Entrainment and the Receiver — visual flicker, Anil Seth, binaural beats, and the magnet-and-radio question.

1. Flicker as old phenomenon — Brion Gysin's Dreamachine

In 1959 the artist Brion Gysin, working with the mathematician and inventor Ian Sommerville, built a device that consisted of a perforated metal cylinder rotating around an ordinary light bulb. The viewer sat in front of the device with eyes closed. The rotation produced a flicker on the closed eyelids whose frequency fell in the alpha band — roughly eight to thirteen flickers per second. Within a minute or two, most subjects reported visual phenomena: geometric patterns, drifting colours, occasional more elaborate imagery, sometimes a feeling of moving through space. Gysin called the device the Dreamachine, and described it as the only artwork meant to be experienced with the eyes closed.

William Burroughs experimented with the Dreamachine heavily through the 1960s and 1970s. Allen Ginsberg used one. Gysin and Sommerville built more, gave the plans away, and saw the device replicated by anyone with a bulb, a turntable, and the patience to cut a cylinder. The phenomenon was real, reproducible, and largely ignored by mainstream science for several decades. The art world treated it as art. The hippie underground treated it as a free alternative to psychedelics. The neuroscience of the period had not yet developed the conceptual machinery to ask the right questions about what the device was doing.

What the Dreamachine demonstrated was not aesthetic. It was structural. Under the right conditions — eyes closed, uniform visual field, flicker at the brain's resting cortical rhythm — ordinary consciousness produced contents that ordinary consciousness does not produce. The contents differed across subjects but the architecture did not. Something about alpha-band visual stimulation, delivered to the closed eyelid, reliably altered what the brain generated in the absence of ordinary structured input.

2. The neurophysiological mechanism — photic driving and alpha entrainment

The mechanism the Dreamachine exploits is one of the better-characterised phenomena in clinical neurophysiology. The visual cortex, at rest with eyes closed, exhibits dominant electrical activity in the alpha band — roughly eight to thirteen cycles per second. This is the standard EEG finding that defines the alpha rhythm: it appears within seconds of eye closure and disappears just as quickly when the eyes open.

When a flickering visual stimulus is delivered at a frequency near the brain's resting alpha rhythm, the cortical electrical activity locks to the stimulus. This is called photic driving. It is documented in every clinical EEG textbook, has been used diagnostically since the 1940s (notably in evaluations for photosensitive epilepsy), and is a baseline finding of decades of research. The entrainment is particularly strong at alpha-band frequencies because the cortex is, in some structural sense, already running at that rhythm. The stimulus does not impose a foreign rhythm on the cortex; it captures and amplifies the rhythm the cortex is already producing.

With the eyes closed and a uniform visual field, the constraint that ordinary structured visual input places on cortical activity is removed. The entrained cortical rhythm becomes, in a sense, phenomenally accessible. The brain generates imagery the way it generates imagery during light sleep or in the hypnagogic state at the boundary of waking and sleep — but in waking consciousness, while the subject is alert and able to report. This is the structural condition the Dreamachine produces and the contemporary research formalises.

3. Anil Seth's Ganzflicker and the 2022 Dreamachine installation

Anil Seth, professor of cognitive and computational neuroscience at the University of Sussex, runs the Sackler Centre for Consciousness Science. His lab has been formally studying flicker-induced perceptual phenomena for over a decade. The 2022 Dreamachine installation — a collaboration between Seth and the artist collective Collective Act, mounted in London, Cardiff, Belfast, and Edinburgh as part of the UK's Unboxed festival — was a contemporary realisation of Gysin's idea at vastly larger scale and with proper experimental instrumentation.

Participants lay on cushions in a darkened space. White light, generated by carefully calibrated LED panels, was delivered to closed eyelids at alpha-band frequencies, sometimes ramped through the eight-to-thirteen-Hz range. Sessions lasted around thirty minutes. The accompanying scientific protocol — thousands of participants completing pre-session and post-session questionnaires, with subsets contributing detailed phenomenological reports and (in laboratory follow-ups) EEG recordings — generated one of the largest experimental datasets ever assembled on flicker-induced perceptual phenomena.

The published findings, in collaboration with Reshanne Reeder, David Schwartzman, Keisuke Suzuki, and others, document robust effects across thousands of subjects. The phenomena cluster reliably: most subjects report something; the something varies; the variation is not random but predictable from prior individual measures. Two adjacent literatures are also documented — the older Ganzfeld tradition (uniform sensory deprivation without flicker) and the wider Ganzflicker methodology that combines a uniform visual field with controlled-frequency flicker. Both have been studied by the Sackler Centre and adjacent labs. The Dreamachine installation was the most ambitious public–scientific implementation of the basic phenomenon to date.

4. Different brains, different images — the individual-difference finding

One of the most striking findings of Seth's research, and one of the reasons the phenomenon is interesting beyond the merely curious, is the strong individual variation in what subjects experience.

Some subjects report form constants — the geometric patterns Heinrich Klüver catalogued in the 1920s during his mescaline research: spirals, lattices, tunnels, concentric rings, honeycombs. The mathematical neuroscience of Jack Cowan and Paul Bressloff has explained these form constants as natural eigenmodes of the visual cortex's intrinsic dynamics, which surface visibly when the constraint from external structured input is removed. The same form constants appear under psilocybin, under sensory deprivation, in migraine aura, and in flicker stimulation — pointing at a shared substrate (the cortex itself) being expressed through different routes.

Other subjects report colour fields, gradients, and slowly drifting hues. Others report more elaborate imagery — faces, animals, scenes, occasionally narrative content. A small fraction report relatively little at all. The variation is partially predictable from prior individual differences. Aphantasia status — the spectrum of voluntary visual imagery capacity, ranging from people who cannot form a mental image of anything to people whose mental imagery is nearly as vivid as perception — significantly predicts the quality of flicker-induced phenomena. Aphantasic subjects tend to experience the geometric and colour-field clusters; people with strong voluntary imagery tend to experience the more elaborate cluster as well.

This individual variation is not, by itself, evidence for any particular model of consciousness. It is exactly what one would expect from any biological substrate subjected to external perturbation. Different brains with different cortical architectures, different histories of experience, different genetic propensities, and a long list of documented and undocumented individual variables will respond differently to the same stimulus. It is the magnet-and-radio point of §8 below at the cross-subject scale — bringing a strong magnet near ten different makes and models of radio will produce different effects in each, because each radio's internal construction is different. The variation tells us about the radios. It does not, by itself, tell us anything about what the radios are doing. A production-model account of the brain predicts the individual variation; a receiver-model account predicts the same individual variation. Both posit a biological substrate that processes input. Both predict that the processing will vary by substrate. The variation is fascinating empirically, important for any account of cortical machinery, and structurally silent on the question both frameworks would like to use the entrainment evidence to settle.

5. Why this fits Seth's controlled hallucination framework

The result fits Seth's broader theoretical position about perception, articulated at book length in Being You: A New Science of Consciousness (Faber & Faber, 2021). The position, in its compressed form: ordinary perception is what Seth calls controlled hallucination. The brain does not passively receive sensory information and assemble it into experience. The brain generates its best Bayesian model of what is producing the sensory input, and that model — constrained by but not determined by the input — is what we experience.

Under ordinary conditions, sensory input constrains the model tightly. The room I am in is real in the relevant operational sense; my brain is generating a model of the room that the sensory data has very little freedom to depart from. The hallucination is controlled because the sensory data does the controlling.

Under conditions of reduced or ambiguous input — eyes closed, uniform visual field, alpha-band cortical entrainment — the constraint relaxes. The brain's prior expectations get more room to express themselves. Because priors differ across individuals (by personal history, by genetic propensity, by the visual-imagery capacity that varies along the aphantasia spectrum), the resulting phenomena differ as well. Same stimulus. Different brain priors. Different experiences. The individual-difference finding is not a complication of Seth's framework. It is what the framework predicts.

The trilogy's reading of this is not very different from Seth's, in vocabulary, and substantially different in commitment. The framework agrees that the brain is generating its experience under constraint from sensory data. The framework adds that the substrate the brain is generating against includes the consciousness field — that the priors Seth's Bayesian framework treats as internal include, in the trilogy's reading, contributions from coupling to a substrate the brain is partly receiving and partly modelling. Whether this last addition is correct is, on the framework's wager, an empirical question. Seth's results are entirely compatible with the production-model reading. They are also entirely compatible with the receiver-model reading. The frameworks diverge on which features of the architecture they treat as fundamental, not on the architecture itself.

6. The frequency question — yes, the alpha-band

The specific frequency question: the strongest visual entrainment effects in the contemporary literature are reliably produced in the alpha band, eight to thirteen Hz, with strongest effects clustered around ten Hz. The Dreamachine protocols sit there. The Sackler Centre experiments sit there. The classical photic-driving literature sits there. Other frequency bands produce different and generally weaker effects: theta-band flicker (four to seven Hz) can produce different perceptual phenomena and is sometimes associated with drowsy or hypnagogic states; beta-band flicker (fifteen to thirty Hz) tends to produce less imagery and more general arousal; gamma-band flicker (around forty Hz) has been studied for different reasons (most prominently in the Alzheimer's research described in §7 below).

The alpha-band effect works because of resonance. The cortex's resting rhythm matches the stimulus frequency, the entrainment is therefore particularly efficient, and the entrained activity becomes phenomenally available under closed-eye conditions. The phenomenon is not generic. It depends on the structural alignment between stimulus and substrate.

7. The auditory cousin — binaural beats, isochronic tones, and the contemplative traditions

The natural question: if visual flicker at alpha frequencies produces this effect, does auditory stimulation at the same frequencies do something similar? The short answer is yes, weaker, and through a more interesting set of mechanisms.

The complication. A pure tone at ten Hz is below the human hearing threshold of approximately twenty Hz. You cannot just play a ten-Hz tone the way you can flash a ten-Hz light. Three workarounds dominate the literature.

Binaural beats. Gerald Oster's 1973 Scientific American paper introduced binaural beats to a wide audience: play a tone of two hundred Hz in one ear and two hundred and ten Hz in the other; the brainstem (specifically the superior olivary complex) processes the interaural difference and produces a neural ten-Hz signal that the auditory cortex tracks. The signal does not exist acoustically in the room; it is generated inside the head as a difference between two acoustic frequencies. The phenomenon is real and physiologically well-characterised. Whether binaural beats produce consequential cognitive, mood, or EEG effects has been studied for half a century with mixed results.

Monaural beats and isochronic tones. Two frequencies mixed before delivery produce a real acoustic beat that physically exists as amplitude modulation. Isochronic tones — a single tone rapidly gated on and off at the target frequency — produce a clear amplitude-modulated stimulus the auditory system entrains to. Both tend to produce somewhat more reliable EEG effects than binaural beats, though the literature here is also mixed.

Amplitude-modulated noise. White or pink noise modulated at the alpha frequency. Used in some sleep, attention, and meditation research; the entrainment effect on the auditory cortex is real but generally weaker than for periodic-tone stimuli.

The well-documented mechanism. The auditory steady-state response (ASSR) is the auditory cortex's electrical entrainment to amplitude modulation of an acoustic stimulus. It is documented in clinical neurophysiology (including in objective hearing tests used in neonates), and the response can be measured reliably from EEG. The ASSR is strongest at gamma frequencies, around forty Hz — which is why the MIT laboratory of Li-Huei Tsai has been studying 40 Hz auditory and visual stimulation as a potential intervention in Alzheimer's disease across the 2018–2026 period, with promising preliminary results in both mice and early human trials. At alpha-band frequencies (eight to thirteen Hz), auditory entrainment occurs but tends to be weaker than visual entrainment at the same frequencies.

The contested cognitive–mood literature. A 2019 meta-analysis by Garcia-Argibay and colleagues, examining binaural beat studies for anxiety, memory, and attention effects, found small but statistically significant effects across multiple domains, with substantial heterogeneity between studies and a high risk of bias in several individual studies. A more skeptical 2015 critical review by Chaieb, Wilpert, Reber, and Fell in Frontiers in Psychiatry emphasised the methodological weaknesses in much of the field and concluded that, while binaural-beat effects exist, the literature does not yet support the strong claims often made for the technique by commercial vendors. The honest summary: there is signal, the signal is not negligible, the signal is not as strong as enthusiasts often claim, and rigorous methodology continues to refine the picture.

Combined audio-visual entrainment (sometimes called AVE in the older literature, associated with David Siever and others) tends to produce stronger and more reliable effects than either modality alone, presumably because the stimulus reaches the cortex through two independent sensory pathways simultaneously. The Dreamachine installation in 2022 included audio components for this reason among others.

The contemplative-traditions cousins. A number of contemplative traditions have, across centuries, independently developed practices whose acoustic structure looks suspiciously like entrainment by other means. Tibetan singing bowls produce strong amplitude modulation at low frequencies (the slow beating of multiple overtone partials). Gregorian chant uses long rhythmic structures whose periodicity falls in or near contemplative-relevant brainwave ranges. Sufi dhikr — rhythmic recitation of divine names — produces sustained acoustic and motor periodicity. Shamanic drumming at roughly four Hz (in the theta band rather than alpha) has been studied empirically by Michael Harner, Michael Winkelman, and others and is associated, in laboratory replication, with EEG theta-band shifts and reported altered states.

Whether the contemplative traditions are independently exploiting brainwave entrainment is an interesting question that the empirical literature has approached only sporadically. The convergence is suggestive. The framework's reading is that several thousand years of trial and error, selecting for what reliably produces the recognition states the traditions value, would plausibly converge on stimulus parameters the cortex is structurally responsive to — whether or not the traditions named what they were doing in the modern vocabulary.

8. The magnet and the radio — what entrainment shows and what it doesn't

It is worth being clear about what the alpha-band entrainment phenomena actually demonstrate. They demonstrate that the brain's machinery can be perturbed by external electromagnetic stimulation in reproducible ways, and that the perturbation alters what consciousness renders. They do not, by themselves, demonstrate anything about whether consciousness is produced by that machinery or received through it.

A useful analogy. Picture a radio receiving a music broadcast. The radio's electronic circuits transduce the incoming electromagnetic wave into electrical signals that drive the speakers, and music comes out. Now bring a magnet close to the radio's circuitry. The audio output changes — distortions, hums, unexpected tones, sometimes audio coming out of the left speaker that has no obvious relation to anything being broadcast. The magnet has perturbed the radio's machinery, and the perturbation is showing up in the output.

The magnet experiment is fascinating. It tells us a great deal about how the radio's electronic circuits behave under perturbation. It tells us nothing about whether the music being broadcast is real, or whether the radio is generating the music from scratch from its own internal noise. Both interpretations of the radio are consistent with the magnet effect. The magnet acts on the machinery, not on the question of what the machinery is doing.

Brainwave entrainment is structurally the same kind of experiment. Alpha-band visual flicker, binaural beats, isochronic tones — these perturb the cortical machinery in measurable ways, and the perturbation shows up as altered conscious content. The phenomena are real and important. What they do not do is decide the underlying question. A production-model reading takes the result the same way: the brain produces consciousness, the brain's machinery is being perturbed by alpha-band stimulation, and the perturbation produces altered content. A receiver-model reading takes the result the same way: the brain receives consciousness, the brain's machinery is being perturbed by alpha-band stimulation, and the perturbation alters what the receiver renders. Both accounts predict the phenomena. Both accounts predict the individual differences. Both accounts predict the cross-modal extension to binaural beats. Neither account gets unique support from the entrainment evidence.

The same logic applies, sharpened, to the case the production-model literature most often cites as decisive. General anaesthesia abolishes ordinary conscious content reliably and reversibly; this is one of medicine's strongest practical achievements and one of the most-cited pieces of evidence for the brain-produces-consciousness account. The standard argument: if consciousness depends so completely on the integrity of cortical machinery that turning the machinery off turns the consciousness off, then the machinery must be where the consciousness is generated.

The radio-and-magnet framing reveals the gap. Turning a radio off at the switch reliably and reversibly abolishes the music coming out of its speakers. This tells us a great deal about the radio's dependence on its own power supply and circuitry. It tells us nothing about whether the broadcast continues to exist while the radio is off. The broadcast persists or it does not, independently of whether any particular radio can currently receive it; turning off one radio gives us no information either way about the broadcast itself. Anaesthesia is the off-switch version of the magnet experiment, not its refutation. The receiver-versus-production question is no more decided by anaesthesia than it is by alpha-band flicker. Both interventions show us the machinery responds to perturbation. Neither shows us what the machinery is doing.

The trilogy's framework is honest about this. The entrainment phenomena are not, strictly speaking, evidence for the receiver model over the production model. They are evidence about the architecture of the cortical machinery, which both models share. The receiver-versus-production question is decided in other places — in the receiver-signatures the framework predicts and the production model does not (the long catalogue in Why biology? §4), in the substrate-dependence question worked through in Why biology? §7, in the cross-tradition contemplative convergence. The entrainment literature does not decide that question. It demonstrates that the machinery can be perturbed and that conscious content is sensitive to the perturbation. The radio works one way or the other; the magnet only shows us how the machinery itself behaves.

9. What the trilogy's framework actually gets from this literature

What the framework does get from the entrainment literature is something more modest than direct empirical support for the receiver model: the phenomena confirm what the framework treats as the broader architecture. Ordinary conscious content is more contingent than ordinary consciousness suggests it is. The cortical filter — however we want to model it — can be perturbed reproducibly. The everyday rendering is one of several possible renderings the same machinery is capable of producing. These are foundational claims the framework rests on, and the entrainment literature is consistent with them on terms either model can read.

The phenomena also fit cleanly alongside the other conditions the framework names elsewhere: the contemplative discipline that thins the self (see meditation and the receiver), the pharmacological route through psilocybin and adjacent agents (see Carhart-Harris's entropic brain), and the structural class of moments — physical extremity, overwhelming consequence, grief, awe, terminal lucidity — catalogued in the Gnostic essay §8. Entrainment is a technological route to one of the same kinds of perturbation those other cases produce by other means. The Dreamachine has been around for sixty-five years; the framework's broader vocabulary has only recently developed to name what conditions it shares with these other cases.

The Li-Huei Tsai gamma-band (40 Hz) research approaches the same architecture from the opposite direction. Where alpha-band entrainment relaxes the cortical constraint and lets internal priors express themselves, gamma-band entrainment appears to tighten the constraint and, on the Tsai work, to support cellular processes the brain uses to clear pathological proteins. Same underlying architecture, different parameters, different functional consequences. The fact that the same machinery admits both kinds of intervention is, in the framework's reading, additional evidence that what is being perturbed is more architectural than the inherited model treats it as being — without yet telling us, in either case, whether the machinery is generating or receiving.

10. Honest closing

What the entrainment phenomena confirm: that the cortical machinery generating ordinary conscious content can be perturbed reproducibly, that what it produces under perturbation is consistent and informative, that the architecture admits multiple uses across multiple frequency bands and across multiple sensory modalities. What they do not confirm: that consciousness is being received from a field rather than produced by the brain. The radio works one way or the other; the magnet shows us how the machinery behaves regardless.

What the framework adds, on its own terms, is the broader claim that the architecture the entrainment literature is mapping is one piece of a wider architecture named elsewhere on this site. The receiver model treats consciousness as field-coupled and the brain as the receiver. The phenomena Anil Seth's lab studies are entirely consistent with this; they are also entirely consistent with the production-model account Seth himself defends. The trilogy's claim about which model is correct is not decided by the entrainment evidence. It is decided in places the entrainment evidence cannot reach — in the receiver-signatures the framework predicts and the production model does not, in the substrate-dependence question, in the cross-tradition contemplative convergence. The entrainment evidence is corroborative of the architecture both frameworks share. The framework's argument about which model the architecture belongs to is made elsewhere.

That distinction is worth keeping clean. The entrainment phenomena are striking, important, and worth the long contemporary research programme Anil Seth and adjacent labs are conducting. What they are not is unique support for one side of the production-versus-receiver debate. The framework's honest position is that they corroborate the architecture without deciding the question, and that the question itself is decided by evidence the entrainment literature is structurally not in a position to provide.

Reading list

Visual flicker — foundational and contemporary

Brion Gysin & Ian Sommerville, the Dreamachine (1959). The device, the plans, and the surrounding literature in Gysin's archive (Beat Hotel Press; multiple subsequent biographical studies).

Heinrich Klüver, Mescal and Mechanisms of Hallucinations (University of Chicago Press, 1928 / 1966 reprint). The form-constants catalogue.

Jack D. Cowan & Paul C. Bressloff, papers on cortical pattern formation and form constants (multiple, 1979 onward; including the influential Bressloff, Cowan, Golubitsky, Thomas, Wiener 2001 paper on geometric visual hallucinations and the functional architecture of striate cortex).

Anil Seth, Being You: A New Science of Consciousness (Faber & Faber, 2021). The book-length statement of the controlled-hallucination framework.

Keisuke Suzuki, Reshanne Reeder, David Schwartzman, Anil Seth and colleagues, the published papers from the 2022 Dreamachine project and the Sackler Centre's Ganzflicker research programme.

Auditory entrainment

Gerald Oster, Auditory Beats in the Brain, Scientific American 229 (1973): 94–102. The classic introduction.

Leila Chaieb, Elke C. Wilpert, Thomas P. Reber & Juergen Fell, Auditory beat stimulation and its effects on cognition and mood states, Frontiers in Psychiatry 6 (2015): 70. The critical review.

Miguel Garcia-Argibay et al., Efficacy of binaural auditory beats in cognition, anxiety, and pain perception: a meta-analysis, Psychological Research 83 (2019): 357–372. The systematic synthesis.

Li-Huei Tsai laboratory, the 2016–2026 series on gamma-band (40 Hz) auditory and visual entrainment and Alzheimer's disease, including the original Nature paper (Iaccarino et al., 2016) and the human-trial follow-ups.

Contemplative-traditions entrainment

Michael Harner, The Way of the Shaman (Harper & Row, 1980). The originating contemporary anthropology of drumming-as-entrainment.

Michael Winkelman, Shamanism: A Biopsychosocial Paradigm of Consciousness and Healing (Praeger, 2010). The neuroscientifically informed synthesis.

The wider acoustic-ethnomusicology literature on traditional contemplative practices — Tibetan ritual sound, Gregorian chant analysis, Sufi dhikr studies — for the cross-tradition convergence material.

This page is part of the Reading companion essays. For the contemplative practice that does by discipline what entrainment does by stimulus, see meditation and the receiver; for the pharmacological route to the same filter-rerouting architecture, see Carhart-Harris's entropic brain; for the substrate question that frames why this architecture matters at all, see Why biology? — the autopoiesis test for receivership; for the broader Gnostic framing of the structural class of recognition-conditions, see Gnosis, the Pleroma, and the Field; for the wider synthesis, The Evidence.