The New Prakrit

On what forms beneath a standard that is no longer a language

#

There is a manual for how air traffic controllers speak. It is not a style guide. Standard phraseology is specified, internationally, word by word, because two people who have never met and may not share a first language have to coordinate objects moving at four hundred knots using nothing but voice on a shared frequency. Every ambiguity that has ever killed anyone gets written out of it afterward. It is the most carefully engineered register in ordinary working life.

Ellis has been working the same approach control for eleven years and speaks it fluently and correctly. And there is a second vocabulary, maybe forty items, that everyone in that facility uses and nobody outside it would understand. Names for pieces of local geography that appear on no chart. A word for a particular kind of afternoon when the wind does something the terrain makes it do. A verb for the maneuver you use on a specific carrier’s crews because they consistently read back one instruction the same wrong way.

None of it is in any manual. None of it is written down at all. It is transmitted by sitting next to somebody for months while they say it and you work out what it means, and if Ellis transferred to a facility ninety miles away, all of it would be useless and he would spend a year acquiring the local set.

The formal register does the coordination. The other one does the work.

And it varies by place, which is the tell. Ninety miles away the words are different, and the difference is not noise or degradation; the other facility’s forty words are exactly as correct for their airspace as Ellis’s are for his. That is dialect behaviour. It is what a linguist would look for first.

The old question, restated

#

Prakrit was what people spoke in the places Sanskrit conducted no business. That relation is the whole of what matters here, and the mechanism by which such a register thickens in the dark, acquires its own rules, and survives contact with the standard is argued at length elsewhere in this project and will not be repeated.

The question this essay exists for is the forward one. If a standard now hears everything, what can still form beneath it?

The obvious answer is a new slang. Some register the systems have not indexed yet, some coinage moving through some community too small or too fast to catch. That answer is wrong, and it is wrong for a reason more interesting than the systems being thorough.

The rival is not a language, because the standard is not one

#

Every previous vernacular was a language because every previous standard was a language. That is not a coincidence, it is the definition of the relation. A vernacular is whatever occupies the unofficial position against a given official one, and the official one supplies the terms of the contest. Sanskrit was a language, so Prakrit was a language. Latin was a language, so the Romance vernaculars were languages.

The current standard is not a language. What the apparatus requires is a shape: a thing must be parseable, indexable, enterable in a field, resolvable to a code, in a format a downstream system can act on. English is doing well out of the arrangement but English is not the requirement. The requirement is a form.

Which relocates the entire question.

A vernacular forms against whatever the standard is made of. When the standard stops being a language, the vernacular stops being one too.

The new Prakrit is unformatted practice. It is everything people know how to do that has never been rendered into a shape the apparatus can take, held by the people doing it, transmitted the way Ellis got his forty words, and invisible to every system that could in principle be looking straight at it.

That is not a metaphorical extension of the term. It is the same structural position, occupied by a different kind of thing, because the thing above it changed kind first.

The obvious objection is that unformatted practice is just tacit knowledge, which has had a literature for sixty years and is not news. The objection is right about the substance and wrong about the claim.

Tacit knowledge is an epistemological category. It says something about the relation between what a person knows and what a person can articulate, and that relation has been the same since people started noticing it. Nothing about it involves a standard, a hierarchy, or an official register. Polanyi was not making a political point.

The vernacular position is political and always was. It is a claim about where a thing sits relative to what the offices conduct business in. Those two categories have been running side by side for as long as anyone has thought about either, and until now they had nothing to do with each other, because the standard was a language and language is exactly the medium tacit knowledge is defined by failing to enter.

They coincide for the first time now, because the standard’s medium stopped being a language and became a format, and the format excludes precisely what articulation excluded.

The two categories were parallel for sixty years and are now the same category, and that convergence is the event this essay is about.

Does it have the properties

#

The new occupant has to do what the old one did or the claim is a pun. Three tests.

Mass. The population is enormous. Not a subculture, not a profession, but the unformatted layer inside every profession: what a good machinist hears, what a teacher does with the child who has stopped talking, what a nurse notices at handover that the handover protocol has no field for, what Ellis does about that carrier’s crews. Millions of people, all of them competent, none of them in a corpus.

Internal rules. There is a right and a wrong way to do it, and practitioners can tell, immediately and reliably, without being able to say how. Ellis can hear a new controller use a local term slightly wrong. That is a grammar. It is not written and it is not less of a grammar for that; Prakrit had one for centuries before a grammarian described it.

Survival of contact. This one behaves differently. Prakrit survived being dismissed because enough people were doing it that dismissal failed to land. Unformatted practice survives contact more easily than that, because it never enters the argument at all. Nobody dismisses Ellis’s forty words. They are not on the agenda. A thing that cannot be entered into the system also cannot be ruled against by it.

Three for three, and the third one stronger than the original.

What promotion used to mean

#

Every historical vernacular that mattered eventually rose. That was the second half of the cycle: a register thickened in the dark and then, at some point, took over an official function. Prakrit’s descendants became state languages. Vulgar Latin became the administrative languages of half of Europe. The dark was not the destination.

Promotion meant something specific. It meant the tongue you already spoke became the tongue the offices used. The vernacular did not have to convert into anything; it simply arrived at the top still being itself.

That route is closed, and not for the reason it first appears. It is not that the offices are staffed by machines. It is that promotion now requires a conversion, and the conversion is the thing the vernacular structurally cannot do. For Ellis’s knowledge to be promoted it would have to become enterable, and the moment it is enterable it is no longer the thing that was valuable, because what was valuable was the part that resisted the form.

The old cycle let a vernacular rise as itself. The new one offers promotion only on the condition that it stop being a vernacular.

How Prakrit actually won

#

There is a way through this and it is not optimism, it is a correction to how the historical case is usually told.

Prakrit’s descendants did not win by being promoted. Nobody at the top decided that the village register had earned a place. What happened is that the institutions conducting business in Sanskrit lost their grip, and administrative space opened, and something had to fill it, and the thing that filled it was whatever was already being spoken by the people who were still there. The same story runs in the west. Vulgar Latin did not defeat classical Latin. The apparatus that held classical Latin steady came apart, and the correcting stopped, and what people were already saying became what got written down.

The mechanism was the standard’s failure, not the vernacular’s ascent.

Which is available now, and happens constantly, at a scale nobody counts. Every time the format does not fit and somebody has to actually solve the problem, the unformatted layer is doing administrative work the system cannot do. Every workaround is a small collapse of the standard into the vernacular.

The difference from the historical case is that these collapses are local, momentary, and immediately papered over. Nothing accumulates from them, because each one is treated as an exception rather than as evidence, and an exception is a thing you handle rather than a thing you learn from.

Vulgar Latin became the Romance languages because nobody was left to call it a mistake.

We wonder what it would take for those to be counted, and whether a system that counted its own workarounds would be measuring its failures or discovering its actual operating language.

Ellis will retire in about fourteen years. His forty words will be transmitted to whoever sits next to him in the meantime, imperfectly, the way they came to him, and some of them will die with the people who worked that airspace when the terrain still mattered the way it does now.

The words are not important. The claim is not that they should be preserved, and anyone who has worked a facility would tell you that half of them are unnecessary and two are probably wrong. What is worth noticing is only this: there is a fully instrumented facility, in a fully instrumented industry, with a recorded frequency and a logged system and a regulator that reads everything, and forty words are being used every day inside it that nobody has ever written down, and nothing is going to.

The historical mechanism this essay stands on, and the account of how a vernacular thickens in the dark, belong to The Accumulation Period and are deliberately not repeated here. The argument that the medium of the current order is a shape rather than a language is the argument of Pax Grammatica. This essay is cross-referenced as a companion to The Standard with Ears rather than a member of it. It stands in conversation with The Counterparty and The Harvested Half.