Orienting of Attention

Abstract

Orienting is the covert allocation of visual attention to a location in space, shifted independently of where the eyes point. The spatial-cueing paradigm measures it as the reaction-time benefit at a validly cued location and the cost at an invalidly cued one, showing that attention can be summoned to a place before any movement of the eyes. Two modes are distinguished: endogenous orienting, directed voluntarily by a symbolic cue, and exogenous orienting, captured reflexively by a peripheral event, which follow different time courses and can reverse into inhibition of return. Orienting is realized in a dorsal frontoparietal network that aims attention and a ventral network that reorients it toward salient events. Three interactive demonstrations model the cueing benefit and cost, the facilitation-to-inhibition time course, and the attentional-network scores.

Keywords: spatial attention, spatial cueing, covert orienting, inhibition of return

Orienting is the alignment of attention with a source of sensory input, and in vision it can proceed covertly, without any accompanying movement of the eyes or head (Posner, 1980). A reader can hold central fixation on one word while attending to a location in the periphery, and the attended location is processed faster and more accurately than an unattended one even though the retinal image is unchanged. This dissociation between the direction of gaze and the direction of attention is the defining fact of the field, and the selective advantage it confers is measured in tens of milliseconds of reaction time. This article traces orienting from the spatial-cueing paradigm that isolates it, through the distinction between voluntary and reflexive control and the inhibition that follows a reflexive shift, to the frontoparietal networks that implement it.

What Orienting Is

Orienting is the process of directing attention toward a location or object so that stimuli there are processed with priority. In the visual case the crucial refinement is that this direction can be covert: attention moves to a peripheral location while the eyes remain fixed elsewhere, so the advantage that follows cannot be attributed to a better retinal image (Posner, 1980). Covert orienting is distinguished from overt orienting, in which a saccade brings the fovea to the target, and the two normally cooperate, but the experimental separation of covert attention from gaze is what made orienting measurable as a mental operation rather than a movement.

The benefit conferred by orienting is not merely faster responding. Directing covert attention to a location raises contrast sensitivity and sharpens spatial resolution there, so the attended stimulus is in a real sense seen better, not just reported sooner (Carrasco, 2011). Attention can also be allocated to an object rather than a bare region of space: when a cue appears on one end of an object, the benefit spreads along that object to its other end more readily than to an equally distant location on a different object, evidence that the units of orienting can be perceptual objects as well as locations (Egly et al., 1994). Orienting is therefore best understood as a flexible prioritizing of visual processing, aimed sometimes at a point, sometimes at a region, and sometimes at a structured object.

The Spatial-Cueing Paradigm

The paradigm that isolates orienting is spatial cueing. On each trial a cue indicates a location, a target appears after a variable interval, and the participant makes a speeded detection or discrimination response while holding central fixation. The cue is valid when the target appears where it indicated, invalid when the target appears elsewhere, and neutral when it conveys no location. Comparing reaction time across these conditions decomposes the effect of attention into a benefit, the speeding at the valid location relative to neutral, and a cost, the slowing at the invalid location relative to neutral (Posner, 1980; Posner et al., 1980). Figure 1 shows the characteristic pattern, and the demonstration that follows lets the reader select a cue condition and read off the modeled reaction time.

The cost-benefit logic is what gives the paradigm its power. A benefit alone could reflect a general readiness that the cue induces, but a cost, a genuine penalty for having attended to the wrong place, implies that a limited resource was committed to the cued location and had to be withdrawn and redeployed (Posner et al., 1980). The size of the validity effect, the difference between the invalid and valid conditions, indexes how strongly attention was concentrated at the cued location. Because the response is made without an eye movement, the whole pattern is attributable to the covert shift rather than to foveation of the target.

Covert and Overt Orienting

Overt orienting moves the eyes so that the fovea, the retina's high-resolution center, lands on the object of interest; covert orienting shifts the attentional priority without moving the eyes. The two are ordinarily coupled, because attention typically precedes and guides a saccade to its target, but they are separable, and the cueing paradigm exploits that separability by requiring fixation to be held while attention is drawn elsewhere (Posner, 1980). The covert shift is faster than a saccade and can be deployed to several candidate locations in sequence while gaze stays still, which is why it is treated as the more fundamental operation for the study of attention.

The relationship between the two is close enough that the premotor theory of attention proposed covert orienting to be the planning of an eye movement that is not executed, tying attention to the oculomotor system. Whatever the merits of that strong identity, the systems that direct gaze and covert attention overlap substantially, and orienting a covert focus and preparing a saccade recruit common frontoparietal and collicular structures (Corbetta & Shulman, 2002). The practical upshot for the paradigm is that keeping the eyes fixed is not a trivial control but the very manipulation that lets covert attention be studied apart from the movement it usually accompanies.

Endogenous and Exogenous Orienting

Orienting comes in two modes that differ in what drives them. Endogenous orienting is voluntary and goal-driven: a symbolic cue at fixation, such as an arrow pointing left, tells the participant where the target is likely to appear, and attention is directed there deliberately because it is useful. Exogenous orienting is reflexive and stimulus-driven: an abrupt event in the periphery, such as a sudden flash beside a possible target location, captures attention automatically whether or not it predicts anything (Muller & Rabbitt, 1989). The two are dissociated by their triggers, a central informative symbol versus a peripheral salient event, and by their obedience to intention: endogenous orienting can be withheld if the cue is uninformative, whereas exogenous capture occurs even when the participant knows the peripheral cue is useless and tries to ignore it.

The distinction is more than a taxonomy of cues, because behavioural and neural evidence indicates that the two modes are partly separate systems rather than one mechanism engaged two ways (Chica et al., 2013). Reflexive orienting develops quickly and decays, resists interruption once triggered, and is followed by inhibition; voluntary orienting builds slowly, is sustained as long as it remains useful, and is not followed by inhibition (Muller & Rabbitt, 1989). The time courses of central and peripheral precues confirm the split: a peripheral cue produces its maximal benefit within about a tenth of a second, while a central symbolic cue needs several times longer to reach full effect (Cheal & Lyon, 1991). Table 1 sets the two modes side by side, and the next demonstration lets the reader move the interval between cue and target and watch them diverge.

The Time Course of Orienting

The interval between the cue and the target, the cue-target stimulus-onset asynchrony, governs how much orienting has developed and, for reflexive orienting, whether it has already reversed. An exogenous peripheral cue produces facilitation that rises sharply, peaking within roughly one hundred milliseconds, and then falls away; an endogenous central cue produces facilitation that rises gradually over several hundred milliseconds and is then sustained (Cheal & Lyon, 1991; Muller & Rabbitt, 1989). Reading the effect of a cue therefore requires knowing when the target arrived, because the same peripheral cue can help, do nothing, or hinder depending on the interval.

The most striking feature of the exogenous time course is that facilitation does not merely decay to zero but crosses into inhibition. Beyond an interval of roughly two to three hundred milliseconds, responses to a target at the previously cued location become slower than responses to an uncued location, so the early benefit becomes a late cost (Posner et al., 1985). Endogenous orienting shows no such reversal, which is one of the clearest markers separating the two modes. The demonstration models both curves against a common interval axis, so that the exogenous curve can be seen to fall from facilitation into inhibition while the endogenous curve rises and holds.

Inhibition of Return

Inhibition of return is the delayed responding to a location or object that was recently the focus of a reflexive shift of attention. After attention is drawn exogenously to a peripheral location and then withdrawn, a target appearing there some hundreds of milliseconds later is detected more slowly than a target at a fresh location, reversing the earlier facilitation (Posner et al., 1985). The effect requires that attention have left the location, and it is tied to the reflexive rather than the voluntary system, appearing after peripheral cues and after saccades but not as a consequence of sustained endogenous attention.

The function usually ascribed to inhibition of return is the promotion of efficient sampling of the environment. By tagging recently inspected locations with a bias against returning, it discourages attention from cycling back to places it has just left and nudges it toward locations not yet examined, an advantage for visual search and foraging (Klein, 2000). On this account the late inhibitory phase of exogenous orienting is not a failure of attention but a complement to it, a novelty-seeking pressure that makes search less redundant. The effect has a partly subcortical basis and survives some forms of cortical damage, consistent with its role as a relatively automatic orienting mechanism (Posner et al., 1985).

The Orienting Networks

Orienting is one component of a broader account in which attention is carried by several partly separable networks rather than a single faculty. In the original three-network framework, an orienting network directs attention to locations, an alerting network maintains readiness, and an executive network resolves conflict, each with its own anatomy and its own signature in behaviour (Posner & Petersen, 1990; Petersen & Posner, 2012). The orienting network was tied from the outset to posterior parietal cortex, the pulvinar of the thalamus, and the superior colliculus, structures implicated respectively in disengaging, moving, and engaging attention at a new location.

A complementary anatomy divides orienting into two cortical systems by the source of control (Corbetta & Shulman, 2002). A dorsal frontoparietal system, including the intraparietal sulcus and the frontal eye fields, carries goal-directed, endogenous orienting, generating and maintaining the attentional set that biases processing toward expected locations and features. A ventral frontoparietal system, lateralized to the right hemisphere and including the temporoparietal junction and ventral frontal cortex, acts as a circuit-breaker that detects salient or unexpected events and reorients the dorsal system toward them, corresponding to exogenous capture. The Attention Network Test operationalizes the orienting network as a single reaction-time score, the difference between trials with an uninformative central cue and trials with a valid spatial cue, and the demonstration that follows computes that score alongside the alerting and executive scores (Fan et al., 2002).

The Neural Basis

The clearest early evidence for the anatomy of orienting came from patients with parietal damage. Lesions of the posterior parietal lobe produce a specific deficit in covert orienting: patients are disproportionately slow to respond to a target on the side opposite the lesion when attention has first been drawn to the same side as the lesion, as though they cannot disengage attention from its current location to move it across the midline (Posner et al., 1984). This disengage deficit dissociates the components of orienting and locates the disengaging operation in parietal cortex, and it underlies the clinical syndrome of hemispatial neglect, in which the contralesional side of space is effectively ignored.

Functional imaging in healthy participants confirmed and refined this picture. Preparatory endogenous cues activate the dorsal frontoparietal system, including the frontal eye fields and intraparietal sulcus, before any target appears, showing that top-down orienting is a genuine anticipatory biasing of sensory cortex rather than a change in how the target is processed once it arrives (Hopfinger et al., 2000). The reorienting driven by unexpected targets recruits the right-lateralized ventral system (Corbetta & Shulman, 2002). The inhibitory tagging of visited locations depends in part on the superior colliculus, a midbrain structure central to gaze control, consistent with the tight coupling of covert orienting to the oculomotor system (Posner et al., 1985; Posner & Petersen, 1990).

Applications

The orienting paradigm has become a standard clinical and developmental probe because a single reaction-time contrast isolates a well-defined operation. The disengage deficit revealed by invalid cues is a sensitive marker of parietal dysfunction and of hemispatial neglect, and the cueing task is used to characterize the attentional consequences of stroke and of degenerative disease (Posner et al., 1984). The decomposition of attention into orienting, alerting, and executive components allows these to be measured separately in the same brief test, so that a disorder can be localized to one network rather than attributed to attention in general (Fan et al., 2002).

Beyond the clinic, orienting bears on any setting in which salient events compete for a limited focus. The reflexive capture of attention by abrupt peripheral onsets is the mechanism a warning signal exploits and the mechanism an irrelevant distractor abuses, and its automaticity means that a sufficiently salient event will draw attention even against the observer's intention (Muller & Rabbitt, 1989). Inhibition of return, in turn, shapes how efficiently a display is searched, since the bias against re-inspecting visited locations is what keeps visual search from revisiting the same items (Klein, 2000). Understanding orienting therefore informs the design of alarms, displays, and interfaces in which attention must be guided or protected.

What the Paradigm Does and Does Not Show

The cueing paradigm cleanly isolates a spatial shift of attention, but several of its interpretations have been contested. The assumption that orienting moves through space like a spotlight, with a fixed focus that travels from one location to another, is a metaphor that the data only partly support: attention can be split, its focus can vary in size, and it can be allocated to objects that are not spatially contiguous (Egly et al., 1994). The object-based benefits show that a purely location-based account of orienting is incomplete, and that the units the system prioritizes are not always regions of the visual field.

A second caution concerns what the reaction-time cost measures. A slower response at an invalid location is often read as the time to disengage, move, and re-engage attention, but part of the effect can reflect decision and criterion changes rather than a shift of a perceptual resource. The finding that covert attention alters contrast sensitivity and apparent contrast establishes that at least part of the benefit is genuinely perceptual (Carrasco, 2011), yet disentangling perceptual enhancement from decisional bias in any given cueing experiment requires more than the mean reaction time. The paradigm is thus a precise instrument for detecting that attention was oriented, and a blunter one for deciding exactly which stage of processing the orienting changed.

Worked Example

Consider the cueing model behind the first demonstration, with its default constants. The neutral baseline reaction time, when the cue carries no location, is 350 ms. A valid cue lowers it to 315 ms and an invalid cue raises it to 395 ms. The benefit of a valid cue is the neutral reaction time minus the valid reaction time, 350 minus 315, or 35 ms. The cost of an invalid cue is the invalid reaction time minus the neutral, 395 minus 350, or 45 ms. The overall validity effect, the single number most often reported, is the invalid minus the valid reaction time, 395 minus 315, or 80 ms, which is exactly the sum of the benefit and the cost. That the cost here exceeds the benefit is the typical pattern for reflexive cueing, because withdrawing attention from the wrong location and moving it to the target costs more than a correct head start saves.

The same arithmetic of differences defines the network scores in the third demonstration. With a center-cue reaction time of 530 ms and a valid spatial-cue reaction time of 500 ms, the orienting score is the center-cue minus the spatial-cue value, 530 minus 500, or 30 ms, the benefit of knowing where rather than merely when the target will appear. The alerting score, from a no-cue reaction time of 550 ms and a double-cue reaction time of 515 ms, is 35 ms, and the executive score, from an incongruent reaction time of 620 ms and a congruent one of 520 ms, is 100 ms. Each network is thus summarized by one subtraction, and the demonstration lets the reader vary the condition means and watch the three scores respond independently.

Discussion

Orienting has held a central place in the study of attention because it turned a metaphor into a measurement. The claim that attention can be directed independently of the eyes became testable the moment the cueing paradigm attached a benefit and a cost to a covert shift, and the orderly reaction-time differences that followed have anchored theories of attention for four decades. The distinction between endogenous and exogenous orienting, the reversal of reflexive facilitation into inhibition of return, and the mapping of these operations onto dorsal and ventral frontoparietal systems are among the most reproducible and most generative findings in cognitive psychology.

The trajectory of the field shows the familiar movement from a single faculty toward a set of interacting mechanisms. The early spotlight gave way to a system that can be object-based as well as location-based, that operates in two modes with different dynamics, and that is realized in separable networks with distinct anatomies and distinct failures. The disengage deficit and hemispatial neglect connect the laboratory measure to clinical reality, and the Attention Network Test packages the whole decomposition into a few minutes of testing. Read this way, orienting is not a beam that sweeps across a scene but a structured set of operations for prioritizing where and what the visual system processes next, and its value lies precisely in how finely those operations can be separated and measured.

Glossary

Alerting network.

The attentional system that achieves and maintains a state of readiness for incoming stimuli, measured in the Attention Network Test by the benefit of a warning cue.

Benefit.

The speeding of a response at a validly cued location relative to a neutral baseline, reflecting the head start conferred by correctly oriented attention.

Cost.

The slowing of a response at an invalidly cued location relative to a neutral baseline, reflecting the time to withdraw and redeploy misdirected attention.

Covert orienting.

A shift of attention to a location without a corresponding movement of the eyes, isolated by requiring fixation during a cueing task.

Cue-target asynchrony.

The interval between the onset of the cue and the onset of the target, the variable that governs how far orienting has developed and whether it has reversed.

Disengage deficit.

The difficulty, after parietal damage, of releasing attention from a current location to move it toward the side opposite the lesion, a component failure of orienting.

Dorsal attention network.

A frontoparietal system, including the intraparietal sulcus and frontal eye fields, that carries goal-directed, endogenous orienting.

Endogenous orienting.

Voluntary, goal-driven direction of attention in response to a symbolic cue, slow to develop and sustained while useful.

Exogenous orienting.

Reflexive, stimulus-driven capture of attention by a salient peripheral event, fast to develop, transient, and followed by inhibition.

Inhibition of return.

The slowing of responses to a location or object that was recently attended reflexively, biasing attention toward novel locations.

Neutral cue.

A cue that conveys no information about the target location, providing the baseline against which benefits and costs are measured.

Object-based attention.

The allocation of orienting to a perceptual object, so that a benefit spreads within an object more readily than to an equidistant location on another object.

Orienting network.

The attentional system that directs attention to locations, tied to parietal cortex, the pulvinar, and the superior colliculus, and measured as a spatial-cue benefit.

Overt orienting.

A shift of attention accompanied by a saccade that brings the fovea onto the object of interest, normally coupled to a preceding covert shift.

Spatial-cueing paradigm.

The experimental design in which a cue indicates a location and reaction time to a subsequent target is compared across valid, neutral, and invalid conditions.

Validity effect.

The difference in reaction time between invalidly and validly cued targets, indexing the strength of the spatial orienting produced by the cue.

Ventral attention network.

A right-lateralized frontoparietal system, including the temporoparietal junction, that detects salient events and reorients attention toward them.

Key Researchers

Michael I. Posner. Professor Emeritus of Psychology at the University of Oregon; devised the spatial-cueing paradigm and the covert-orienting construct, distinguished endogenous from exogenous control, and framed orienting as one of three separable attention networks. ORCID - Faculty Page - Google Scholar - Wikipedia)

Steven E. Petersen. James S. McDonnell Professor of Cognitive Neuroscience at Washington University in St. Louis; with Posner articulated the three-network model of attention and the anatomy of the orienting system, and later updated it with two decades of imaging evidence. ORCID - Faculty Page

Maurizio Corbetta. Professor and Chair of Neurology at the University of Padua; with Shulman mapped orienting onto a dorsal frontoparietal system for goal-directed control and a ventral system that reorients attention toward salient events. ORCID - Faculty Page

Raymond M. Klein. Professor Emeritus of Psychology and Neuroscience at Dalhousie University; synthesized the evidence that inhibition of return is a foraging facilitator that biases attention toward locations not yet inspected. ORCID - Faculty Page - Google Scholar

Marisa Carrasco. Julius Silver Professor of Psychology and Neural Science at New York University; showed that covert attention does not merely speed responses but alters appearance, sharpening contrast sensitivity and spatial resolution at the attended location. ORCID - Faculty Page

Ana B. Chica. Professor of Experimental Psychology at the University of Granada; marshalled the behavioural and neural case that endogenous and exogenous spatial attention are two partly independent systems rather than one mechanism engaged two ways. ORCID - Faculty Page - Google Scholar

Frequently Asked Questions

What is the orienting of attention?
It is the covert direction of visual attention to a location or object, so that stimuli there are processed with priority, and it can occur without any movement of the eyes (Posner, 1980).

How is orienting measured?
The spatial-cueing paradigm compares reaction time to targets that are validly, neutrally, or invalidly cued, decomposing attention into a benefit at the cued location and a cost at an uncued one (Posner et al., 1980).

What is the difference between endogenous and exogenous orienting?
Endogenous orienting is voluntary and driven by a symbolic central cue, developing slowly and sustained while useful, whereas exogenous orienting is reflexive, driven by a peripheral event, fast, and transient (Muller and Rabbitt, 1989).

What is inhibition of return?
It is the slowing of responses to a location that was recently the focus of a reflexive shift of attention, a late reversal of facilitation that biases attention toward locations not yet inspected (Klein, 2000).

Can attention move without the eyes moving?
Yes; covert orienting shifts attentional priority to a peripheral location while fixation is held, and the resulting benefit shows that attention and gaze are separable (Posner, 1980).

Which brain systems carry orienting?
A dorsal frontoparietal network directs goal-driven orienting and a right-lateralized ventral network reorients attention toward salient events, complementing the parietal, pulvinar, and collicular structures of the classic orienting network (Corbetta and Shulman, 2002).

Why are patients with parietal damage slow to reorient?
Posterior parietal lesions produce a disengage deficit, a difficulty releasing attention from its current location to move it across the midline, which underlies hemispatial neglect (Posner et al., 1984).

What does the Attention Network Test measure?
It derives separate reaction-time scores for the alerting, orienting, and executive networks from one brief cued-flanker task, isolating the orienting score as the benefit of a valid spatial cue over an uninformative central cue (Fan et al., 2002).

References