Stage 3 — Grammar and notation
Particles and notation marks are introduced together because you encounter them
together in every real Tonesu sentence. la-mi pa-re-mu lo-de ka-se ta-ti-de
has five particles and encodes agent, location, patient, action, and time — all in
one line. You've been reading sentences like this since Stage 0. Now the grammar
behind them becomes explicit.
Cluster 1 — The particle system
Every participant in a Tonesu sentence carries a particle that declares its role. Word order carries no grammatical weight — the particle does the work.
| Particle | Role |
|---|---|
la- |
perspective anchor — agent in action clauses; stance-holder in epistemic clauses |
lo- |
patient — the entity acted on, or in a state |
ro- |
instrument — the means or tool |
pa- |
location — where the action takes place |
ta- |
time reference — when |
ka- |
action marker — the action or process |
ne- |
relation / recipient — who receives, or the relational partner |
na- |
proper name marker — signals an identifier, not a compositional noun |
Particles prefix directly to what they mark. A full corpus sentence, role by role:
I examined the decay in the generator room — past. (C001 A1)
The order above is: agent — location — patient — action — time. Any order is valid.
lo-de la-mi ta-ti-de ka-se pa-re-mu says exactly the same thing.
Most particles are transparent. pa- (location) derives from pa (space/place).
ka- (action) derives from ka (intentional action). ne- (recipient) derives from
ne (relation). You already know these roots from Stage 1 — particles are the same
semantic atoms pressed into grammatical service.
Exercise 1 — Which particle marks the patient?
In la-mi pa-re-mu ___-de ka-se ta-ti-de, which particle marks the decay —
the thing being examined?
Explanation
lo- marks the patient — what the action is directed at. lo-de = the decay, as
the entity being examined. la-mi = I (as agent). pa-re-mu = at the generator
(location). ka-se = examine (action). ta-ti-de = past time.
Five particles. One sentence. Unambiguous parse regardless of which order they appear in.
Cluster 2 — Predication types
Three distinct ways to attribute a quality to an entity. They are not interchangeable.
Type 1 — lo-X Q — contingent state
lo-{entity} {quality} = "X is in state Q"
The entity is the patient — it holds or enters a state. The claim is contingent: the state can change.
lo-si-mu ru → The relay is stable. (C002 A3)
lo-pa ha-vo → The room is warm. (S033)
lo-li vo → The person is (socially) valued. (S163)
Type 2 — la-X Q — structural property
la-{entity} {quality} = "X has quality Q"
The entity is the agent/anchor — the structural bearer of the quality. The claim is intrinsic.
Type 3 — ka Q — manner adverbial
ka {quality} = "the action is performed with Q"
The critical minimal pair
Same noun, same quality, different particle, different ontological claim:
la-li vo → A person has worth. (S162) — intrinsic; cannot be revoked
lo-li vo → The person is valued. (S163) — contingent social esteem; can reverse
la-li vo makes worth structural — a property of personhood. lo-li vo makes it a
current social state. The language requires you to decide which claim you are making.
Exercise 2 — Intrinsic or contingent?
The claim: a person has worth — structurally, not revocably by social consensus.
Fill in the blank: ___-li vo
Explanation
The choice is an ontological commitment, not a stylistic one.
la-li vo = worth is a structural feature of personhood.
lo-li vo = this person is currently in a valued-state.
The same distinction applies to any quality: la-si-mu no-ru (the relay
structurally lacks coherence — by design) vs lo-si-mu no-ru (the relay is
currently in an incoherent state — it might recover).
Cluster 3 — Questions and negation
Questions
All questions use to-si (W026 — knowledge-seeking signal). Placement determines
the type.
Content questions — to-si in the argument slot
Put to-si where the unknown belongs. The proposition has a gap; to-si fills it.
Polar questions — to-si fronted, before the proposition
to-si — lo-pa-ra be-now → Is the field active? (C006 A1)
to-si — la-tu ki pa-li-pu? → Are you approaching the gathering? (C003 A1)
Casual register: ku? clause-final
In informal speech, ku? follows the complete proposition instead:
Answers to polar questions:
| Form | Meaning |
|---|---|
ru |
Yes — that coheres |
ru — {proposition} |
Yes + elaboration |
no |
No — that does not hold |
no — {proposition} |
No + elaboration |
ru and no are the primitive roots for unity and negation, used sentence-initially
as discourse responses. Their meaning is transparent — ru = it coheres; no = it
is absent.
Negation
Four scope levels:
| Level | Form | Example |
|---|---|---|
| Root prefix | no-X |
no-de (intact), no-ha (cold) |
| Compound prefix | no-{compound} |
no-ka-ki (don't go) |
| Clause negation | no {ka-clause} |
no {ka-se} (cannot be examined) |
| Contrast coordinator | A no B |
lo-to-re-su be no lo-wi-to (followed doctrine, not the plan) |
The contrast coordinator marks the first constituent as what actually occurred and the second as the rejected alternative. Both must be the same grammatical type.
Exercise 3 — Content or polar?
Which form asks "what quality of damage?" — seeking the unknown argument, not the truth value of the proposition?
- A content question puts
to-siwhere the unknown belongs (in the argument slot). - A polar question fronts
to-sibefore the full proposition.
Explanation
Position is everything. to-si inside the proposition = content question (unknown
argument). to-si before the proposition = polar question (unknown truth value).
de vo to-si? fills the quality slot with to-si: "damage of what quality?"
The fronted form to-si — de vo? asks: "is it the case that the damage has this
quality?" — seeking a yes/no.
Cluster 4 — The epistemic frame and notation
Personal epistemic modality
The epistemic frame places a confidence level between speaker and embedded proposition:
| Level | Root | Meaning | Entailment |
|---|---|---|---|
| Perceptual floor | se |
I have signal / I perceive | — |
| Hypothesis | si |
I am assessing / I hypothesize | entails se |
| Established | to |
I hold as known / I am certain | entails si, se |
The floor denial is the strongest claim: no-se (no perceptual basis) forecloses
everything above it. Denying only the ceiling (no-to) is consistent with still
having a hypothesis or a perceptual signal.
Corpus cluster:
la-mi si {lo-de no-ru} → I hypothesize the decay is unstable. (C001 A3)
la-mi to {lo-ze se} → I hold as established: that signal is perceptual. (C005 B1)
la-mi to {la-tu no-se lo-ne-ra} → I hold as established: you have no perceptual
basis for the resonance. (C007 A5)
The last example is a nested epistemic frame: the outer la-mi to certifies the
inner la-tu no-se {…} as an established proposition. No additional grammar required —
the inner frame is itself a well-formed clause in the embedded slot.
go — causal frame
go {premise}, result asserts a necessary connection — the premise produces the
result.
; — sequential connector
; connects two clauses as a sequence without asserting the mechanism. A ; B = "A,
and then B" — constant conjunction, not necessary connection. This is Hume's
distinction: go = necessary connection; ; = constant conjunction.
Use go when the mechanism is the point. Use ; when the connection is observed but
the mechanism is not being asserted.
/ — parallel partition
/ marks a formally paired bi-clausal structure. The relationship between the flanking
clauses (antithetical, complementary, causal) is supplied by content, not the mark.
Notation marks
| Mark | Name | Read as | Spoken form |
|---|---|---|---|
(clause) |
evidential frame | reportedly / unattributed | vund … vunds |
[text] |
aside frame | annotation — removable | zeld … zelds |
~X |
approximation | approximately X | ven |
— |
prosodic suspension | held in suspension | el |
"" |
quotation / mention | direct speech; use-mention | sild … silds |
: (clause) |
topic frame | as for {topic}, — | helm |
Evidential frame (): marks content as reported or unattributed. Use when citing
hearsay, a contested claim, or inference not directly asserted by the speaker.
(la-Yeshua ra-no-fe) = it is reportedly said that Yeshua is all-powerful.
~ approximation: pre-positional hedge. ~to-su = approximately a knowledge
system. ~tonesu = approximately truth = working conjecture.
Aside frame []: annotation that does not alter truth conditions. Removing every
[…] must leave the core argument unchanged — this is a self-policing constraint.
Discourse markers
| Marker | Position | Function |
|---|---|---|
he |
before name | vocative: he na Re'ka! = Re'ka! |
ya, |
clause-initial | attention: attend to what follows |
ke, |
clause-initial | pivot: prior claim denied (contextually), new claim advancing |
ke! |
clause-initial | heated pivot |
ru-fe, |
clause-initial | exclusive scope: only / solely: |
ya and ke can stack: ya, ke, {clause} = now pay attention — the prior position
was wrong and here is the correction.
Exercise 4 — Epistemic level
Complete: la-mi ___ {lo-de ru} — I hold it as established that the decay
is stable.
Explanation
to is the ceiling: the speaker asserts this as settled knowledge, not perception
or hypothesis.
The embedded proposition {lo-de ru} is Type 1 predication: the decay is in a
stable state. The full sentence la-mi to {lo-de ru} = I [as agent], hold-as-
established, [the-decay is-stable].
Grammar introduced in this stage
| Form | Type | Meaning |
|---|---|---|
la- |
particle | perspective anchor / agent |
lo- |
particle | patient |
ro- |
particle | instrument |
pa- |
particle | location |
ta- |
particle | time reference |
ka- |
particle | action marker |
ne- |
particle | relation / recipient |
na- |
particle | proper name marker |
go {…}, … |
frame | causal frame — necessary connection |
; |
connector | sequential connector — sequence only |
/ |
marker | parallel partition |
he |
particle | vocative |
ya, |
marker | attention signal |
ke, / ke! |
marker | pivot |
ru-fe, |
marker | exclusive scope |
(…) / vund … vunds |
notation | evidential frame — reported / unattributed |
[…] / zeld … zelds |
notation | aside / annotation |
~ / ven |
notation | approximation |
— / el |
notation | prosodic suspension |
Next
Stage 4 — Derived vocabulary works through the derived registry by root family. You'll parse 30–40 registered forms from their roots, see the operator patterns from Stage 2 extended to new domains, and drill longer compounds including four-root chains.