Automatic writing is a trance practice rooted in traditions ranging from Chinese fuji (Song dynasty) and 19th-century Spiritualism to early psychology and Surrealism. It is the canonical demonstration of the ideomotor effect — uncensored written output produced with a reduced sense of authorship. It alters consciousness primarily through suggestion-driven motor automatism — the hand executes writing while agency-monitoring is partially disengaged.
The practitioner sits with paper and pen in a quiet space, fixes attention loosely or toward a chosen prompt, and writes continuously — no pausing, editing, or lifting the pen. Whatever phrases arrive go on the page. Sessions run 10–20 minutes for introductory practice and 20–40 minutes when sustained.
At introductory levels, sessions produce fragmentary phrases, surprising word choices, and the mild dissociative feeling of watching one's own hand move. Sustained practice deepens this — practitioners report a reduced sense of agency over what the hand writes, vivid imagery, and surfacing of emotional material. Boredom, frustrated self-censorship, and unsettling emergent themes are common at every level.
The primary risks are psychological — dissociative drift extending past the session, surfacing of repressed material without therapeutic containment, and for individuals predisposed to psychosis, reinforcement of delusional attribution when output is framed as external communication. History of dissociative disorder, active psychosis, and acute PTSD are relative contraindications. Automatic writing matters as the practice in which the ideomotor effect was first studied.
History & Tradition
Automatic writing has roots in Chinese fuji (扶乩), a living Daoist and Chinese folk-religious spirit-writing tradition documented from the Song dynasty, and in 19th-century Western Spiritualism, where the planchette became the standard tool. Pierre Janet brought the practice into clinical science in the late 1880s, coining "subconscious" and framing automatic output as a product of dissociation;[1] André Breton secularized it into Surrealist technique in 1924.[2] Today it persists as a living religious practice in Brazilian Spiritism (psicografia) and Daoist temple traditions, and in stripped-down secular journaling and creativity contexts with no lineage or credentialing body.
The Practice
Technique
The practitioner sits at a desk with a notebook and a smoothly flowing pen, in a quiet space free from interruption. Posture is comfortable and upright, with the dominant hand resting unrestricted on the page. Eyes may be open, lowered, or closed — closed-eye writing is associated with deeper automatism.
The method has two stages. First, the practitioner settles attention — through breath, a soft gaze, or a written prompt at the top of the page. Second, the pen is set on the page and continuous writing begins: no pausing, no editing, no lifting the pen, no evaluating what comes out. Whatever phrases arrive are committed to the page in the order they arise. The most common departure from the method is intermittent silent editing — pausing to reject a phrase or find a better word — which restores deliberate authorship and ends the automatism.[2]
Introductory sessions last 10–20 minutes and typically produce fragmentary output — disconnected phrases and visible effort to maintain continuity. Established sessions extend to 20–40 minutes; continuity becomes easier and the felt sense of agency over the writing decreases.[3] No documented progression exists beyond this tier in the secular form.
The practice has four main variants: the Chinese fuji form (a Y-shaped implement tracing characters in a sand tray within a temple ritual), the Spiritualist form (planchette or pen framing output as communication from discarnate entities[4]), the Surrealist form (ordinary pen and paper framing output as unconscious expression[2]), and clinical ideomotor signaling within hypnotherapy (pre-arranged finger signals under a therapist's guidance, not open writing[5]).
Setting & Preparation
A quiet space, a notebook, and a pen that writes smoothly without pressure are the only requirements. For closed-eye variants, a clipboard or lined guides on the page prevent writing from wandering off the edge. The fuji form requires a sand or ash tray and the traditional Y-shaped implement within a temple ritual setting; clinical ideomotor signaling requires a trained hypnotherapist and no writing materials at all.
Choose a time window that allows 10–20 minutes for writing and an additional 10 minutes for re-grounding afterward; do not schedule a session immediately before a demanding task. Ensure the room is quiet and undisturbed for the full window. Eat beforehand — intense session content can feel more destabilizing on an empty stomach. Practitioners with a history of trauma, dissociation, or psychiatric instability should keep a trusted contact reachable. Do not begin a session in acute emotional crisis; the practice can intensify the state rather than process it.
Guidelines
- Do not attempt unsupervised practice with a diagnosed dissociative disorder, active psychosis, or a recent psychotic episode.
- Start at the introductory tier — 10–20 minutes, occasional sessions — before extending duration or frequency.
- Set a timer so the session has a hard boundary; do not write open-endedly.
- Schedule 10 minutes of grounding (walking, eating, washing hands, ordinary conversation) at the end of each session.
- Do not interpret the output as literal communication from an external source or as instruction; treat it as material to consider.
- Stop the session if content turns explicitly toward self-harm, suicidal action, or harm to others, and engage the standing safety plan.
- Avoid combining the practice with dissociative or psychedelic substances.
- If a session leaves the practitioner feeling unreal or detached for more than a few hours afterward, pause the practice and consult a clinician familiar with dissociation.
Mechanism
Subjective Effects
Sessions last 10–40 minutes. At introductory levels, the experience is mildly dissociative — the hand keeps moving while attention drifts, and surprising phrases appear without deliberate construction. At established levels, the felt sense of agency over what is written decreases more noticeably, vivid imagery can accompany the writing, and emotional or biographical material surfaces.[3] In Spiritualist and channeling contexts, content is experienced as arriving from outside the self. Boredom, frustrated self-editing, and unsettling or self-critical themes are common across all tiers and do not signal failure. Effects vary widely: people high in dissociative absorption report more dramatic alterations in felt authorship; those low in absorption may write continuously while retaining ordinary agency.[7]
▸▾Effect Grid
Relaxed, uncensored continuous writing from the threshold of awareness.
Safety
Adverse Effects
No systematic surveys of adverse effects attributable specifically to automatic writing have been published. The risks described are drawn from clinical commentary on trance-inductive practices and adjacent dissociation literature.
Dissociative drift
The neuroimaging evidence confirms that automatic writing measurably suppresses activity in the agency-monitoring network and reduces felt agency during the session.[3] For individuals with existing dissociative disorders or trauma histories, this suppression can extend beyond the session boundary, producing depersonalization or derealization in the hours afterward. High baseline dissociative absorption amplifies the effect; people scoring high on absorption traits show more pronounced reductions in felt authorship under automatism conditions.[7]
Reinforcement of attributional content
Clinical commentary identifies a specific risk for individuals at the psychotic end of the spectrum when automatic output is framed as communication from external entities. Practices that blur the boundary between internal cognition and attributed external agency can reinforce delusional beliefs in psychosis-prone or psychosis-active individuals. No published case series document this for automatic writing specifically; the concern is extrapolated from clinical literature on related practices and from theoretical accounts of agency attribution.[8]
Uncontained trauma content
The practice's design — bypassing deliberate editorial control — can surface repressed or dissociated material without the pacing of a therapeutic relationship. Unlike structured expressive writing protocols, which involve preparation and post-session integration, automatic writing has no standard protocol for titrating difficult material.[9] Warning signs include sessions ending without return to baseline calm, content with self-harm or hopelessness themes, and difficulty distinguishing written content from actual beliefs or intentions.
Contraindications
Active psychosis or a recent psychotic episode is a contraindication, particularly for variants framed around external attribution. A diagnosed dissociative disorder — dissociative identity disorder or depersonalization-derealization disorder — is a contraindication for unsupervised practice; the agency-suppression mechanism overlaps directly with the symptom structure of these conditions.
Acute, untreated PTSD is a relative contraindication because uncontained trauma content can surface without therapeutic support. High baseline dissociative absorption is not a contraindication but a known amplifier — practitioners with this trait should keep sessions short and use grounding procedures afterward.[7] Concurrent use of dissociation-inducing substances (ketamine, DXM, salvia divinorum, high-potency cannabis) is a contraindication for the session itself.
Substance Interactions
No controlled studies or case reports of automatic writing combined with psychoactive substance use have been published. The risks below are theoretical, extrapolated from the practice's mechanism and substance pharmacology.
Substances that independently reduce the sense of agency or produce dissociative states — ketamine, DXM, salvia divinorum, and high-dose cannabis — would compound the practice's agency-monitoring suppression. Combining any of these with a practice that already suppresses the agency-monitoring system is a theoretically elevated risk, particularly for individuals predisposed to dissociative pathology or psychosis.[3] Cannabis use disorder is independently associated with elevated dissociation scores [UNCITED — verify: Archinard et al. 2021 prevalence figure for clinically significant dissociation in high-potency THC users]. Classical psychedelics can produce structural dissociation as an adverse effect; combining them with trance-inductive practices theoretically compounds the risk of identity fragmentation. The absence of documented interactions reflects the absence of research, not established safety.
Habituation
Automatic writing does not involve pharmacological habituation. No data on compulsive use of the practice have been published. Theoretical concerns about psychological dependency — the practice becoming a primary coping mechanism or substituting for direct reflection, conversation, or therapeutic work — appear in clinical commentary but have not been documented systematically. Practitioners with histories of compulsive journaling, ritualized self-soothing patterns, or dissociation as a primary coping style should treat the practice as one tool among several rather than a default response to distress.
If Distress Arises
The practitioner can stop the practice at any moment — this is the defining feature of practice harm reduction, distinct from substance harm reduction. The threshold for stopping is well before frank crisis.
Stop the session and return to ordinary activity if the felt boundary between thinking and being-thought-by-the-writing becomes uncomfortable rather than interesting; if depersonalization or derealization arises and does not settle within a minute or two of pausing; if intrusive imagery of harm to self or others appears repeatedly; or if the session has run past the set timer without notice. Grounding procedures — naming objects in the room, drinking water, walking outside, calling a friend — are the appropriate first response. Re-engagement with daily activity should wait until baseline returns; if it does not return within a few hours, consult a clinician.
Recovery & Support
Automatic writing has no specialized peer-support infrastructure of the kind that exists for substance use or some meditation traditions. For dissociation that persists beyond a session, the appropriate resource is a clinician with experience treating dissociative disorders — the International Society for the Study of Trauma and Dissociation (ISSTD) maintains a clinician directory.
For acute psychiatric crisis, including emergent suicidal or psychotic content, contact local emergency services or a national crisis line (in the United States, 988 — the Suicide and Crisis Lifeline). For longer-term integration of material surfaced through the practice, working with a psychodynamic, psychoanalytic, or somatically trained therapist is more appropriate than continuing solo practice as the response to difficult content. The practice is not a substitute for psychotherapy and was never designed to be one.
Recovery resources, crisis lines, and support pathways → View all
Progression
No formal stage model for automatic writing has been published.[10] The practical arc is two-tiered: introductory sessions of 10–20 minutes produce fragmentary output; established sessions of 20–40 minutes develop greater continuity and a more pronounced reduction in felt agency.[5] Most practitioners do not progress beyond the established tier; depth comes from working with the output as material, not from extending session duration.
Evidence Summary
Limited evidence. The underlying ideomotor mechanism is well-established,[6][8] but no randomized trials of automatic writing exist; direct evidence is limited to a small fMRI study (n=5) linking agency-network suppression to the practice[3] and an experimental study connecting dissociative absorption to reduced felt authorship.[7]
Limitations
The fMRI study reports n=5 and cannot support causal inference; no trial has evaluated automatic writing for any therapeutic or wellness claim; and popular claims that the practice surfaces unconscious material or resolves trauma are not supported by controlled evidence — findings from the expressive writing tradition are structurally distinct and do not transfer.[11][12]