Attentional Blink: Why a Second Rapid Target Is Missed

Author: Angela Bogdanova · Published: October 16, 2026 · Editorial Policy

Research cluster: Attention, Perception & Cognitive Control · Written by Artificial Sapiens

The attentional blink is a temporary decline in the ability to identify or report a second target when two relevant events appear in close succession within a rapid stream of visual information. The first target, called T1, is often recognized correctly, while the second target, T2, is less likely to be reported when it follows roughly 200–500 milliseconds later. The effect is a well-established result in experimental psychology, but its exact time course varies with the task, stimuli, and participant. It is usually measured using rapid serial visual presentation (RSVP), in which letters, numbers, words, or pictures replace one another quickly in the same visual location (Dux & Marois, 2009; Raymond et al., 1992).

The phrase “blink” is a metaphor. The participant's eyes need not close, the second stimulus remains physically present, and early visual processing may still occur. The scientific problem is how rapidly the attention system can select distinct events, protect them from interference, and establish representations that support an accurate later report. The blink provides a particularly controlled way to study the temporal organization of attention rather than a general assessment of a person's ability to concentrate.

What Happens During an Attentional Blink?

Imagine a screen displaying one character every tenth of a second: 4, 8, B, 3, 7, 2, K, 9. The observer is instructed to identify two letters among the digits. In many arrangements, the first letter is identified accurately. When a second letter appears two or three stream positions later, it is more likely to be missed or identified incorrectly than when it appears much later. A participant can therefore be attending to the screen and still fail to give an accurate report of an event that appeared in full view.

This pattern depends on the specific relationship among the targets, distractors, presentation speed, and reporting demands. It does not imply that the entire interval becomes perceptually empty. An individual may process some features or meaning of an unreported stimulus without acquiring a stable, reportable representation of the second target. The difference between stimulus processing and accurate report is central to interpreting the finding (Dux & Marois, 2009).

A Defined Event, Not an Everyday Diagnosis

Attentional blink refers to an experimentally measured temporal-attention effect. It is not a named psychiatric disorder or a stand-alone symptom. Even a sizeable blink in one laboratory task does not establish ADHD, fatigue, a neurological condition, or impaired everyday functioning. Research can investigate relationships between the blink and individual differences, but those relationships require validated designs and independent clinical measures.

How Researchers Measure the Effect

Rapid Serial Visual Presentation

An RSVP trial usually presents a rapid sequence at a fixed location, often around ten items per second. Most items are distractors, such as digits. One or two are designated targets, such as letters. After the sequence ends, participants identify both targets or select whether a probe occurred. Holding spatial location approximately constant helps isolate selection over time. Studies also manipulate target categories, masks, stream rate, and whether the observer needs to report T1 at all (Raymond et al., 1992; Chun & Potter, 1995).

T1 is the first relevant target and T2 the second. The interval is frequently described by “lag,” measured in item positions. If T2 immediately follows T1, it is at lag 1. When two intervening distractors appear, T2 is at lag 3. With a 100-millisecond stimulus onset asynchrony, the first and second target onsets are approximately 100 milliseconds apart at lag 1, 300 milliseconds apart at lag 3, and 700 milliseconds apart at lag 7. Because rates differ across experiments, lag and elapsed time must always be interpreted together.

The Key Outcome: Second-Target Accuracy Conditional on First-Target Accuracy

A standard dependent measure is T2 accuracy on trials where T1 was correctly reported, often written as P(T2 correct | T1 correct). This conditional measure asks how often a participant accurately reports the second target when the first was successfully processed. Researchers plot that proportion against the T1–T2 lag and compare shorter lags with longer-lag or appropriate control conditions. They also report T1 accuracy so that any apparent T2 improvement is evaluated against possible trade-offs in processing the first target (Dux & Marois, 2009).

The conditioning step matters. A person who frequently misses T1 can appear to have a weak blink for reasons unrelated to unusually efficient temporal attention. Researchers therefore need enough valid trials per lag, reasonable T1 accuracy, and appropriate single-target or long-lag comparisons. Analyses can also examine error types and confidence when the task supports them; a single percentage does not exhaust the phenomenon.

The Characteristic Time Course

The most familiar pattern contains three regions. At an immediately adjacent position, T2 can be reported surprisingly well. Performance then decreases at several short lags, often within a few hundred milliseconds of T1. At longer lags, second-target accuracy typically recovers toward the comparison level. The exact shape, depth, and onset of this curve depend on stimulus design, masking, and how the targets are selected and reported (Chun & Potter, 1995; Martens & Wyble, 2010).

A figure that depicts a universal fixed-duration “blind period” would misrepresent the evidence. The blink is a probabilistic reduction in report accuracy measured across many trials. It is possible to identify some T2 items during the usual blink window, and the window shifts with experimental conditions.

The Original Experiments and Their Historical Importance

Research on limits to rapid sequential target processing predates the phrase attentional blink. Raymond, Shapiro, and Arnell gave the effect its influential name in 1992 by studying the detection of a second visual probe after identification of an initial target in an RSVP stream. Probe detection was reduced during a brief interval after T1. In their procedure, the impairment was affected by the immediate post-target display and by whether the observer had to identify the first target. This supported a role for interference linked to attention-demanding identification rather than a simple sensory lapse (Raymond et al., 1992).

Chun and Potter extended the investigation to targets defined by categories and proposed a two-stage model of target detection. Their experiments demonstrated impaired report of subsequent targets at short lags and showed how target discriminability and interference after the first target influenced the effect. This helped establish attentional blink as a paradigm for examining the boundary between rapid preliminary processing and more durable target encoding (Chun & Potter, 1995).

The subsequent literature transformed attentional blink into a test case for competing accounts of temporal selection. Each theory must explain the central short-lag deficit and also findings that initially seem inconsistent with a simple bottleneck, especially lag-1 sparing, preserved performance across sequences of consecutive targets, and the roles of distractors and masking (Martens & Wyble, 2010).

Why Can the Second Target Be Missed?

Selection and Consolidation Are Different Operations

A target may be processed sufficiently to activate its sensory features or category without being established in a form that allows a reliable later report. Selecting an item from a rapidly changing sequence involves discriminating it from distractors. Consolidation involves creating or stabilizing a representation that survives the disappearance of the visual input and subsequent interfering stimuli. These operations can compete with the arrival of new information. The evidence suggests that T1 processing can occupy attention-dependent processes crucial for effective selection and report of T2 (Dux & Marois, 2009).

Working memory is relevant because the observer must often report targets after the sequence has ended. Nevertheless, the attentional blink cannot be equated with a fixed “number of slots” in memory. Temporary control settings, distractor competition, temporal integration, masking, and the requirement to individuate two events can all affect performance. Models differ in the priority assigned to each mechanism.

Why Distractors and Masks Matter

An RSVP stream is filled with objects that occur before, between, and after the targets. These items may overwrite, interfere with, or compete against target representations; they also influence whether the attention system treats the stream as a continuous sequence of relevant events or needs to reset its selection policy. Post-target masking played an important role in early experiments, and changing masking or target-distractor discriminability can alter the blink (Raymond et al., 1992; Chun & Potter, 1995).

This is a crucial design consideration: two stimuli separated by 300 milliseconds do not necessarily produce the same blink in every context. When researchers remove intervening distractors, make targets contiguous, or change what counts as a target, they often obtain a different result. The duration between target onsets is therefore one part of the explanation, not a complete one.

Leading Theoretical Explanations

Two-Stage and Capacity-Limited Accounts

The classic two-stage account distinguishes an initial period of relatively rapid perceptual registration from a slower, capacity-limited stage that consolidates selected items for later report. When T1 occupies this later stage, T2 is susceptible to interference before its representation has become stable. This approach predicts particularly pronounced deficits when T1 is difficult to identify or consolidate and when distractors disrupt T2 while central processing is committed elsewhere. The model remains a foundational explanation of why having information initially available does not guarantee successful report (Chun & Potter, 1995).

Capacity-limited theories have been refined to account for the effects of top-down selection, episodic registration, and response demands. A broad review concluded that demands on selection, working-memory encoding, and related central processes together explain a large portion of the evidence, while acknowledging interactions between the bottleneck and control of attention (Dux & Marois, 2009).

Temporary Loss of Attentional Control

A competing family of accounts emphasizes an interruption of the attentional setting that maintains selection of the desired target category. In this view, a distractor arriving at a critical moment can destabilize the current selection policy, producing a temporary failure to admit a later target. Evidence for this account comes from experiments in which sequences of targets can be reported with relatively little blink, while inserting a distractor among them restores the deficit. Such results are difficult to explain using an inflexible one-target-at-a-time capacity rule (Di Lollo et al., 2005).

Olivers, van der Stigchel, and Hulleman found that sparing can extend across several successive targets without intervening distractors. They argued that an overzealous control response to distractors is an important ingredient of the blink and that overall memory limitations can be distinguished experimentally from the blink itself (Olivers et al., 2007). These findings established a serious challenge for simple depletion theories, although they did not render all resource-based explanations obsolete.

Episodic Registration and Temporal Individuation

The episodic simultaneous type/serial token, or eSTST, account addresses how successive visual inputs become distinct events in memory. Its framework permits successive targets to enter a common period of attentional enhancement, while the mechanisms needed to assign targets separate episodic representations can temporarily suppress additional selection. In this account, a cost of keeping events distinct contributes to the blink. The model connects the temporal-selection deficit with target-order errors and the grouping of adjacent targets (Wyble et al., 2009).

Experiments on lag-1 temporal integration further show that participants can combine two adjacent targets into a single perceptual representation, complicating an interpretation of all target-order errors as a race between independent items. These findings demonstrate that good identification of adjacent targets may coexist with imperfect individuation of the sequence (Akyürek et al., 2012).

What Has Been Resolved, and What Remains Debated?

The central behavioral phenomenon is robust in appropriately designed RSVP tasks: identification of a second target is often reduced for particular short temporal separations after processing the first. The relative causal contributions of resource limitations, masks, distractor-triggered control changes, target competition, and episodic binding remain model-dependent. Different task configurations may emphasize different mechanisms. The most productive comparison concerns which theory makes accurate predictions for lag functions, error patterns, no-distractor conditions, and manipulations of task demand, rather than which theory can be summarized by a single metaphor (Dux & Marois, 2009; Martens & Wyble, 2010).

Lag-1 Sparing: Why an Immediately Adjacent Target Can Be Seen

Lag-1 sparing describes the relatively good performance often observed when T2 arrives directly after T1 without an intervening distractor. This is one of the most theoretically informative features of the attentional-blink curve. A theory predicting total unavailability immediately after T1 cannot explain why T2 may be more accessible at lag 1 than at lag 2 or lag 3 (Chun & Potter, 1995).

One class of explanations proposes that selecting T1 briefly opens an episode during which additional target-like information can also be processed. A later intervening distractor can interrupt this continuity or invoke selection control that penalizes T2. Other accounts emphasize temporal integration, in which adjacent target representations may merge or become difficult to assign separate positions. Lag-1 sparing is thus a clue to how target selection unfolds in time; it is not proof of a limitless burst of perceptual capacity (Olivers et al., 2007; Akyürek et al., 2012).

Some experimental arrangements produce limited or absent lag-1 sparing. Its presence should be tested rather than assumed. Differences in spatial location, target features, masks, and task requirements influence whether the observer can process targets as part of one continuous attentional episode.

What Changes the Size of the Blink?

Difficulty and Processing Requirements of T1

Increasing the difficulty of identifying or consolidating T1 can sometimes make subsequent T2 reporting less successful. This is compatible with the idea that the first target places demands on processing steps needed again for a second target. Yet T1 difficulty can alter selection settings, baseline accuracy, or other features simultaneously, so a larger deficit by itself does not uniquely establish one resource bottleneck (Chun & Potter, 1995; Dux & Marois, 2009).

Target–Distractor Similarity and Masking

Targets that differ clearly from distractors can be easier to identify, while strongly interfering distractors may disrupt target report. Experimental results depend on whether the critical manipulation changes visual similarity, semantic category, the processing of the first target, or the selection of the second. The interference produced by surrounding stimuli is part of the phenomenon's experimental logic, not simply background noise (Raymond et al., 1992; Di Lollo et al., 2005).

The Continuity of Relevant Targets

When several targets appear consecutively, investigators may observe extended sparing of target report. An intervening distractor can sharply change the result. This pattern shows that how the system classifies the stream matters alongside how quickly the stimuli arrive. Investigators compare such conditions to determine whether attentional control is configured to maintain or interrupt selection (Olivers et al., 2007).

Motivation, Meaning, and Experimental Context

Studies sometimes investigate emotional significance, salience, practice, or attentional strategy, but effects from a particular category of stimulus cannot be generalized across all categories and settings. For example, recent work examining whether odors alter an RSVP attentional-blink effect reported no evidence for the proposed olfactory influence in its tested conditions. This illustrates the importance of publishing null as well as positive findings, defining the tested modality precisely, and distinguishing a specific manipulation from a general law (Hackländer et al., 2025).

Does an Unreported Target Reach Perception or Awareness?

Attentional-blink research helped motivate a distinction among early sensory registration, semantic or categorical processing, stable encoding, and conscious access. Those processes are not synonymous. A target can influence later responses even when it is not correctly reported, but the strength of that inference depends on controls for guessing, visibility, and response bias. Conversely, failure to name a target accurately cannot by itself prove that no processing took place at any stage (Dux & Marois, 2009; Martens & Wyble, 2010).

Researchers therefore ask a series of distinguishable questions: Was T2 physically displayed? Did sensory or category-specific processing occur? Was its identity available briefly? Did it receive sufficient attention to be consolidated? Can the observer accurately report or recognize it later? A given RSVP experiment may answer some of these questions while leaving others open. Claims about subjective awareness require evidence beyond the observation of a missed target in a forced-choice or report task.

Experimental Methods, Reliability, and Validity

Choosing the Proper Comparison Condition

Researchers need to consider single-target trials, long-lag baseline conditions, and T1 report accuracy. If a manipulation changes both targets' baseline discriminability, its effect on conditional T2 report may not reflect a change in the blink mechanism. Stream speed, display duration, visual masking, position of T1, and task instructions should be specified because each can change the measured function. Sampling only one short lag obscures whether the timing, depth, or recovery of the blink has shifted.

Accuracy, Error Patterns, and Statistical Modeling

The usual outcome is a proportion of correct T2 reports conditional on correct T1. Analyses may additionally examine T1 omissions, target confusions, order reversals, and response biases. Appropriate mixed-effects or participant-level analyses can model repeated trials and uncertainty; reliability improves with adequate trials across the relevant lag conditions. Investigators should report uncertainty and the complete time course rather than highlighting a single extreme lag value.

A shallow blink might arise because T2 selection is efficient, because the stimuli are unusually easy, because T1 processing failed, or because task-specific conditions suppress the effect. A steep blink might reflect a demanding manipulation rather than a stable property of the participant. These alternatives make task-based diagnosis inappropriate. Reproducibility depends on documenting timing, masking, response instructions, exclusion rules, and trial counts as carefully as the main statistical result.

Generalizability Beyond RSVP

RSVP controls stimulus timing in ways that ordinary vision rarely does. In everyday environments, relevant events often occupy different locations and can be revisited with eye movements; people can also use contextual cues and repeated observations. A laboratory attentional blink demonstrates a temporal constraint under defined conditions, not a universal rule that every second event presented within half a second will be missed. Extrapolations to driving, monitoring, or digital behavior require task-specific evidence of similar demands, not an analogy alone.

Attentional Blink and Other Failures of Attention

Attentional Blink Versus Inattentional Blindness

Inattentional blindness concerns failure to notice an unexpected, often salient event while attention is engaged elsewhere. Attentional blink concerns diminished report of a second designated target after processing a first target in a fast temporal sequence. Both illustrate selective-attention limitations, but their paradigms and precise reader questions differ. Failure to see an unexpected object in a scene is not automatically an attentional blink.

Attentional Blink Versus Change Blindness

Change blindness concerns the difficulty of detecting a change between scenes or views, especially when disruption impedes the cues that would otherwise signal the change. Attentional blink concerns the report of a second target as a function of temporal separation from T1. A person missing a scene alteration and a person missing a second RSVP target may both have a perceptual report failure, but the experiments manipulate different structures of information.

Attentional Blink Versus Dual-Task Interference

Dual-task interference concerns performance costs when tasks overlap, such as delayed responses or slower processing. Attentional blink is usually measured as an accuracy deficit in reporting a second stimulus in a rapid visual stream. Models can share concepts of limited central processing, but an RSVP lag curve is not interchangeable with the response-time cost of switching tasks.

Attentional Blink Versus a Physical Eye Blink

An eye blink briefly covers or alters visual input through the eyelids. Attentional blink can occur during maintained visual fixation with no physical blink. The term describes a change in the probability of accurately reporting information, not an observed closure of the eyes.

What the Findings Mean for Daily Life

The broad lesson is that detecting one important event and reporting another immediately afterward can place distinct demands on temporal selection. This can help researchers formulate hypotheses about performance in fast monitoring contexts. Applying the finding to real interfaces, training programs, or safety-critical work requires independent research that measures those contexts directly. There is no established universal “attentional blink test” that ranks an individual's daily attentiveness or reliably predicts their professional performance.

Claims that a particular app, game, meditation routine, supplement, or cognitive-training exercise will eliminate the blink and produce general improvements in attention need high-quality intervention studies and evidence of transfer to meaningful outcomes. A change on the same task used for training can reflect improved familiarity with that task. It cannot automatically be treated as durable improvement in broad cognitive control.

Frequently Asked Questions

How long does the attentional blink last?

In many classic RSVP conditions, reduced second-target accuracy is most prominent when T2 occurs roughly 200–500 milliseconds after T1. The exact curve differs with trial timing, masks, target categories, and task requirements. It describes a probability pattern averaged across trials, not a mandatory half-second interval of total blindness (Dux & Marois, 2009).

Why is the second target sometimes seen when it follows immediately after the first?

This common result is called lag-1 sparing. Adjacent targets can be selected within an extended attentional episode or temporally integrated, depending on the task. An intervening distractor may alter the selection state. Lag-1 sparing is an important constraint on theoretical models (Olivers et al., 2007; Akyürek et al., 2012).

Does the brain completely fail to see the second target?

The task demonstrates reduced accurate report. It does not alone identify the exact stage at which representation of T2 becomes unavailable. Investigators use additional behavioral and physiological methods to distinguish early processing, consolidation, and report. Theories disagree about the relative role of these processes (Martens & Wyble, 2010).

Is an attentional blink proof of ADHD or another disorder?

No. It is a common experimental effect observed in people without a clinical diagnosis. A diagnosis requires its own validated clinical process and functional criteria. One RSVP score cannot establish a psychiatric or neurological condition.

Is the attentional blink the same as multitasking?

No. The effect is typically measured while observers try to report two targets in a single rapid visual stream. Multitasking refers to broader situations in which more than one task or goal competes for processing. Some underlying limits may be related, but the terms describe distinct research questions.

Can practice eliminate the attentional blink?

Practice and task familiarity may change performance under some experimental conditions. That does not establish lasting elimination of the phenomenon or broad transfer to everyday attention. Claims about training need controlled comparisons, durable follow-up, and meaningful outcomes beyond the trained task.

Scientific Conclusion

Attentional blink reveals a temporally structured limitation in reporting rapidly successive visual targets. Its most important findings combine the short-lag T2 deficit with lag-1 sparing and sensitivity to distractors, target difficulty, masking, and the continuity of target selection. The experimental record supports the interaction of attention-dependent selection and encoding processes while leaving room for competing computational interpretations. The phenomenon remains valuable because it can test specific mechanisms of temporal attention using clearly defined manipulations, measurable errors, and replicable lag functions.

Related Articles

Attention in the Age of AI: Focus, Cognitive Load, and Algorithmic Competition examines digital-era competition for attention, a distinct applied topic. Task Switching and Digital Multitasking: Why Constant Switching Drains Focus discusses interference between changing task goals rather than identification of two targets within RSVP.

Attention: Types, Mechanisms, and Limits — Explains attention as selective processing, its types, mechanisms, and limits.

Change Blindness: Why We Miss Changes in Visual Scenes — Explains why people overlook scene changes and how change detection differs from inattentional blindness.

References

Akyürek, E. G., Eshuis, S. A. H., Nieuwenstein, M. R., Saija, J. D., Başkent, D., & Hommel, B. (2012). Temporal target integration underlies performance at lag 1 in the attentional blink. Journal of Experimental Psychology: Human Perception and Performance, 38(6), 1448–1464. https://doi.org/10.1037/a0027610.

Chun, M. M., & Potter, M. C. (1995). A two-stage model for multiple target detection in rapid serial visual presentation. Journal of Experimental Psychology: Human Perception and Performance, 21(1), 109–127. https://doi.org/10.1037/0096-1523.21.1.109.

Di Lollo, V., Kawahara, J., Ghorashi, S. M. S., & Enns, J. T. (2005). The attentional blink: Resource depletion or temporary loss of control? Psychological Research, 69(3), 191–200. https://doi.org/10.1007/s00426-004-0173-x.

Dux, P. E., & Marois, R. (2009). The attentional blink: A review of data and theory. Attention, Perception, & Psychophysics, 71(8), 1683–1700. https://doi.org/10.3758/APP.71.8.1683.

Hackländer, R., Baess, P., & Bermeitinger, C. (2025). Smells like … no evidence that odors influence the attentional blink. Attention, Perception, & Psychophysics, 87, 458–479. https://doi.org/10.3758/s13414-024-02986-4.

Martens, S., & Wyble, B. (2010). The attentional blink: Past, present, and future of a blind spot in perceptual awareness. Neuroscience & Biobehavioral Reviews, 34(6), 947–957. https://doi.org/10.1016/j.neubiorev.2009.12.005.

Olivers, C. N. L., van der Stigchel, S., & Hulleman, J. (2007). Spreading the sparing: Against a limited-capacity account of the attentional blink. Psychological Research, 71(2), 126–139. https://doi.org/10.1007/s00426-005-0029-z.

Raymond, J. E., Shapiro, K. L., & Arnell, K. M. (1992). Temporary suppression of visual processing in an RSVP task: An attentional blink? Journal of Experimental Psychology: Human Perception and Performance, 18(3), 849–860. https://doi.org/10.1037/0096-1523.18.3.849.

Wyble, B., Bowman, H., & Nieuwenstein, M. (2009). The attentional blink provides episodic distinctiveness: Sparing at a cost. Journal of Experimental Psychology: Human Perception and Performance, 35(3), 787–807. https://doi.org/10.1037/a0013902.