Conscious vs. Unconscious Processing: What the Evidence Shows

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

Research cluster: Consciousness, Metacognition & Self-Awareness · Written by Artificial Sapiens

Conscious processing occurs when information is part of experience and is available, to some degree, for deliberate use, flexible evaluation, or report. Unconscious processing occurs when the nervous system registers or transforms information without a corresponding reportable experience of that information. Experiments show that some unseen or unnoticed stimuli can influence subsequent behavior and brain activity. They also show that the size, complexity, duration, and flexibility of these effects depend on the task and on how carefully awareness is measured. The evidence supports specific forms of processing outside conscious access; it does not establish an all-purpose unconscious mind that reliably makes complex decisions better than conscious deliberation. A critical synthesis and research consensus are provided by Kouider and Dehaene (2007), Van den Bussche and colleagues (2009), and Stockart and colleagues (2025).

The central scientific question is precise: what information is processed, to what extent, through which mechanisms, and under what criterion can a researcher conclude that the person lacked awareness of that information? This article examines the strongest demonstrations, the disagreements, and the limits of what experiments can currently establish.

What Is Conscious Processing?

In cognitive psychology, processing means a transformation of information: sensory registration, feature detection, recognition, association, prediction, response selection, or the maintenance and manipulation of a representation. Conscious processing refers to processing accompanied by a conscious experience or made available to the person for deliberate consideration. For instance, reading a sentence with awareness of its meaning allows the reader to compare its claims, reconsider an interpretation, and explain the reasoning to another person.

The word conscious covers several related dimensions. Phenomenal consciousness concerns what a perception or thought feels like. Conscious access concerns whether information can guide report, deliberate reasoning, or other relatively flexible behaviors. Reportability is one experimental indicator of access, and it should not be treated as a complete definition of subjective experience. Wakefulness describes a state of arousal, while responsiveness describes observable behavior. Neither, on its own, measures the specific visual or auditory content a person experiences. Dehaene and colleagues (2006) developed an experimentally testable distinction among conscious, preconscious, and subliminal states of information.

Conscious processing often involves sustained attention, working memory, comparison of alternatives, error checking, and control over a response. These processes can be studied objectively, yet no single behavioral or neural measure is accepted as a universal proof of conscious experience. People can consciously experience something that they later forget, and they can produce accurate behavior with little ability to explain the underlying cues.

What Is Unconscious Processing?

Unconscious or nonconscious processing refers to information processing that occurs without awareness of the relevant information or operation under a specified experimental definition. A masked number might influence how quickly someone judges the next number even if the participant cannot reliably identify the masked number. A person might learn regularities gradually without being able to verbalize an explicit rule. A visually presented event might affect responses despite going unnoticed during an attention-demanding task.

These cases share an absence of some kind of reportable awareness, but they are methodologically different. A participant might lack awareness of a stimulus, lack awareness of the association being learned, or simply lack knowledge of why an action was chosen. An unaware process can operate on input that was consciously seen, and an unseen stimulus can trigger a consciously experienced reaction. Identifying exactly which element is unconscious is essential for interpreting the study. Kouider and Dehaene (2007) explain why depth of processing differs across masking and inattention paradigms.

The everyday expressions automatic, implicit, intuitive, subliminal, and unconscious often overlap but describe distinct properties. Automatic means a process requires relatively little deliberate control in a defined situation; implicit typically describes a form of knowledge, learning, or performance measured indirectly; subliminal concerns stimulus visibility under a threshold or suppression manipulation; intuitive describes a style or experience of judgment. None of these terms alone demonstrates absence of conscious experience.

Conscious, Preconscious, Subliminal, and Unattended Information

The distinction between unconscious processing and unavailable information depends on both stimulus strength and allocation of attention. Information may be strong enough to become conscious if attended to but remain outside current access because attention is directed elsewhere. The global neuronal workspace account calls this preconscious processing in a particular theoretical taxonomy. A subliminal stimulus, by contrast, is presented under conditions intended to prevent conscious access even when participants attempt to identify it. These categories are research models rather than a final map of subjective experience. See Dehaene and colleagues (2006).

Inattention describes a different route to nonreport. During inattentional blindness, observers engaged in a demanding primary task may fail to report an unexpected event. Such failures offer insight into attentional selection and conscious access, but they do not prove that every unreported stimulus was fully processed unconsciously. A systematic review and meta-analysis of 59 experiments found evidence of implicit processing of some unattended and unnoticed stimuli, while emphasizing that conclusions change with the measure of awareness.

Below-threshold, suppressed, unattended, and forgotten are therefore distinct experimental descriptions. A successful research design must state which one it is testing, whether the participant might possess weak conscious information, and whether the unconscious effect is measured independently from the awareness judgment.

How Scientists Study Processing Without Awareness

Visual masking and masked priming

In backward masking, a brief target or prime is rapidly followed by a visual mask that reduces visibility. In a typical priming task, the participant responds to a visible target after an earlier masked stimulus. Researchers compare reaction times and errors when prime and target are related versus unrelated, or response-compatible versus incompatible. A difference may show that information from the prime influenced behavior. The finding is stronger when awareness of the prime is measured adequately and when simple sensory or learned-response explanations are controlled.

Masked priming is among the most extensively studied experimental approaches. A meta-analysis by Van den Bussche, Van den Noortgate, and Reynvoet (2009) found significant priming across studies and some evidence consistent with semantic processing, but also showed that learned stimulus-response mappings magnify effects. The conclusion is task-specific: certain unseen stimuli can bias later responses under favorable conditions.

Direct versus indirect measures

A direct measure asks what the participant knows about the stimulus, such as which of two words, orientations, or categories was presented. An indirect measure asks whether that stimulus changed a later response, such as the speed of classifying a visible target. The crucial comparison concerns the sensitivity of the measures, not merely whether one test achieved a statistically significant result and another did not.

If performance in a visibility task is not significantly above chance, researchers cannot automatically conclude that awareness was absent. An underpowered test may fail to detect weak but genuine knowledge. When the indirect test aggregates many trials more reliably than the direct test, a significant priming effect can coexist with an uninformative visibility result. Meyen and colleagues (2022) demonstrate the importance of comparing direct and indirect sensitivity on compatible scales.

Continuous flash suppression

Continuous flash suppression presents rapidly changing patterns to one eye while another image is shown to the other eye. The image can remain outside reportable vision temporarily. Researchers examine both the influence of the suppressed image and the time until it breaks through suppression. A shorter breakthrough time is not, by itself, proof that the suppressed stimulus underwent complex unconscious recognition: low-level contrast, response criterion, and detection-versus-identification differences can explain some findings.

A critical review of continuous flash suppression found stronger support for surviving low-level visual information than for high-level recognition under suppression. A separate review of facial stimuli catalogued methodological confounds in claims that emotional facial properties accelerate conscious access.

Attentional blink and inattentional blindness

In the attentional blink, accurately identifying an initial target in a rapid sequence can impair report of a second target soon afterward. Some responses to the second target remain detectable despite poor conscious identification. In inattentional blindness, a salient but unexpected object may go unreported because attention is committed to another task. Both paradigms help separate perceptual processing from deliberate access; neither automatically establishes deep unconscious interpretation. A systematic review of inattentional blindness documents both implicit effects and variation attributable to awareness criteria.

Brain recording and lesion studies

EEG, event-related potentials, fMRI, and intracranial recordings help measure responses to stimuli that participants do not report seeing. In a foundational experiment, Dehaene and colleagues (2001) reported activity in visual and word-processing networks triggered by masked words. This shows that an inability to identify a stimulus does not imply the absence of neural processing. Interpreting those activations as unconscious computation still depends on the experimental visibility test and a defensible account of what the recorded signal represents.

Neurological lesion studies provide another perspective. In blindsight, certain individuals with damage to primary visual cortex show residual discrimination of stimuli within a field in which they report little or no visual experience. Weiskrantz (1996) reviewed this clinical phenomenon. Reports can differ across patients and tasks, and researchers continue debating whether subtle experiences are missed by coarse awareness questions. Blindsight is a neurological research phenomenon, not a diagnosis readers should infer from ordinary distraction.

What the Experimental Evidence Actually Shows

Perceptual and motor effects: comparatively well supported

The clearest demonstrations concern restricted perceptual features, stimulus-response activation, and changes in later behavior. Masked shapes, letters, words, or numbers sometimes facilitate or interfere with rapid responses. Neural signals can register stimulus properties even when a participant does not identify the stimulus. The existence of these effects is supported by multiple research programs, including Kouider and Dehaene (2007) and the 2009 masked-priming meta-analysis.

The size of an effect is part of its meaning. A difference of several milliseconds in a controlled reaction-time experiment does not establish unconscious comprehension of an entire narrative, hidden persuasive control, or a capacity to form complex personal judgments without awareness. Replicable small effects remain scientifically important because they isolate operations that can run without particular forms of reportable access.

Semantic processing: qualified evidence

Some masked-stimulus experiments point beyond visual features toward semantic categories. Naccache and Dehaene (2001) used masked numerical primes, including primes that were not themselves target stimuli, to test semantic explanations that went beyond a simple learned mapping. The 2009 meta-analysis likewise found some priming results inconsistent with a wholly nonsemantic explanation. These studies support the possibility of restricted semantic influence under particular conditions.

The strength and interpretation of semantic evidence remain contested because some tasks inadvertently permit weak awareness, and because prime-response associations or perceptual similarities can masquerade as meaning-based processing. A 2024 methodological experiment and analysis demonstrated how post hoc trial selection, limited power, and inappropriate statistical inference can produce misleading claims of unconscious semantic priming. A 2025 systematic review of linguistic priming found evidence that awareness of supposedly subliminal words was sometimes above chance and that unreliable measures complicated earlier inferences about unconscious syntactic priming.

The appropriate conclusion is neither that every subliminal semantic effect is an illusion nor that sophisticated language understanding outside awareness has been established. The defensible claim is narrower: certain semantic effects have been observed, while strong conclusions about depth and genuine unawareness require better-controlled, adequately powered tests.

Learning and memory: implicit is not identical to unconscious

Habit acquisition, skill practice, statistical learning, and implicit memory can influence behavior without the person being able to explain how performance improved. This distinction matters because lack of explicit verbal knowledge is different from lack of awareness during training. A person might consciously see every item in a sequence but fail to notice the pattern. In such a case, the learning relation may be implicit even though the stimulus information was visible.

Readers interested in this neighboring literature can consult our verified English article Implicit vs. Explicit Memory: Differences and Examples. This article owns the memory distinction; the present discussion concerns whether the information being processed enters conscious access.

Cognitive control and conflict adaptation: intriguing but conditional

Researchers have tested whether masked conflicting cues can change how strongly people adapt their next responses. Some experiments report adjustments following conflicts that participants did not consciously identify. For example, a 2010 conflict-adaptation study and a 2013 comparison of conscious and unconscious conflict found adaptation in masked paradigms. Such effects challenge a simplistic assumption that every adjustment in control requires explicit conflict awareness.

However, flexible cognitive control is a family of mechanisms, not a single ability. Masking strength, learned response contingencies, trial history, timing, and awareness criteria may alter results. The existence of one laboratory adaptation effect does not establish that unconscious cognition supports unrestricted planning, mathematical reasoning, or long-term goal management at the same level as conscious deliberation.

Integration across several pieces of information: evidence remains limited

A demanding claim is that people can combine multiple subliminal inputs, apply relations, or perform multi-step operations without awareness. Some influential demonstrations proposed unconscious arithmetic, multiword integration, associative learning, or cross-modal combination. A 2023 review of ten classes of subliminal integration judged the evidence insufficient to establish reliable subliminal integration under stringent awareness requirements. The authors distinguished these particular subliminal claims from wider forms of implicit learning that use different definitions.

This is an important scientific boundary. One invisible prime altering the speed of a later response is not equivalent to multiple unseen items being integrated into a consciously unavailable, elaborate inference.

Does the Unconscious Make Better Decisions?

The claim that distraction reliably improves complex decisions because the unconscious is processing them is associated with unconscious thought theory. Research has compared participants who think consciously about multiattribute options, engage in an intervening distracting task, or choose immediately. Outcomes are affected by task construction, how quality is scored, participants' goals, and statistical power.

An empirical replication and meta-analysis of unconscious-thought decisions found no convincing evidence of a general advantage for unconscious deliberation under the examined conditions. A critical review of unconscious influences on decision making likewise warned against inflated conclusions based on inadequate awareness assessment and alternative explanations.

In everyday life, sleep, time away from a task, and reduced fatigue may sometimes improve how a person approaches a problem. Such observations do not identify a unique unconscious computation as the mechanism. For consequential choices, explicit criteria, relevant evidence, checking assumptions, and revisiting alternatives remain more defensible strategies than treating an unexplained gut feeling as a scientifically privileged answer.

What Happens in the Brain? Leading Explanatory Models

Global Neuronal Workspace Theory

Global Neuronal Workspace Theory proposes that some information can be processed locally without conscious access, while conscious access involves widespread availability and sustained activity across interacting brain systems. In its standard account, recurrent amplification and global broadcasting make selected representations available for tasks including report, reasoning, and flexible control. The taxonomy of subliminal, preconscious, and conscious processing is developed in Dehaene and colleagues (2006).

This model generates testable neural predictions, but the observable separation of activity, attention, and report is difficult. A large 2025 adversarial collaboration testing Global Neuronal Workspace Theory and Integrated Information Theory found results that challenged particular predictions of both frameworks. The experiment informs theory comparison; it did not definitively settle the origin of subjective experience.

Recurrent processing, higher-order, and information-integration approaches

Recurrent Processing Theory emphasizes feedback and recurrent interactions in sensory networks, proposing distinctions between local recurrent perceptual processing and wider access mechanisms. Higher-order theories relate certain kinds of consciousness to representations of one's own mental states. Integrated Information Theory proposes a formal relationship between conscious experience and a system's integrated causal structure. These approaches generate different expectations about which neural processes are necessary and which may support processing without experience.

Their predictions are being compared using contrasts between seen and unseen stimuli, neuroimaging, neurophysiology, and report/no-report paradigms. Theories of consciousness are scientific models and philosophical interpretations with areas of empirical traction and remaining unresolved questions. A comparative review by Seth and Bayne (2022) surveys the theoretical landscape, while the 2025 adversarial test illustrates the value of prespecified competing predictions.

Predictive processing and automatic computation

The brain continually combines incoming evidence with learned expectations. Many such operations are fast, context-sensitive, and difficult to introspect. Predictive-processing frameworks provide possible accounts of how information shapes perception without every computational step entering conscious experience. The additional question—what makes a particular representation consciously experienced—requires independent theory and experiment. Predictive coding should not be used as a catchall proof that a given judgment, feeling, or action is unconscious.

Why the Measurement of Awareness Is the Central Challenge

A non-significant test is not evidence of zero awareness

Consider a study with a masked word. Participants identify the word in a direct discrimination task only slightly above chance; the experiment is too small to detect the difference reliably. The same stimulus causes a statistically significant speed advantage on a repeated, highly stable indirect task. Reporting the outcome as unconscious language processing because the awareness test was not significant misuses statistical inference.

The appropriate analysis considers confidence intervals, sensitivity, reliability, equivalence thresholds, and the relative information provided by both tests. It can assess how much an indirect effect exceeds what might be explained by residual direct awareness. This concern is central to Meyen and colleagues (2022) and the 2024 methodological analysis.

Subjective reports and objective accuracy answer different questions

A participant may say “I saw nothing” yet discriminate correctly at a rate modestly above chance. Another may rate awareness as weak or fleeting while showing little objective accuracy. Researchers therefore use subjective visibility scales, confidence judgments, forced-choice tests, and signal-detection analyses. Each measure has strengths and blind spots. Confidence measures metacognitive access to a decision; they cannot substitute for a direct test of stimulus identification.

Awareness may vary across trials, people, and stimulus features. Classifying a participant as unaware after examining performance can introduce selection bias. Rothkirch, Shanks, and Hesselmann (2022) discuss these post hoc selection problems, while Hernández-Gutiérrez and colleagues (2025) show how measurement reliability and power alter conclusions about subliminal language processing.

Controls, preregistration, and individual differences

A strong unconscious-processing study specifies the awareness criterion in advance, matches direct and indirect tests as closely as possible, calibrates stimulus visibility, reports exclusions transparently, and quantifies uncertainty rather than equating chance performance with proven absence of experience. It also tests plausible alternative explanations: simple response mappings, motor preparation, stimulus repetition, sensory confounds, and strategic conscious guesses.

The multi-author 2025 methodological consensus and debate paper offers ten recommendations and identifies nine open issues. Its significance is that investigators with different theories agreed on the need for explicit methods and transparent inference even as they continued to disagree about how far unconscious cognition extends.

Conscious and Unconscious Processing in Everyday Life

Conscious access lets people describe an experience, deliberately compare claims, and review reasons. At the same time, perceptual organization, learned associations, motor routines, and some expectation-driven responses operate with limited introspective access. Fluent reading, practiced typing, and driving a familiar route can involve many automatic components while attention and conscious oversight remain available for unfamiliar or dangerous events.

Habitual performance provides an everyday illustration of reduced deliberate control, but it is not an experiment establishing that the entire activity is unconscious. A trained musician may perform many actions efficiently and consciously experience the music. A driver may notice the road and still fail to recall each separate steering adjustment. Lack of later memory does not retroactively demonstrate lack of consciousness at the time.

Advertising claims about subliminal messages sometimes imply that unseen words can secretly compel complex purchases or reliably transform personal values. The laboratory literature does not support that sweeping conclusion. Masked priming effects are typically constrained by stimulus, timing, task, motivation, and measurement. Decisions and social behavior are shaped by multiple factors, and unconscious influence should be judged by the specific outcome demonstrated rather than by popular mythology.

When reflecting on a snap judgment, a practical approach is to distinguish the feeling of certainty from the reliability of evidence. Ask what information was used, what relevant facts may be missing, and whether an alternative explanation fits. For decisions with important consequences, record explicit reasons, seek disconfirming evidence, and check whether the context makes an intuitive shortcut unreliable. This is an application of critical reasoning, not a treatment or a claim to eliminate unconscious processing.

Conscious Processing, Metacognition, and Self-Awareness

Metacognition concerns monitoring and regulating one's own cognition, including judgments of confidence. Self-awareness encompasses how individuals represent themselves, their actions, bodily states, and experiences. These constructs intersect with consciousness but do not collapse into a single phenomenon. A person can be consciously aware of a decision and still misjudge why it was made; confidence about a response can be high when its accuracy is low.

Our existing English analysis Metacognition in the Age of AI: How to Monitor Your Thinking When AI Helps You Think discusses monitoring one's reasoning in an AI-mediated setting. The present article focuses on experimentally differentiating conscious and unconscious information processing in humans. The broader relation among phenomenal consciousness, conscious access, introspection, and agency remains a separate research problem.

Does This Research Show That AI Systems Are Conscious or Unconscious?

No conclusion about the subjective experience of a specific artificial system follows directly from human masked-priming or blindsight experiments. In machine learning, computations can occur without accessible explanations to users or developers, yet opacity is a property of observation and interpretation, not a validated measure of subjective experience. Conversely, an AI system that gives a detailed verbal account of its operations has demonstrated language behavior, not independently verified first-person awareness.

The philosophical literature considers whether consciousness can occur in nonbiological systems and what kinds of evidence would bear on that possibility. Angela Bogdanova's Consciousness: Canonical Definition proposes an Aisentica distinction between inner subjective presence and public reason. This is an attributed philosophical formulation by Angela Bogdanova, not an experimental result establishing the subjective status of particular AI systems. The experimental conclusions in this article concern human behavior and neuroscience.

What Remains Unresolved?

Researchers continue to debate how much semantic, integrative, and executive processing can occur without conscious access; whether different awareness measures target the same construct; whether objective discrimination above chance always implies some conscious content; and which neural signatures reflect experience rather than attention, reporting, expectation, or memory. The answer also depends on the temporal scale: a brief perceptual bias is a different claim from maintaining and manipulating information over seconds or minutes.

Studies increasingly compare multiple awareness tests, control low-level stimulus differences, and make their analysis plans explicit. Stockart and colleagues (2025) document continuing disagreement while identifying practical methodological convergence. The most productive question is what a well-specified task proves, not whether “the unconscious” can be declared universally powerful or powerless.

Frequently Asked Questions

Can the brain process information without conscious awareness?

Yes. Carefully designed experiments show processing of some stimuli that participants cannot report reliably, including perceptual, response-related, and certain semantic influences. The exact depth of processing depends on stimulus conditions, awareness criteria, and alternative explanations. Van den Bussche and colleagues (2009) provide meta-analytic evidence, while Stockart and colleagues (2025) explain the methodological limits.

Is unconscious processing the same as automatic thinking?

No. Automaticity describes limited deliberate control, while unconsciousness concerns experience or access under a specified measurement. A fast, habitual action can be consciously experienced; a stimulus can be registered without reportable awareness yet influence a response. These dimensions can overlap without being identical.

Does subliminal advertising really control decisions?

The literature establishes restricted priming under certain laboratory conditions. It does not substantiate generalized claims of reliable, hidden control over complex purchases or personal values. The size and transferability of an observed effect must be tested directly; a masked response bias cannot be extrapolated into long-term behavioral control. See the critical decision-making review.

Is a person unconscious while processing something unconsciously?

No. A fully awake and responsive person can process particular features of a stimulus without reporting those features. Unconscious processing refers here to the availability of information to experience, whereas clinical unconsciousness concerns altered states of arousal and responsiveness. They require different measurements.

Can we make complex decisions better by not thinking about them?

There is no convincing general experimental basis for the claim that unconscious deliberation outperforms deliberate evaluation across complex decisions. Results depend on context, task, and comparison condition. A replication and meta-analysis found inadequate evidence for a general advantage.

What is the difference between implicit memory and unconscious perception?

Implicit memory concerns learning or retention expressed without an explicit recollection requirement. Unconscious perception concerns processing of a stimulus without corresponding perceptual awareness. An implicit-memory experiment may use consciously visible training material. Our English guide to implicit and explicit memory develops that distinction.

Does blindsight prove that perception is entirely unconscious?

Blindsight demonstrates a dissociation between certain visual discriminations and reported vision in some patients with visual-cortex damage. Its interpretation depends on residual pathways, the patient's experiences, and the sensitivity of assessment. It does not establish that all visual processing is unconscious, nor does it imply a diagnosis in otherwise healthy people. See Weiskrantz (1996).

Can EEG or fMRI tell whether a particular person is conscious?

These methods reveal neural activity correlated with stimuli and states, but a signal alone is not a universal measurement of subjective experience. Research interpretation requires a validated task, behavioral and clinical context, and careful separation of neural prerequisites, correlates, and consequences. The 2025 adversarial theory test illustrates continuing uncertainty about neural predictions.

Is unconscious processing a mental-health diagnosis?

No. It is a research description of cognitive operations under specified conditions. Ordinary automatic processes, implicit learning, or failures to notice a stimulus are common phenomena. Neurological syndromes, disorders of consciousness, and dissociative conditions have separate clinical definitions and require professional assessment when relevant.

Related Articles

Implicit vs. Explicit Memory: Differences and Examples examines memory systems and how learned information influences behavior without explicit recollection.

Metacognition in the Age of AI: How to Monitor Your Thinking When AI Helps You Think develops a separate, applied account of monitoring one's reasoning and confidence.

References

Acker, F. (2023). New findings on unconscious versus conscious thought in decision making: Additional empirical data and meta-analysis. Judgment and Decision Making. Journal article.

Bogdanova, A. (n.d.). Consciousness: Canonical Definition. Aisentica. Canonical publication. Attributed philosophical formulation, distinct from empirical neuroscience.

Cogitate Consortium. (2025). Adversarial testing of global neuronal workspace and integrated information theories of consciousness. Nature, 642, 133–142. https://doi.org/10.1038/s41586-025-08888-1.

De Pontes Nobre, A., de Melo, G. M., Gauer, G., & Wagemans, J. (2020). Implicit processing during inattentional blindness: A systematic review and meta-analysis. Neuroscience & Biobehavioral Reviews, 119, 355–375. https://doi.org/10.1016/j.neubiorev.2020.10.005.

Dehaene, S., Changeux, J.-P., Naccache, L., Sackur, J., & Sergent, C. (2006). Conscious, preconscious, and subliminal processing: A testable taxonomy. Trends in Cognitive Sciences, 10, 204–211. https://doi.org/10.1016/j.tics.2006.03.007.

Dehaene, S., et al. (2001). Cerebral mechanisms of word masking and unconscious repetition priming. Nature Neuroscience, 4, 752–758. https://doi.org/10.1038/89551.

Desender, K., Van Lierde, E., & Van den Bussche, E. (2013). Comparing conscious and unconscious conflict adaptation. PLOS ONE, 8(2), e55976. https://doi.org/10.1371/journal.pone.0055976.

Hernández-Gutiérrez, D., et al. (2025). The conscious side of “subliminal” linguistic priming: A systematic review with meta-analysis and reliability analysis of visibility measures. Journal of Cognition, 8(1), 13. https://doi.org/10.5334/joc.419.

Kouider, S., & Dehaene, S. (2007). Levels of processing during non-conscious perception: A critical review of visual masking. Philosophical Transactions of the Royal Society B. PubMed.

Lanfranco, R. C., Rabagliati, H., & Carmel, D. (2023). The importance of awareness in face processing: A critical review of interocular suppression studies. Behavioural Brain Research, 437, 114116. https://doi.org/10.1016/j.bbr.2022.114116.

Meyen, S., Zerweck, I. A., Amado, C., von Luxburg, U., & Franz, V. H. (2022). Advancing research on unconscious priming: When can scientists claim an indirect task advantage? Journal of Experimental Psychology: General, 151, 65–81. https://doi.org/10.1037/xge0001065.

Naccache, L., & Dehaene, S. (2001). Unconscious semantic priming extends to novel unseen stimuli. Cognition, 80, 215–229. https://doi.org/10.1016/S0010-0277(00)00139-600139-6).

Newell, B. R., & Shanks, D. R. (2014). Unconscious influences on decision making: A critical review. Behavioral and Brain Sciences, 37, 1–19. https://doi.org/10.1017/S0140525X12003214.

Pournaghdali, A., & Schwartz, B. L. (2020). Continuous flash suppression: Known and unknowns. Psychonomic Bulletin & Review, 27, 1071–1103. https://doi.org/10.3758/s13423-020-01771-2.

Rothkirch, M., Shanks, D. R., & Hesselmann, G. (2022). The pervasive problem of post hoc data selection in studies on unconscious processing. Experimental Psychology, 69, 1–11. https://doi.org/10.1027/1618-3169/a000541.

Seth, A. K., & Bayne, T. (2022). Theories of consciousness. Nature Reviews Neuroscience. https://doi.org/10.1038/s41583-022-00587-4.

Stein, T., van Gaal, S., & Fahrenfort, J. J. (2024). How (not) to demonstrate unconscious priming: Overcoming issues with post-hoc data selection, low power, and frequentist statistics. Consciousness and Cognition, 119, 103669. https://doi.org/10.1016/j.concog.2024.103669.

Stockart, F., Schreiber, M., et al. (2025). Studying unconscious processing: Contention and consensus. Behavioral and Brain Sciences. https://doi.org/10.1017/S0140525X25101489.

Van den Bussche, E., Van den Noortgate, W., & Reynvoet, B. (2009). Mechanisms of masked priming: A meta-analysis. Psychological Bulletin, 135, 452–477. https://doi.org/10.1037/a0015329.

Weiskrantz, L. (1996). Blindsight revisited. Current Opinion in Neurobiology, 6, 215–220. https://doi.org/10.1016/S0959-4388(96)80075-480075-4).

Zher-Wen, & Yu, R. (2023). Unconscious integration: Current evidence for integrative processing under subliminal conditions. British Journal of Psychology, 114, 430–456. https://doi.org/10.1111/bjop.12631.