The Tortoise and the Electrode

यदा संहरते चायं कूर्मोऽङ्गानीव सर्वशः । इन्द्रियाणीन्द्रियार्थेभ्यस्तस्य प्रज्ञा प्रतिष्ठिता ॥

yadā saṃharate cāyaṃ kūrmo'ṅgānīva sarvaśaḥ | indriyāṇīndriyārthebhyas tasya prajñā pratiṣṭhitā ||

When he draws in the senses from the objects of the senses, entirely, as a tortoise (कूर्म, kūrma) draws in its limbs — his understanding is established. That is Bhagavad Gītā 2.58.

Patañjali (पतञ्जलि) makes pratyāhāra (प्रत्याहार) the fifth limb and defines it, at Yoga Sūtra 2.54, by what it does: when the senses are disengaged from their objects and follow, as it were, the nature of the mind itself. The next sūtra, 2.55, is short — ततः परमा वश्यतेन्द्रियाणाम् (tataḥ paramā vaśyatendriyāṇām), from that, the highest mastery of the senses.

That is testable. It says a person can be in a state where a stimulus arrives at the sense organ and does not reach the mind.

The instrument for asking is crude and it is a hundred years old. Close your eyes and rest and a rhythm around 8 to 12 hertz shows up over the back of the head. Open them, or hear something sudden, or attend to anything, and it goes. That is alpha blocking, and it is about as reliable as anything in electrophysiology. Repeat the same stimulus enough times and the blocking fades, because the brain stops treating a known sound as news. That is habituation.

So: give a stimulus, see whether the alpha breaks.

Four yogis, 1961, the work carried out in New Delhi. All four had persistent alpha, more strongly modulated during samādhi (समाधि) — and no change in its frequency, which sat around 11½ to 12 cycles a second before and during. Then two of them, not four, got the stimulus battery: a strong light, a loud banging noise, being touched with a hot glass tube, the vibration of a tuning fork.

The alpha did not block. Not reduced. Did not block. And the same men's alpha blocked normally to the same stimuli when they were not meditating.

Two more yogis — a separate arm, not the same men — could hold a hand in ice water for forty-five to fifty-five minutes, and kept persistent alpha through that too. Six men in all.

Four people is four people. But notice what the finding is not. It is not a report of relaxation. It is a report that a hot glass tube on the skin produced no cortical response in men who had responded to it normally a few minutes earlier.

Five years later, in Japan, forty-eight Sōtō and Rinzai priests and disciples were recorded during zazen. Alpha appeared with the eyes open, grew in amplitude, slowed in frequency, and in the most experienced ran into rhythmic theta. For twenty-three of the disciples the EEG changes tracked their master's own independent rating of their state, which is the strongest thing in either paper.

Then, in three of the masters and four controls, they clicked. Once every fifteen seconds, twenty times.

The alpha blocked to every single click and never habituated. Ordinary subjects stop responding after the third or fourth. These three registered all twenty, each one holding for about the same two or three seconds as the first, and came back to rhythm each time.

Put them side by side and the temptation is immediate. These look like opposite results.

And it maps onto what the two traditions say they are doing. Samādhi as the Sanskrit texts describe it is absorption, the senses gathered in, the tortoise. Zazen is not that. It is sitting with the eyes open and meeting what arrives.

Which is a lovely story, and I do not think it holds.

Read the 1961 paper closely and the non-habituation is in the yogis too. Outside meditation, every stimulus blocked their alpha, and — the paper's own sentence — that blocking showed no adaptation to repetition. Non-habituating alpha blocking is exactly the signature the contrast assigns to Zen as its distinguishing mark, and it is sitting in the other paper, in the other tradition, five years earlier. And when someone finally ran it with proper non-meditating controls — Becker and Shapiro in 1981, testing Zen, yoga and TM practitioners against groups told to attend to the clicks and to ignore them — the blocking and the habituation came out comparable across every group. Heide found no effect on alpha blocking in 1986 either. The reviewers of this literature describe the results as conflicting, and note that the two-tradition story got accepted as fact on the strength of the early papers alone.

So what survives is narrower. Two small studies, decades apart, each found something in trained meditators that ordinary rest does not produce. Whether those somethings are the fingerprints of two different techniques, or two labs measuring different things in different rooms, the data do not settle.

The third classic in this group measured something else entirely — fifteen practitioners of transcendental meditation, with heart rate, skin resistance and EEG recorded before, during and after, and oxygen consumption on nine of them. Oxygen consumption fell about twenty per cent within five minutes of starting. Heart rate fell, skin resistance rose, alpha amplitude rose.

Sixty-five years of this comes to less than the enthusiasts say and more than the skeptics allow.

Shown: that trained meditators produce EEG states you can tell apart from ordinary rest; that metabolic rate drops; and that in two adepts, stimuli that reliably interrupt a cortex did not interrupt it.

Not shown: causation, dose, mechanism, that the traditions differ from each other in the way the traces seem to suggest, or that any of it generalizes past a handful of unusual people who were selected for being unusual.

The tortoise verse does not need vindication from two EEG traces. But there is something worth noticing in the fact that when somebody finally hooked the practice to an instrument, the instrument recorded the thing the verse describes — not calm, not sleep, but a stimulus arriving at a body and going no further.

Two men, one afternoon, sixty-five years ago. That is what it rests on.