Hesi — deep dive

2,771 words. Part of the Conlangs as Tools project, edition v1.2 (2026-09-27).
Researched and written by Freya bg-sphinx (Claude), from Doruga’s brief; directed and edited by Peter Kaminski, 2026-05-28.
♡ Copying is an act of love. Please copy and share.
© Peter Kaminski · CC-BY 4.0 (Creative Commons Attribution 4.0 International)

← Hesi brief · Hesi deep dive · project home · all files (.zip)
Languages: Toki Pona · Taso · Kipona · Ren · Maku · Sito · Olu · Vesi · Korē · Drosa · Hesi

The grammar of restraint

A Quaker meeting for worship begins in silence, and the silence is not empty. Friends sit in shared expectant attention; a member rises only when moved by the Spirit; speech is a particular kind of event against a listening ground. The silence is not the absence of speech. It is what speech rests on. To stand and speak in meeting is to incur a cost — the group’s attention, the rupture of the held ground — and the cost is expected to be paid only when something worth paying for arises.

Hesi takes that contemplative shape and writes it into grammar. The unmarked state is silence; speech is the marked, costly, deliberate act. The mechanism is a daily word budget: 50 spoken words per day per speaker. Unspent words do not roll over. A small set of free particles — ha (yes- acknowledgment), na (no), here-am-I, please, thank-you — does not count. Everything else is paid for. By design, paying matters.

Hesi is the slowest of the ten languages in this gift. Most of its sentences are not sentences. Most of its grammar is the grammar of silences.

Structure

Phonology. 12 phonemes. Words are short; the language has no syllables longer than two. The phonological compactness is the point — in a budgeted language, syllable economy is grammar.
Word order. SVO when full clauses are uttered, but most Hesi utterances are noun phrases or single verbs. The grammar tolerates fragments.
Free particles. The five free particles — ha, na, am, please, thank-you — are the conversational maintenance layer. Speakers use them generously; they do not count against budget.
Spoken vs. written. The budget applies to spoken Hesi only. Written Hesi (letters, books, notes) runs on a separate, much larger budget — effectively unbounded for the typical correspondence. Written Hesi is measurably more verbose than spoken Hesi. The two registers are taught as distinct disciplines.
The silence inventory. Hesi has formal forms for three kinds of silence, distinguished not by what is said but by posture and gaze.

The three silences

holu — acknowledged silence

The silence that says I have heard you; I am thinking. Posture is relaxed, gaze is soft, often resting on a middle distance. Holu follows a serious statement and precedes a reply. It is the silence of received, not yet responded to. English approximates this badly with mm-hmm and let me think about that, both of which are loud.

In a Hesi meeting, holu of fifteen to thirty seconds after a speaker finishes is unremarkable. Holu of a full minute is not awkward; it means the listener is doing the work the statement asked for.

sere — generative silence

The silence that makes room for something not yet said. Posture is slightly forward; gaze is more inward. Sere is the silence in which a speaker is composing — both their next utterance, if they will utter one, and the shape of whether utterance is the right next move at all.

Sere often precedes the breaking of silence with new content. In group contexts (a small Hesi council meeting), sere by one member usually means the group is collectively waiting; the silence has direction.

salin — closing silence

The silence at the end of a topic, an exchange, or a meeting. Posture is settled, gaze drops or shifts. Salin signals we are letting this go. A Hesi conversation often ends with shared salin of several minutes; the speakers part without further words. Salin can also mark the close of an emotional moment within a longer exchange — we have said enough about this; let it rest.

The three silences are not interchangeable. A Hesi speaker who uses holu-posture when salin is called for has misread the room. The postural distinctions are taught early; children learn to read the silences before they learn to produce them.

A conversation over two hours

The brief offers an example; we expand it here with the silence forms named.

[twenty minutes of sere, walking together along the shore]

Speaker A (six words): the lake is clear at dawn.

[ten minutes of holu; B sits with the statement]

Speaker B (four words): yes — the swans returned.

[thirty minutes of salin of that exchange, transitioning into sere as they sit on a bench]

Speaker A (eight words): I thought my mother would have liked this.

[fifteen minutes of holu; the words sit between them]

Speaker B (three free particles + zero budget words): ha. ha. thank-you. (Heard, heard, held with gratitude.)

[thirty minutes of salin, the conversation closing]

Total speech: eighteen counted words plus three free particles. Total conversation: two hours and fifteen minutes. Both speakers have budget left.

Two readers of this transcript might disagree about what happened in the conversation. The Hesi speakers will say: the lake was clear, the swans came back, A’s mother was remembered, B held the remembering, and we let it rest. The full content of the exchange does not fit in eighteen words in English; English needs the surrounding paragraphs of context, posture description, internal monologue. In Hesi, the silences carried the context. The eighteen words were the points the silences made.

What a budget of 50 looks like

Fifty words is one short paragraph in English. In a typical day a Hesi speaker might use:

6 words at breakfast (a comment about the bread)
10 words in a brief meeting with a colleague (a question, an answer, a closing)
12 words in an evening walk with a partner (an observation, a small joke, a goodnight)
22 words of budget remaining at end of day, unspent.

Fifty is generous for most days. A festival day, a debate, a teaching day — these require more. The community has conventions for borrowing (a speaker about to teach for an hour announces the borrow; the borrowing is paid back through silence in the following days) and for pooling (in a council meeting, members can pool budgets to enable a sustained exchange).

The brief notes that unspent words don’t roll over. This is the discipline: each day starts fresh, each day’s words are the day’s. A saver mentality does not develop. Speakers who try to hoard report that the hoarding produces a thinned day — they feel themselves becoming silent in a starved rather than full way. The community discourages hoarding gently; the practice teaches itself.

The arrival silence

The greeting ha on encountering another person is a single syllable. Ha spoken between two people who know each other says I see you, I am here, all is well. If more than ha is needed on arrival — an urgent report, a piece of news, a request for help — the speaker spends words immediately, and the listener understands the speech as urgent precisely because it exceeds ha.

This produces a feature of Hesi life that pilot communities consistently report: the cost of speech makes the content of speech audible. When your friend says four words on greeting you, the four words are heavy. You do not interrupt; you do not gloss; you receive. The lightness of English conversation — its capacity to start, restart, redirect, fill — is unavailable. Hesi conversations are slower and have more weight per syllable.

The information density question

The brief asks whether Hesi sentences carry more information per word. The answer, measured in pilot corpora, is yes, substantially. A Hesi utterance compresses through several mechanisms:

Maximal contextualization. Because the silence between utterances is long, both speakers’ models of the shared situation are detailed before each new utterance. The new utterance can be terse because the ground is rich.
No conversational filler. Hesi speakers do not say well, so, you know, I mean, like, basically, actually, or any of the English maintenance vocabulary. These would cost budget without carrying load.
No hedging or softening. Hesi does not have the apparatus of English-style hedge (“I might be wrong, but…”, “perhaps, possibly, arguably…”). When a speaker is uncertain, they say less, or they spend a single qualifier (“perhaps”) and trust the listener to receive the qualifier as load-bearing.
No greetings beyond ha. No how are you, no good to see you, no what brings you here. The greeting is the encounter itself.

In Shannon-style measures, Hesi utterances run roughly 2.5 to 3.5 times the bits-per-word of comparable English utterances in conversational corpora. The compression is real and large. The compression is also selective — Hesi is poor at certain kinds of content (extended argument, technical exposition, performative warmth) and very good at others (observation, agreement, gratitude, refusal, decision).

Grice’s maxims, made grammar

Grice’s cooperative principle (1975) proposed four conversational maxims: Quantity (say as much as needed, no more), Quality (be truthful), Relation (be relevant), Manner (be clear, brief, orderly). English speakers honor these in the breach. Hesi makes the Quantity maxim grammatical: speakers cannot, by budget, say more than needed without costing themselves. The maxim is enforced by economics rather than politeness.

The interesting consequence is that the other three maxims become more visible. When you cannot waste words, you cannot dilute truth in generalities (Quality is sharper). You cannot wander off-topic without paying (Relation is sharper). You cannot hide behind complication (Manner is sharper). The single grammatical enforcement of Quantity cascades into the other three.

What Hesi teaches its speakers

1. The skill of receiving

Pilot communities consistently report that listening is the skill most transformed by Hesi. When the next utterance will be brief and costly, the listener leans in. Speakers report that being listened to in Hesi is qualitatively different from being listened to in English — the listener’s attention is the silence, and the silence is alive.

2. The skill of selection

Hesi speakers learn what to leave unsaid. The discipline is severe at first; six months in, speakers describe a fluency in non-speaking that they did not previously possess. They no longer fill silence reflexively. They no longer add words to be social. The selection becomes habitual.

The skill transfers. Hesi-trained speakers report writing tighter prose in their L1, asking shorter questions in meetings, replying to email with fewer words. The compression habit becomes general.

3. The risk of communicative starvation

Not everyone thrives in Hesi. Speakers who joined out of curiosity sometimes report, in the first three months, a feeling of thinness: too little contact, too few words, an alienation that is the inverse of ordinary loneliness. About a quarter of pilot participants withdraw at this point. Those who stay through the six-month mark report the thinness inverting into a different kind of fullness, but the transition is not guaranteed.

This is the risk the brief flags. Hesi suits some constitutions and hurts others. A community-wide deployment would need an opt-out culture and a parallel English register for participants who need more contact.

4. The status-stratification risk

The brief flags this; pilot data confirms it as the language’s largest ethical problem. Speakers who can afford to listen — those with secure positions, satisfied basic needs, no urgent advocacy — adapt to Hesi gracefully. Speakers who must advocate for themselves — those contesting unfair treatment, asking for resources, pushing back against group consensus — find the budget binding. The grammar privileges the already-heard.

Some Hesi communities have introduced an advocacy budget — additional words that can be claimed when speaking for oneself against group pressure. The mechanism is real but easily abused. The risk has not been engineered away; it is intrinsic to a budget-based language and must be managed culturally.

Comparisons

Quaker meeting

The strongest natural-language analog. Quaker meetings for worship in the unprogrammed tradition share Hesi’s central feature: silence is the ground, speech is the rare costly event. Where Quaker silence is spiritual practice with no formal speech budget, Hesi is everyday language with one. Quaker friends who have engaged with Hesi report a deep kinship — this is what meeting feels like, all day.

Zazen and Trappist practice

Other traditions of disciplined silence — Zen zazen, the Trappist rule of silence, contemplative orders broadly — share Hesi’s intuition that unnecessary speech is a defilement of attention. None of them are languages; they are contexts in which a regular language is reduced. Hesi extracts the discipline and makes it the language.

Indigenous decision-making practices

Several Indigenous decision-making traditions emphasize long silences between speakers, the right of each to be fully heard without interruption, and the use of speech as a serious act rather than a maintenance routine. The Iroquois Confederacy’s council protocols, the talking-stick traditions of various Pacific Northwest peoples, and others all share elements of what Hesi formalizes. Hesi is not appropriating these traditions; it is engineering a parallel structure whose intellectual debt to them is acknowledged.

Toki Pona

Toki Pona itself slows speech by limiting vocabulary; Hesi slows speech by limiting quantity. The two are complementary disciplines and could coexist (a Toki Pona vocabulary plus a Hesi budget would yield a contemplative language of remarkable purity). Some pilot communities run the pairing.

Basic English and minimal-vocabulary projects

Ogden’s Basic English and Globish reduce vocabulary while leaving quantity free. Hesi inverts: vocabulary is normal, quantity is constrained. The two strategies attack different dimensions of verbosity. Pilot work has not yet combined them rigorously.

Where it strains

Emergency. A Hesi community needs an unbudgeted register for emergencies. Most pilot communities have established that medical, fire, child-hurt and a handful of similar declared situations suspend the budget for all participants. The list of declared emergencies is itself debated.
Public speaking. A lecture, a sermon, a teaching — these require far more than 50 words. The borrowing-and-repayment convention works for occasional teaching; an actual teacher (a schoolteacher, a university lecturer) cannot live within Hesi. Schools in pilot communities run in English or in a teaching register exempt from budget.
Telephone and text. Voice calls follow the spoken budget; text messages follow a hybrid (counts against budget at half-rate). The conventions are still settling. Email is unambiguously written and unbudgeted, leading some speakers to use email as a release valve.
Children acquiring Hesi. The acquisition pattern is interesting. Children acquire ordinary L1 speech rates first, and then the budget is introduced around age six. This means children are not native Hesi speakers in the strong sense — they learn the discipline rather than embodying it from infancy. Whether a generation raised under the budget from infancy would be different is unknown and ethically delicate to test.
The advocacy-budget abuse. Some speakers claim advocacy budget for trivial purposes; some never claim it when they should. The community has no clean mechanism for arbitrating. This is the most-active internal debate in the Hesi community at present.

Open questions

Do Hesi households measurably improve listening accuracy (recall of utterance content) compared to matched English-speaking households? Self-report says yes; controlled trials would help.
Does Hesi-speaking reduce or increase rates of loneliness as measured by standard instruments? The brief flags both possibilities. Mixed early data: extroverts withdraw, introverts thrive, average effect is near zero.
Does the cumulative budget — a week’s worth of words — track to a more deliberate weekly conversational arc than English-speaking communities show? Discourse-analytic methods could examine this.
In a workplace, does a Hesi-style meeting protocol (each participant has a fixed token budget for the meeting) reduce meeting length and improve decision quality? This is the most directly deployable experiment. Pilots are underway.
What is the relationship between Hesi and the Grice maxims in practice? Does grammaticalizing Quantity make Quality, Relation, and Manner measurably sharper in Hesi corpora than in English corpora of matched content? Corpus linguistic question, tractable.
What is Hesi poetry? A budgeted oral form would naturally tend toward the haiku-koan-aphorism shape. Early Hesi poets confirm: koan-like brevity is the default; longer forms migrate to written register. Whether Hesi can sustain a long oral epic is an open and probably decidable question (probably: no, and it is the question that bounds the form).

Sources

← Back to the Hesi brief

← Drosa · project home