Awareness

Abstract

Awareness is the state of a stimulus being consciously perceived, the difference between a signal that reaches experience and one that drives behavior without being noticed. This article treats awareness as an object of measurement: how a fixed stimulus comes to be seen or missed, and how an experimenter can tell which happened. It separates perception without awareness from the failure to perceive at all, distinguishes the objective threshold of discrimination from the subjective threshold of report, and reviews blindsight as the sharpest dissociation of the two. It surveys the instruments that quantify awareness, subjective scales, wagering, confidence, and metacognitive sensitivity, and weighs whether awareness arrives gradually or all at once. Three demonstrations drive the thresholds apart, grade a percept on a perceptual awareness scale, and watch wagering track accuracy only as far as metacognition allows.

Keywords: awareness, perception without awareness, subjective measures

Awareness is the narrowest and most tractable of the words that cluster around consciousness. Where consciousness names at once a global state, a relation to content, and the felt quality of experience, awareness picks out just the middle term: the condition of a particular stimulus being consciously registered rather than processed in the dark (Seth et al., 2008). Framed that way, awareness becomes something an experiment can hold in view. Present a stimulus twice under identical physical conditions, arrange that on some trials it is noticed and on others not, and the difference between the two is a difference in awareness that can be probed, timed, and measured. The sections below build the concept from that operational core: they distinguish awareness from mere sensory processing, separate the objective threshold at which a signal begins to inform behavior from the subjective threshold at which the observer reports seeing it, examine the blindsight patient in whom the two come dramatically apart, catalog the measures that put a number on awareness, and ask whether the transition into awareness is graded or sudden.

What Awareness Is

The concept of awareness earns its keep by marking a line that behavior alone blurs. A great deal of what the brain does with a stimulus, extracting its features, priming a related word, biasing a choice, happens whether or not the stimulus is ever consciously seen. Awareness is the further fact that, on some occasions, there is something it is like to perceive the stimulus, that it enters experience and can be reported as seen. To study awareness is therefore to study a dissociation: the same input, processed to the same first stages, sometimes accompanied by experience and sometimes not (Merikle et al., 2001). The scientific question is not the metaphysical one of why any processing is felt, but the empirical one of what, in the stimulus and in the brain, decides whether it is.

Awareness in this sense is content-bound and graded across contents. One is aware of a face, a motion, an ache, and the awareness of each can be present or absent independently while the organism remains fully awake. Even the awareness of the body's own internal state has a distinct substrate: A. D. Craig argued that interoceptive feelings, the sense of the body's physiological condition, are made available to awareness through the anterior insula, a re-representation that turns visceral signals into felt states (Craig, 2009). What unites the visual, auditory, and interoceptive cases is the same operational signature. In each, a stimulus that can be shown to have been registered, by its effect on discrimination or on later behavior, may or may not cross into the state the observer will report as awareness, and the whole science of the topic consists in measuring where that crossing falls and what governs it (Lamme, 2006).

Perception Without Awareness

The claim that a stimulus can be perceived without being consciously seen is old, contentious, and now well supported, but only under a demanding standard of evidence. The difficulty is that any single measure of awareness can fail for a trivial reason: an observer who denies seeing a masked word may have seen it dimly and set a conservative bar for saying so. Philip Merikle's resolution was to insist on a qualitative dissociation rather than a mere absence of report. Perception without awareness is demonstrated when a stimulus the observer cannot consciously detect nonetheless produces an effect that conscious perception would not, or produces it in the opposite direction, so that the unconscious influence cannot be dismissed as weak conscious perception (Merikle et al., 2001). An unseen prime that speeds recognition of a related target, while the observer performs at chance on the prime itself, is the paradigm case.

How far such unconscious processing reaches is a matter of degree, fixed by how the stimulus is rendered invisible. Sid Kouider and Stanislas Dehaene, reviewing visual masking, concluded that masked stimuli are processed well beyond their physical features, up to the level of meaning, but that the depth falls off sharply as the mask grows stronger, and that the richest claims of high-level unconscious cognition are the least robust (Kouider & Dehaene, 2007). This graded picture maps onto the taxonomy Dehaene and colleagues drew between subliminal stimuli, too weak to be reported even with attention, preconscious stimuli, strong enough to be seen but denied access by inattention, and consciously reported stimuli (Dehaene et al., 2006). The categories matter because they warn against a single dividing line: a stimulus can be invisible for want of strength or for want of attention, and the two routes to unawareness leave different traces.

Objective and Subjective Thresholds

The measurement of awareness turns on a distinction between two thresholds that ordinary language runs together. The objective threshold is the stimulus strength at which an observer's forced-choice discrimination of the stimulus first rises above chance, the point below which the signal carries no information the observer can use. The subjective threshold, which sits higher, is the strength at which the observer begins to report being aware of the stimulus at all. Jim Cheesman and Philip Merikle drew the line and showed that it is real: between the two thresholds lies a band in which observers discriminate a stimulus better than chance while insisting they see nothing, and they demonstrated that qualitatively different effects occur below versus above the subjective threshold, not merely weaker versions of the same effect (Cheesman & Merikle, 1986). Awareness, on this analysis, is not the presence of information in the system but the crossing of the higher, subjective bar.

Signal detection theory makes the two thresholds quantitative and, in doing so, exposes what a report actually measures. An observer's ability to tell signal from noise is captured by d-prime, the separation of the two internal distributions, which is independent of how willing the observer is to say yes. Whether the observer reports awareness depends in addition on a criterion, a placement of the bar that can be cautious or liberal for reasons that have nothing to do with what was seen. Perception without awareness is exactly the regime of positive d-prime with the subjective criterion unmet: discrimination is above chance, yet the observer sets the bar for reporting awareness beyond the signal's reach. Figure 1 lays out the three regimes along a stimulus-strength axis, and the demonstration that follows lets the reader move a stimulus through them and read off the divergence of objective sensitivity from subjective report.

Blindsight

The dissociation that laboratory masking engineers in fractions of a second occurs permanently, and far more dramatically, in blindsight. Lawrence Weiskrantz and colleagues studied patients with damage to primary visual cortex who reported a region of the visual field as simply blank, yet who, when pressed to guess about a stimulus placed in that region, discriminated its position, motion, and orientation far above chance (Weiskrantz et al., 1974). The patients experienced no visual sensation and had to be coaxed to respond at all, insisting they were merely guessing; their discrimination nonetheless demonstrated that visual information was being processed and used without any accompanying awareness. Blindsight is the objective threshold crossed while the subjective threshold is not merely high but abolished, a natural demonstration that awareness is separable from the visual competence it normally accompanies, and it implicated pathways bypassing the striate cortex on which conscious vision depends.

The logic of blindsight can be recreated in the intact brain, which shows that the phenomenon is about awareness rather than about brain damage as such. Hakwan Lau and Richard Passingham matched a masked stimulus at two points that produced identical discrimination performance but differed in how often observers reported seeing the stimulus, a relative blindsight in the normal observer: the same objective sensitivity with different levels of subjective awareness (Lau & Passingham, 2006). Because task performance was held constant, whatever brain activity distinguished the high-awareness from the low-awareness condition was a correlate of awareness itself rather than of the discrimination, and the difference implicated the dorsolateral prefrontal cortex. The experiment turns blindsight from a clinical curiosity into a general method: fix objective performance, vary subjective report, and isolate the neural difference that awareness makes.

Measuring Awareness

If awareness is the crossing of a subjective threshold, then measuring it means asking the observer, and the whole difficulty is that asking can be done in ways that measure different things. The most direct instrument is a subjective scale. Thomas Ramsoy and Morten Overgaard, finding that a simple seen-or-not question forced a false dichotomy, had observers rate their experience on a graded perceptual awareness scale running from no experience through a brief glimpse and an almost clear image to a clear image, and found that observers used the intermediate points reliably (Ramsoy & Overgaard, 2004). The scale takes the observer's introspective report as the measure of awareness, which is its strength, it asks about experience directly, and its risk, it inherits whatever biases attend introspection. The demonstration below builds a perceptual awareness scale and shows how the distribution over its four ratings shifts as a stimulus strengthens.

To escape the suspicion that observers simply differ in how readily they say they saw something, a second family of measures makes awareness pay. In post-decision wagering, the observer makes a perceptual judgment and then bets on it; if awareness of the stimulus is genuine, high wagers should follow correct judgments, and Navindra Persaud and colleagues proposed the resulting winnings as an objective, bias-resistant index of awareness, on the reasoning that no rational observer conceals money-making knowledge (Persaud et al., 2007). Confidence ratings work on the same principle without the payoff: Craig Kunimoto and colleagues showed that when observers discriminate above chance their confidence tracks their accuracy, and that this coupling, rather than the mere denial of awareness, is the operational mark of conscious perception (Kunimoto et al., 2001). What both measures quantify is metacognitive sensitivity, the degree to which an observer's own confidence discriminates their correct from their incorrect judgments. Stephen Fleming and Hakwan Lau formalized it as meta-d-prime, the type-1 sensitivity that would be required to produce the observed confidence-accuracy coupling, expressed relative to actual sensitivity so that a ratio of one denotes an ideal metacognitive observer (Fleming & Lau, 2014). The demonstration that follows shows wagering paying off only to the extent that metacognitive sensitivity allows.

Table 1 sets the four instruments side by side, with what each asks of the observer and the characteristic weakness each carries.

Note. The first three measures ask the observer for a report and differ in how far they resist bias; meta-d′ is computed from the confidence-accuracy relationship rather than asked for directly (Fleming & Lau, 2014).

No single measure is decisive, which is why the field has moved to comparing them head to head. Anil Seth and colleagues argued that a credible science of awareness must combine behavioral measures with neurophysiological ones and hold each to explicit criteria of reliability and validity, since a measure that is sensitive but biased, or unbiased but insensitive, will misplace the threshold in a characteristic direction (Seth et al., 2008).

The same problem of timing an inner event arises when what is measured is awareness not of a stimulus but of an intention. Asking observers to note, against a rapidly rotating clock, the instant they first felt the urge to move, Benjamin Libet and colleagues fixed the moment of conscious awareness of an intention and found that it followed the onset of the brain's readiness potential by several hundred milliseconds (Libet et al., 1983). The clock method and its bearing on voluntary action are taken up in the companion article on the readiness potential and the timing of conscious intention.

Gradual or Dichotomous

Whether awareness arrives all at once or by degrees is not merely a question about experience but a test of the measures themselves, because a yes-or-no report cannot detect a graded transition even if one exists. Overgaard and colleagues addressed the question by comparing report methods directly, having observers describe the same stimuli with a dichotomous seen-or-not judgment and with the graded perceptual awareness scale, and found that the graded scale revealed a continuum of intermediate awareness states that the dichotomous measure collapsed into an artificial step (Overgaard et al., 2006). On this evidence the appearance of all-or-none awareness is partly an artifact of forcing a binary answer: allowed a richer vocabulary, observers report seeing things partly, dimly, and in fragments.

The finding does not settle the matter, because a measure can also invent gradations that experience lacks, and the comparison of measures is the only way to tell. Kristian Sandberg, Overgaard, Cleeremans, and colleagues put the leading scales into competition, asking which best captured the presence of awareness against an objective performance criterion, and found the perceptual awareness scale to be the most exhaustive, tracking small differences in stimulus visibility that confidence ratings and wagering missed (Sandberg et al., 2010). The convergence is instructive: the same instrument that reveals graded awareness also proves the most sensitive index of awareness overall, which suggests the gradations are real rather than an artifact of a lax scale. Whether the underlying transition is truly continuous or a fine-grained staircase remains open, but the methodological lesson is settled, that the shape one finds for awareness depends on the resolution of the instrument used to look.

Worked Example

The gap between the objective and subjective thresholds can be made concrete with signal detection theory, which turns a pattern of hits and false alarms into a bias-free measure of sensitivity. Consider a masked-detection task in which a faint target appears on half the trials and the observer reports present or absent. At an intermediate stimulus strength the observer scores a hit rate of 0.69, correctly reporting the target when it is there, and a false-alarm rate of 0.31, reporting it when it is absent. Sensitivity is d-prime equals z of the hit rate minus z of the false-alarm rate, where z is the inverse normal. Because 0.69 and 0.31 are symmetric about one half, z of 0.69 is 0.4959 and z of 0.31 is its negative, so d-prime equals 0.4959 minus negative 0.4959, about 0.99. The observer is discriminating the target close to a full unit of sensitivity above chance. The criterion, computed as minus one half the sum of the two z-scores, is exactly zero: the observer is unbiased on the detection task itself.

Yet objective sensitivity of one is entirely compatible with the observer reporting no awareness. Suppose the subjective threshold, the strength at which the observer starts to say the target was consciously seen, sits well above this point, so that on these same trials the probability of an awareness report is only about 0.10. This is the regime the demonstration labels perception without awareness: at a stimulus strength of 45 on its scale, objective d-prime is 0.99 while subjective report stands near 0.095, and the observer discriminates a stimulus they decline to call seen. Push the strength to 60 and the report probability reaches one half, the subjective threshold by definition, while d-prime has climbed to about 1.98; by strength 75 the report probability is 0.905 and d-prime is near its ceiling of 3, the fully conscious regime. The single lesson in the arithmetic is that discrimination and awareness are measured on different axes: a positive d-prime establishes that the information is in the system, but only the crossing of the subjective criterion establishes that it is in awareness, and the two need not coincide.

Discussion

Awareness matters to cognitive psychology because it is the point where the study of consciousness becomes an experimental science with instruments and error bars. By defining awareness operationally, as the crossing of a subjective threshold that can be dissociated from objective sensitivity, the field converts an intractable question about experience into a tractable one about measurement (Seth et al., 2008). The payoff is a body of robust dissociations: priming from stimuli discriminated at chance, blindsight in which vision survives the loss of visual awareness, and relative blindsight in the intact brain where the same performance carries different levels of report (Merikle et al., 2001; Weiskrantz et al., 1974; Lau & Passingham, 2006). Each depends on separating what the brain does with a stimulus from whether the stimulus is experienced, and each is only as convincing as the measure of awareness it rests on, which is why the qualitative-dissociation standard and the comparison of measures have done more for the field than any single clever paradigm (Cheesman & Merikle, 1986; Sandberg et al., 2010). The open questions are correspondingly methodological as much as substantive. Whether the transition into awareness is continuous or discrete is unresolved and may differ across stimuli and tasks (Overgaard et al., 2006); how deeply an unseen stimulus is processed depends on how it is hidden and is contested at the upper levels (Kouider & Dehaene, 2007); and the neural difference that awareness makes, whether it demands prefrontal involvement or can be read off posterior activity, is the same dispute that divides the theories of consciousness. What is settled is the reframing: awareness is a variable to be measured, and the quality of the measurement sets the quality of everything built on it (Fleming & Lau, 2014).

Common Misconceptions

If a stimulus influences behavior, it must have been consciously perceived.

A stimulus discriminated at chance can still prime a related response, and a blindsight patient can point to a target they report as invisible (Merikle et al., 2001; Weiskrantz et al., 1974). Behavioral influence establishes that the stimulus was processed, not that it was experienced; the two are measured on different axes and routinely dissociate.

Denying that one saw a stimulus proves it was unconscious.

A no answer can mean the stimulus was not seen or that it was seen dimly and fell short of a conservatively placed subjective criterion (Cheesman & Merikle, 1986). This is why a bare absence of report is weak evidence, and why the field demands a qualitative dissociation or a bias-resistant measure such as wagering before concluding that perception was truly unconscious (Persaud et al., 2007).

Awareness is all-or-none.

Whether awareness looks binary or graded depends on the instrument. Forced to answer seen or not, observers produce an apparent step; allowed a graded scale, they reliably report intermediate states of partial and unclear awareness (Ramsoy & Overgaard, 2004; Overgaard et al., 2006). A dichotomous measure cannot detect a continuum even where one exists.

Glossary

Awareness.

The state of a particular stimulus being consciously perceived, as opposed to being processed without entering experience.

Blindsight.

Residual above-chance visual discrimination within a field the patient reports as blind, following damage to primary visual cortex.

Confidence rating.

A judgment of how sure one is of a perceptual decision, used to index awareness through its coupling to accuracy.

Criterion.

In signal detection theory, the placement of the bar for responding yes, independent of sensitivity and driven by bias and payoff.

d-prime.

A bias-free measure of sensitivity, the standardized distance between the internal signal and noise distributions, computed as z(hit) minus z(false alarm).

Interoceptive awareness.

The felt awareness of the body's internal physiological state, associated with re-representation in the anterior insula.

Metacognitive sensitivity.

The degree to which an observer's confidence discriminates their own correct from incorrect judgments, formalized as meta-d-prime.

Objective threshold.

The stimulus strength at which forced-choice discrimination first exceeds chance, marking the entry of usable information.

Perception without awareness.

Processing of a stimulus that influences behavior while the observer reports no conscious experience of it; the regime between the two thresholds.

Perceptual awareness scale.

A graded report scale, from no experience to a clear image, used to capture intermediate states of visual awareness.

Post-decision wagering.

A bias-resistant measure of awareness in which the observer bets on a perceptual judgment, so that winnings index conscious knowledge.

Preconscious.

A stimulus strong enough to be seen but denied awareness for want of attention, as opposed to one too weak to be seen at all.

Subjective threshold.

The stimulus strength at which the observer begins to report conscious awareness, sitting above the objective threshold.

Subliminal.

Below the threshold of report even under full attention; too weak to be consciously perceived.

Visual masking.

The rendering of a stimulus invisible by a second stimulus close to it in time or space, a standard tool for controlling awareness.

Key Researchers

Axel Cleeremans (b. 1963). Brussels psychologist who has compared the leading subjective measures of awareness against one another, helping establish which instrument most faithfully tracks conscious perception.
Universite libre de Bruxelles - ORCID - Google Scholar - Wikipedia

Stephen M. Fleming (University College London). Cognitive neuroscientist who, with Hakwan Lau, formalized meta-d-prime as a measure of metacognitive sensitivity, giving the confidence-accuracy coupling a rigorous, bias-corrected form.
University College London - ORCID - Google Scholar

Philip M. Merikle (b. 1942). Waterloo psychologist who set the evidentiary standard for perception without awareness, arguing that only a qualitative dissociation, an unconscious effect that conscious perception would not produce, can establish that a stimulus was processed without being experienced.
University of Waterloo - Wikipedia

Morten Overgaard (Aarhus University). Cognitive neuroscientist who developed the perceptual awareness scale and used it to argue that visual awareness is graded rather than dichotomous, showing that the shape of the transition depends on the report method.
Aarhus University - ORCID - Google Scholar

Anil K. Seth (b. 1972). Sussex neuroscientist who has pressed for a rigorous science of awareness that combines behavioral and neurophysiological measures under explicit criteria of reliability and validity.
University of Sussex - ORCID - Google Scholar - Wikipedia

Lawrence Weiskrantz (1926-2018). Oxford experimental psychologist who coined the term blindsight and established that patients with striate-cortex damage could discriminate stimuli in a field they reported as blind, the founding dissociation of visual competence from visual awareness.
Royal Society - Wikipedia

Frequently Asked Questions

What is awareness in cognitive psychology?
Awareness is the state of a particular stimulus being consciously perceived, the difference between a signal that enters experience and one that is processed without ever being noticed. It is narrower than consciousness, which also covers the global state of wakefulness and the felt quality of experience. Treating awareness operationally, as the crossing of a threshold that can be dissociated from objective sensitivity, is what lets it be measured (Seth et al., 2008).

What is perception without awareness?
Perception without awareness is the processing of a stimulus that influences behavior while the observer reports no conscious experience of it. It is demonstrated most convincingly not by a bare denial of seeing but by a qualitative dissociation, an unconscious effect that conscious perception would not produce, such as priming from a stimulus the observer discriminates at chance (Merikle et al., 2001).

What is the difference between the objective and subjective thresholds?
The objective threshold is the stimulus strength at which forced-choice discrimination first rises above chance; the subjective threshold, which is higher, is the strength at which the observer begins to report seeing the stimulus. Between them lies a band in which the observer discriminates a stimulus better than chance while reporting no awareness of it (Cheesman & Merikle, 1986).

What is blindsight?
Blindsight is residual visual ability in a region of the visual field that a patient reports as blind, following damage to primary visual cortex. Such patients discriminate the position, motion, and orientation of stimuli they insist they cannot see, demonstrating that visual processing can survive the loss of visual awareness (Weiskrantz et al., 1974).

How is awareness measured?
Awareness is measured by asking the observer, in ways designed to resist bias. Subjective scales such as the perceptual awareness scale ask directly about experience; post-decision wagering makes the observer bet on a judgment so that winnings index conscious knowledge; and confidence ratings quantify metacognitive sensitivity, the degree to which confidence tracks accuracy (Ramsoy & Overgaard, 2004; Persaud et al., 2007; Fleming & Lau, 2014).

Is awareness gradual or all-or-none?
It depends on the stimulus and, crucially, on the measure. A yes-or-no report produces an apparent step, but a graded scale reveals intermediate states of partial awareness that the binary measure hides, and the graded scale also proves the most sensitive index overall (Overgaard et al., 2006; Sandberg et al., 2010).

Does an unseen stimulus get processed for meaning?
Sometimes, but the depth depends on how the stimulus is hidden. Masked stimuli are processed beyond their physical features, up to the level of meaning, but this falls off sharply as the mask strengthens, and the strongest claims of high-level unconscious cognition are the least robust (Kouider & Dehaene, 2007).

How does awareness relate to consciousness?
Awareness is the content-bound component of consciousness, the being-conscious-of a particular stimulus, held apart from the global state of being awake and from the general question of why any processing is felt. Studying awareness is how the broader problem of consciousness is made experimentally tractable (Dehaene et al., 2006).

References