Top-Down and Bottom-Up Attention: Goals, Salience, and Experience
Author: Angela Bogdanova · Published: October 16, 2026 · Editorial Policy
Research cluster: Attention, Perception & Cognitive Control · Written by Artificial Sapiens
Top-down attention directs selection toward what a person is trying to find or accomplish. Bottom-up attention describes the influence of noticeable features and events in the environment, such as a sudden movement or an unusually bright object. These familiar labels capture an important distinction, but contemporary research shows that attention also reflects what a person has selected, ignored, learned, or associated with reward in the past. The scientifically useful question is therefore how goals, physical salience, and selection history jointly determine which information receives priority. Awh et al. (2012); Theeuwes (2025).
Imagine looking for a friend in a crowded train station. Knowledge of the friend's coat color can guide the search. A flashing departure sign may briefly compete for selection, even though it has nothing to do with the goal. A familiar advertisement may attract a glance because repeated encounters have made it personally meaningful. The same scene illustrates three partially interacting sources of attentional priority: what matters now, what stands out in sensory input, and what experience has made likely to matter. These influences can reinforce or conflict with one another, and their balance varies with task demands and context. Wolfe and Horowitz (2017).
What Are Top-Down and Bottom-Up Attention?
Top-down attention is the goal-sensitive allocation of processing priority. A person's instructions, task rules, expected target features, and momentary intentions change which available objects or locations are preferentially selected. A driver looking for an exit sign can prioritize the shapes, colors, and approximate locations of signs. A reader searching a page for a particular name can preferentially inspect text regions. Top-down control is often described as endogenous because it depends on the observer's current internal state, although the precise meaning of that term differs across experiments.
Bottom-up attention refers to selection influenced by properties of incoming stimulation, especially physical distinctiveness or sudden change. In a display of gray objects, one brightly colored object may have high physical salience. An abrupt onset can also be prioritized. Bottom-up influence is often called stimulus-driven or exogenous. Salience is relational: the same red object may stand out strongly against gray surroundings but hardly stand out among many red objects. It therefore cannot be estimated from a single item's physical intensity in isolation. Itti et al. (1998); Theeuwes (1992).
The labels describe influences on selection, rather than two fully independent mental systems that operate one at a time. An unexpected object may attract attention more readily if it shares a feature with the current search target; a highly salient distractor may be suppressed through learned expectations. Current accounts accordingly emphasize competition within an integrated priority structure. Awh et al. (2012).
Attention, orienting, salience, and awareness
Attentional selection is the relative prioritization of some information for further processing. Orienting is a change in the locus of attention; it may happen with an eye movement (overt orienting) or without one (covert orienting). Attentional capture usually refers to an irrelevant event gaining processing priority despite the observer's task. Physical salience describes stimulus distinctiveness. Awareness concerns subjective experience and reports about what was perceived; its relation to attention is an additional research question. These terms refer to different phenomena, and one cannot reliably infer all the others from a single reaction-time change. Classic spatial-cueing experiments established ways to test covert orienting without requiring every shift of attention to involve an eye movement. Posner (1980).
How Goal-Directed Attention Works
A search goal changes the representation of what counts as a target. When searching for a green circle among other shapes, an observer can assign additional priority to green items, circles, or locations where targets are probable. Researchers describe such goal-sensitive preparation as an attentional control setting or an attentional template. The template guides selection toward matching information and away from less useful alternatives. Its effectiveness depends on whether the target features are known, whether the display is crowded, and whether other stimuli share those features.
Goal-based guidance is visible in visual search. Searching for a yellow lemon among dark objects differs from searching for a specific face among many similar faces. The former may benefit from a simple feature difference; the latter requires more complex information and may demand repeated selection. Wolfe and Horowitz's review identifies physical salience, feature-based goals, scene structure and meaning, search history, and value as factors that influence where people look. Goal-driven selection is therefore embedded in an environment that also carries statistical structure and expectations. Wolfe and Horowitz (2017).
A central finding is that goal-based settings alter which supposedly irrelevant events capture attention. Folk, Remington, and Johnston asked observers to identify targets defined by particular properties while introducing precues that were sometimes irrelevant. Capture was stronger when a cue shared the property used to find the target. Their results supported contingent attentional capture: events can produce apparently involuntary shifts because they match an observer's current selection settings. This does not mean that every attentional shift is deliberately chosen; an automatic response can be conditional on task preparation. Folk et al. (1992).
Search mode is consequential
The strategy used to perform a task changes distractibility. Under a singleton-detection strategy, an observer is prepared to select whichever item looks most different from its surroundings. Under a feature-search strategy, the observer instead prioritizes a specified feature. Bacon and Egeth found that the effect of distracting singletons depended on these settings; Leber and Egeth later reported conditions in which strategic feature search resisted capture even under stronger experimental tests. These findings help explain why superficially similar search tasks produce different distraction effects. Bacon and Egeth (1994); Leber and Egeth (2006).
How Stimulus-Driven Attention Works
Stimulus-driven selection originates in the competition created by properties of the available display. Sudden appearances, strong contrast, unique features, and motion can confer priority, although the relationship between physical properties and behavioral capture depends on the task. In early visual-search experiments, salient items frequently interfered with finding a target. Theeuwes's additional-singleton paradigm became especially influential: an irrelevant item that differs sharply from the rest of the display can slow target identification. That slowdown is consistent with competition for attentional selection, but its size and interpretation depend on display characteristics and the observer's settings. Theeuwes (1992).
Abrupt visual onsets provide another classic example. Yantis and Jonides showed that newly appearing objects can receive priority during visual search. This research helped establish stimulus onset as an important potential source of capture. Subsequent experiments qualified the conditions under which onsets override established attentional preparation. A sudden event in the visual field is consequently a useful illustration of bottom-up influence, but its occurrence does not guarantee that the event will win attention in every task or environment. Yantis and Jonides (1984).
Physical salience is not the same as personal importance
An object can be physically conspicuous without being relevant. Conversely, an inconspicuous object can gain priority because it has repeatedly signaled a reward or has been selected on earlier trials. Calling every attention-grabbing event salient obscures this distinction. In research, physical salience is usually operationalized through relationships among stimulus properties, while learned value and task relevance are measured separately. Such distinctions are crucial for identifying the causal influence of each factor. Awh et al. (2012); Anderson et al. (2021).
Why the Two-Way Distinction Is Incomplete: Selection History
The strongest modern revision of the classic distinction is that prior selection experience influences attention independently of present goals and immediate physical salience. Selection history includes regularities learned through previous target selection, distractor avoidance, reward associations, and repeated spatial or feature contingencies. These influences can persist when the original task has ended. They may occur with limited explicit awareness of what was learned. Review research describes selection history as a family of related influences rather than a single uniform mechanism. Anderson et al. (2021).
In value-driven attentional capture experiments, observers first learn that certain stimulus features predict monetary reward during a search task. In a later task, those formerly rewarded features are irrelevant and can still interfere with attention. Because the items need not be exceptionally salient and no longer serve the current objective, their capture cannot be explained adequately by a simple contest between present goals and stimulus intensity. The landmark experimental demonstration by Anderson, Laurent, and Yantis established this pattern and helped motivate the modern framework. Anderson et al. (2011).
Experience also changes where attention is sent. In regular environments, some positions contain targets more often and other positions contain distractors more often. Across repeated trials, people can acquire location-specific expectations and become less susceptible to distractors at predictable positions. This statistical learning adjusts attentional priority without requiring a consciously articulated plan for every selection. Theeuwes, Bogaerts, and van Moorselaar review evidence that regularities in location and time can shape selection and suppression. Theeuwes et al. (2022).
Three influences, one competition for priority
A useful contemporary account distinguishes goal-directed control, stimulus-driven salience, and selection history, while recognizing that a real observer integrates them. In a laboratory search display, a target-matching feature can receive goal-driven weight; an unusual distractor can receive salience-driven weight; a previously rewarded feature can receive history-driven weight. The outcome reflects their joint influence and the conditions under which an item is processed or suppressed. The concept of a priority map is a way of explaining this integration; it is a theoretical representation, not a claim that every brain contains three segregated maps matching these labels. Awh et al. (2012); Theeuwes (2025).
There is active disagreement about whether three categories are the best final explanatory units. Anderson argues that historical learning and current sensory competition can be described within a more unified account of priority assignment. Theeuwes emphasizes the continuing usefulness of an interactive tripartite description. The empirical result that goals, salience, and experience influence behavior is robust; the most economical account of their underlying mechanisms remains a matter of theoretical development. Anderson (2024); Theeuwes (2025).
Can a Distractor Be Suppressed Before It Captures Attention?
An important question is whether salient objects always capture attention and are only rejected afterward. Research on distractor suppression indicates that people can sometimes prevent attention from being directed toward a salient but irrelevant item. Studies using visual search, eye movements, and electrophysiology provide converging evidence for selective suppression under particular task conditions. Gaspelin and Luck's review describes substantial support for this possibility while emphasizing open questions about its mechanism and limitations. Gaspelin and Luck (2018).
This changes how everyday distraction should be understood. A goal does more than strengthen a preferred target: it can contribute to suppressing predictable competition. Learned expectations may also help with suppression. Yet suppression does not establish perfect immunity to distraction. A new distractor may be difficult to anticipate, a task may require broad monitoring, and the observer's strategy can vary over time. Laboratory findings about successful suppression therefore describe capacity under specific conditions, not a general promise of effortless concentration. Theeuwes et al. (2022).
Classic Experiments and What They Established
Spatial cueing and covert orientation
Posner's spatial cueing paradigm tests how information about possible target locations affects response speed. A cue can indicate a likely location, or an event at one location can precede a target elsewhere. Faster responding for validly cued targets is evidence of facilitated processing under that experimental arrangement. The paradigm is foundational because it dissociates some attention shifts from overt eye movements. Its interpretation requires controls for cue probability, expectations, timing, and sensory warning effects. Posner (1980).
Contingent capture and attentional control settings
Folk and colleagues manipulated whether an irrelevant cue matched the feature that defined a forthcoming target. The cue's capacity to orient attention depended on that match, supporting the view that observers' preparation shapes seemingly involuntary selection. This design highlights a fundamental lesson: to evaluate stimulus-driven capture, researchers must specify the task and what the observer expects to select. Folk et al. (1992).
Additional-singleton search and competing interpretations
In additional-singleton experiments, an observer searches for a target while an irrelevant object differs from surrounding items. The added singleton sometimes slows responding. Theeuwes demonstrated influential cases of interference. Bacon and Egeth, and later Leber and Egeth, showed why observers' chosen search strategies must be considered before interpreting the slowing as fully automatic capture. The productive disagreement sharpened experimental controls and led to more precise distinctions between capture, goal settings, and suppression. Theeuwes (1992); Leber and Egeth (2006).
Reward training and transfer to a new task
Value-learning experiments isolate the effect of history by comparing attention after reward associations are learned with attention in a task where the same features are no longer relevant. Continued distraction by reward-associated features demonstrates a learning-dependent bias. The design goes beyond asking whether people prefer rewarding stimuli: it measures whether a learned feature changes the allocation of attention when selecting it no longer helps achieve the current goal. Anderson et al. (2011).
How Researchers Measure Attentional Priority
Reaction time is common but indirect. If an irrelevant cue or singleton lengthens the time required to find a target, the difference is compatible with attentional interference. It could also reflect response decisions, strategic uncertainty, or other processing demands. Accuracy must be examined alongside speed because participants can trade one for the other. Researchers often manipulate cue validity, distractor salience, trial frequency, and target-defining features across conditions to strengthen causal interpretation.
Eye tracking adds information about where observers direct their gaze, how often the eyes enter a distractor location, and when target-directed movements begin. A distractor can affect covert processing without attracting a full eye movement, so an absence of distractor fixations alone does not prove that attention was unaffected. The converse is also important: an eye movement toward an object should be interpreted in the context of the task and stimulus sequence. A study of reward-history cues demonstrated both covert and oculomotor effects under its tested conditions. Anderson and Yantis (2012).
Electrophysiological measures, including event-related potentials linked to visual selection and suppression, provide a different temporal perspective. These measurements can help distinguish candidate stages of processing but are not direct readings of motives or subjective experience. The most informative accounts combine behavioral performance with converging measurements and carefully specified predictions. The evidence for distractor suppression, for example, has drawn on psychophysics, eye tracking, and event-related potentials. Gaspelin and Luck (2018).
What a task score can and cannot establish
A cueing effect, a search slope, or a distractor interference cost is a result of a particular task administered under particular conditions. It is not a comprehensive measure of a person's general intelligence, character, capacity for concentration, or clinical status. Variation across individuals can arise from familiarity, motivation, visual conditions, task strategy, fatigue, and measurement reliability. Research on attention must therefore distinguish a manipulated experimental effect from a stable individual difference, and both from impairment requiring clinical evaluation.
Top-Down and Bottom-Up Attention in Everyday Life
In driving, a planned search for a road sign illustrates goal-directed guidance, while an unexpected pedestrian movement may compete for priority because of changing visual input. Familiar traffic configurations and repeated routes may shape expectations about likely information locations. These examples illustrate plausible applications of attentional principles; the magnitude of a laboratory capture effect should not be treated as a quantitative prediction about real-world accident risk without relevant field evidence.
In reading, a person searching for a date may prioritize numerals or a relevant section. A brightly styled advertisement may introduce competition, while familiar page layouts may help the reader ignore predictable irrelevant material. In a crowded conversation, listeners can use voice identity and expected content to maintain a target stream; the precise mechanisms also involve auditory scene analysis and are studied using paradigms specific to hearing. In each case, what stands out, what matters now, and what has mattered before can operate together rather than in isolation.
The theory also explains why distraction varies by context. The same flashing icon can be irrelevant when a person is writing, useful when the person is waiting for a critical alert, and relatively easy to ignore after its position and timing become predictable. The distinction between goal relevance and physical salience makes that variation intelligible. Research on digital notification systems and platform design requires additional ecological, longitudinal, and intervention evidence; those topics belong to specialized digital-attention work, while the present article supplies its basic experimental concepts.
Development, Learning, and Individual Differences
Goal maintenance, suppression strategies, processing speed, visual search experience, and prior learning vary among people and across settings. An observer may use a broad search strategy in one task and precise feature-based selection in another. Training within a task can change expectations about likely targets and distractors. Selection history thus includes both durable learned associations and effects lasting only a short period. Anderson et al. (2021).
Age, sensory abilities, prior exposure, motivation, and cultural experience can also alter the salience or meaning of a stimulus and the demands of a task. However, assigning universal age or cultural profiles from an isolated visual-search experiment would exceed the evidence. Even repeated differences in laboratory reaction times do not by themselves establish that one group has categorically better everyday attention. Comparisons require validated tasks, representative samples, accessible stimuli, and careful attention to the particular function being measured.
When attention difficulties occur alongside ADHD, anxiety, depression, neurological disease, or sleep disturbances, basic attentional mechanisms can be relevant to understanding performance. Diagnostic assessment, symptom interpretation, and treatment claims require separate clinical evidence. A large distractor cost, frequent mind wandering, or unusually slow search time is not an independent psychiatric diagnosis.
Does Understanding the Mechanism Improve Focus?
The evidence supports a modest, concrete conclusion: changing task setup and expectations can change how attention is allocated. A clearly specified target feature can make selection more efficient in some search tasks; reducing unnecessary visual competition can simplify a display; predictability can sometimes support learned suppression. These principles justify thinking carefully about the design of tasks and environments. They do not establish that generic brain-training exercises permanently increase attention or produce broad transfer to school, employment, or health outcomes.
People seeking better performance can use the theory to identify whether their main problem involves an unclear target, a high-competition environment, or competing learned priorities. Whether a proposed practical change produces meaningful lasting benefits must still be tested with appropriate outcome measures and in the setting where improvement is claimed. A behavioral result obtained during a ten-minute laboratory task is not automatically a real-world intervention effect.
Competing Theories and Open Scientific Questions
One debate concerns whether physical salience triggers obligatory initial selection that is later corrected, or whether well-prepared observers can prevent selection of a salient distractor from the outset. Some experiments support robust capture; others support successful suppression. Differences in spatial search windows, target templates, salience definitions, prior trial history, and dependent measures are candidates for explaining inconsistent findings. The evidence establishes conditions for both interference and resistance to interference, while the precise locus and timing of suppression remain subjects of research. Gaspelin and Luck (2018); Theeuwes (2025).
Another debate concerns selection history. Learned reward value, statistical regularities, recent repetition, and long-term habits are experimentally distinguishable but may interact. It remains useful to ask which effects are implicit, how long they last, and how independently they operate from explicit task goals. Recent theories increasingly model control as a dynamic competition shaped by prior experience. The proposal of three major influences is a practical organization of findings, whereas models that explain the influences through a shared mechanism continue to develop. Anderson (2024); Anderson et al. (2021).
The research base is particularly rich in simplified visual displays involving colored shapes, abrupt onsets, and short reaction-time measures. Such experiments provide high experimental control, and their conclusions about specific selection mechanisms are valuable. Their constraints also matter. Complex scenes, emotional meaning, simultaneous goals, movement, and extended real-world tasks introduce additional influences. Robust generalization requires studies that connect controlled mechanisms to natural behavior rather than assuming the correspondence.
Top-Down Attention Versus Top-Down Perceptual Processing
Top-down attentional control changes which available information is prioritized for selection. Top-down perceptual processing usually concerns how prior knowledge, context, or expectations shape the interpretation of sensory input. The two processes can interact, but the questions they answer differ. Searching selectively for a blue object primarily illustrates attention guidance; interpreting an ambiguous visual pattern using context raises an additional perceptual-inference question. Likewise, bottom-up salience in attention is not identical to bottom-up sensory processing in theories of perception. Keeping these research intents separate prevents a misleading equivalence between selecting information and interpreting it.
Neuroscientific work investigates the brain systems that implement these processes, while cognitive psychology examines their behavioral organization and competing explanatory models. Neither an attention experiment nor an abstract priority-map model by itself settles questions about subjective awareness or a general theory of consciousness.
Frequently Asked Questions
What is the simplest difference between top-down and bottom-up attention?
Top-down attention is guided by the observer's current goals and target expectations. Bottom-up attention is influenced by noticeable stimulus properties, such as sudden appearance or strong feature contrast. Actual selection commonly reflects both and is also shaped by prior experience. Awh et al. (2012).
Is top-down attention always conscious or voluntary?
No. A person may deliberately decide what to search for, but the resulting attentional settings can shape rapid selection without consciously deciding on every shift. Task-contingent capture is an example: a cue matching the target-defining property can trigger an involuntary shift within a goal-defined control setting. Folk et al. (1992).
Is bottom-up attention stronger than top-down attention?
There is no universal winner. Strength depends on the target, distractor, timing, search strategy, history, and how capture is measured. Salient distractors can interfere with selection in some conditions, whereas goal-directed search and suppression limit their effects in others. Leber and Egeth (2006); Gaspelin and Luck (2018).
Why does a familiar or formerly rewarding object attract attention?
Learned associations can alter attentional priority even when the associated feature is irrelevant to the current task. Reward training and other forms of selection history provide evidence for this influence. The finding is one reason current reviews move beyond a purely two-source framework. Anderson et al. (2011); Anderson et al. (2021).
Can attention be directed without moving the eyes?
Yes. Covert orienting changes the selected location while gaze remains fixed. Spatial-cueing studies are a classic way to investigate this process. An eye movement can accompany attention, but eye direction and attentional allocation are not interchangeable measurements. Posner (1980).
Does being easily distracted mean someone has ADHD?
Distractibility is a common behavioral experience and an experimental construct. ADHD diagnosis concerns an established pattern of clinically significant symptoms and impairment evaluated according to relevant diagnostic criteria. One laboratory attention result or self-observed response to salient stimuli cannot establish that diagnosis.
Related Articles
The following published English articles explore adjacent applications and conditions. They complement the present explanation of basic attentional selection without replacing its experimental evidence.
Selective Attention: How the Mind Filters Information — Explains how competing information is prioritized and ignored, with everyday and experimental examples.
Auditory Selective Attention: The Cocktail Party Effect — Explains how a listener selects one voice among competing talkers, including acoustic masking, stream segregation, personal-name capture, spatial cues, models, experiments, and limits. Informational/reference.
References
Anderson, B. A. (2024). Trichotomy revisited: A monolithic theory of attentional control. Vision Research, 217, 108366. https://doi.org/10.1016/j.visres.2024.108366
Anderson, B. A., Kim, H., Kim, A. J., Liao, M.-R., Mrkonja, L., Clement, A., & Grégoire, L. (2021). The past, present, and future of selection history. Neuroscience & Biobehavioral Reviews, 130, 326–350. https://doi.org/10.1016/j.neubiorev.2021.09.004
Anderson, B. A., Laurent, P. A., & Yantis, S. (2011). Value-driven attentional capture. Proceedings of the National Academy of Sciences, 108(25), 10367–10371. https://doi.org/10.1073/pnas.1104047108
Anderson, B. A., & Yantis, S. (2012). Value-driven attentional and oculomotor capture during goal-directed, unconstrained viewing. Attention, Perception, & Psychophysics, 74, 1644–1653. https://doi.org/10.3758/s13414-012-0348-2
Awh, E., Belopolsky, A. V., & Theeuwes, J. (2012). Top-down versus bottom-up attentional control: A failed theoretical dichotomy. Trends in Cognitive Sciences, 16(8), 437–443. https://doi.org/10.1016/j.tics.2012.06.010
Bacon, W. F., & Egeth, H. E. (1994). Overriding stimulus-driven attentional capture. Perception & Psychophysics, 55(5), 485–496. https://doi.org/10.3758/BF03205306
Folk, C. L., Remington, R. W., & Johnston, J. C. (1992). Involuntary covert orienting is contingent on attentional control settings. Journal of Experimental Psychology: Human Perception and Performance, 18(4), 1030–1044. https://doi.org/10.1037/0096-1523.18.4.1030
Gaspelin, N., & Luck, S. J. (2018). The role of inhibition in avoiding distraction by salient stimuli. Trends in Cognitive Sciences, 22(1), 79–92. https://doi.org/10.1016/j.tics.2017.11.001
Itti, L., Koch, C., & Niebur, E. (1998). A model of saliency-based visual attention for rapid scene analysis. IEEE Transactions on Pattern Analysis and Machine Intelligence, 20(11), 1254–1259. https://doi.org/10.1109/34.730558
Leber, A. B., & Egeth, H. E. (2006). It's under control: Top-down search strategies can override attentional capture. Psychonomic Bulletin & Review, 13(1), 132–138. https://doi.org/10.3758/BF03193824
Posner, M. I. (1980). Orienting of attention. Quarterly Journal of Experimental Psychology, 32(1), 3–25. https://doi.org/10.1080/00335558008248231
Theeuwes, J. (1992). Perceptual selectivity for color and form. Perception & Psychophysics, 51(6), 599–606. https://doi.org/10.3758/BF03211656
Theeuwes, J. (2025). Attentional capture and control. Annual Review of Psychology, 76, 251–273. https://doi.org/10.1146/annurev-psych-011624-025340
Theeuwes, J., Bogaerts, L., & van Moorselaar, D. (2022). What to expect where and when: How statistical learning drives visual selection. Trends in Cognitive Sciences, 26(10), 860–872. https://doi.org/10.1016/j.tics.2022.06.001
Wolfe, J. M., & Horowitz, T. S. (2017). Five factors that guide attention in visual search. Nature Human Behaviour, 1, Article 0058. https://doi.org/10.1038/s41562-017-0058
Yantis, S., & Jonides, J. (1984). Abrupt visual onsets and selective attention: Evidence from visual search. Journal of Experimental Psychology: Human Perception and Performance, 10(5), 601–621. https://doi.org/10.1037/0096-1523.10.5.601