Aggarwal’s Law of Revirting

Revirting

Revirting (verb): the turning of an experience back into its representation, and the settling of preference there.

The term is a compression, and it is working on more than one level at once. Each reading is intended.

The word is not a joke at the expense of the idea. It is the shortest available statement of it.

Ask people what they prefer and the answers are consistent and sincere. They prefer seeing a friend to texting one. They prefer a concert to a playlist. They prefer the room to the call, the trip to the photograph, the conversation to the thread. Then watch what they do. The overwhelming majority of their friendships run on text, their music arrives through a speaker, their colleagues appear in a window, and their travel is largely conducted through a screen.

Aggarwal's Law of Revirting states the pattern formally: for any experience that can be represented, people will choose the representation over the experience more and more often as time passes, until the representation is what they do by default; and the reason is rarely that it is better, but that it is easier.

The law is not an accusation, and it is not about deception. Nobody is confused about which is which. It is a claim about frequency: not what a person would choose in the abstract, where reality still wins, but what they actually choose on a Tuesday, where it loses. Economists would call the first a stated preference and the second a revealed preference. This law says the two diverge, that the divergence widens, and that revealed preference is the one that ends up shaping a life.

Three positions, and the state they produce

The law is usually stated as one proposition, which understates it. It is better read as an ascent. There are three positions a person can occupy with respect to any given experience, each strictly stronger than the last, and there is a fourth condition which is not a position at all but what the ascent leaves behind when nobody is watching it. The ascent is not measured in what a person feels about a category, which may not change at all. It is measured in how often they choose the copy.

The three positions, in ascending order, and the terminal state:

Preference and reversal are not the same rung

The two are easily conflated, because both end with the copy being chosen. The difference is in what happens when the original disappoints, and it is a clean test.

At preference, disappointment attaches to your choice. You went to the concert instead of staying in with the record, and you regret the trip: the ticket, the parking, the man who talked through the quiet song. Reality remains the benchmark, and you have simply decided it is not worth its price.

At reversal, disappointment attaches to reality itself. The tempo is wrong. The arrangement is different. She cannot hit the note any more. You are no longer weighing whether the original was worth the trouble; you are marking it against the copy, and it has failed the comparison. The copy has stopped competing with the original and has taken over the position of the thing that is compared to.

One sentence separates them. At preference you say the real thing was not worth it. At reversal you say the real thing was not right. The first is a judgment about cost. The second is a judgment about fidelity, and only a standard can be unfaithful to.

The positions form a ratchet rather than a slope. They are easy to climb and difficult to descend, because each one reorganises the memory the previous one depended on. Once the record is the version you know, the concert cannot be heard innocently again.

Two limits keep the law from collapsing into a truism. It applies to experience that can be represented at all, which is to say experience whose value is substantially perceptual or informational. And it predicts that categories move up and do not move back down, while saying nothing about the rate at which any particular category moves.

Why the ladder is climbed and never descended

A law of preference needs a mechanism, otherwise it is a mood. Three forces drive the ascent. They are listed in order of how much work they do, and the first does most of it.

The division of labour is straightforward. Convenience produces the choice, and produces it again tomorrow. Editing ensures the choice costs little enough that nobody audits it. Repetition, helped along by primacy where primacy happens to apply, turns ten thousand unaudited choices into the standard. Convenience explains why people choose the recording. Repetition explains why the concert eventually sounds wrong.

The forces compound, which is why the drift accelerates rather than settling. Each substitution manufactures another representation (the photograph of the overlook, the clip of the concert, the screenshot of the call) which is in turn cheap, edited, and available to somebody else before the original is. Representations multiply faster than originals. Meanwhile every convenient choice makes the next one more convenient still, because the infrastructure thickens around the path that gets used.

When the copy becomes the original

The clearest demonstration is not exotic, and it is complete. It is recorded music.

A studio recording is, in the plain sense, a copy. It is an artefact assembled after the fact out of performances that happened in a room, frequently out of several attempts at the same bar. Yet the recording is the version an audience learns, which makes it the version that constitutes the work. When the same musicians play the same song on a stage, the audience is not hearing a performance that the record once approximated. It is hearing a rendition of the record, and it is grading the rendition.

Observe what has happened to the direction of authority. The performance is prior in time and posterior in status. The copy did not win the comparison; it became the thing that comparisons are made against. Once this is seen in music it is difficult to stop seeing.

The same inversion, in categories that have travelled at different speeds:

Walter Benjamin argued that mechanical reproduction strips the original of its aura, the authority conferred by singular presence. The case described here is stronger and stranger. The aura does not evaporate. It migrates. It settles on the reproduction, which is now the singular thing that everybody knows, while the original becomes one instance among many, and a poorly lit instance at that.

Aura in Benjamin's sense attaches to the here and now of the work. The claim here is that our sense of where the here and now is located has moved.

Why this is deeper than the cave

The obvious ancestor is the allegory of the cave, and the comparison is worth making carefully, because the differences are where this law earns its keep.

In the Republic the prisoners are bound. They face a wall, they have never seen anything but shadows cast on it, and they take the shadows for the world. The ascent into daylight is painful and is resisted, because eyes accustomed to firelight are hurt by the sun. The tragedy is captivity compounded by ignorance, and the remedy follows from the diagnosis: turn the prisoner around, and endure his complaints until he can see.

Revirting differs from that picture in three ways, and each one makes it the harder problem:

The consequence is that Plato's remedy is unavailable. Turning a prisoner around accomplishes nothing when the prisoner has already been turned around, has looked at length, and has walked back on purpose. There is no ignorance here for education to correct and no captivity for liberation to end. What there is instead is a preference, formed slowly, mostly unobserved, and pointing one way.

So the cave was the easier case. Plato described a world in which reality is the standard and we are prevented from reaching it. This law describes a world in which we can reach reality whenever we like, and have quietly stopped treating it as the standard. The shadow on the wall is no longer a degraded copy of the thing outside. The thing outside has become a degraded copy of the shadow.

Plato records the objection twice. In the Phaedrus, Socrates has the god Thamus warn that writing will not strengthen memory but replace it, producing the appearance of wisdom without its substance. It is the first recorded statement of this law, and it was made about the first medium.

The law behind a great deal of what we call progress

If the law holds, it is not a curiosity about tourism and concerts. It describes the mechanism of a large share of technological history, because an enormous proportion of what gets called innovation consists of substituting a representation for a physical process.

Consider the pattern in the record:

Memory

→

Writing

Barter of goods

→

Money, then credit

Territory

→

The map

Sitting for a portrait

→

Photography

Performance

→

Recording

Travel to speak

→

Telephony, then video

Building to test

→

Simulation

Consulting a person

→

Consulting a model

Each line is the same move: an experience whose value could be carried in a channel was moved into the channel, whereupon the channel version became cheaper, better edited, and earlier, and preference migrated. The industrial history of the last two centuries can be read, in substantial part, as the systematic manufacture of Revirting.

The claim should be kept to its defensible size. Not every advance fits: anaesthesia, antibiotics, refrigeration, sanitation and the shipping container changed physical facts and represented nothing. The honest version of the claim is narrower and still striking, namely that most of what is distinctive about informational and communicative progress is this law being industrialised, and that the share of innovation which takes this form has been rising for a century.

Read that way the law becomes an instrument rather than an observation, and this is its most useful property. To find the next category, ask which experiences are still met physically by default, and among those, which carry their value in a channel. That is the frontier, and it is where the next decade of building will happen whether or not anybody names the reason.

What resists, and for how long

A law that predicted the replacement of everything would predict nothing. This one is more useful, because the ordering of resistance is not arbitrary. Categories climb in proportion to how much of their value is perceptual and how little is physiological, consequential, or dependent on the presence of somebody who can be surprised.

Categories ordered by observed resistance:

The ordering yields the law's most testable consequence. The frontier advances category by category in roughly this order, and each technology annexes the next band rather than all bands at once. Photography took landscape. Recording took performance. Telephony and then text took conversation. Video conferencing took collegial presence. Conversational machines are, at the time of writing, at the boundary of counsel and companionship.

The law is therefore falsifiable in a specific way. Should a category that has reached reversal descend, durably and in aggregate, without the loss of the technology that carried it there, the law is wrong. The return of vinyl is not such a case: it is a preference for one representation over another, which is movement within a position rather than descent from it.

The categories that resist longest share a property. Their value is not carried in the channel. A meal must be metabolised. An injury must heal. Sleep, physical labour, childbirth, and the death of somebody loved are not compressible into a signal, and every attempt to represent them yields an artefact that nobody mistakes for the thing. The frontier is bounded by the body.

That bound should be held loosely rather than triumphantly. The history of this law is a history of categories confidently placed beyond the frontier which then crossed it. A great deal of what was once thought irreducibly bodily, including the presence of a teacher, the examination of a patient and the company of a friend at distance, has already moved.

Two serious replies, at full strength

Two objections deserve their strongest form rather than a caricature.

Half of that is right and should be conceded plainly: the phenomena are not newly observed, and this essay claims no priority over any of them. What those accounts describe, though, is a property of the artefact (aura), a condition of the culture (hyperreality), or a thought experiment about value (the machine). This law makes a narrower and more awkward claim than any of them. It is about an individual's ordering of preferences; it asserts that the ordering moves in one direction over time; it names the mechanism, which is editing, cost and primacy; it names the frontier, which is the body; and it specifies what would refute it. Hyperreality cannot be checked against a person. This can.

There is a second difference, and it is the one developed in part four. The tradition treats the copy as a deficient substitute and reality as the standard it falls short of. Reversal says the offices have been exchanged. That is not a restatement of Plato. It is the case Plato did not consider.

This mistakes what the law predicts. It does not predict that people stop having original experiences. It predicts a change in what those experiences are for and how they are judged. A crowd at an overlook is not evidence against Revirting when much of the crowd has come in order to produce a representation, is facing away from the canyon to do it, and will encounter the afternoon principally through the artefact it carries home. Attendance is not the measurement. The measurement is which of the two holds the office of the standard, and on that measurement the queue at the overlook and the phone raised above it point the same way.

One part of the objection survives and should be kept. This is a gradient, not a fate. Gradients can be walked against, and everybody who walks against this one reports that it takes deliberate effort, which is exactly what a gradient implies.

An observation, and the one thing it asks

At its shortest: people like the real thing and choose the copy, they choose it more often every year, they choose it because it is easier rather than because it is better, and eventually the choosing stops being a choice. That is the whole law, and two examples carry it: everybody prefers seeing a friend and almost every friendship runs on text; everybody prefers live music and almost all music arrives through a speaker. The third position is the one worth carrying, because it inverts the relation the whole tradition assumed: after enough repetition the copy becomes the original, and the original is left auditioning for its own part.

This is offered as a law and not as a lament, and no argument is made here that the drift should be resisted in general. A great deal of it is straightforwardly good. Almost everybody who will ever know the canyon will know it through an image, and an image is not nothing; it is the widest distribution of a beautiful thing that has ever existed. The same frontier has carried medicine into villages, teaching to the poor, and the voices of the dead to their grandchildren. A law of nature is not a grievance.

What it asks is narrower, and it is the only thing a law of this kind can ask. Notice the substitution while it is happening. A preference that forms unobserved is not a preference, it is a default, and defaults compound in the dark. Noticed, there is a choice available every day: the record or the room, the thread or the call, the photograph or the afternoon. Unnoticed, the account settles itself, one small trade at a time, in the direction the gradient was always running.

Locating yourself on the ladder

The law is only useful if it can be applied without instruments. Take any category (a friendship, a city, a piece of music, a sport, a body of water) and ask the questions in order. Answer as your conduct answers, not as you would like to answer. The first rung that lands is your position.

Then answer the same questions as you would have answered them twenty years ago. The difference between the two sets of answers is the quantity this law is about.

Aggarwal's Law of Revirting, and the term Revirting, are Ashok and Natalie Aggarwal's, coined jointly in the early 2000s. The law carries the name they share.

There was no single moment of discovery, and none is claimed. It accumulated between the two of them over years of noticing the same shape in unrelated places, comparing notes, and finding that the pattern held: a view already seen, a meeting that added nothing, a preference that had quietly changed sides. A law noticed this way is on firmer ground than one deduced, because it was never looking for confirmation.

This essay is the law's first extended statement. The original formulation named the indistinguishability, indifference and preference cases and observed that experience tends over time toward the preference. Developed here are the ladder and its terminal state, the mechanism, the inversion, the comparison with the cave, the reading of technological history, the frontier, and the self-test.

This essay argues from canonical sources rather than from new data, and it deliberately reports no statistics. Where a work is cited it is cited for its argument, never for a figure. These are the accounts the essay stands on or against.