Dedifferentiation

Merleau-Ponty's term for the structural feature of sleep that distinguishes it from both waking consciousness and pure unconsciousness: not absence of articulation, but the collapse of the diacritical system by which waking consciousness holds its terms apart. "Wakefulness is differentiation or diacritical system; sleep is dedifferentiation and passage not to absence of consciousness, not to consciousness of an indeterminacy, but to indeterminacy of consciousness" (Institution and Passivity 136). Developed in the Passivity course as the positive phenomenon that lets MP argue: sleep is not the absence of consciousness but its modalization into a mode in which "the subject is not sovereign, without the subject being inserted in it."

Key Points

Dedifferentiation, not absence — sleep does not turn off the subject; it turns off the articulation of the subject's perceptual field into distinct terms
The diacritical formula — "consciousness of... something is always consciousness of a difference between terms that are not given positively" (136). Wakefulness is this differentiation; sleep is its provisional collapse
Indeterminacy of consciousness, not consciousness of indeterminacy — the crucial distinction. In sleep I do not become conscious of a vague field; my consciousness itself becomes vague
The body holds the place — even in dedifferentiation, the sleeping body maintains an "open sensory field" and a minimal relation to the surroundings that makes reawakening possible
To sleep is a verb I undergo — "I call upon sleep but it is sleep which comes" (143). Sleep is not something I do, nor something done to me; it is a status I lend myself to
Vertigo as analogue — dedifferentiation is structurally like vertigo, where the levels by which the world is organized collapse from the inside and produce "fulfillment by suppression of the levels" (136)
Not Cartesian, not Sartrean — Descartes treats sleep as illusion and Sartre treats it as captive imagining-consciousness; MP accepts Sartre's phenomenology of the hypnagogic but denies it exhausts sleep's structure

Details

The Diacritical Setup

MP's argument proceeds from a claim about waking consciousness: consciousness of something is always consciousness of a difference between terms. "Consciousness of... something is always consciousness of a difference between terms that are not given positively, and that appear as terms only through forgetfulness, interruption of analysis" (136). This is the Saussurean structural principle generalized to perceptual consciousness: meaning is diacritical, a matter of differences rather than positive content.

If waking consciousness is a diacritical system, then sleep can be described without recourse to either "loss of consciousness" (which would posit consciousness as a positive presence that goes out) or "a second consciousness" (which would posit sleep as a different kind of consciousness with its own content). Sleep is simply the provisional collapse of the diacritical system. The terms that had been held apart are no longer held apart; articulation fails; but there is neither void nor substitute.

MP makes this precise: "Wakefulness is differentiation or diacritical system; sleep is dedifferentiation and passage not to absence of consciousness, not to consciousness of an indeterminacy, but to indeterminacy of consciousness" (136).

Why "Indeterminacy of Consciousness"

The distinction between "consciousness of indeterminacy" and "indeterminacy of consciousness" is central. In "consciousness of indeterminacy," the subject remains a clear pole of experience and becomes aware of a vague field. In "indeterminacy of consciousness," the subject itself becomes vague — the pole dissolves, and the polarity through which content is organized collapses on both sides at once. Sleep is the second, not the first.

This distinguishes MP's sleep from Husserlian phenomena of the horizon or the vague. A Husserlian horizon is indeterminate content for a determinate consciousness; MP's sleep is the collapse of the determinate-indeterminate structure altogether. Similarly, it distinguishes MP's sleep from Sartrean fascination: in fascination, the subject remains present to its imaginings (however captive); in sleep, the presence-structure itself fails.

The Body Holds the Place

The strongest objection to an "indeterminacy of consciousness" account is that if consciousness truly dedifferentiates, how does reawakening become possible? An absolutely vague consciousness would have no grip on the world from which to reawaken.

MP's answer: the body holds the place. "The body becomes an indistinct being, and yet it can reconstitute itself. It is even the body that reawakens us. In a certain way, it stays awake (open sensory field — invasion and resuscitation of consciousness by stimuli)" ([125]). The sleeping body maintains a minimal relation to the surroundings — a sensory openness, a postural anchoring — that is itself the site of possible reawakening.

The body's "holding the place" is not a backup consciousness. It is the body's own way of being-in-the-world, which does not require thematic consciousness to maintain its grip on the world. This lets MP say: consciousness dedifferentiates, but being-in-the-world does not fail. The sleeper is still in the world — through the body — even when the subject is no longer differentiating it.

This is a significant ontological claim. It means being-in-the-world is not co-extensive with thematic consciousness. The subject can be absent (dedifferentiated) while the body-world relation continues. Reawakening is then not a creation ex nihilo of consciousness but a reactivation of the diacritical system the body has been holding in abeyance.

The Verb "To Sleep"

MP makes a grammatical-phenomenological observation. "To sleep" is a verb — it names an action — but the action is neither something I do nor something done to me. "It is necessary to grasp what it is to sleep — falling asleep — in a sense an act, expressed by a verb. When I lie down I do something, I not only await sleep, I lend myself to sleep — indulgence. But I do not cause sleep; the will to sleep prevents sleep" (143).

This is the formal structure of lateral passivity applied to sleep. I do not decide to sleep, but sleep does not happen to me from outside. I "lend myself" to the status of sleeping — the verb names a transition whose subject is neither a pure agent nor a pure patient. The reflexive form "s'endormir" in French captures this: I sleep myself, neither actively nor passively.

The deeper phenomenological point: if sleep were purely something I do, it would be under my decisional control, and insomnia would be impossible. If sleep were purely something done to me, I would have no say in it, and "lending myself to sleep" (the complex preparatory act of settling into bed, releasing tension, etc.) would be incoherent. The reality is that sleep is accepted, and acceptance is the verbal form of lateral passivity.

Not the Sartrean Fascination

Sartre's The Imaginary treats sleep through the lens of imagining-consciousness. The hypnagogic state is a "captive consciousness" fascinated by the images it produces; the dream is an imagining-consciousness unable to disengage from its own productions. MP quotes Sartre at length on this in the Passivity course (139–141) and accepts the phenomenology of the hypnagogic state as accurate. But he denies that it exhausts sleep.

Sartre's fascination is still a mode of consciousness present to its contents, however helplessly. MP's dedifferentiation is the collapse of the presence-structure itself. The hypnagogic state is real and is what Sartre describes; but it is a transition into sleep, not sleep itself. "When sleep comes, it is something other than this consciousness which divides itself. It is a status (in the sense of a grâce d'état or devoir d'état)" (143).

This is why MP's account is not simply Sartre's amended. Sartre still thinks of sleep as a content the subject is given over to; MP thinks of sleep as a mode the subject itself enters. The difference is structural, not merely phenomenological.

Vertigo as Structural Analogue

MP draws a surprising analogue: vertigo. "Cf. vertigo or fulfillment by suppression of the levels" (136).

Vertigo is dedifferentiation of the spatial levels through which the body normally organizes its orientation. In vertigo, the vertical and horizontal cease to be held apart, and the spatial framework collapses from the inside. This is not a deprivation (the body is not missing some information); it is a suppression of the levels by the body itself, and it has the peculiar phenomenology of both plenitude and loss.

The analogue illuminates sleep: sleep is to the temporal-intersubjective framework what vertigo is to the spatial framework. Both involve the collapse of the diacritical levels by which orientation is maintained. Both produce a peculiar phenomenology of plenitude and indeterminacy — in vertigo, the body is fully present to a world that has lost its coordinates; in sleep, the subject is fully immersed in a field that has lost its articulations.

The Passivity Course Summary's Formulation

The concept's most compact statement, from the course summary:

"To sleep is not, despite how it sounds, an act, an operation, the thought or consciousness of sleeping; it is a modality of perceptual progression — more precisely, it is the provisional involution or dedifferentiation of consciousness; it is the return to the unarticulated, the withdrawal to a global or pre-personal relation to the world. In sleep, the world is not truly absent, but rather distant, a distance in which the body marks our place and with which it continues to entertain a minimum of relations, which will render reawakening possible." (Passivity Course Summary)

Three moves are packed in: (1) sleep is a "modality of perceptual progression" — it is continuous with perception, not opposed to it; (2) it is "provisional involution" — a folding-in, a return to the unarticulated, not a going-out; (3) the body maintains a "minimum of relations" that preserves the possibility of reawakening.

Positions

Descartes treats sleep as a state in which the mind is fed with illusions distinguishable from reality only by the failure of the coherence-with-waking-experience test. Sleep is a privative state.
Sartre (The Imaginary) treats sleep through fascination and captive imagining-consciousness. Sleep is the subject's surrender to its own imagining productions.
MP (Institution and Passivity) rejects both: sleep is neither privative (Descartes) nor captive-imagining (Sartre) but a dedifferentiation in which consciousness itself becomes indeterminate while the body holds the place. Sartre's hypnagogic phenomenology is accepted as accurate for its stage but denied exhaustiveness.
Freud treats sleep as a regression that satisfies the subject's wish by sheltering it from external excitations. MP is closer to Freud than to Sartre here — accepting that sleep has a function and an intentionality — but denies that sleep presupposes a second agency doing the wish-fulfilling work.

Connections

is the structural form of passivity as it applies to sleep
is continuous with the perceptual-unconscious — both are cases of consciousness without positive content
is the general structure exhibited pathologically by Schneider — Schneider's "limp" intentional arc is a permanent, selective dedifferentiation: concrete motor grasp preserved, abstract projection lost. Sleep is a provisional, total dedifferentiation. Both cases show the body continuing to function while a higher-order articulation fails, converging on the thesis that being-in-the-world is not coextensive with thematic consciousness
contrasts with Sartre's captive consciousness / imagining-consciousness analysis
presupposes the diacritical account of sense — "sense is divergence"
illustrates lateral (not frontal) passivity — the sleeper is not acted upon but enters a status
has the structural form of vertigo — both are dedifferentiation of a framework of levels
anticipates the analysis of sleep and dream in the later ontology as cases of mutual inherence (ineinander)
relates to the body's grasp of space-time discussed in the problem of memory — the body "holds the place" in sleep as it "holds the time" in memory
converges with the PhP pathological method and the Passivity course's sleep analysis on a single thesis about the pre-personal body — what V&I will call flesh

Open Questions

How does dedifferentiation relate to contemporary theories of sleep (REM vs. NREM, sleep stages, memory consolidation)? MP's account is phenomenological, not physiological, but the two ought to be compatible
Is dedifferentiation a specifically sleep-phenomenon, or a more general structural feature of consciousness (present, e.g., in deep concentration, meditation, trance)?
Does the "body holds the place" claim require that the body have its own thetic awareness, or is the body's holding a pre-thetic structural feature?
How is dedifferentiation related to the symbolic-matrix that organizes the unconscious? Both are structural features of a non-thematic consciousness, but one is active (organizing) and the other is passive (collapsing)
Does MP's account handle dreamless sleep, or only the sleep that produces dreams? The Passivity course treats both but leaves the relationship between them underdeveloped
See also: Is Schneider a limiting case of sleep?

Sources

merleau-ponty-2010-institution-and-passivity — the primary source. The Sleep chapter of the Passivity course (137–146) is the main locus. Key passages: dedifferentiation formula at 136; body-holds-the-place at [125]; "to sleep is a verb" at 143; the Sartre quotation and MP's acceptance-with-reservation at 139–141; Passivity Course Summary for the most compressed statement
kaushik-2019-matrixed-ontology — Matrixed Ontology Ch. 3–4 develops sleep as "being in the divergence" — a positive ontological event, not the absence of phenomenality. Sleep "constricts non-being so that it does not amount to a radical negation" and is "the internal possibility" of being, making it endo-ontological. Kaushik reads sleep against Heidegger/Fink (sleep = absorption into the clearing/sun), Sartre (sleep = nothingness of imagination), and Nancy (sleep = "tomb of sleep," absolute self-enclosure). On Kaushik's reading, sleep is the concrete phenomenological site where the matrixed ontology is most directly available — it is where the symbolic matrix generates the difference between sense and dream perceptions. The sleeping body is "possible everywhere and at all times" — it inhabits no definite site but has the capacity for all situations. MP calls this ontology "a sort of time of sleep" (V&I Working Note)