Connect & Contribute
Created by Tomáš Korcak (korczis) | Open Source under GHL