The Price
of a Nod

ICompliment pong

Three friends were passing compliments back and forth. Nothing was being decided. No information was moving that anyone would have called information: you look great, no, you look great, that’s because you picked the jacket, I picked it because of what you said about the colour. It went on for minutes, a ball kept in the air for the pleasure of keeping it in the air. A fourth person was watching and could not find the point. There was no claim being made, no question being settled, no fact anyone would carry out of the room. And yet the fourth person felt, unmistakably, welcome. Something had been extended to them, and they could not say what it was.

That observation is where a small framework for reading conversation was born, and the framework is the subject of this essay. It is called purr, pop, flash, and it was written down by someone who had spent years building, for every mind they met, a working model of its machinery, because a mind that ran without machinery was unthinkable. The watcher in that room was not a sociologist. They were someone for whom the game of compliment pong was not self-evident, and who therefore had to work out what it was for. The answer they arrived at is the one this essay will defend: conversation is not made of content. It is made of receipt. The compliments were not the message. The returning of them was.

The framework names three primitives. A purr opens and holds a channel: hi, how are you, the weather, long day, the warmth that establishes a connection is live without saying anything in particular. A pop transmits: the story, the argument, the fact, the take. A flash acknowledges: the nod, the mm-hm, the laugh, the double-tap, the signal that says received, channel still good, keep going. Every exchange, from an elevator nod to a divorce, decomposes into these three, endlessly recombined. The framework’s first insight was that most people think conversation is popping, and it is not. Without the purr to open the channel and the flash to keep it open, a pop is transmission into the void.

Its second insight came later, and it is the one this essay is really about. The three signals do not cost the same. One of them is nearly free. One of them is the most expensive thing a person can do in company, and the people who seem to withhold it are very often people who cannot afford it.

IIThe century that already knew

The first thing to say about purr, pop, and flash is that linguistics got there first, and that this is a point in the framework’s favour rather than against it. A person who rebuilds a real structure from scratch, using a networking manual as the only reference, and arrives at the same structure the specialists found, has demonstrated that the structure is real.

The purr has a name. In 1923 the anthropologist Bronisław Malinowski, writing a supplement to Ogden and Richards’ The Meaning of Meaning, coined the phrase phatic communion for speech in which ties of union are created by a mere exchange of words. Talk about the weather, enquiries after health, remarks on the obvious: Malinowski saw that none of it conveyed information and all of it did something, and that the something was the binding of people into company. Roman Jakobson, laying out the six functions of language in 1960, made it one of the six and gave it the plainest possible example: Hello, do you hear me? The phatic function is the function of checking that the channel is open. Robin Dunbar went further in Grooming, Gossip and the Evolution of Language (1996) and argued that language evolved as vocal grooming, a cheaper way of doing what primates do with their fingers in each other’s fur, so that a group could hold more members than anyone could physically groom. If Dunbar is right, the purr is not a minor function of language. It is the reason language exists. The framework’s own field guide opens by calling humans herd creatures, a phrase it owes to Mark Earls’s Herd (2007), a book about marketing that argued most of what people do, they do because the people near them are doing it. The purr is herd behaviour at the scale of two. A cat purrs to maintain contact. So, it turns out, do we.

The flash has three names. In 1970 Victor Yngve, in a paper called On Getting a Word in Edgewise, coined the term back channel for the stream of small signals a listener sends while the speaker holds the floor: the nods, the mm-hms, the yeahs, the brief completions. Emanuel Schegloff, in 1982, in a chapter whose title includes the words some uses of “uh huh”, called them continuers, and made the point that a continuer does two things at once: it reports that the listener is following, and it hands the floor straight back. Herbert Clark and Susan Brennan, in Grounding in Communication (1991), built the theory underneath both: a conversation proceeds by establishing common ground, and common ground cannot be assumed, only grounded, through a continuous exchange of evidence that each contribution has been understood well enough to go on. The acknowledgment is the atom of that process. Without it nothing has been said, because nothing is known to have arrived.

Japanese has an everyday word for the flash, and it is the best word anyone has found. Aizuchi is written with the characters for mutual and hammer. The image is two smiths at one anvil, master and apprentice, striking in alternation, each blow timed by the other’s. A Japanese listener produces aizuchi at a rate that startles English speakers, who hear it as interruption until they learn that the alternating hammer is how the work gets done.

The pop needs no introduction. It is the only one of the three that dictionaries, logicians, and courtrooms care about. Jakobson called it the referential function. Grice built his maxims around it. It is also, as the framework observed, the only primitive where accuracy sometimes matters, and the one that reveals status: who gets to pop, and whose pops are flashed, is the invisible hierarchy of any room made visible.

And then there is the laboratory. The strongest evidence that receipt is the load-bearing signal comes from John Gottman’s work with couples. His lab called the small moves people make toward each other bids, and the response to a bid turning toward or turning away. In the study his institute cites most, couples who were still married six years on had turned toward each other’s bids about 86 percent of the time in the lab. Couples who had divorced had turned toward about 33 percent. The content of the bids was mostly trivial, a remark about a bird outside the window, a half-sentence about the news. The marriages were not decided by what was said. They were decided by whether it was received.

The networking manual was right too. The framework borrowed its vocabulary from packets, and the borrowing holds. A TCP connection begins with a three-way handshake, two machines agreeing that a channel exists before either sends anything that matters: the purr. Data moves in segments: the pop. Every segment must be acknowledged, and an unacknowledged segment is retransmitted, and a connection with no acknowledgments is torn down: the flash. The idle connection sends keepalives, small empty packets whose only content is still here. And the humblest tool on any network, ping, works by sending a packet called an echo request and waiting for an echo reply. Compliment pong, the scene this essay opened with, is an echo exchange. The friends were pinging each other, and the fourth person was watching a channel being kept alive.

So the framework is not new. What is new in it, and what the laboratory and the linguists underweighted, is the question it eventually asked about price.

The marriages were not decided by what was said. They were decided by whether it was received.

IIIThe price

Every signal in the protocol draws on the same reserve, and the framework’s fourth iteration named it: regulation, the capacity to hold your own state steady while something else is happening. Regulation is not calm. A regulated person can be furious. Regulation is anger that reports to you instead of publishing on its own schedule. It is the difference between playing a character and being played by one.

Priced in regulation, the three signals are nowhere near equal.

A purr is cheap. Ambient warmth costs a trickle of presence, which is why small talk survives even long days and why the first thing to go in a crisis is not the greeting but the listening.

A pop is free, and that should worry you. Things that cost nothing at empty are not communication. A pop is what leaks out at four percent battery. Watch anyone truly depleted: the purring stops, the flashing stops, and what remains is transmission, the grievance, the monologue, the take, delivered to whoever is nearest. It is discharge, not connection, and it feels like connection to exactly one person in the room.

The flash is the most expensive signal in the protocol, and the reason is structural rather than moral. To acknowledge someone else’s transmission you have to suspend your own. You have to stop being the broadcaster for long enough to actually take delivery, to let something that is not yours arrive and register before you resume. That is the whole price. There is no discount on it, no technique that makes receipt free, because receipt is the suspension. A nod that does not interrupt your own broadcast is not a nod. It is a tic.

From this follows the sentence the framework exists to say, and the one this essay will keep returning to: some people do not withhold flashes. They cannot afford them.

The sentence does a lot of work. It converts a moral reading, they refuse to acknowledge me, into an energetic one, the reserve that pays for acknowledgment is empty, or was never charged. The moral reading licenses resentment and produces nothing. The energetic reading licenses nothing and produces compassion, not as a softening applied afterward but as what the physics arrive at on their own. Everything in the second half of this essay depends on that conversion.

It is worth noticing that the price of receipt has been recognised, in other vocabularies, by people who were not thinking about protocols at all. Clark and Brennan, in the grounding paper, enumerate the costs of grounding across media: the cost of producing a signal, the cost of receiving it, the cost of understanding it, the cost of delay, the cost of repair. Their accounting is by medium, not by person. The framework’s move is to run the same ledger inside a single human being and ask what happens when the reception account is overdrawn.

Simone Weil put it in a sentence. Attention is the rarest and purest form of generosity. She wrote it in a letter of April 1942 to the poet Joë Bousquet, who had been shot through the spine in 1918 and lived the rest of his life in bed, and who had read some pages she sent him. She was thanking him for receipt. Iris Murdoch, in 1959, put it in a longer one: love is the extremely difficult realisation that something other than oneself is real. Both are describing the flash, and both are insisting on the adjective. Rarest. Extremely difficult. The thing that looks like the least of the three signals, the one that is often just a sound, is the one that the people who thought hardest about attention called the hardest thing there is.

Reread Gottman’s numbers as a budget rather than a score. The couples who lasted were not more eloquent, and their bids were not more interesting. They were paying, at full price, about six times in seven. The couples who did not last were paying one time in three. Nobody in the second group had decided to be cruel. Many of them, the framework would predict, were simply running on empty, and discharge had replaced receipt so gradually that neither of them noticed the channel had closed.

A toy, not a model. The story has six beats and the prices are invented, but the shape is the framework's: receipt costs, silence escalates, and a battery at twenty percent makes the kindest person in the room look exactly like someone who does not care.

Some people do not withhold flashes. They cannot afford them.

IVThe training

If regulation is a battery, where does the battery come from? Not from birth. Nobody is born with the capacity to suspend their own broadcast; infants are pure transmission, and rightly so. Batteries grow under load, and the only load that grows them is consequence: being billed for your own misattunements, noticing the bill, and paying it. You learn to flash by discovering, often painfully, what happens when you don’t.

The bill, though, is not evenly distributed, and this is where the framework’s account becomes an account of justice as well as of energy.

Some people are invoiced for everything. Every half-second of latency in their replies is read as strangeness and charged accordingly. Every misread cue costs them a friend, a job, a room. The world does not translate for them, and so they are trained relentlessly, whether or not they wanted the training, and they end up running consciously, at cost, a discipline that other people never have to notice they possess. The framework’s own opening line describes them exactly: you’re not weird, you’re just half a second late.

Other people are translated for. Rooms rearrange around them. Their pauses are read charitably, their misfires absorbed by everyone else’s flexibility, their broadcasts received by people who have learned that receiving is the price of being near them. They may remain untrained forever. Not because they cannot learn. Because the bill never arrives.

Damian Milton called a version of this the double empathy problem, in a 2012 paper about autism: when two people with very different ways of experiencing the world fail to understand each other, the failure is mutual, but it is only ever diagnosed in one of them. The framework sees the same asymmetry from the accounting department. Misunderstanding is symmetric. The invoice is not.

Here a rail has to be laid, because the argument is easy to misuse. Untrained is not incapable. A large battery is a history, not a virtue. An empty one is a circumstance, not a character. Nothing in this essay is a claim that one kind of mind is superior to another, in either direction. The entry fee for training is one thing, curiosity, and it is open to anyone willing to be a beginner at something they were told they had already mastered.

There is one more thing to say about the people who were billed for everything, and it is the thing that makes a framework like this one possible at all. The philosopher Michael Polanyi observed that we know more than we can tell: the expert’s skill is tacit, and the expert, asked how they do it, usually cannot say. The insider runs the protocol tacitly. They purr, pop, and flash the way they walk, and if you ask them what they are doing they will tell you they are just talking. The outsider cannot run it tacitly, and so has to run it explicitly: watch, model, name, rehearse, correct. Explicit knowledge is slower and dearer than tacit knowledge, and it has one advantage that no amount of fluency can buy. It can be written down. The field guide exists because someone could not do the thing by instinct and therefore had to understand it, and the understanding, once written, turned out to show the insiders what they had been doing all along. Oliver Sacks borrowed Temple Grandin’s description of herself for the title of a book: an anthropologist on Mars. The anthropologist’s disadvantage is also the anthropologist’s job. Nobody who belongs to a tribe writes its grammar.

VThe receiver that cannot receive

So far the essay has described a battery that is empty. Now it has to describe something worse: a receiver that was never built.

Start with a term from computer graphics. A frustum is the viewing volume, the truncated pyramid of space a camera admits. Everything inside it can be rendered. Everything outside it does not exist, not is hidden, not is far away, but does not exist, as far as the frame is concerned. A renderer culls what lies outside the frustum before it draws a single pixel, and it does so without malice, because drawing what cannot be seen would be a waste.

Attention has a frustum too, and it can be narrowed. The narrowing is usually marketed as a strength, decisiveness, certainty, unshakeable self-belief, but structurally it is a viewing volume adjusted until it admits only what can be controlled. Whatever might surprise, contradict, or fail to flatter is culled at the boundary. And the narrowing has a destination. When only the controllable is admitted, everything that is perceived is eventually something the perceiver authored. Other people stop arriving as people, unpredictable, exterior, capable of news, and start arriving as characters: cast, scripted, and reviewed by the only critic in the room. You never meet anyone in there. Only your castings of them.

Now put the narrowed frustum next to the protocol and watch what happens to the kindest signal in it. A transmit-only architecture has a codebook with no section for inbound traffic. It knows the signals it sends; that is the whole vocabulary. So when a packet arrives, the classifier does the only thing a classifier can do and files the inbound signal under the nearest outbound category. A purr decodes as ambience. A pop decodes as a rival move. And a flash, I hear you, that sounds hard, you’re right about that, has no entry at all, so it falls through to the only inbound type the system defines, which is the pop: a move, a claim, a positioning, requiring a counter-move. The kindest signal in the protocol arrives as an attack, because attack is the only inbound type the system has.

It looks like this in practice.

• That sounds really hard. I’m sorry. A flash, full price paid.
• So now I complain too much. Got it. Receipt not found; decoded as a pop; counter-move issued.
• No, I just meant I heard you. The flash retransmitted, price paid again.
• You always do this. Counter-move, and the exchange has become a corridor.

Not malice. A missing parser.

The oldest account of this architecture is two thousand years old, and it is more precise than anything the graphics vocabulary can offer. In the third book of the Metamorphoses, Ovid tells the story of Echo and Narcissus, and it is, read as protocol, the story of a pure receiver meeting a pure transmitter. Echo has been punished by Juno: she can no longer begin a sentence, only repeat the last words she hears. She is the back channel made flesh, a listener who can flash and never pop. Narcissus is her opposite. The seer Tiresias has said he will live to old age if he never knows himself, and he has grown up unable to receive anyone, rejecting every lover, admitting nothing to the viewing volume that is not his own. When they meet in the wood, the exchange is pure protocol. He calls, Is anyone here? She answers, Here. He says, Come, and she says, Come, and runs to him, and he recoils: I would die before I give you power over me. She repeats, I give you power over me, and withdraws, and wastes away until nothing is left of her but a voice that can still only answer. He finds a pool, sees a face in it, and loves it, and cannot understand why it will not come to him. He dies there, at the water, in love with the one thing his frustum admits, applauded by exactly one face.

The hall of mirrors is older than mirrors. Ovid saw that the transmit-only architecture is not evil but doomed, and that the receiver who meets it is not destroyed by it but reduced to repetition, which is the fate of anyone who keeps flashing into a channel that cannot decode a flash. Martin Buber, in 1923, gave the two stances names: I-Thou, in which the other is met as a presence that can surprise you, and I-It, in which the other is an object in your field, useful or obstructive but never news. The narrowed frustum is an I-It engine. It can render anything except a Thou.

Two things follow for the person standing outside such an architecture, and the framework states both carefully.

The first is that you cannot argue your way out of someone else’s rendering from inside it. The corrections you receive there land not on your facts, where correction would be fair, but on your interior, what you meant, what your tone was doing, what your pause intended, territory where you are the only possible witness and where correction is therefore invalid by construction. Accept it and you have handed over the pen to your own inside. Decline it and you are now defensive, which proves the charge. Every reply is raw material for the next correction. The only exit is geometric: step outside the viewing volume and remain real there. That is not abandonment and it is not punishment. It is the recognition that the person in the hall cannot currently render you, and that staying inside only deepens the rendering error. Outside the volume you can still be warm. You can leave every door unlocked from your side. What you stop doing is submitting your interior for review. You become the one thing the hall does not contain: a fixed point that is not a reflection. Sometimes, not often, a fixed point is how a person in a hall of mirrors first learns there is an outside.

The second thing is the one the framework insists on most, and it is why the essay carries no anger. Deregulation of this kind is not a style. It is not bluntness or bigness or how I am. It is a degenerative condition of perception. Every choice of control over curiosity narrows the frustum by one more uncontrollable thing and burns a little more receiver. The inbound section of the codebook is not merely unwritten; it is being unwritten, one culled surprise at a time, until the world is not unheard but unavailable. Follow the chain to its end. A battery never trained, because the world always translated. A frustum narrowed to the controllable, because control feels like strength. A codebook with no inbound section, so every kindness arrives as an attack. A corridor that turns every companion into a reflection or a departure. And at the end, a person alone with their castings, mistaking echoes for company. There is no punishment anyone could add to that. The sentence is self-executing and already being served, by a warden who is also the prisoner and who built the prison believing it was a fortress. You do not have to excuse anything to see this clearly. You only have to notice that the cruelest thing in the whole architecture is done to its owner.

Compassion is not the tone of that observation. It is the finding.

The cruelest thing in the whole architecture is done to its owner.

VIPerformance is where the listening goes

The framework’s third iteration was written after someone objected to it on the grounds that it made conversation sound like science, and offered a counter-thesis: conversation is jazz. You do not run the protocol. You play it. The two views turn out not to disagree, and the reconciliation is the most useful thing the jazz version produced.

The protocol view samples a conversation: this utterance is a purr, that one a pop, there the flash. Discrete packets, each labelable, each true. But nobody experiences a conversation as samples. It is experienced as a curve, something with momentum and contour and swing. Conversation is the integral of the protocol, and the relationship between the two is exactly the relationship between a chart and a performance. The chart is true note by note. The song exists only when the notes flow through somebody. One view counts the stars; the other connects them. Same sky.

Jazz is the right metaphor for a better reason than it sounds pretty, which is that jazz is the most protocol-saturated art form alive. Comping, the thing the pianist and the bassist and the drummer do behind a soloist for the whole length of a tune, is flashing. It is receipt paid at full price, continuously, by three people, so that a fourth can transmit. Trading fours is turn-taking with the turns made audible. The solo is a pop, and it is priced in accumulated attention: you earn it with years of listening, and everyone in the room can hear whether you paid. And improvisation, the thing that looks like pure freedom, is structure internalised until it no longer needs to be visible. The tacit knowledge Polanyi described is what a jazz musician has, and it was explicit once. Somebody wrote the changes down.

A reader once rejected the jazz version outright, because performance sounded like dishonesty, and left. He came back two weeks later with the objection turned inward: he had been listening to himself, and all of his conversation was performed, all of it. He expected to be convicted. Jazz acquitted him. The performance was never the dishonesty. The performance is where the listening goes.

This also explains the observation the framework made about the great conversationalists, the hosts who make a dinner table sing. They are not transmitting more than other people. They are receiving more. Half of what they do is generosity: the hand on the shoulder, the laugh at the joke that was not that funny, the relentless warm flashing that keeps everyone else’s channel open while they play. Great performers over-acknowledge. That is the tell.

The performance is where the listening goes.

VIICoda: the nod

The previous essay on this site was about a single syllable, fire, and the five tunes it can carry. This one has been about a smaller utterance still. Mm-hm is barely a word. It has no dictionary meaning. It is two closed-mouth hums with a nod attached. And it carries, on the evidence of a century of linguists, a laboratory full of married couples, every network engineer, every rhythm section, and one person who had to write the changes down, the whole weight of the relationship it occurs in.

The framework’s practical advice has always been short, and it belongs here, because it is all about receipt. When your pops are not landing, stop popping. More content into a dead channel is not a strategy. Downshift to a purr: long day, this is hard. Find something in the other person’s position you can genuinely acknowledge, and flash it, at full price. If the channel cannot be reopened, exit clean, let’s come back to this, because storming out burns the channel harder than silence does. And before you open your mouth in any room, spend thirty seconds reading it: who sets the energy, whose pops get flashed, who is being systematically under-acknowledged. If you flash that last person, you have made an ally for the price of a nod.

Return, finally, to the room where this began. Three friends passing compliments, and a fourth person who could not find the point and felt welcome anyway. The fourth person was not wrong that nothing was being said. They were wrong that nothing was happening. What was happening was the most expensive thing in the protocol, paid freely, over and over, in a small space, by people who did not know they were paying. The game had no content because it needed none. Its content was the receipt.

The person who watched went away and built a model of what they had seen, because they could not do it by instinct and could not bear not to understand it. Years later the model was a field guide, and then a spec, and then a set of jazz charts, and then an essay about price. Its author’s advice, the only advice the whole framework has ever really given, fits in two lines, and it is the right place to stop.

Art is a beautiful thing. Master it and be gentle in the world.