Register & Colloquial Forms
Tonesu isn't one voice. The same grammar serves formal proceedings, casual conversation, and institutional record-keeping — but the choices speakers make shift across contexts. This page shows what changes and what stays fixed.
Formal vs casual: polar questions
The clearest register marker is how you ask a yes/no question.
Formal
Are you going to the city soon?to-si — fronted before the proposition. Neutral, explicit, unambiguous. Appropriate in any context.
Casual
Going to the city soon?ku replaces to-si — and moves to clause-final position. Same question, different social signal. ku says: I know you, this is informal, we're peers.
The rule
Both are grammatically correct. Neither is "more correct." The choice signals social proximity (ku) or neutrality/formality (to-si —). Using ku in an arbitration hearing would be as jarring as using to-si — with a close friend over a meal.
Evidential frames across register
The evidential frame () wraps a clause to mark it as reported, inferred, or unattributed:
Spoken realization
In casual speech, prosody alone marks the evidential frame — a slight pitch shift or pause boundary. No explicit spoken token is needed.
In formal, legal, and dictation register, the spoken forms vund (opening) and vunds (closing) are required:
The stakes are different. In conversation, a missed evidential frame is a social misstep. In a legal record, it's the difference between "I assert this" and "I'm told this" — which can be the difference between perjury and testimony.
ze in formal proceedings
After genuine ze ambiguity is produced in a proceeding (see C008), subsequent turns prefer explicit NPs over ze back-reference until the referent set resets.
Casual
Fine —ze is clear enough from context and the conversation can tolerate a moment of ambiguity.
Formal (after ambiguity event)
After B's hypothesis introduces a competing referent, the formal record switches to the full NP (la-to-su-li) rather than risk ze ambiguity. This isn't a grammar rule — it's institutional practice. But it's consistent across the corpus.
ke — stale vs fresh denial
The pivot particle ke ("implied denial + advancing claim") and the explicit denial no — [claim] are not interchangeable. The choice depends on information freshness.
Fresh denial → use no —
When the denial is new information — the listener doesn't already know you disagree:
No — the doctrine has not crossed the epistemic threshold.The denial is performed explicitly because the listener needs to hear it for the first time.
Stale denial → use ke
When the denial is already contextually established and re-performing it adds no information:
[Moving past that:] the doctrine's resonance exceeds the plan.ke acknowledges the prior disagreement without re-litigating it, then advances. In formal registers (clinical, legal, diplomatic), using no — when the denial is stale may be coded as bad-faith re-litigation rather than genuine counter-evidence.
Colloquial contractions
High-frequency compounds of 3+ morphemes can contract to CVC shortforms in casual speech. The formal compound remains the canonical entry; the contraction is a registered alternative.
Registered forms
| Short | Full compound | Meaning | First attested |
|---|---|---|---|
zol |
zo-se-so-li |
canid (dog/wolf) | S269 |
zof |
zo-se-so-fe |
felid/fox | S269 |
zod |
zo-se-so-di |
bird | S284 |
zos |
zo-su |
plant | S292 |
mas |
ma-su |
rock/stone | S299 |
zom |
zo-se-ma |
fish | S313 |
zop |
zo-se-so-pa |
whale/cetacean | S311 |
zon |
zo-se-ne |
herd ungulate | S320 |
Contraction rules
- At least 3 morphemes in the source compound
- Short form is unambiguous within the discourse domain
- Formal compound remains the canonical written form
- CVC shape (matches the lexical atom tier)
Namespace collisions
Not every compound gets a clean short form:
zo-pe(arthropod) →zopcollides with whale → stays as disyllabiczo-pezo-ne(fungal) →zoncollides with herd animal → stays as disyllabiczo-ne
When the CVC namespace is already occupied, the compound doesn't contract. The collision is documented and the form stays long. No ad-hoc workarounds.
Epistemic nesting as institutional register
In formal institutional contexts, one speaker can certify another speaker's epistemic state:
I hold as established: you have not perceived the resonance.A knowledge-keeper converts a witness's personal epistemic floor into a formally certified matter of record. This construction — a to frame wrapping another agent's se/si/to state — is characteristic of institutional register. You wouldn't use it in casual speech because you wouldn't normally claim authority to certify someone else's perceptions.
Notation symbols in speech
Several written notation symbols have spoken equivalents. In casual speech these are rarely used (prosody serves instead), but in formal contexts they become mandatory:
| Symbol | Spoken form | Register |
|---|---|---|
' (juncture) |
peld |
pedagogical, dictation |
~ (approximation) |
ven |
all registers (also poetic) |
- (morpheme boundary) |
feld |
metalinguistic, teaching |
() (evidential) |
vund / vunds |
formal, legal, dictation |
"" (quotation) |
sild / silds |
formal, legal |
[] (aside) |
zeld / zelds |
dictation, pedagogical |
{} (structural slot) |
suld / sulds |
teaching, formal |
: (topic/definition) |
helm |
all (also predicate use) |
:: (identity) |
helms |
formal, poetic, philosophical |
/ (parallel) |
vel |
formal |
; (sequential) |
teld |
dictation |
— (suspension) |
el |
all (verse, poetic, formal) |
The spoken forms are always available but the expectation of their use is register-dependent. In a casual conversation, nobody says vund. In a courtroom dictation, everybody does.
Register summary
| Feature | Casual | Formal |
|---|---|---|
| Polar question | ... ku? |
to-si — ... |
| Evidential frame | prosody only | vund / vunds required |
ze after ambiguity |
continue using ze |
switch to full NPs |
| Stale denial | ke or no — (flexible) |
ke preferred; no — risks bad-faith reading |
| Long compounds | colloquial CVC forms OK | full formal compound |
| Notation symbols | silent (prosody) | spoken forms expected |
The grammar doesn't change across registers. The particle choices and disambiguation strategies do. Register in Tonesu is about how much precision and explicitness the context demands — not about different rules.