Three Years in the Zendo

 

Copyright: Sanjay Basu

Why I sat down

For three years I sat on a black cushion for 40 minutes every morning before the rest of the house woke up. I was not looking for peace. I was looking for a way out of my own head, which at the time was running a debugger on everything, including itself.

The years were mid 1993 to early 1997 and the place was Singapore. I had taken a contract as a communications software engineer, which in practice meant one thing: designing a TCP/IP stack and retrofitting it onto OS/2, which by the autumn of 1994 had become OS/2 Warp. IBM was betting the operating system on networking, and the networking was late. I spent my days inside RFC 793 and RFC 1122, arguing with retransmission timers and a 16-bit device driver layer that had never expected to carry a 32-bit protocol stack.

I had come to Zen the way engineers usually come to anything: sideways, through a book. I was reading about the measurement problem in quantum mechanics and kept running into the same word in the footnotes of physicists who had gone to Japan in the 1970s. The word was practice. Not belief, not theory. Practice.

That distinction is the whole thing, and it took me a year to see it. Zen does not ask you to accept a proposition. It asks you to do something, daily, with your body, and then to notice what happens. For someone trained to demand a hypothesis before an experiment, this was backwards. I did it anyway.

The irony was not lost on me later. I was building a protocol whose entire design assumed the network would lose, reorder and duplicate what you sent, and whose only answer was to keep paying attention and retransmit. I just had not yet applied that design to myself.

The zendo was a converted garage 11 minutes from my apartment. There were 9 cushions, one bell, and a teacher who said almost nothing for the first six weeks I was there. This article is about what those three years did to how I think, and about why I eventually stopped going, which turned out to be part of the lesson.

Year one: the body is the instrument

The first year was almost entirely about knees. Nobody warns you about this. The literature talks about emptiness and the sound of one hand. The first six months talk about whether your left hip will survive 40 minutes in half lotus.

I was told to count breaths from 1 to 10 and start over. If I lost count, back to 1. For about four months I never reached 10. I would get to 6 or 7 and find myself designing a caching layer or rehearsing an argument with a colleague. Then I would notice, feel a flash of irritation at myself, and return to 1.

The specific thing I kept drifting to, that first monsoon season, was a bug in slow start. Under load on a Token Ring segment in the lab, the stack was collapsing its congestion window to one segment and never recovering. I would be at breath 5 and suddenly see the whole state machine laid out, convinced I had found it. I never had. The fix, when it came three weeks later, came at my desk at 11 at night and had nothing to do with anything I had worked out on the cushion.

The instruction about that irritation was the first real teaching I received. My teacher told me the noticing was the practice. Not the counting, not the reaching 10. The moment where attention catches itself drifting and comes back. Everything else was scaffolding.

This reframed failure for me in a way no management book ever had. Every lapse in attention became a repetition, like a set at the gym. The more I drifted, the more reps I got. By month eight I had stopped keeping score, and that was the point where I began to reach 10 regularly.

The other thing year one taught me is that thoughts have a texture, and the texture can be observed before the content is believed. An anxious thought arrives with a tightening across the chest a fraction of a second before the words show up. A resentful one arrives with heat in the jaw. Once you have sat with this 300 mornings in a row, you cannot unsee it. You start to catch the body’s vote before the mind casts it.

I first used this at work in early 1995, during the push to ship the Warp Connect networking bundle. A test lead walked up to my desk with a trace showing my stack sending a reset on a perfectly good connection. I felt the heat in the jaw before I had read the first line of the trace. I waited for it to pass, which took about the length of one breath, and then read the trace. He was right. A lifetime earlier I would have spent 20 minutes defending the code before looking at it.

I did not know it then, but this was the beginning of a shift in where I located myself. I had assumed I was the thinker. Year one suggested I might be the one watching the thinker, and that the thinker was a somewhat unreliable subcontractor.

Year two: nothing holds still, including the one who notices

In the second year I was given a question to sit with instead of numbers. I will not repeat it here because the question is not the point and reading it does nothing. What matters is that it could not be answered by thinking, and I spent roughly nine months trying to think my way through it anyway.

This was humbling for a person with three degrees. I built frameworks. I drew diagrams. I brought my teacher clever answers and he sent me back to the cushion without comment, 14 times by my count. On the fifteenth visit I had nothing prepared, said so, and he smiled for the first time in two years.

What the question eventually did was expose how much of my mental life was storytelling. I told myself stories about who I was, what I deserved, how a meeting should have gone, what a person had meant by a remark. The stories ran continuously and I had mistaken them for perception.

The central observation of that year was impermanence, and I want to be precise about what I mean because the word has been worn smooth. I do not mean the poster version, that all things pass. I mean the operational version: if you watch any single emotion closely for about 90 seconds without feeding it a story, it changes shape. Anger becomes fatigue. Fatigue becomes a kind of curiosity about the anger. Nothing I observed ever stayed what it was.

That included the observer. Somewhere in month 20 it became clear that there was no fixed Sanjay sitting behind the eyes, auditing experience. There was a process that kept calling itself Sanjay, and the calling was itself one more event in the stream. This was not a mystical insight. It was closer to noticing that a variable you had treated as a constant was being reassigned on every loop.

I had a ready analogy and it came from the day job. A TCP connection feels like a thing. People talk about opening one, holding one, dropping one. But there is no connection anywhere on the wire. There are two endpoints, each holding a small block of state, each agreeing to behave as if something joined them. Take away the state blocks and nothing remains to point at. Twenty months of sitting had shown me that the self was built the same way: a control block that persists because it keeps getting updated, not because it was ever there.

The practical fallout showed up at work before it showed up anywhere else. Disagreements lost their urgency. Not because I cared less about outcomes, but because I had seen, hundreds of times, that the heat of a disagreement at 10 a.m. was gone by 2 p.m. whether or not I acted on it. I started waiting. I became measurably better at my job by doing less of it in the moment.

The clearest case was a dispute in mid 1995 over whether to implement Nagle’s algorithm by default on Warp. The product side wanted it off because an interactive terminal application felt sluggish in a demo. I wanted it on because turning it off would flood the network with 1-byte segments from every chatty application in the catalog. The old me would have escalated that afternoon. Instead I sat on it for two days, built a 40-line test that showed the flood in numbers, and sent the trace with no commentary. The argument ended without anyone having to lose it.

The second year also taught me to distrust peak experiences. Twice during long sits I had episodes that I would once have called transcendent: a dissolving of the boundary between the room and whatever was doing the sitting. They lasted perhaps a minute each. My teacher’s response when I reported them was to ask whether I had done the dishes. He was not being flippant. He was warning me that chasing those states is just desire wearing a robe, and that an insight you cannot carry into the kitchen is not yet an insight.

Year three: the mess is universal

The third year was quieter and, in a sense, more disappointing. I had expected the practice to keep producing revelations. Instead it produced something closer to maintenance. I sat. I noticed. I got up and went to work.

What changed was my regard for other people. Two years of watching my own mind had shown me a place that was chaotic, petty, generous, frightened and kind in no particular order, often within the same minute. The third year delivered the obvious corollary: everyone else’s mind is like this too. The colleague who snapped at me in a review was running the same noisy process I was, with the same unreliable subcontractor doing the talking.

By 1996 the stack was shipping and I was fielding defect reports from customers in six countries. The reports came in angry, and several were addressed to me by name. Reading them became a practice of its own. Behind each one was a systems administrator at 2 in the morning, somewhere in Europe, with a LAN that had stopped and a manager who did not care whose fault it was. Once I could see that person, the tone of the report stopped registering. Only the packet traces did.

This is what the tradition calls compassion, and I had misunderstood the word for most of my life. I thought it meant feeling sorry for people. It turned out to mean something more like accurate modeling. When you know from the inside how a human mind actually behaves under load, you stop expecting it to behave otherwise. You cut slack not as a kindness but as a correction to a bad prior.

It also turned out that most of my suffering was optional. Not the pain itself, but the second layer: the commentary about how unfair the pain was, how it should not be happening, how someone was to blame. The practice did not remove a single difficulty from my life in those three years. It removed a great deal of the commentary, and the commentary had been most of the weight.

I should be honest about attachment too, because the practice gave me a new definition of it. I had assumed attachment meant caring about things. It means gripping a fixed view of things. I was deeply attached, in that sense, to being right, to being the smartest person in the room, to a particular story about my career. Loosening the grip on those did more for my relationships than any amount of effort to be a better communicator.

The test case for this arrived in the form of a junior engineer, two years out of university, who told me in a design review that my sliding window implementation was wasting a buffer copy on every inbound segment. I had designed that path myself and had been quietly proud of it for a year. He was right, and the fix improved throughput on a 16 megabit ring by close to 12 percent. Three years earlier I would have found a reason he was wrong. That afternoon I thanked him, took the patch, and noticed how little it cost.

And then, at the end of the third year, I stopped going to the zendo.

There was no crisis. My teacher moved to another city. The garage was sold. I could have found another group within 20 minutes of home, and I looked at two. But I noticed that the looking had the same flavor as my earlier chasing of peak experiences. I was trying to hold on to the form of a thing whose entire content had been about not holding on.

The timing lined up with the work. The retrofit was done, Warp’s networking was in the field, and my contract wound down in the first weeks of 1997. I packed up a flat in the east of the island, a shelf of RFCs and one black cushion, and left Singapore at about the same time I left the zendo.

So I let the form go and kept the sitting. I still do it, most mornings, for 25 minutes now rather than 40. I no longer call myself a Zen practitioner, partly out of respect for people who have given their lives to it and partly because the label had started to feel like one more story about who I was. Three years was the dose. Whatever it was supposed to do, it did.

What survived the cushion

Four things came with me when I left, and they are the reason I am writing this in a newsletter about technology and philosophy rather than in a journal.

First, a working definition of attention. Paying attention to what is actually present, without a running verdict on it, is a skill, and it degrades without use. I now treat it the way I treat any perishable capability: I practice it or I lose it. Most of the failures I see in engineering leadership are failures of attention dressed up as failures of strategy.

Second, a distrust of the finished self. I used to think of a person, including me, as something with stable properties that could be known and optimized. Three years of watching the properties shift under observation cured me of that. When I later encountered QBism, where a quantum state is an agent’s expectation rather than a fact about the world, it did not feel like a new idea. It felt like the cushion again: the measurement and the one measuring are not cleanly separable, and pretending otherwise produces tidy equations and bad predictions about people.

Third, a method for difficulty. When something is unbearable, the instinct is to focus on it or to push it away. The practice taught a third option: widen the frame until the difficulty is one object among many. The pain in the knee is still there, and so is the sound of traffic, the breath, the light in the room, the 40 years of mornings that came before this one. Nothing is removed. The proportions change. I have used this in hospital waiting rooms and in budget meetings, and it works in both.

Fourth, and this is the one I hold most loosely, a suspicion of anyone who claims to have arrived. Enlightenment, whatever it is, obeys the same rule as everything else: it comes and goes. The most awake people I met in those three years were the ones who would tell you, without embarrassment, that they had been a fool at breakfast. The one person I met who claimed to be permanently clear was, as far as I could tell, permanently selling something.

I do not recommend Zen. I am not sure it can be recommended any more than a particular childhood can. What I can say is that three years of sitting still every morning, for no stated goal, with a teacher who mostly declined to explain, was the most rigorous education in epistemology I have had. It taught me that the instrument has to be calibrated before the measurement means anything, and that the instrument is me.