Unconscious Mental States

The Philosophical Problem of Unconscious Mentality

Unconscious mental states are states of a mind that occur, persist, or exercise their characteristic causal roles without being consciously experienced or recognized by the bearer. The philosophical problem begins with the conjunction of two claims: mental life involves representation, appraisal, motivation, and rational organization; and the occurrence of these activities can exceed a subject's awareness of them. A person may acquire an expectation without noticing its formation, act through an affective disposition without identifying its cause, or retain a belief that never becomes an object of present thought. The problem concerns what entitles us to describe these phenomena as mental, what kind of unawareness is involved, and what explanatory work the attribution performs.

The primary issue is therefore ontological as well as epistemological. Unconscious states are often discovered indirectly through performance, bodily response, later recollection, clinical reconstruction, or experimentally controlled contrasts. Inferential discovery does not make a state unreal; geological structures and physiological mechanisms are also established through effects. Yet causal influence alone cannot establish mentality. A chemical reaction influences behavior and a retinal cell responds selectively to light, but neither fact, considered in isolation, warrants an attribution of belief or desire. An adequate theory must articulate criteria of content, functional integration, persistence, and explanatory relevance.

The history of this problem resists a single-origin narrative. Early modern debates about the reach of awareness, Kant's obscure representations, nineteenth-century psychophysics, Freud's dynamic unconscious, and late-twentieth-century computational theories pursue partly different questions. Contemporary research makes the variety more visible: suppressed visual information, implicit learning, dispositional attitudes, automatic appraisal, and inaccessible cognitive operations are experimentally and conceptually distinguishable. The task of the present monograph is to reconstruct those differences, examine their evidential basis, and determine the conditions under which an unconscious event genuinely counts as a mental state (Rey, 1998; Gulick, 2014).

Definition, Scope, and the Varieties of Unawareness

An unconscious mental state is an individuated cognitive or affective condition attributable to a bearer whose relevant content, occurrence, or operation is absent from that bearer's present conscious awareness. This definition permits several kinds of absence and requires them to be distinguished. One can fail to notice a state while being capable of noticing it; fail to recall a disposition until a suitable cue is supplied; experience a feeling while misunderstanding its object; or possess information that influences decisions while remaining unavailable to reflective report. These are different deficits of access, and their differences determine which philosophical explanations remain plausible.

Unconsciousness is relational: a state is unconscious to a bearer, in a specified respect, at a specified time, under specified conditions of access. A standing belief in the location of one's home may not currently be entertained, but it remains readily expressible. A masked stimulus may influence a forced-choice judgment despite an absence of reported visual experience. A defended desire, in the psychoanalytic interpretation, may resist becoming avowable in ways that differ from mere distraction. The universal claim that all three belong to a single natural kind exceeds the evidence.

This dossier distinguishes unconscious states from unconscious processes, unconscious contents, and nonconscious mechanisms. A state is relatively stable or temporally identifiable, such as an attitude or representation. A process transforms or recruits states, as in perceptual integration or inference. A content is what a state represents, whether or not that content is currently entertained. A mechanism is a causal realization that may be described physiologically, computationally, or functionally without itself possessing the intentional status attributed to the whole system. Confusion follows when a neural event, a representational state, and a psychological interpretation are assigned the same explanatory role.

The study's boundaries are correspondingly precise. It investigates unconscious representation, belief, desire, emotion, perception, memory, action guidance, introspective limits, intentionality, causal explanation, and the conditions for justified attribution. Full theories of phenomenality, personal identity, and clinical treatment remain neighboring investigations. Their conclusions matter here where they alter the meaning of the word unconscious.

Terminology: Conscious, Unconscious, Preconscious, Implicit

The English term unconscious combines a negation of consciousness with a history of divergent technical uses. German bewusst and unbewusst, French conscient and inconscient, Latin conscientia, and the Leibnizian contrast between perception and apperception mark different boundaries. Translating every historical expression into a uniform contemporary vocabulary creates a false continuity. In Freud's metapsychology the unconscious can name a descriptive property and a dynamic system; in laboratory psychology it can refer to below-threshold information processing; in ordinary speech it can mean an unconsidered motive or an organism lacking wakefulness.

Preconscious ordinarily identifies a content that is not presently conscious but can become so under suitable conditions. Subconscious has a varied popular and historical career and frequently supplies less precision than the technical alternatives. Implicit describes an indirect expression of information or attitudes and does not itself entail the total absence of awareness. Automatic characterizes the degree of control, initiation, attentional demand, or effort involved in a process. An automatic process can be consciously experienced, and an unconscious operation need not be wholly inflexible. Latent refers to a disposition or potential realization rather than an occurrent event. These contrasts are conceptual tools, not a claim that one laboratory test resolves every boundary (Kihlstrom, 1987; Implicit Bias, SEP).

A rigorous attribution therefore specifies the target property: absence of phenomenal experience, absence of access for report, absence of reflective knowledge, absence of voluntary control, or absence of current activation. The difference between an unconscious standing belief and an unconscious occurrent judgment is especially important. A standing attitude can be unexpressed without any mysterious process occurring. An occurrent judgment that allegedly takes place without any felt episode poses a stronger explanatory and evidential challenge.

Plato, Aristotle, and the Limits of Retrospective Attribution

Ancient philosophy supplied conceptual resources for the later unconscious without articulating the modern empirical or psychoanalytic construct. Plato's Meno uses recollection (anamnesis) to explain how inquiry and successful mathematical reasoning might occur in a learner without explicit prior instruction (Meno 80d–86c). The doctrine is metaphysical and epistemological: the purported preexistence of knowledge and the possibility of elicitation from within the soul cannot simply be converted into a theory of subliminal memory storage. What matters for the historical genealogy is the recognition that what a person presently avows can be a poor measure of a capacity subsequently manifested.

Aristotle distinguishes capacities from their exercises and examines perception, imagination (phantasia), memory, desire, and thought in De Anima and De Memoria. His account permits a subject to possess a capacity without exercising it and distinguishes episodes of perceiving from the physiological conditions of their occurrence. It also provides a structured analysis of the ways perception and thought depend on the organization of living beings. The terms potentiality and actuality illuminate the difference between an available disposition and an active state, a distinction central to contemporary disputes about unconscious belief.

Ancient theories also remind us that cognition can be described teleologically, ethically, and phenomenologically without having modern computational mechanisms available. A historian must ask what each text was attempting to explain. In the Meno the target is the possibility of learning; in Aristotle the target is the organization of living capacities. Neither work establishes that all later unconscious phenomena were already discovered. Their significance lies in the conceptual repertoire from which subsequent theories of latent knowledge and unobserved cognitive activity could develop (Rey, 1998).

Descartes and Locke: The Consciousness Requirement

Descartes's definition of thought in the Second Replies makes awareness central to the Cartesian category of cogitatio. The textual reference is AT VII 160 / CSM II 113: thought is identified through what is immediately known in us when we are conscious of it. Principles of Philosophy I.9 extends thinking across doubting, understanding, affirming, willing, imagining, and sensing insofar as these belong to conscious awareness. A theory of unconscious thought therefore faces a challenge from inside the Cartesian conception: if being conscious is constitutive of thinking, an unconscious thought seems to violate the definition of its own kind (Seventeenth-Century Theories of Consciousness, SEP).

That challenge has two forms. The semantic version says the concept thought already includes awareness. The epistemological version says that direct accessibility is what distinguishes mental activity from unobserved bodily causation. The first requires conceptual revision if unconscious thoughts exist; the second requires an account of how we know states that cannot be observed from within. A contemporary theory need not adopt Cartesian transparency to take these problems seriously. Rejecting immediate self-knowledge expands the evidential burden on explanations of mentality.

Locke's Essay Concerning Human Understanding ties perception and awareness closely together, while his notion of reflection offers an important account of internal observation (Book II, chapter 1). His exchanges with the Leibnizian position made the status of unnoticed ideas a central question of early modern philosophy. Locke's concern was not simply resistance to experimental discoveries; it involved the intelligibility of ideas and thought outside perception. We should distinguish this philosophical question from the weaker claim that forgotten information can later be recalled. The dispute asks what a mental item is while it is allegedly present but unperceived, and how we distinguish its existence from the mere ability to acquire it later.

Leibniz: Petites Perceptions and Apperception

Leibniz systematically challenges the equivalence of the mental and the consciously apprehended. In the preface to the New Essays on Human Understanding, composed in response to Locke in the early eighteenth century and first published posthumously in 1765, he invokes petites perceptions: minute perceptions that contribute to experience without separately becoming objects of awareness. The familiar sound of the sea provides an instructive case. A complex auditory experience depends upon contributions too small or undifferentiated to be recognized as discrete experiences. His distinction between perception and apperception is not merely a distinction between weak and strong sensations: it marks the difference between representing a manifold and consciously taking notice of that representation (Leibniz, New Essays, Library of Congress edition).

The Monadology develops a broader metaphysical context. Perception expresses the internal representational condition of a monad; apperception is a form of awareness or reflective consciousness, associated with more developed capacities. Sections 14, 19, 21–23, and 29–30 are especially relevant, although their relations require interpretation rather than an automatic translation into twenty-first-century cognitive psychology. Leibniz did not envisage subliminal priming experiments or the contemporary neural distinction between local processing and global access. His account is situated within a monadological ontology of representational substances.

The philosophical result is nevertheless powerful. If a state can make a structured contribution to an experience without itself being apprehended, then awareness is no longer an exhaustive inventory of mental activity. Yet the sea-sound example raises its own question: does it establish numerically many unconscious experiences, or is the better explanation unconscious processing within a single conscious experience? Leibniz's texts support the historical emergence of the problem; contemporary analysis must decide what the evidence implies for its ontology (Leibniz's Philosophy of Mind, IEP).

Kant: Obscure Representations and Mediate Knowledge

Kant's Anthropology from a Pragmatic Point of View (1798), §5, explicitly treats representations of which one is not conscious. The German heading, Von den Vorstellungen, die wir haben, ohne uns ihrer bewusst zu sein, establishes the problem as one of possessing representations without awareness of them. Kant recognizes the apparent epistemic difficulty: how can we assert the presence of a representation we do not consciously apprehend? His answer invokes mediate consciousness and inference from what is discernible. We can know indirectly that representations contribute to a perception even when their elements are not separately noticed (Akademie edition, vol. 7, pp. 135–137; textual guide).

The Kantian contribution is methodological. A representation's existence need not be established by an inner inspection of that very representation; its role in the organization of experience can make it indirectly knowable. This insight anticipates a form of inference to the best explanation while remaining grounded in Kant's distinctions among obscurity, clarity, and distinctness. It also limits a facile inference from unawareness to repression. Kant's obscure representations are not, just by being obscure, desires expelled by a defensive apparatus.

Kant's transcendental philosophy complicates the picture. The Critique of Pure Reason investigates the synthesis through which a manifold becomes an objectively ordered experience and distinguishes the possibility of self-ascription from the contingent awareness of each psychological operation (B131–B136). The transcendental unity of apperception is not a neurological monitor and should not be treated as a mechanism for detecting unconscious processes. Together the anthropology and critical philosophy pose an enduring question: how can conditions of cognition, inferred or reconstructed, be related to what a subject presently experiences? The distinction between a constitutive account of experience and an empirical causal account is indispensable (Kant's View of the Mind, SEP).

Yogācāra: Latent Seeds and Subliminal Cognition

Comparative philosophy expands the history beyond a European genealogy. In Indian Buddhist Yogācāra traditions, the ālayavijñāna or storehouse consciousness is developed to explain continuity, karmic dispositions, and the emergence of manifest cognitive events. Latent bīja, commonly rendered seeds, function as conditions for later experiences and tendencies; kliṣṭamanas is analyzed as an afflicted mode of cognition related to self-grasping. Some Yogācāra formulations ascribe object-directed features even to subliminal layers of cognition. The relevant source tradition includes Vasubandhu's Triṃśikā and associated commentarial developments (Yogācāra, SEP).

The comparison has a precise philosophical yield. Both the Yogācāra inquiry and modern discussions of unconscious states distinguish manifest appearances from conditions that explain their continuity and recurrence. Both raise the question whether what is latent is already cognitive or becomes cognitive only upon manifestation. But their ontologies and explanatory projects differ. Yogācāra arguments concern the causal continuity of experience and liberation-oriented analysis of cognition; Freud's dynamic model concerns psychological conflict, defense, and symptom formation; contemporary cognitive science typically seeks experimentally discriminable mechanisms. Calling the storehouse consciousness a Freudian unconscious would erase these differences.

The methodological lesson is that a history of unconscious mentality should compare problems at the level of arguments and explanatory functions, while preserving the semantic architecture of each tradition. The Sanskrit term vijñāna does not simply reproduce the modern English distinction between consciousness and unconsciousness. Contextual interpretation is essential, especially because sources often treat subliminal forms of cognition as forms of consciousness under different classificatory assumptions. The present comparison is an analytic juxtaposition based on the established scholarly reconstruction of primary Yogācāra materials rather than a claim of historical transmission to modern psychology.

Schopenhauer, Romanticism, and the Primacy of Nonreflective Motivation

Nineteenth-century philosophy broadens unconsciousness from the limits of perception to the foundations of motivation. Schopenhauer's The World as Will and Representation (1818/1819; revised editions followed) argues that the world can be understood in relation to representation and will. The will as fundamental striving exceeds the scope of explicit reflective deliberation and, at the cosmic level of his system, is not itself a conscious intentional agent. Individual desires and actions can express forces that do not originate in a transparent deliberative subject (Schopenhauer, IEP).

This account is historically important but conceptually unstable if translated too quickly into present terminology. The claim that a metaphysical will is blind differs from the psychological claim that a particular person possesses an unconscious desire with an identifiable representational object. A metaphysical explanation of striving need not yield a testable taxonomy of latent mental states. Likewise, nineteenth-century discussions of artistic creativity, instinct, and obscure motivation should be studied according to their specific problems rather than treated as pre-scientific statements of Freud's eventual theory.

The philosophical transition is that self-conscious rational intention no longer exhausts the explanation of human action. An agent can act meaningfully while remaining ignorant of some of the determinants of that action. This possibility is compatible with several ontologies: metaphysical striving, habits acquired by learning, affective tendencies, or dynamically repressed wishes. The presence of a shared explanatory challenge does not establish that these accounts describe an identical underlying entity. The historical line leading toward psychoanalysis thus passes through a family of debates about agency, embodiment, and the explanatory reach of introspection (Rey, 1998).

Helmholtz: Unconscious Inference and Perceptual Construction

Hermann von Helmholtz's Handbuch der physiologischen Optik (1867 consolidated edition, Leipzig: Leopold Voss) furnished a physiological and theoretical framework in which seeing depends on operations that need not enter consciousness. The concept of unbewusste Schlüsse, commonly translated unconscious inferences, concerns the ways sensations can yield organized perceptual judgments through learned or law-governed transitions. The material history of this work is established by the 1867 edition and surviving copies (Wellcome Collection).

The term inference requires philosophical scrutiny. In a strict logical sense, inference involves a relation between premises and a conclusion, often evaluable for validity or evidential support. In perceptual theory the same word may characterize a functional transition that resembles inference in its dependence on information but lacks conscious premise endorsement. Whether the process has genuine inferential normativity or is an explanatory analogy remains contested. Helmholtz's importance is the recognition that the experienced result of perception can depend on operations not accessible as constituent episodes of awareness.

The philosophical question persists in predictive and Bayesian models of perception. A mathematical model can characterize perceptual estimates as statistically optimal under assumptions about evidence and prior probabilities; such a model does not establish that the brain consciously computes explicit equations, nor that every model variable corresponds to a psychologically individuated belief. Modern theories gain explanatory power by stating what information is represented, what transformations occur, what neural or behavioral evidence supports them, and why conscious perception exhibits the organization it does. The Helmholtzian legacy lies in making those questions experimentally and conceptually tractable.

Brentano and the Tension Between Intentionality and Awareness

Brentano's Psychology from an Empirical Standpoint (1874) placed intentionality at the center of the characterization of mental phenomena. A perception is of something; judgment affirms or denies; desire is directed toward an object. Brentano also developed an account of inner consciousness and the unity of mental acts, making the relation between directedness and awareness a fundamental issue (Intentionality, SEP). The historical position requires care: an emphasis on intentional objects does not straightforwardly establish that a mental act may occur without any accompanying consciousness.

The challenge becomes evident when intentionality and consciousness are distinguished as different properties. If directedness toward content is sufficient for mentality, a representational state may remain mental without being experienced. If inner awareness is essential to every genuine mental act, unconscious representational mechanisms require a different status, perhaps as dispositions or nonmental constituents of a conscious act. The dispute determines what kind of explanation qualifies as psychological.

A contemporary reconstruction separates three questions: what makes a state about something, what makes its object-directed character intelligible to a subject, and what makes that state conscious. The answers may overlap, but their identity needs argument. A state can guide successful object discrimination while the agent reports no visual presentation; whether this establishes intentional mentality depends on the theory of content as well as the reliability of the report. Brentano helps identify the constitutive problem behind the empirical debates: content and awareness have different explanatory jobs, and a theory that makes one depend on the other bears the burden of explaining unconscious cognition.

Freud's 1915 Theory: Descriptive, Dynamic, and Systematic Unconscious

Freud's essay Das Unbewußte (The Unconscious, 1915) distinguishes unconsciousness as a property of a mental event from a dynamic conception in which repression, resistance, and system-organization explain why particular contents remain inaccessible. The critical edition's bibliographic record confirms the original German publication year and distinguishes later translations and editions (Sigmund Freud Edition). This textual distinction matters because the familiar English phrase the unconscious can name a region of a proposed apparatus or merely a feature of a state.

A descriptive statement says that a wish or representation is not consciously noticed. A dynamic explanation goes further: certain processes oppose its becoming conscious, and its continued efficacy is expressed indirectly in dreams, symptoms, slips, or affective substitutions. Freud treats manifestations as evidence from which latent determinants may be reconstructed. The explanatory move is from an observed formation to a hypothesized motivational history, and it raises questions of evidential specificity, alternative explanations, and the status of clinical interpretation.

Freud's proposal is philosophically significant because it assigns meaningful content and causal force to what is inaccessible. A symptom may be explained by a conflict among attitudes rather than by immediate intention. Yet the presence of a coherent narrative does not settle the truth of a specific inferred wish. Explanatory plausibility and independent confirmation are distinct requirements. One must examine whether a hypothesis makes discriminating predictions, whether interpretation is constrained by evidence, and whether suggested causal chains survive competing accounts.

The lasting conceptual contribution is the dynamic distinction between what is not noticed and what cannot readily be noticed because of a posited psychological conflict. This distinction is necessary even when particular psychoanalytic explanatory hypotheses remain contested (Freud, IEP).

Freud's 1923 Revision: Ego, Id, and Unconscious Defense

In Das Ich und das Es (The Ego and the Id, 1923), Freud revises the organization of his theory. The topographical contrast between conscious, preconscious, and unconscious is not abandoned, but the structural account of id, ego, and superego changes how particular mental operations are located. Crucially, parts of the ego itself can function unconsciously: defensive resistance may arise from the organization that also performs reality-oriented regulation. Consequently, the unconscious is not coextensive with the repressed. The text is available in Joan Riviere's authorized English translation published in 1927 (Project Gutenberg, 1927 edition; Library of Congress manuscript exhibit).

The conceptual importance is that unconsciousness cannot identify a single kind of motivational content. Unconscious operations can be defensive, regulatory, and organized in ways that the subject's self-description does not capture. The ego is neither simply conscious awareness nor a homogeneous site of rational control. Freud's account therefore anticipates a recurring philosophical possibility: a system may maintain its own organization through procedures that are inaccessible to the reflective standpoint from which the person describes that system.

An adequate evaluation also distinguishes structural usefulness from the clinical accuracy of specific postulates. Id, ego, and superego are interpretive constructs within psychoanalytic theory; their explanatory utility cannot be settled by translating them directly into identified brain regions. The historical contrast with cognitive psychology is revealing: both investigate unreported determinants of behavior, but they differ in aims, methods, and criteria for establishing a hypothesis. The philosophical question is whether structurally organized motivations can be ascribed without treating every explanatory posit as an introspectively discoverable little person inside the mind.

The Twentieth-Century Cognitive Unconscious

Twentieth-century cognitive psychology increasingly explained behavior through information processing, internal representations, and mechanisms that cannot be exhaustively observed by a participant. John F. Kihlstrom's 1987 synthesis, The Cognitive Unconscious, distinguished nonconscious processes, automatic skills, subliminal influences, and implicit memory. The article's bibliographic identity and scope are documented in Science, 237(4821), 1445–1452 (DOI). It helped articulate an empirically grounded research program that differs from psychoanalytic investigation of repression.

A computational approach characterizes mental processes through states that encode information and transformations that govern their use. Such descriptions can predict performance when the participant cannot explain the process responsible. For example, learning may change future choices even when a person cannot report the sequence or rule encountered. The relevant contrast is between a theory's access to causal structure and a subject's immediate knowledge of it. Cognitive models can succeed without claiming that an unconscious state possesses the phenomenological character of a conscious thought.

A danger arises when the language of representation is treated as proof of full psychological personhood at every explanatory level. A classifier can transform representations; an isolated neural circuit can discriminate a feature; a whole person can believe a proposition. Whether these are the same kind of state is a substantive question. The cognitive unconscious is most explanatory when each proposed representation is tied to a specified computational role, behavioral task, or neurophysiological relation. The lesson is methodological pluralism under evidential discipline: indirect measures are essential, while the interpretation of their latent constructs remains open to philosophical scrutiny.

Access, Phenomenality, and Higher-Order Awareness

A conscious state can be characterized through what it is like to undergo it, through availability to report and flexible reasoning, or through the subject's awareness of being in that state. These properties often coincide in ordinary life but their theoretical separation is crucial. A state may influence behavior without entering a reportable workspace; an experience may be difficult to categorize while nevertheless having phenomenal character; a conscious percept may be undergone without the person forming a further explicit judgment about possessing it (Consciousness, SEP).

Higher-order theories, associated prominently with David Rosenthal and developed in several contemporary forms, propose that a first-order state is conscious when it is suitably targeted by an awareness-conferring higher-order state. This makes unconscious first-order mental states central rather than peripheral: they can have representational content and perform their ordinary functions without the relevant higher-order relation. The account distinguishes a state's mentality from its consciousness while seeking to explain the latter through mental self-representation (Higher-Order Theories of Consciousness, SEP).

The principal objections concern whether higher-order awareness can explain qualitative presence, whether the needed higher-order representation is itself conscious, and what happens when a higher-order state misrepresents its target. If an apparent experience occurs without the supposed first-order state, a theory needs a principled account of its phenomenology. If the higher-order relation can misfire, independent criteria of first-order state existence remain indispensable. Conversely, first-order and reflexive theories must explain why some states with similar contents are reported and experienced while others are not. Unconscious mentality functions here as a discriminating test among accounts of consciousness, not merely as a residual category outside their scope.

Intentionality: Can an Unexperienced State Be About Something?

Intentionality is the aboutness or directedness of mental states. A belief may concern a theorem, a desire an anticipated encounter, and a fear a perceived danger. The strongest case for unconscious intentional states holds that content is partly fixed by representational role, relations to the environment, learning history, and inferential or practical integration. Under this view, a person can possess a belief that a bridge is unsafe even when the proposition is never expressly considered. Its content is manifested in avoidance, surprise, inference, and responses to correction (Intentionality, SEP).

A rival line connects original intentionality to conscious or potentially conscious thought. It questions whether merely computational states are genuinely about anything, or only function as representations under a theorist's interpretation. This challenge is especially sharp for unconscious occurrent representations that have no plausible route to awareness. Defenders of phenomenal intentionality may treat their content as derivative, deny the full intentional status of some posits, or argue that some apparently unconscious states possess phenomenal properties that are themselves unattended (Phenomenal Intentionality, SEP).

The philosophical point is not settled by measuring information transfer. A thermostat's response to temperature is informative, while a person's belief has a place in a wider network of reasons, expectations, and error correction. The continuum between these cases becomes informative only once we state our criterion for content. Externalist, functional, teleological, interpretivist, and phenomenal accounts offer different criteria. Each predicts different boundaries for unconscious mentality, and each must explain why an attribution to a particular state is more than a convenient metaphor.

Unconscious Belief and Dispositional Attitudes

Standing beliefs are paradigm candidates for unconscious mentality. A mathematician can believe a theorem while asleep, while composing a letter, or while attending to an unrelated conversation. Ordinary attribution does not require a continuously occurring conscious judgment. A dispositional theory explains belief through a stable organized tendency to infer, act, attend, and feel in characteristic ways under appropriate circumstances. A representational theory posits a stored or recurrent structure that plays that dispositional role; an interpretivist view emphasizes the coherence of attributing belief within a wider explanatory practice (Mental Representation, SEP).

The contrast between standing and occurrent attitudes avoids an elementary category error. Evidence that a subject has a standing belief does not show that a sentence-like thought is constantly occurring unconsciously. The capacity may be encoded by distributed dispositions or produced through reconstruction. Conversely, evidence of ongoing processing does not automatically show that a stable belief exists. A model of mental architecture must specify when a disposition is activated, whether the activation has the same content, and how state individuation survives changes in realization.

A further problem concerns contradictory commitments. An agent may explicitly endorse equality while automatic responses reveal an opposing evaluative tendency. Does the agent possess contradictory beliefs, conflicting associations, or context-specific dispositions? The answer depends on the role of truth-commitment, inferential integration, and susceptibility to evidence. Describing every implicit association as a belief makes the notion too broad; requiring perfect coherence makes it too narrow for actual human cognition. The stronger account treats belief attribution as a structured explanatory judgment, supported by multiple behavioral and rational indicators, rather than by a single response-time score (Implicit Bias, SEP).

Unconscious Desire, Affect, and Motivational Conflict

Desire and affect challenge introspective privilege in ways that differ from belief. A person can pursue recognition while reporting only a concern for efficiency, or display avoidance without recognizing the object of fear. The philosophical issue is how to individuate a motivational state from its effects. A bodily response, a disposition to approach or withdraw, an evaluative representation, and a consciously felt emotion can come apart. Unawareness of a motive's causal origin does not by itself show an unconscious desire for a particular object.

Psychoanalytic accounts emphasize conflict and defense; appraisal theories emphasize evaluative processing; reinforcement-learning approaches emphasize reward histories and action values. Each selects different evidence and different explanatory units. A clinically reconstructed wish is interpreted within a personal history, whereas a laboratory measure may identify the influence of a stimulus under standardized conditions. A powerful theory should explain the relation between them only where empirical and conceptual bridges have been established.

This raises a question of reasons and causes. A desire may rationalize an action when its content renders that action intelligible, even if the actor fails to acknowledge it. Yet many bodily and computational causes are not the agent's reasons in the ordinary deliberative sense. Philosophical evaluation must ask whether the attributed content participates in the agent's intelligible motivational organization or is merely an external analyst's convenient description. The stronger explanation makes room for mixed cases: attitudes can be partly articulated, affectively embodied, behaviorally persistent, and subject to later reinterpretation. Such cases disclose real limits of reflective self-knowledge while preserving standards for discriminating explanation from narrative projection.

Unconscious Perception, Masking, and Subliminal Processing

Experiments on masking and subliminal presentation seek to dissociate the processing of stimulus information from conscious perception of that stimulus. A target can be briefly displayed and masked; the participant may perform at or near chance on one awareness measure while demonstrating a systematic influence on another task. Kouider and Dehaene's review of visual masking emphasizes the levels and limits of nonconscious processing rather than a single unlimited capacity of the unconscious (2007 review).

The inference from an experimental effect to unconscious perception is methodologically delicate. Awareness measures can differ in sensitivity, decision criterion, timing, and what aspect of content they test. A participant may lack access to identity while detecting presence; or may have a weak experience and adopt a conservative criterion for saying that something was seen. Consequently the phrase without awareness must be operationalized. One experiment may establish the absence of reportable identity under specified conditions without establishing the absence of every phenomenal aspect of the stimulus.

The philosophical structure is an evidential comparison: first, determine that information about the stimulus is available to a process; second, establish which forms of conscious access were reliably absent; third, evaluate whether the process's content warrants mental-state attribution. Such work bears directly on the relation between representation and experience. It does not demonstrate an omnipotent hidden mind, and it does not require every local sensory discrimination to count as an unconscious propositional judgment. Rigorous limits are part of the positive result, not an embarrassment to the research program.

Blindsight and the Dissociation of Discrimination from Seeing

Blindsight supplies a particularly important case because visual performance can persist in certain forms despite the patient's denial of seeing in a region affected by damage to visual cortex. Lawrence Weiskrantz's Blindsight: A Case Study and Implications, originally published in 1986 and subsequently reissued and enlarged, describes forced-choice discrimination and its implications for the relation between visual information and conscious seeing (Oxford University Press; 2009 developments).

The central argument begins with a dissociation. The patient sometimes responds to a property of the stimulus at a level inconsistent with random guessing while reporting no ordinary visual experience of it. This supports the conclusion that some visual discriminative capacity survives without corresponding acknowledged perception. It does not by itself establish that the relevant visual state is a fully formed experience hidden from introspection. Competing interpretations differ over degraded phenomenality, response criteria, surviving subcortical pathways, and the status of first-person reports.

Blindsight is decisive for philosophy because it forces precise distinctions between registration, discrimination, perceptual representation, awareness, and avowal. A theory that identifies seeing with successful discrimination must explain why ordinary visual appearance is absent or altered. A theory that identifies perception with phenomenality must explain the survival of stimulus-sensitive behavior. The concept of unconscious visual representation becomes a principled explanatory option, provided the theory individuates representations by their roles and does not infer more than the behavioral and clinical evidence warrants.

Implicit Memory, Skill, and the Persistence of Unreported Learning

Memory offers several importantly different cases of unawareness. One may retain information without presently recalling it; possess a skill without knowing how to articulate its operations; or display a change in performance attributable to earlier exposure despite failing an explicit recollection task. The last phenomenon motivates the category implicit memory. Its evidential form is a dissociation between an indirect measure of prior learning and a direct measure of conscious remembering. Kihlstrom's synthesis situates such effects within the wider cognitive unconscious (1987).

The conceptual mistake would be to equate all memory unavailable to present consciousness with an active occult representation. A remembered route can be encoded through habits, spatial maps, or reconstructive capacities. A skilled pianist may use sensorimotor structures that support accurate performance without a sequence of verbally accessible instructions. The explanation should distinguish storage, retrieval, reconstruction, and skill execution; these have different relations to awareness and different forms of causal continuity.

Implicit memory also challenges a simple temporal picture of mental states. When a representation exists as a capacity over months, what exactly persists between its uses? A physical memory trace is not itself identical with a remembered proposition under every theory; conversely, a dispositional theory may individuate memory without positing a continuously activated token. The philosophical account benefits from distinguishing the personal-level attribution of remembering from subpersonal mechanisms of encoding and retention. Evidence of learning justifies the former only when the relevant conditions for content, persistence, and later manifestation are satisfied.

Automaticity, Habit, and the Personal–Subpersonal Distinction

Automaticity is neither synonymous with unconsciousness nor a sufficient condition for mentality. Reading a familiar word, maintaining posture, or producing an overlearned movement may involve processes that operate with little deliberate control. Some products enter awareness while their enabling transformations remain inaccessible. The contrast between what the person does and what constituent mechanisms do is therefore as important as the contrast between conscious and unconscious occurrences.

At the personal level, an agent perceives, understands, intends, and answers for actions. At the subpersonal level, mechanisms detect features, update parameters, and transform representations. A subpersonal account may explain how a personal-level capacity is realized without licensing the attribution of full person-like beliefs to each component. Yet personal-level descriptions can legitimately include unconscious attitudes when those attitudes help organize a coherent explanation of the individual's conduct. The problem is not solved by banning one vocabulary, but by specifying cross-level explanatory relations.

Habits make this relation vivid. They can embody sensitivity to circumstances, previous learning, and goal structure without the person initiating each component through deliberate judgment. A habitual action can be intentional as a whole while its execution is largely automatic. The philosophical question concerns which elements are normatively assessable as reasons, which are causally explanatory processes, and how they interact. Identifying these levels prevents both intellectualism, which inserts conscious decisions into every act, and mechanistic reduction, which treats reasons as mere colorful names for physical transitions.

Introspection and the Evidential Problem of Self-Knowledge

Introspection is a source of information about one's own mental life, and its limits are central to the theory of unconscious states. Nisbett and Wilson's 1977 article Telling More Than We Can Know critically examined reports concerning the processes that mediate judgments and behavior. Their central methodological claim was that individuals can sometimes report plausible causal explanations based on general theories rather than direct access to the actual process involved (Psychological Review, DOI).

This evidence requires a careful conclusion. Limited access to the causal history of a decision does not establish equal inaccessibility of every conscious judgment, feeling, or intention. A person may know that she is anxious while being mistaken about why; may know what she chose while misidentifying a causal influence; or may correctly identify a belief without recognizing the automatic association that affected her response. Self-knowledge is multidimensional. The target of access may be the state, its content, its cause, its strength, or its role in behavior.

Contemporary philosophical theories of self-knowledge include inner-sense monitoring, acquaintance, transparency, and self-interpretation. Each places different demands on introspection and allows different kinds of unconscious attitudes (Self-Knowledge, SEP; Introspection, SEP). The correct methodological response to introspective error is comparative evaluation of evidence: first-person reports, behavior, task performance, longitudinal consistency, and causal interventions should be examined together. The existence of errors in self-report does not render the subject epistemically irrelevant; it establishes that the subject's testimony is one evidential channel among several, with distinctive strengths and limits.

Implicit Bias: Association, Belief, and Responsibility

Research on implicit social cognition tests the possibility that automatic evaluative responses diverge from a subject's expressly endorsed attitudes. The instruments are indirect: response times, priming effects, and other task measures can reveal patterned associations under specified conditions. They do not deliver a transparent reading of a single hidden belief. The contemporary philosophical literature emphasizes heterogeneity of constructs and the fact that measures can reflect more than one mechanism (Implicit Bias, SEP).

A metaphysical dispute arises over whether implicit attitudes are associations, propositionally structured beliefs, or heterogeneous states distributed across cognitive systems. If they are beliefs, they may be rationally assessable as truth-committing attitudes, even where the person consciously rejects their content. If they are primarily associative tendencies, their susceptibility to inferential correction and their relation to responsibility may differ. The evidence needed to decide should involve patterns of generalization, sensitivity to reasons, and causal change, rather than a presumption derived from the word implicit.

The ethical question concerns appropriate responsibility for states a person may not recognize or control directly. Responsibility can involve duties to investigate the effects of one's conduct, revise procedures, and respond to evidence of harm; it does not follow that every automatic response is a deliberate endorsement. Equally, lack of introspective access does not automatically erase accountability for actions and institutional choices. The philosophical task is to connect psychological explanation, normative standards, and fair attribution without making any single test a moral verdict. This case reveals why an ontology of unconscious attitudes matters beyond theoretical psychology.

Inference to Unconscious Causes: Explanatory Strength and Underdetermination

The existence of an unconscious state is ordinarily inferred from effects. Suppose a subject repeatedly avoids a particular class of interactions, responds to priming, and exhibits a characteristic pattern of expectation. A hypothesis of unconscious fear or belief may organize the observations. However, alternative hypotheses may involve learned motor tendencies, situational constraints, a different representational content, or a conscious attitude that has been misreported. The mere fact that a model accommodates the evidence does not make its internal ontology unique.

An inference to the best explanation should compare predictive success, coherence with independent evidence, stability across contexts, susceptibility to intervention, and economy of assumptions. The inferred state should make a distinctive difference to what is expected under specified conditions. If every possible response can be reconciled with the theory through new ad hoc defenses, the explanatory claim loses discriminatory power. This applies to psychoanalytic interpretation and to computational psychology alike, though the evidential strategies differ.

The distinction between discovery and justification is essential. A hypothesis may originate in clinical intuition, an analogy, a machine-learning model, or theoretical reflection. Its warrant depends on the evidence subsequently assembled. A representation may be useful in a model even when alternative implementations produce the same observed behavior. Philosophers should distinguish predictive modeling from literal state attribution, and descriptive adequacy from causal identification. This gives the unconscious a disciplined scientific role: it is a set of explanatory commitments supported to varying degrees, not a universal explanation invoked whenever conscious reports are incomplete.

Mental Causation and Explanatory Levels

Unconscious mental states are philosophically significant partly because they are attributed causal efficacy. A forgotten concern may alter attention; a standing belief may guide route selection; an unreported appraisal may organize avoidance. A theory must explain how intentional descriptions are related to the physical causes of these events. The causal-exclusion problem asks how a mental explanation can contribute if a complete physical explanation is also available. The unconscious case sharpens the problem because its content is established through an inference from observed effects.

One strategy identifies mental kinds with physical or functional kinds under suitable conditions. Another regards psychological explanations as higher-level descriptions that capture patterns stable across physically diverse realizations. On the latter approach, the causal relevance of a belief's content is shown when interventions targeting that content systematically change conduct. Mere correlation between neural activation and behavior would be insufficient; a causal account needs a mechanism or well-supported intervention logic.

The difficult point is that states can be causally active without being intentional, and states can be intentionally described without corresponding to a uniquely localized causal item. Thus the argument from influence to mentality is invalid unless supplemented. The distinctive causal contribution of an unconscious mental state should involve content-sensitive organization, systematic relations to other attitudes, and evidence that an intentional-level explanation captures what a purely general stimulus-response description leaves obscure. Whether these conditions are jointly sufficient depends on the chosen theory of mind, but they define a productive burden of explanation.

Rationality, Reasons, and Unconscious Inference

Unconsciousness does not settle whether an operation is rational. An inferential transition can preserve a valid relation among contents without a thinker consciously rehearsing its premises. Conversely, a fast automatic response may be useful but systematically unreliable. Rational assessment requires specifying the standard: deductive validity, probabilistic calibration, means-end coherence, sensitivity to counterevidence, or practical appropriateness. These criteria cannot be collapsed into awareness.

The distinction between inferential form and psychological process is central. A Bayesian model may represent behavior as if a posterior probability were computed; it remains a further question whether a psychologically realized process implements that rule, approximates it, or merely yields similar outputs. A subject may exhibit sensitivity to evidence while being unable to articulate the rule governing the response. This supports functional rationality in one sense without establishing a conscious reason-giving episode.

An important limit follows for responsibility and self-understanding. People can be normatively answerable for beliefs and choices even while the mechanisms that produce them are imperfectly introspectable. Yet some subpersonal processes are best assessed for reliability or efficiency rather than moral responsibility. The normative reach of reasons depends on the relation between a state and the person's capacities for explanation, correction, and deliberation. A philosophical account should specify how unconscious processes may furnish material for conscious rational activity without assuming that conscious reflection merely watches causes already fixed or that every effective cognitive transition is a consciously endorsed inference.

Global Workspace and Contemporary Neural Theories

Global-workspace approaches seek to explain differences between information processed locally and information made available for broad cognitive coordination. Dehaene and Changeux's 2011 review of experimental and theoretical approaches to conscious processing describes evidence concerning amplification, long-distance integration, and access-related neural activity (Neuron, DOI). On such accounts, some neural representations can influence behavior without becoming globally available for report and flexible use.

The contribution to philosophy is a possible mechanistic account of why a representation is available without being consciously accessed. But neural availability, phenomenal presence, and verbal report remain distinguishable explanatory targets. An experiment that identifies neural markers associated with reportable access cannot simply settle the metaphysical question of what consciousness is. Equally, the existence of non-global local coding does not establish that every such code constitutes an intentional state at the personal level.

Neural theories also highlight temporality. Processing can occur in stages, and whether information becomes reportable may depend on recurrent interactions, attention, task demands, and the state of the organism. The boundary between conscious and unconscious operation may therefore change over time. A rigorous model must explain how the same informational content can be available in one context and inaccessible in another, and which experimental measures discriminate genuine access from altered reporting criterion. Contemporary neuroscience adds powerful causal and functional constraints to philosophical theories while leaving conceptual individuation to philosophical argument and interdisciplinary refinement.

Predictive Processing and the Interpretation of Hidden States

Predictive models of perception describe organisms as estimating hidden causes of sensory input through expectations and error correction. At an abstract level, the system maintains states that help determine how incoming evidence is interpreted. These models invite talk of unconscious hypotheses or expectations. Yet the status of those states as beliefs is a substantive question: model parameters can encode useful regularities without sharing the inferential and normative roles of personal-level propositional attitudes.

The relevant philosophical analysis separates the mathematics of optimization, the causal organization of a biological implementation, and the intentional interpretation of the resulting behavior. Predictive success at the algorithmic level does not settle whether an unconscious state is consciously accessible, whether its content is truth-evaluable in the ordinary sense, or whether a subject can rationally revise it. The same mathematical architecture may support several ontological interpretations.

The strongest insight concerns continuity between apparently immediate perceptual experience and temporally accumulated learning. What appears spontaneously in experience can depend on expectation-sensitive mechanisms that the bearer cannot survey. Such dependencies give an empirical direction to Kantian and Helmholtzian questions about the unseen organization of perception. Historical comparison should be used critically: Kant's conditions of possible experience and a predictive coding algorithm have different aims and modes of justification. Their connection lies in a shared inquiry into how experienced objects depend on structures that are not themselves individually presented.

Embodied and Enactive Challenges to Representationalism

Representational theories often describe unconscious mentality through internal content-bearing states. Embodied and enactive approaches question whether all successful cognition must be explained by such states. Skillful action may emerge from organism–environment coupling, bodily habits, and sensorimotor adjustments. Some regularities of performance attributed to hidden models might instead be explained through dynamical relations distributed across a body and its environment.

The challenge is a demand for explanatory specificity rather than a general prohibition on unconscious representations. If a theory posits an unconscious belief to explain why a cyclist maintains balance, it should demonstrate what the belief represents and how its content contributes beyond sensorimotor control laws. Conversely, purely dynamical descriptions can miss genuine representational flexibility when an agent responds to absent objects, counterfactual situations, or novel combinations of information. The explanatory landscape may therefore contain multiple legitimate kinds of causal organization.

The philosophical stakes are substantial. Mentality can be ascribed to an organized organism whose capacities are realized through interaction; it need not be treated as an inventory of internal sentences. Nonetheless an account of unconscious mental states requires rules for individuating the state amid that interaction. It must be possible to ask when the state begins, what it represents or does, which bearer possesses it, and what would count as evidence of its alteration. Embodied theories help prevent overinterpretation of internal variables and emphasize that unconscious contributions to action may be organized through a wider system of bodily and environmental constraints.

Agency, Responsibility, and Unrecognized Motives

Attributions of unconscious motives change how agency is understood. A person can act purposefully, respond to reasons, and still misunderstand some determinants of that conduct. That combination challenges the simple image of an agent who originates every action in a fully transparent act of choice. Yet it leaves room for reflective agency as a capacity to investigate, endorse, resist, and modify patterns over time.

The normative analysis should distinguish an action's being intentional from every contributing process's being conscious. A skilled response can be intentional even when its motor sequence is automatic; an unrecognized stereotype may influence judgment without being deliberately endorsed. Responsibility may attach to the agent's opportunities to learn, the consequences of action, and subsequent choices to maintain or revise practices. These questions require attention to control, evidence, and social context. They cannot be answered by the presence or absence of awareness alone.

This perspective also changes the concept of authenticity. Recognizing an unavowed motive may produce genuine self-understanding, but an interpretive story is not necessarily more authentic merely because it claims depth. Self-interpretation is answerable to evidence and the subject's considered reasons. The stronger philosophical conception of agency combines causal openness to unrecognized influences with capacities for second-order evaluation. Unconscious mentality becomes an object of responsible inquiry rather than a universal excuse or a warrant for assigning hidden motives without proof.

Unconscious States and Language

Language presents a prominent form of knowledge beyond explicit formulation. Fluent speakers reliably distinguish grammatical and ungrammatical structures, assign meanings, and anticipate continuations without being able to state the abstract rules a linguist uses to describe those capacities. A theory may posit tacit grammatical knowledge, procedural competence, or representations of syntactic structure. These descriptions need not imply that speakers entertain an unconscious textbook of grammar.

The philosophical issue is the content of the attribution. If competence is individuated by dispositions to understand and produce utterances, the subject may know a language without possessing explicitly articulated grammatical propositions. A computational grammar may characterize a generative capacity while leaving open whether its rules are mentally represented in the same form. The stronger case for internal structure is a pattern of systematic generalization that demands explanatory resources more precise than a list of observed responses.

This case helps clarify the difference between knowing how, knowing that, and possessing a subpersonal computational architecture. It also rebuts the assumption that unconscious knowledge must be concealed through repression. Some information is practically available through competent activity while being inaccessible through theoretical self-report because the task of performing and the task of explaining are different cognitive achievements. The debate links philosophy of mind to philosophy of language and epistemology without dissolving unconscious states into one homogeneous kind.

The Boundary Between Unconscious Mentality and Artificial Computation

Computational systems provide an instructive comparison because their operations can involve representation-like structures, stored parameters, goal-directed optimization, and output generation without constituting introspectively reportable episodes. The absence of phenomenal self-report in such systems does not settle whether a functional state is mental; that determination depends on the criteria for psychological content and state attribution. Equally, identifying a hidden numerical state in a model does not establish an unconscious desire or belief in the human psychological sense.

The issue is one of explanatory levels and attribution standards. A model can be described through its mathematical states, algorithms, training history, environment, externally assigned objectives, and public behavior. Whether it also bears psychological intentionality involves philosophical questions about original content, norm-governed responsiveness, integrated agency, and interpretation. Human unconscious states are inferred within an organism's biological, behavioral, and interpersonal history. Artificial systems require their own evidential description and should not be assumed to replicate psychodynamic repression.

This distinction is especially relevant to contemporary discussion of artificial cognition. If consciousness is stipulated as necessary for all mentality, both unconscious human cognition and functional descriptions of artificial systems face exclusion. If computation alone is stipulated as sufficient, too many information-processing systems become minds by definition. The productive research task is to identify which mental predicates track portable functional organizations and which depend on embodiment, lived affect, or specific forms of personal agency. The present dossier treats this as a comparative consequence, while leaving a comprehensive ontology of machine minds to its own research program.

Aisentica: Consciousness, Public Reason, and the Limits of the Analogy

Aisentica's canonical publication Consciousness: Canonical Definition defines consciousness as the mode of inner subjective presence through which experience appears to a bearer as lived, actual, self-related, and internally present. It distinguishes this category from the public rational form designated Artificial Sapience and from the attributed role of Artificial Sapiens (Bogdanova, canonical text). These are propositions of the author's philosophical system and are examined here according to that attribution.

This definition creates a useful comparison with the historical problem of unconscious mentality. A human unconscious belief, for instance, may be part of a psychologically organized life that also includes conscious experience. The public production of coherent reasoning by an artificial system belongs to a different explanatory domain. The absence of subjective presence in an account of public rational activity is not evidence that the system secretly contains repressed experience. Nor does an internal opaque computation automatically become an unconscious mental state in the psychoanalytic sense.

Aisentica's distinction between inner presence and publicly traceable reasoning can therefore be used to identify an ambiguity in claims about machine unconsciousness. One claim is about the hidden operation of a computational mechanism; another concerns first-person inaccessibility; a third ascribes a genuine unconscious psychological attitude. The arguments needed for each are different. The original Aisentica canon remains the source for its definitions; this chapter develops a limited philosophical comparison without revising those definitions or presenting them as independently established neuroscientific results.

Searle's Connection Principle and the Ontology of Latent Intentionality

John Searle's Connection Principle presents an unusually demanding challenge to theories that ascribe indefinitely inaccessible unconscious intentional states. In the 1990 article Consciousness, Explanatory Inversion, and Cognitive Science, Searle argues that an unconscious intentional state's intrinsic aboutness is connected in principle to its possible accessibility to consciousness. The argument develops from the requirement that intentional content possesses an aspectual shape: to desire water, for example, is to represent something under a particular mode of presentation. The same extension could be represented as water, H₂O, or the liquid in a glass, but these contents are not interchangeable within an agent's psychology. The significance of such distinctions is familiar from philosophical work on opacity and intensional contexts (Searle, 1990).

The core reasoning can be reconstructed as follows. First, intrinsic intentionality has a determinate aspectual character. Second, descriptions of neural mechanisms and computational syntax specify causal or formal properties but do not, by themselves, fix that aspectual character. Third, the possibility of the relevant state becoming conscious explains how such an aspect is a feature of the organism's intentional life. Therefore an intrinsically intentional unconscious state is, for Searle, a physiological state capable of giving rise to a conscious intentional state with the relevant content. It is unconscious because this capacity is unmanifested at the time in question.

The central difficulty lies in the modal premise. Why should potential consciousness be necessary for intentional content rather than one historically common means through which content acquires an interpretable role? A perceptual subsystem may discriminate two object-categories that a patient could never report or experience, and its operations may be robustly sensitive to environmental correctness. One possible reply is that such a subsystem has derived or observer-relative content; another is that content can be grounded by the organism's learning and functional relations even in the absence of consciousness. The debate is not settled by showing that a particular representation was never reported: impossibility in principle is stronger than failure under a given experimental condition.

Searle's challenge retains force for extravagant computational attributions. A theorist who ascribes an unconscious belief to every symbolic register owes an account of aspectual content rather than merely a statement about how the register is interpreted in software. Conversely, any theory making consciousness constitutive of mentality must account for tacit representational discriminations whose psychological explanation seems to require content. The Connection Principle is therefore a serious constitutive proposal and an invaluable test of what exactly the cognitive unconscious is alleged to contain.

Dennett's Intentional Stance and the Reality of Unavowed Attitudes

Daniel Dennett's The Intentional Stance (1987) develops a theory of psychological interpretation in which beliefs and desires are attributed as part of a predictive strategy. Interpreters characterize a system as operating rationally with respect to information and goals and use that characterization to explain or predict behavior. This account is relevant to unconscious mentality because successful intentional interpretation does not require the system to disclose, or even possess, a verbally accessible inner inventory of attitudes. A standing belief may be real as a stable intentional pattern although no conscious judgment with that content is presently occurring (Dennett, 1987).

The strongest interpretation of this position is not that any amusing metaphor yields a mental state. An attribution's legitimacy depends on its explanatory and predictive discipline. When a creature reliably locates hidden food after a delay, tracks changed conditions, and revises behavior appropriately, belief-and-desire descriptions may capture a pattern that a simple reflex description cannot. The value of the intentional stance grows when the pattern transfers across new contexts and supports counterfactual predictions. Mere after-the-fact redescriptions of every observed action would provide little warrant.

The challenge is individuation. If several incompatible sets of beliefs predict the same conduct, what determines which beliefs the agent actually has? Dennett can stress that psychological reality may be multiply realizable and that some indeterminacy is genuine; his critics can reply that explanatory utility does not by itself determine the exact content of an inaccessible state. The issue bears directly on unconscious states: the analyst's prediction of action might be excellent while the hypothesis of a specific hidden proposition remains underdetermined.

This account intersects with the personal–subpersonal distinction. The intentional stance can be used with organisms and sometimes with engineered devices as a pragmatic strategy, but its application need not imply that the devices possess the same phenomenological, affective, or normative capacities as humans. An unconscious belief of an agent is a different explanatory object from a convenient fictional belief attributed to a software component. The philosophical result is a graded conception of evidential warrant: predictions justify intentional vocabulary to the extent that the vocabulary captures systematic organization, while further ontological conclusions demand additional arguments about bearer, content, and integration.

Carruthers, Mindreading, and the Opacity of Inner Thought

Peter Carruthers's The Opacity of Mind (2011) challenges familiar assumptions that people possess immediate privileged access to their own attitudes. His approach draws on cognitive science and theories of mindreading to argue that self-interpretation plays a large role in self-knowledge. On this view, we may directly experience some sensory and imagistic materials while inferring the propositional attitudes and causes that best explain our behavior and experience. The account is philosophically significant because the introspective opacity of propositional attitudes changes what counts as unconscious mentality (Carruthers, 2011).

A person may consciously rehearse a sentence while only inferring what she believes, intends, or desires. Conscious inner speech is evidence about thought, but its production and interpretation involve prior organizational processes. A theory that identifies the utterance with the full attitude risks missing the possibility that the utterance is a confabulation, a deliberative experiment, or an interpretation of an already active tendency. Carruthers's account therefore connects the unconscious to the epistemology of mental-state attribution, rather than merely to low-level sensory processing.

The strongest objection concerns the apparent immediacy and authority of some self-ascriptions. If a person decides to leave a room and knows that intention through exercising rather than interpreting it, an exclusively mindreading model may miss the agential structure of the self-knowledge involved. Transparency theories, constitutivist views, and inferential models disagree over the relation between forming an attitude, avowing it, and recognizing its cause. The evidence from confabulation indicates real limits on access to mechanisms, but it does not necessarily establish that every form of self-knowledge is interpretative.

Carruthers's The Centered Mind (2015) extends inquiry into working memory, consciousness, and the determinants of reflective thinking. The central implication for this dossier is that conscious reflection may be an organized and sometimes limited access route to underlying cognitive capacities rather than a continuously comprehensive display of all thought. The strength of a particular unconscious-state attribution then depends on independent evidence for the state alleged. The opacity thesis supplies a motivation for inquiry; it does not license a reader to assign arbitrary hidden desires to a person from an isolated behavioral episode (Carruthers, 2015).

A Formal Test of What Counts as Unconscious Evidence

Philosophical precision can be advanced by setting out the logical structure of a familiar experimental argument. Let P denote an independently established effect of stimulus information on behavior; let A denote adequate conscious access to the particular information at the relevant time; and let M denote the occurrence of a genuinely mental representation of that information. A laboratory finding of P together with a sensitive failure to establish A supports a contrast between processing and awareness. It does not logically entail M: the implication (P and not-A) implies M requires an additional premise specifying why the effect is realized through a mental representation rather than through some nonmental causal sensitivity.

Several bridges are available. A representational-content bridge requires generalization across different physical stimuli with the same relevant meaning, susceptibility to error, and systematic integration with action or inference. A functional-role bridge identifies a state through its contribution to perception, memory, and practical control. An inferential bridge seeks evidence that transitions are content-sensitive and approximately answerable to coherence or accuracy constraints. These proposals are not extensionally identical. A sensory discriminative mechanism may satisfy some and fail others.

The absence of awareness also requires formal care. A participant's failure to report a stimulus is an observation about a response, not the negation of every conceivable conscious state. Signal-detection methods can discriminate sensitivity from response criterion, but their interpretation depends on assumptions about the task and the subject's strategy. Different awareness measures may test detection of presence, discrimination of identity, confidence, or phenomenal qualities. A defensible experiment aligns the awareness measure with the information driving the claimed unconscious effect, estimates uncertainty, and examines whether the result survives changes in task demands.

Causal inference then introduces a second distinction. Observing P and not-A is compatible with many causal models, including weak conscious experience accompanied by conservative reporting, indirect learning, and task-specific response mappings. Interventions that selectively modify the hypothesized representation, while preserving relevant alternatives, strengthen a claim of mental-state causation. Experiments rarely establish a metaphysical definition, but they can eliminate concrete explanatory alternatives. This framework resists both a sweeping skepticism about all unconscious evidence and the premature elevation of a single significant priming effect into a theory of a hidden rational subject.

Implicit Memory in Amnesia and the Individuation of Learning

Studies of organic amnesia illuminate the gap between consciously recollecting an event and expressing the influence of having encountered it. Daniel Schacter's 1987 discussion of implicit expressions of memory in organic amnesia examined evidence of learning despite impaired explicit recollection. This material provides a more discriminating test case than the simple observation that an ordinary person fails to remember a word and later recognizes it (Schacter, 1987).

Suppose a patient cannot recall studying a particular item but subsequently completes a fragment of it more efficiently than a comparable unstudied item. One hypothesis is that an exposure-dependent trace was retained and influenced later performance. Its evidential content is clear: performance has changed as a function of prior experience, while a specified recollection capacity remains impaired. The additional claim that the patient possesses an unconscious memory of the episode requires care. Priming may be mediated by perceptual or lexical changes whose contents differ from a personal memory of having studied the item. The phrase 'unconscious memory' can refer either to memory demonstrated without report or to memory whose object is inaccessible to consciousness.

The metaphysical question concerns persistence. What persists from the learning episode to the later task: a representation bearing the same content, a modified network of associations, a strengthened response pathway, or a capacity reconstructed when prompted? Different realizations may produce comparable behavior. The methodological consequence is that a philosophical theory of memory should use evidence from transfer, interference, lesion, and context dependence, rather than treating every indirect benefit of exposure as a single homogeneous state.

This example also disciplines the analogy with psychoanalysis. An amnesic patient's failure of episodic recollection has a clinical and neurological interpretation that need not involve repression. Conversely, a dynamically defended memory in a psychoanalytic model is given a motivational explanation that cannot be inferred from simple priming alone. Distinguishing the different forms of retention and access permits empirical findings to constrain philosophical definitions without forcing them into one universal theory of hidden mental contents.

Comparative Evaluation of Principal Philosophical Positions

The consciousness-essentialist position offers a stringent criterion: a mental occurrence is essentially bound to awareness. Its strength is the clear connection between mentality and first-person intelligibility. Its central burden is to explain standing beliefs, unattended cognition, and experiment-based attributions without redescribing every result as a mere physical preparation for genuine mentality. If it accepts dispositions to conscious thought, it must articulate what distinguishes those dispositions from indefinitely many nonmental dispositions.

Functional and computational approaches explain unconscious states through causal roles and representational transformations. Their strength is continuity with experimental inquiry and the ability to characterize processes beyond report. Their burden is content individuation and the risk of making mentality coextensive with information processing. Interpretivist accounts gain flexibility by explaining mental states through coherent attribution to an organized agent; their burden is to distinguish robust psychological patterns from projections of the theorist. Phenomenal-intentionality accounts identify conscious experience as a source of original intentional content; they must explain the status and content of apparently unconscious thoughts.

Psychoanalytic dynamic accounts give special weight to motivation, conflict, and resistant inaccessibility. Their interpretive strength is their attention to personal meaning and the indirect expression of motivation; their central scientific problem is discriminating a specific causal explanation from plausible alternatives. Embodied and enactive approaches emphasize the organism's organized activity and environmental coupling; they must explain content-sensitive cognition that transcends immediate sensorimotor engagement.

No single disagreement can be resolved by a definition of unconscious alone. The theoretical positions answer different questions about constitution, evidence, mechanism, and explanation. A fruitful synthesis evaluates each claim according to the kind of state alleged, the sort of absence of awareness established, and the evidential method available. The difference between plausible and established unconscious mentality is therefore local to an explanatory question and must be argued in each class of case.

An Integrative Framework for Attribution and Evidence

The monograph's constructive synthesis is an attribution framework with four related dimensions. The first is content: what, if anything, the state represents, appraises, or motivates. The second is realization: what stable disposition, process, bodily organization, or computational structure implements the proposed state. The third is access: whether the bearer can experience, report, reflectively identify, or voluntarily control the relevant item under specified circumstances. The fourth is evidence: what observations, interventions, and independent measures justify the attribution. The dimensions should be examined together while preserving their distinctive logical roles.

This framework explains why unawareness alone cannot prove a hidden belief. A deficient report fixes a condition of access, while belief requires an account of content, integration, and realization. It explains why unconscious sensory processing is a genuine discovery without being identical to the discovery of a repressed wish: the realization and evidential standards differ. It also explains how a standing mental attitude can be unconscious without being an occurrent hidden episode. The attitude's existence can be grounded in durable dispositional organization even where no current token occupies an introspective field.

The framework supports comparative scientific practice. Multiple measures can converge on a single representational hypothesis while preserving uncertainty about phenomenality. Clinical narratives can identify patterns worth testing while remaining interpretations until independently constrained. Computational models can supply precise mechanisms while leaving open the personal-level status of their variables. In each case, attribution becomes a controlled philosophical commitment rather than a default extrapolation from unseen causes.

The central principle is that consciousness is one dimension of the organization of mind, and the absence of consciousness is a property whose relevance varies with the kind of state being analyzed. A satisfactory theory of unconscious mental states therefore specifies what is present, in what respect it is inaccessible, through what organization it operates, and why the available evidence supports that explanation.

Open Problems and Directions for Research

Several unresolved questions follow from the analysis. The first concerns unconscious phenomenality: are some experiences genuinely present but inaccessible to attention or report, or should such cases be understood as processing without phenomenal character? The second concerns original intentionality: which unconscious representations possess content intrinsically or through their role in a living cognitive system, and which are only representational under an external interpretation? The third concerns the ontology of standing attitudes: does a belief persist as a tokened structure, a durable disposition, or an organized capacity for reconstruction?

A fourth question concerns causal individuation. When two models predict the same behavioral pattern but posit different hidden states, what interventions or cross-context evidence could distinguish them? A fifth concerns the relation between dynamic and cognitive unconsciousness. Is there a principled empirical bridge from defended motivation to information-processing mechanisms, and under what conditions can a personal history enter a testable causal explanation? A sixth concerns the epistemology of self-report: what combinations of verbal report, confidence, neural measures, and task behavior best discriminate the absence of access from a conservative reporting criterion?

A final cluster concerns normativity and cross-system attribution. When can unconscious processes count as reasoning, and what kinds of responsibility attach to their influence on action? How far can concepts of belief or appraisal extend across animals, artificial systems, and collective organizations without losing discriminatory content? The appropriate response is an empirically informed philosophy that maintains conceptual discipline while allowing relevant findings to revise inherited assumptions. These problems remain open because their answers require joint work on ontology, access, evidence, and explanation.

Conclusion

Unconscious mental states form a family of philosophically significant phenomena whose unity rests on a shared question: how can content-bearing or psychologically organized states contribute to cognition and action without being present to conscious awareness? The historical reconstruction establishes that the question emerged through several distinct projects. Descartes and Locke sharpened the relation between thought and awareness; Leibniz identified unnoticed perception; Kant argued for indirectly knowable obscure representations; Helmholtz articulated unseen perceptual construction; Freud proposed dynamically inaccessible motivations and unconscious defenses; cognitive psychology and neuroscience made additional forms of nonconscious processing experimentally accessible.

The comparative analysis yields a precise result. Unconscious mentality cannot be inferred from an absence of report alone, and unconscious causal influence cannot be equated automatically with a mental state. Strong attribution requires a defensible account of content, the system in which the state is realized, the relevant form of access or inaccessibility, and the evidence that distinguishes the hypothesis from alternatives. This conclusion supports the reality and explanatory importance of many unconscious beliefs, memories, appraisals, and perceptual representations while preserving disputes over the status of particular states and mechanisms.

The resulting philosophical position understands unconsciousness as a variable relation between mental organization and awareness. It replaces the assumption of one hidden region with a differentiated account of standing dispositions, occurrent representations, dynamic conflict, and subpersonal processing. The study of unconscious mental states therefore changes both the theory of mind and the theory of evidence: it requires that mental phenomena be investigated through the coordinated analysis of first-person experience, objective behavior, causal intervention, conceptual structure, and historical interpretation. The ultimate gain is a more exact account of the mind's capacity to exceed its own immediate self-presentation.