Skip to content

Fallacy-Resistance Corpus

Theme: Grammar & syntax · 33 sentences.

← Grammar & syntax · ← Corpus

FAL-001 · Fallacy-Resistance Corpus

Purpose: First attestation of the () evidential frame and : topic frame operators, both formalized in spec/grammar.md and spec/phonology.md. Each sentence demonstrates how Tonesu's grammar either structurally prevents a classic logical fallacy, or — when prevention is not possible — makes the fallacious move structurally visible and therefore contestable. Ten fallacy types tested: equivocation (S364), straw man (S365), false dichotomy (S366), circular reasoning (S367), ad hominem (S368), appeal to authority (S369), post hoc ergo propter hoc (S370), composition/division (S371), is/ought (S372), modal fallacy (S373).

S364 P1: lo-ra to-fe ne P2: lo-wi-fe lo-la-wi-li go ne C: (lo-ra lo-la-wi-li go ne) P1: Force obeys an epistemic regularity. P2: Normative rules require a willer. C: (So it is claimed: force requires a willer.)

Notes

  • to-fe (W028) = epistemic boundary / natural regularity = "law of physics." wi-fe (W100) = will-imposed limit / normative rule = "law as obligation." These are structurally distinct compounds in Tonesu — no single lexical item spans both meanings. The equivocation requires explicitly substituting one compound for the other mid-argument.
  • The () frame on C marks the conclusion as an unattributed claim — a reported inference rather than a demonstrated consequence. A speaker presenting C bare (without ()) would be asserting first-person certainty for what is actually a non-sequitur, making the epistemic overreach visible to any interlocutor.
  • Structural prevention: the fallacy exploits a single English word ("law") covering two distinct concepts. Tonesu assigns these to to-fe and wi-fe — structurally non-identical forms. The equivocating inference to-fe → wi-fe → willer-required is impossible without an explicit identity bridge (to-fe ne wi-fe), which a speaker would have to assert and defend.
  • la-wi-li = willer-person = an agent whose nature is will-directed. Compositional: la- (agent prefix) + wi (will) + li (person) — used here as a noun for "a rule-making agent."

S366 go lo-zu ne-li no , lo-zu wi-de-li ne Because you are not a bonded ally, you are an adversary.

Notes

  • Three-way vocabulary: ne-li = bonded relational person (neighbor / ally); no-ne-li = person without bond (stranger); wi-de-li (W124) = person of opposing will (adversary/enemy). The false dichotomy collapses the three into two by treating no-ne-li as equivalent to wi-de-li. Tonesu's compositional roots make the three categories lexically distinct.
  • go lo-zu ne-li no , … = "because [you are not a ne-li]": the fallacious move is the go (causal frame) — grounding adversary-status in the absence of bond.
  • no-go in the correction = not-therefore = explicit denial of the causal frame. no-go lo-zu wi-de-li ne = "it does not follow that you are an adversary." This directly targets the go-relation the false dichotomy relied on.
  • Structural exposure: the vocabulary gap between no-ne-li (merely un-bonded) and wi-de-li (actively opposing) is intrinsic to the compounds. The inference from "not ne-li" to "therefore wi-de-li" skips no-ne-li, and a Tonesu interlocutor can insert it: you claimed ne-li no → wi-de-li, but ne-li no = no-ne-li, and no-ne-li ≠ wi-de-li.

S367 A: lo-to-re-su to , go la-be-go-li lo-ze si-past B: la-be-go-li lo-ze si-past , go lo-to-re-su si A: The scripture is confirmed — because the creator attested it. B: The creator attested it — because the scripture says so.

Notes

  • Each sentence uses go (causal/grounding particle) to ground one claim in the other. A-grounds-B (scripture's truth depends on creator's attestation); B-grounds-A (creator's attestation depends on scripture's claim). The loop is fully explicit: two go-frames pointing back at each other.
  • Grammar surfacing the circularity: in Tonesu, go [cause] [effect] asserts a genuine grounding relation — the go claim can be challenged on its own merits. When two go-relations mutually ground each other, both can be laid side by side and the absence of any independent go-ground becomes visible. An interlocutor asks: is there a third sentence — one that grounds either A or B without using the other?
  • Well-formed alternative: break one link. For example, replace B's go lo-to-re-su si with an independent grounding: la-mi se lo-be-go-li = "I perceived [the creator directly]" — a first-person perceptual claim (se) outside either circular premise. Whether that claim is defensible is a separate question; it at least breaks the loop.
  • This sentence also demonstrates the to-re-su (W064) / be-go-li theological vocabulary pair from the derived registry in a critical-reasoning rather than devotional context.

S369 (la-to-li lo-to-su no-to) , (du lo-to-su no-be) (The scholar reportedly does not accept the model.) (Therefore, reportedly: the model does not hold.) [humble/correct form]

Notes

  • to-li (W003) = scholar/expert/knower. to-su (W030) = theory/model. no-to = does not confirm (negation of the epistemic certainty predicate). no-be = does not hold / does not exist.
  • The correct form (first sentence) keeps both the cited authority's stance and the derived conclusion inside (). Both are reported. Neither is the speaker's first-person certainty. An honest speaker using a cited authority wraps both the report and any inference in ().
  • The problematic form (second sentence) asserts la-mi to (I confirm) as the matrix predicate, then grounds it with a go-clause wrapping () evidential content. The go here bridges a () premise to a la-mi to conclusion — moving from reported/uncertain to personally-certain. The grammar does not block this, but the visible structure makes the move explicit: the speaker is claiming certainty from an anonymous report.
  • Principle: in Tonesu, go (premise) → la-mi to [conclusion] is always a suspicious pattern. Valid grounding flows from first-person perception, signal from confirmed sources, or demonstrated reasoning — not from anonymous () content.

S370 ta [la-mi lo-ma-no-de mu-ka-past] , la-mi zo-de-fe-past When I used medicine, my illness ended. [temporal frame: ta]

Notes

  • ma-no-de (W046) = medicine / treatment substance. mu-ka (W022) = use/operate an artifact. zo-de-fe = organism-fatigue-boundary = the endpoint of organism-fatigue = recovery.
  • ta and go are structurally distinct clause introducers in Tonesu (§Subordinate Clauses). ta introduces a temporal co-occurrence frame without causal implication. go introduces a genuine causal or grounding claim.
  • The fallacy surfaced: the post hoc argument substitutes go for ta. In English, "when X happened, Y happened, therefore X caused Y" — the causal claim is implicit and smuggled through the temporal conjunction. In Tonesu a speaker must commit to one of the two clause introducers. Using go is an explicit claim of causation, disputable on its merits. Using ta is an honest assertion of temporal co-occurrence without causal weight.
  • A speaker who uses go here has incurred an epistemic debt: they must now be able to defend the causal claim (by ruling out spontaneous recovery, other interventions, etc.). ta incurs no such debt. The grammar surfaces which kind of claim is being made.

S371 lo-ko-mu : nu-de ne As for each component: it has small mass. [part claim]

Notes

  • ko-mu (W052) = container-artifact = the bounded component unit. su-ru (W085) + mu = su-ru-mu = integrated-whole-artifact = the assembled machine as a unified structure. Two lexically distinct topic NPs.
  • : (topic frame) forces the speaker to name the referent of each claim. The part claim has topic lo-ko-mu; the whole claim has topic lo-su-ru-mu. The structural shift between the two : declarations makes the composition move explicit: you changed what you're talking about.
  • The () frame on the conclusion marks it as an unestablished inference — not a first-person certainty. Composition from parts to whole requires justification beyond the part-property, and () signals that the speaker is not personally vouching for the inference.
  • Division fallacy (reverse): lo-su-ru-mu : nu-de ne , (du lo-ko-mu : nu-de ne) — the whole is lightweight; therefore (allegedly) each part is lightweight. Same structure in reverse; both are exposed by the : topic-shift.
  • nu-de = quantity-decrease = small amount/mass. Compositional: nu (digit/quantity) + de (decrease) = quantitatively small.

S372 la-wi-zo lo-wi-zo de [descriptive] Wild animals kill each other. [de — descriptive fact]

Notes

  • The descriptive sentence uses de (primitive: decrease/decay) as the bare predicate — a natural process, no agent intentionality marker. The normative term is no-ka-de-zo (W128 negated) = not-deliberately-kill = the prohibition against murder. no-ne = does not hold (negation of the relational predicate).
  • The inferential move is wrapped in (): the claim that the descriptive fact grounds the normative conclusion is presented as a reported assertion, not something the speaker is directly vouching for.
  • Hume's guillotine: in Tonesu, de/be/ki are descriptive predicates (process, growth, motion); vo, wi-fe, ka-to-fe, and no-ka-de-zo are normative predicates (value, rule, judgment, prohibition). A go clause bridging a descriptive predicate to a normative conclusion is an explicit claim that the fact grounds the norm. This bridge itself requires justification — it cannot be taken for granted. Tonesu forces the speaker to state the go claim explicitly; natural language often leaves the bridge implicit.
  • The () frame signals: "this is a reported/contested inference." A speaker who makes the bridge bare (without ()) has committed to the claim that natural behavior is normatively binding — a substantive and contestable premise.

S373 la-mi si lo-ti-be lo-ma-de I hypothesize: in the future, material resources will decline. [si — proposed/possible]

Notes

  • la-mi si lo-X = I signal/propose X — the speaker's epistemic predicate is si (hypothesis / proposal / possibility). la-mi to lo-X = I confirm X — the epistemic predicate is to (organized knowledge / certainty). These are structurally distinct predicates on la-mi.
  • ti-be = future time (W040). ma-de = matter-decrease = resource depletion. Compositional.
  • The modal fallacy is the bare du (therefore) claiming to elevate si to to. In Tonesu, moving from la-mi si to la-mi to requires genuine new grounding — additional perception (la-mi se lo-X), a signal from a confirmed source, or a logical derivation from established knowledge. The du particle alone does not provide this.
  • The () frame wraps the fallacious elevation: the speaker acknowledges this inference is not directly warranted and flags it as a reported/ungrounded claim.
  • Well-formed epistemic progression: la-mi si lo-X , go [new evidence] , la-mi se lo-X , du la-mi to lo-X = I proposed X; because of [new evidence], I perceived X; therefore I now confirm X. Each step incurs and discharges an epistemic commitment.

Batch Summary

Confirmed results:

  1. () first attestation. ✅ S365 (straw man), S369 (appeal to authority), and S373 (modal fallacy) provide the first corpus uses of the evidential frame. All three function as designed: reported/cited content in (), speaker-certified content bare. The contrast in each pair is immediately parseable.

  2. : first attestation. ✅ S368 (ad hominem) provides the first corpus use of the topic frame. The mechanism works cleanly: two sentences with two different : topic NPs make the topic shift structurally explicit. S371 (composition/division) provides a second attestation within the same batch.

  3. ta vs go as fallacy exposure. ✅ S370 (post hoc) demonstrates the sharpest structural prevention in the batch: the grammar forces a speaker to commit to either temporal co-occurrence (ta) or causal grounding (go). There is no temporal-to-causal drift in Tonesu.

  4. Structural prevention (non-(), non-: mechanisms): S364 (equivocation) — distinct compounds to-fe / wi-fe prevent same-word ambiguity. S366 (false dichotomy) — three-way vocabulary ne-li / no-ne-li / wi-de-li exposes the excluded middle. Both show that lexical compositional transparency does substantial fallacy-resistance work prior to any meta-operator.

  5. Descriptive/normative gap. S372 (is/ought) confirms that Tonesu has no structural bridge from de/be/ki predicates to vo/wi-fe/ka-to-fe predicates. The gap is not enforced syntactically but is visible at the predicate level — a go-bridge across this gap incurs an explicit contestable commitment.

  6. (du …) pattern. ✅ First attestation of (du …) = "reportedly-therefore" — the () frame wrapping a du-introduced derived claim. Marks the inference as unestablished. Pattern: (premise) , (du [conclusion]) = the speaker neither confirms the premise nor endorses the conclusion, presenting both as in circulation.

Open grammar questions (FAL-follow): - [x] FAL-F-001: ~~Is there a well-formed Tonesu expression for "either/or" (inclusive/exclusive disjunction)?~~ Resolved. No disjunction particle needed. False dichotomies are better exposed by no-go + excluded-middle vocabulary than by a logical OR operator. A formal disjunction particle would serve symbolic logic but would reduce fallacy visibility in natural reasoned discourse — it would let a speaker assert a disjunction without committing to why the alternatives are exhaustive. The current no-go-based approach forces the causal denial to be explicit. Policy: do not introduce a disjunction particle. - [x] FAL-F-002: ~~Does Tonesu need an explicit rule blocking go (reported-premise) → la-mi to [conclusion]?~~ Resolved. Do not block grammatically. (1) Sometimes legitimate: (scientists report X) can warrant la-mi to X when the listener endorses the source. (2) Structural visibility is sufficient: the move is explicit — a go-chain from a () premise to a bare la-mi to conclusion is legible and challengeable. An interlocutor can ask: why does reported evidence yield personal certainty? Blocking syntactically would remove a valid inference mode. Policy: the move is legal, it is marked, and the reader holds the burden of challenge.

FAL-001 · first : attestation · Topic Frame First Attestation

S368 lo-to-si-ze : la-mi lo-ze ka-to-fe , no lo-ze As for her proposition — I judge it, not her. [correct form]

Notes

  • S368 is the first corpus attestation of the : (topic frame) operator.
  • The first sentence sets the topic as lo-to-si-ze (her proposition / the claim she made). ka-to-fe (W122) = deliberate epistemic bounding = to judge/evaluate. The topic frame declares: I am evaluating the proposition, not the person.
  • The second sentence shifts the topic to lo-ze (her, as person). ki-pa-mu-be = large vehicle (compositional: ki-pa-mu W086 = vehicle + be = growing/large). The irrelevant personal fact is now explicitly framed as the topic.
  • Grammar doing the work: the : line requires a sentence-initial topic NP declaration. A speaker who opens with lo-to-si-ze : and later introduces lo-ze : has made two structurally distinct topic commitments. The shift is audible/visible. An interlocutor can flag it: "your topic changed — you switched from the proposition to the person. How does the latter bear on the former?"
  • This is structurally stronger than natural-language fallacy detection, which relies on pragmatic inference to notice the shift. In Tonesu, the topic is an explicit grammatical commitment at each sentence.

FAL-002 · Rhetorical Fallacy Corpus

Purpose: FAL-001 demonstrated Tonesu's structural resistance to logical fallacies (equivocation, false dichotomy, post hoc, circular reasoning, modal inflation). FAL-002 tests rhetorical fallacies — moves that exploit pragmatic rather than purely inferential gaps, and which interact more directly with the : topic frame and () evidential frame. Five cases: motte-and-bailey (topic-switch under challenge), moving the goalposts (retroactive condition change), slippery slope (unjustified causal cascade), appeal to emotion (affective-to-normative leap), loaded question (embedded presupposition). The first three stress-test : and wi-fe; the last two stress-test () and the go-chain requirement.

S374 Bailey (strong claim, bare assertion): la-ka-li-su : lo-zo-li to-no-fe ka Motte (retreat when challenged): la-ka-li-su : lo-zo-li wi-ne ne Correction: go [la-ka-li-su lo-zo-li wi-ne ne] , no la-ze du lo-to-no-fe ka Motte-and-bailey: strong claim retreating to weak claim.

Notes

  • : forces explicit predicate commitment at every defensive move. Bailey predicate = to-no-fe ka (unlimited-knowledge-action = total authority); Motte predicate = wi-ne ne (will-relation holds = has intentional engagement). These are structurally distinct predicates.
  • A listener can demand at any point: which predicate are you defending? The : framing makes this a grammatically answerable question, not a rhetorical dispute.
  • no la-ze du lo-to-no-fe ka = "it does not produce unlimited-authority-action" — establishing wi-ne ne yields only wi-ne ne, no more.
  • wi-ne = will-relation = intentional/purposive engagement (compositional wi will + ne relation; new compound, ⚠️ proposed).

S375 (1) Original condition (attributed, past): (go [la-mi lo-ra se] , la-mi lo-to-si-ze to-ki) (2) Condition met: la-mi lo-ra se-past (3) New goalpost (bare, present): la-mi wi-fe be : lo-ra-ne se-to Moving the goalposts: retroactive condition change.

Notes

  • wi-fe be = W100 wi-fe + be inchoative = rule-comes-into-existence. This is the structural tell: the standard has been replaced. The old condition was inside () (attributed, reported); the new condition is bare-asserted in the present. No grounds are given for the change.
  • The asymmetry is legible: () premise → met condition → ungrounded bare wi-fe be. An interlocutor can ask: what warrants the new condition entry?
  • to-ki = knowledge-movement = epistemic approach to certainty (compositional to + ki; new compound, ⚠️ proposed).
  • ra-ne = signal-relation = corroborating/relational evidence (compositional ra + ne; new compound, ⚠️ proposed).
  • se-to = perception-knowledge = perception that has risen to certainty (compositional se + to).

S376 Fallacious chain: go [lo-wi-fe de] , (du lo-ne-ki de) , (du lo-ka-li-su de) , (du lo-ne-su de) Correction: go [lo-wi-fe de] , la-mi no-se lo-ne-ki de Slippery slope: unjustified causal cascade.

Notes

  • The fallacious chain has one bare go [X] (speaker certifies the initial causal premise), then wraps every subsequent step in (du …) — each marking a further alleged-therefore. In English, "then Y follows, then Z, then catastrophe" sounds like a continuous causal chain. In Tonesu, the speaker must use go [X] , Y for each certified step. Here the speaker wrote go [X] once and stacked four (du...) consequences — the uncertainty structure is directly visible.
  • The correction uses no-se (W086: no perceptual basis) to mark the first inferential gap: even granting the initial premise, the next step has no grounding.
  • For a legitimate causal chain, each step would need its own independent go-grounding before being asserted bare.
  • ne-ki = relation-inchoative = bond-forming / social connection. First corpus attestation S013 as ne-ki verb; here lo-ne-ki de = bond-forming decays = social bonds weaken.
  • ne-su = relation-structure = social/relational fabric (compositional ne + su; new compound, ⚠️ proposed).

S377 Illegitimate (emotional fact → specific policy without warrant chain): la-zo-ne-du lo-fa de-past , (du la-mi lo-zo-ne-du pa-ki) Legitimate (full warrant chain stated): go [la-zo-ne-du lo-fa de-past] , go [la-mi lo-zo-ne-du vo-wi] , la-mi lo-zo-ne-du pa-ki Appeal to emotion: affective fact to normative conclusion.

Notes

  • The illegitimate form presents an affective fact (lo-fa de) and wraps the specific responsive action in (du …), marking the causal leap as alleged. The emotional state alone does not specify which response is warranted — multiple responses are possible, none generated automatically by fa de.
  • The legitimate form surfaces the minimum chain: (a) the affective state as fact, and (b) the agent's declared value commitment (vo-wi). Only with both in play does the action receive warrant. A fully explicit form would add a third go-link: the empirical claim that this specific action reduces fa-de.
  • Design insight: the go-chain requirement forces a speaker to distinguish the state that triggers feeling from the warrant for a specific response. Emotional facts belong in the chain; they do not bypass it.
  • zo-ne-du = organism-lineage-result = offspring/child (compositional zo-ne lineage + du result; new compound, ⚠️ proposed).
  • pa-ki = space-movement = relocation (compositional pa space + ki movement; see also ki-pa corridor, attested NUM-001).
  • vo-wi = value-will = will directed at value / cherish (compositional vo + wi).
  • fa-de = affect-decay = affective distress (compositional fa + de).

S378 Loaded (presupposition hidden inside question frame): to-si [la-zu lo-to-de de-fe-past] Well-formed interrogation (presupposition extracted and marked): (la-zu lo-to-de ka-past) , to-si [la-ze de-fe-past] Loaded question: smuggled presupposition.

Notes

  • The loaded form embeds la-zu lo-to-de ka-past (you were deliberately distorting knowledge) directly inside the to-si question frame as if it were established. The listener is invited to answer the question, which implicitly accepts the presupposition.
  • The well-formed form extracts the presupposition, wraps it in () (contested/reported, not established), then asks only about the terminus of that alleged practice. A listener can now step back and challenge the () premise before engaging with the question.
  • This confirms the general principle: any to-si question that contains a contestable premise requires that premise to be made explicit as () before the question is well-formed. The mechanism is identical to the one that governs bare assertion hygiene.
  • to-de = knowledge-decay = falsehood/distortion (compositional to + de; new compound, ⚠️ proposed).
  • de-fe = decay-limit = terminus / the point at which a decay process ends (compositional de + fe).

Batch Summary

Confirmed results:

  1. : in motte-and-bailey. ✅ S374 confirms that : holds the speaker to explicit predicate commitment at each defensive move. Bailey (to-no-fe ka) and Motte (wi-ne ne) are structurally distinct — the : frame makes the substitution a visible grammatical event, not an unmarked rhetorical slide. Establishing the Motte yields only the Motte predicate; no la-ze du lo-to-no-fe ka is the explicit refusal of the bailey.

  2. wi-fe be as goalpost marker. ✅ S375 establishes wi-fe be (rule-comes-into-existence) as the structural signal of retreating standards. The old condition lives in () (attributed, past); the new condition is bare-asserted present. The asymmetry is directly readable; no meta-commentary is needed.

  3. (du …) cascade for slippery slope. ✅ S376 demonstrates that a slippery-slope argument naturally produces a tower of (du …) wraps, each certifying one uncertain link. The speaker's commitment level is rendered transparent per step: one bare go launching four (du...) alleged-therefores is immediately parseable as "only the first step is asserted."

  4. go-chain explicitness for appeal to emotion. ✅ S377 confirms that affective states are valid inputs to causal chains but cannot short-circuit them. At minimum a value-commitment (vo-wi) must be stated alongside the emotional fact to warrant any specific action. The (du …) on the policy conclusion in the illegitimate form marks the inferential gap correctly.

  5. to-si presupposition extraction. ✅ S378 confirms that well-formed interrogation requires embedded contestable premises to be extracted and wrapped as () before the question is posed. This is not a new rule — it is the () principle applied to the question-formation domain.

Cross-batch observation (FAL-001 + FAL-002): Rhetorical fallacies engage the same four mechanisms as logical ones. The motte-and-bailey is a : topic-predicate substitution. The loaded question is a () extraction failure. The slippery slope is undisclosed (du …) stacking. The appeal to emotion is an incomplete go-chain. No new grammatical machinery was needed for FAL-002 — the existing system covered all five cases.

Open grammar questions (FAL-follow): - [x] FAL-F-003: ~~wi-fe be registration question.~~ Resolved. Do not register yet. The inchoative composition wi-fe + be is fully self-evident from the existing spec (as are all [root] be / [compound] be inchoatives). Register only if it recurs frequently enough to warrant a canonical gloss — e.g. in formal debate-register or legal-rhetoric corpus entries. Single-attestation compositional forms do not meet the registration threshold. - [x] FAL-F-004: ~~Is a nested presupposition corpus test needed?~~ Resolved: yes. The single-presupposition case (S378) is necessary but not sufficient to establish the principle at scale. The truly difficult human case stacks multiple hidden premises under one surface clause ("have you stopped lying to protect your corrupt friends?" bundles: you lied, your friends are corrupt, your lying protected them). Whether () extraction scales cleanly when there are two or three hidden commitments is a distinct question. Programmed as FAL-003.

FAL-003 · Nested Argumentation Corpus

Purpose: FAL-001 and FAL-002 established Tonesu's defenses against single-layer fallacies. FAL-003 tests depth: stacked, nested, or layered argumentative moves where ordinary languages become "soup." Five cases: nested loaded question (multiple simultaneous presuppositions), double topic-substitution (motte-and-bailey with an intermediate step), compound slippery slope (cascade structure where some links are established and others aren't), nested evidential attribution (reported-within-reported), and nested modal inflation (epistemic ladder climb across multiple agents and frames). The governing question throughout: as argument depth increases, does the () / : / go-chain machinery scale, or does it become unreadable?

S379 Triple-loaded (all three presuppositions hidden): to-si [la-zu lo-wi-fe de-fe-past] Full extraction (three () premises, then clean question): (la-zu lo-ne-ki de-past) (la-zu go lo-mu-su lo-zo-li ka-past) (la-zu lo-ka-li-su lo-mu-su be-past) to-si [la-zu lo-wi-fe de-fe-past] Nested loaded question: three bundled presuppositions.

Notes

  • Each () block is an independent presupposition — independently contestable. The listener can accept any subset and reject the rest before engaging with the question about termination.
  • Extraction has no required ordering — each () block is an independent premise and may appear in any sequence. Discourse convention may prefer a foundational-to-derivative alignment (authority-holding → harmful actions → relational consequences) but this is not a grammatical rule. The ordering here (relational decay → actions → authority) reflects the natural challenge sequence rather than a deduction order. Both orderings are well-formed.
  • Scaling result: () extraction scales to three simultaneous presuppositions without grammatical strain. Each () block is its own clause with standard grammar. The question to-si [...] following three () blocks is unambiguous: the question is about whether the practice described in aggregate ended.
  • mu-su = harm-structure = structured harm / harmful pattern (compositional mu + su; W050 mu-be attested — harm-inchoative; mu-su new ⚠️ proposed).

S380 Claim (A — strong): la-ka-li-su : lo-zo-li to-no-fe ka Challenged. Retreat to B (intermediate): la-ka-li-su : lo-zo-li vo-wi-fe ne-past Challenged again. Retreat to C (undeniable): la-ka-li-su : lo-zo-li ne ne Correction (blocking A from both retreats): go [la-ka-li-su lo-zo-li ne ne] , no la-ze du lo-vo-wi-fe ne go [la-ka-li-su lo-zo-li vo-wi-fe ne] , no la-ze du lo-to-no-fe ka Double topic-substitution: motte-and-bailey with intermediate step.

Notes

  • The double-retreat is the harder case. It is common in practice because a speaker forced off bailey-A does not concede to C directly — they retreat to an intermediate B that sounds like a reasonable position while still being much closer to A than to C.
  • : forces explicit predicate commitment at every step: A, B, and C are structurally distinct. The correction needs two separate blocking statements — one per retreat. This is more expensive rhetorically but unavoidable: each step is a distinct claim and earns a distinct block.
  • Scaling result: : scales to multi-step substitution sequences. Each step requires either defense of that predicate on its merits or an explicit block. The grammar does not allow the speaker to preserve the inferential benefit of A while standing on C — each : frame commits to exactly one predicate.
  • vo-wi-fe = value-will-rule = normative obligation / duty (compositional vo value + wi will + fe limit/rule; new compound ⚠️ proposed).

S381 Fallacious (three bare links + one bracketed, legitimacy gradient hidden): go [lo-wi-fe de] , lo-ne-su de , lo-ka-li-su de , (du lo-ka-no-fe de) , lo-ne-su be-no-fe Corrected (each link labeled by speaker's actual commitment): go [lo-wi-fe de] , la-mi se lo-ne-su de go [lo-ne-su de] , la-mi no-se lo-ka-li-su de go [lo-ka-li-su de] , (du lo-ka-no-fe de) (du lo-ka-no-fe de) , (du lo-ne-su be-no-fe) Compound slippery slope: mixed established and unestablished links.

Notes

  • The fallacious form has an invisible legitimacy gradient: the first link may have some empirical grounding while the last two are fantasy. In English (and most natural languages) these are strung together with "and then" at equal syntactic weight, hiding the gradient.
  • The corrected form uses per-link epistemic marking: la-mi se (I have some basis), la-mi no-se (I have no perceptual basis), then (du …) for the two ungrounded final links. A listener can evaluate each link independently.
  • Scaling result: go-chaining + per-link epistemic predicates handle mixed-reliability cascades. The grammar forces the speaker to either commit to each link or visibly hedge it. There is no syntactic mechanism for distributing weight across a chain uniformly when the weight is uneven.
  • ka-no-fe = action-without-limit = unchecked power / authority-without-boundary (compositional ka + no + fe; new compound ⚠️ proposed).

S382 Standard evidential (one layer): (la-to-li lo-to-su ne) Nested evidential (two layers — attributed report of a report): (la-ka-li-su (la-to-li lo-to-su ne) re-ka) Speaker endorsement breaking the chain: la-mi no-se lo-to-su ne , go (la-ka-li-su (la-to-li lo-to-su ne) re-ka) Nested evidential attribution: reported-within-reported.

Notes

  • () nesting: an outer () makes the entire inner clause evidential; the inner () separately marks the scholar's attribution. The result is a committed two-step provenance chain, completely readable without special notation. The outer () means: this reported-report is itself the content of the evidential frame.
  • Scaling result: () stacks cleanly. (A (B) ka) = "A did some action regarding [the reported content B]." The inner () marks B as uncertified from the outer speaker's perspective; the outer () marks the whole reported-action as uncertified from the ultimate speaker's perspective.
  • re-ka = recurring-action = announced/repeatedly stated (compositional re repetition + ka action; new compound ⚠️ proposed).
  • Significance: this pattern handles hearsay chains (X said Y said Z claims W) without new machinery. Each link is an outer () wrapping an inner (). The speaker can break the chain by asserting the bottom link directly if they have direct knowledge, but cannot claim direct knowledge without removing the brackets.

S383 Agent A's report (attributed hypothesis): (la-na-Ren lo-to-si-ze si) Agent B's elevation (transforms A's hypothesis to B's inference): go [(la-na-Ren lo-to-si-ze si)] , la-na-Kael lo-to-si-ze se Institutional elevation (transforms B's inference to institutional certainty): go [la-na-Kael lo-to-si-ze se] , la-to-fe-su lo-to-si-ze to-su Speaker's diagnosis: la-mi no-to lo-to-si-ze to-su , go [la-to-fe-su lo-to-si-ze to-su] , go [la-na-Kael lo-to-si-ze se , go (la-na-Ren lo-to-si-ze si)] Nested modal inflation: epistemic ladder across multiple agents.

Notes

  • Each elevation step uses go [previous-epistemic-stance] to ground the next. This forces the chain to be fully linearized: there is no way to achieve multi-agent epistemic inflation without writing out the complete sequence of go-links.
  • The speaker's diagnosis sentences exactly retrace the chain in reverse, demonstrating that the chain from si to to-su across three agents has no independent grounding — each link only borrows from the link below it.
  • Scaling result: the go-chain + epistemic predicate system (si/se/to) scales to multi-agent epistemic inflation. The grammar does not prevent the inflation — but it requires each agent to explicitly commit to their elevation step, making the full chain auditable and the borrowed grounding visible.
  • to-fe-su = epistemic standards framework/body (W072 ✅, active). Registered before the FAL series; first corpus use S128 (CF-001-C). FAL-003 S383 is the FAL-corpus first attestation.
  • to-su = knowledge-structure = established-knowledge / institutional fact (compositional to + su).

Batch Summary

Confirmed results:

  1. () scales to three simultaneous presuppositions. ✅ S379 demonstrates that triple-nested presupposition extraction is grammatically clean. Each () block is an independently contestable clause. The following to-si question is unambiguous. The only constraint is practical sequencing: foundational presuppositions before derivative ones. No new mechanism needed.

  2. : scales to multi-step topic/predicate substitution. ✅ S380 shows that each retreat in a double motte-and-bailey requires a separate blocking statement. The grammar cannot be made to give the speaker the inferential benefit of A while standing on C — that benefit is not inherited transitively through : frames. Every step earns only itself.

  3. Per-link epistemic marking handles mixed-reliability cascades. ✅ S381 demonstrates that a compound slippery slope with uneven link quality is fully expressible and fully readable. The speaker commits per link: bare assertion, la-mi se (some basis), la-mi no-se (no basis), (du ...) (alleged). The argument's actual structure — one grounded link launching four unjustified ones — is immediately visible.

  4. () stacks for hearsay chains. ✅ S382 confirms that (A (B) ka) is a clean notation for reported-report. The grammar handles chains of attributed attributions without new operators. Each layer of () marks one step of provenance uncertainty. The speaker breaks the chain by asserting directly only when they have direct epistemic access.

  5. Multi-agent epistemic inflation requires full chain write-out. ✅ S383 demonstrates that inflating a hypothesis to institutional knowledge across three agents forces explicit go [stance] → elevation at every step. The chain is entirely auditable. The speaker's retrospective diagnosis completely exposes the borrowed-grounding structure.

Cross-batch observation (FAL-001 + FAL-002 + FAL-003): Depth does not break Tonesu's defenses — it makes them more expensive to write, which is the correct behavior. A speaker attempting multi-layer manipulation must pay, in explicit grammatical units, for every step. The language does not become soup: it becomes long and hard to dispute without tracking all the links.

Scaling summary:

mechanism single-layer depth-3+
() extraction ✅ S378 ✅ S379 (×3), S382 (stacked)
: predicate commit ✅ S374 ✅ S380 (double retreat)
go-chain grounding ✅ S370 ✅ S381 (mixed), S383 (multi-agent)
epistemic predicates ✅ S373 ✅ S383 (three agents, three levels)

Open grammar questions (FAL-follow): - [ ] FAL-F-005: S382 attests (A (B) re-ka) as the hearsay-chain pattern. Grammar expected to remain well-formed at depth-3+; human readability likely degrades before grammatical validity does — probably before four layers. Three-layer test programmed for FAL-005 (or standalone test). Four-layer attestation would be a deliberate stress test; speakers probably deserve to suffer at that point. - [x] FAL-F-006: ~~to-fe-su registration candidate.~~ Deferred. FAL-003 is first corpus attestation; FAL-004 (S384) will be second. Per Monday review: wait for a third independent domain use before registering. to-fe-su is a known compositional form in use but not yet at registration threshold.

FAL-004 · Political Rhetoric Corpus

Purpose: FAL-001–003 tested fallacies in isolation, in combination, and at depth. FAL-004 tests a realistic composite: a short political speech arc that simultaneously deploys all four established mechanisms — evidential laundering, topic substitution, causal cascade, and normative leap — across five connected utterances. S388 synthesizes all four moves into one text and provides the structural audit ("stack trace"). The batch asks: when four manipulations run in parallel across one short text, does Tonesu still surface each failure independently?

S384 Speech: (la-to-fe-su lo-ka-li-su de) , la-mi to lo-ka-li-su de Diagnosis: la-mi to lo-to-fe-su lo-ze re-ka , no la-mi to lo-ka-li-su de Campaign opening: evidential laundering.

Notes

  • The laundering step: () → bare la-mi to, identical content. The () marks the source as reported; the bare to asserts it as personal certainty without new evidence. The diagnosis restores the distinction: la-mi to lo-to-fe-su lo-ze re-ka = "I am certain [only] that the institution announced it."
  • This is FAL-001 S369 (appeal to authority) compressed into a political opening. The mechanism is identical; only the source type differs (standards body rather than a named scholar).
  • to-fe-su = epistemic standards framework/body (W072 ✅ — already registered; see FAL-F-006 for deferred-registration history).

S385 Challenged: la-zu lo-ka-li-su-mi to-si ne Response (topic substitution): lo-ka-li-su-mi : la-mi lo-ze se ne , na-Kur : la-ze lo-to-de re-ka Policy challenge deflected: topic substitution.

Notes

  • The challenge asks for the predicate of lo-ka-li-su-mi. The response opens with a thin affirmation (se ne = I hold it with perceptual basis) — noticeably weaker than the opening to in S384 — and then immediately substitutes : topic to na-Kur.
  • The : shift is structurally legible: lo-ka-li-su-mi : (policy topic) → na-Kur : (person topic). The original question was about the policy predicate; the response changed the subject. The listener can demand: "return to lo-ka-li-su-mi :."
  • la-ze lo-to-de re-ka is a bare accusation — not inside (). An interlocutor can require evidential wrapping: (la-ze lo-to-de re-ka) before the charge is contestable rather than merely asserted.
  • The thin affirmation (se rather than to) on the policy question is a subtle retreat from S384's opening certainty claim — a micro-instance of moving the goalposts before the topic is even abandoned.
  • ka-li-su-mi = governance-of-me = my-governance (possessive construction ka-li-su W147 + mi perspective anchor; ⚠️ proposed form).

S386 go [la-mi se lo-ra-ma-de se] , (du lo-ne-su mu-be) , (du lo-ka-li-su de-fe) , (du la-ka-no-fe lo-ne-su ka-ki) Catastrophe cascade from weak economic signal.

Notes

  • One grounded input (la-mi se = some perceptual basis, which is honest), then three consecutive (du …) outputs. The political speech does not fabricate its evidence entirely — the cascade's insidiousness is that it borrows one real link to launch three imaginary ones.
  • de-fe used here for governance collapse = "governance reaches its boundary-end." First attested S378 for wi-fe de-fe; second attestation in a different domain confirms productive use.
  • ra-ma-de = food-decay = food-supply decline (W144 ra-ma + de; new compound ⚠️ proposed).
  • Structurally identical to S381 (compound slippery slope) applied to political subject matter. The compound confirms the pattern is domain-general.

S387 Demand (bare): go [la-mi se lo-ra-ma-de se] , la-mi lo-zo-li ka-li-su wi Corrected (minimum warrant chain): go [la-mi se lo-ra-ma-de se] , go [lo-ra-ma-de lo-zo-li mu-be] , go [la-mi lo-zo-li vo wi] , la-mi lo-zo-li ka-li-su wi Normative demand: warrant chain skipped.

Notes

  • The bare demand presents la-mi se (weak perceptual basis) → bare la-mi wi [governance] (will-claim). Two chains are absent: (a) the factual harm-link connecting food-decay to injury to persons, and (b) the speaker's declared value-anchor warranting this specific response.
  • The corrected form still reaches only la-mi wi [governance] — not la-mi to [governance is required]. The normative demand remains the speaker's will-claim; that it is the correct or obligatory response for persons to accept requires further argument.
  • lo-zo-li vo wi = "toward persons, value, will" = I will in the direction of valuing persons = I hold persons' value as my action-anchor. (Compositional: W148 zo-li, vo, wi.)
  • This is S377 (appeal to emotion) with an economic rather than affective input. The warrant-gap structure is identical: the factual state (food-decay, fa-de) alone does not specify which normative response is demanded.
  • Ethical reasoning template. The corrected four-link chain — fact → harm-link → value-commitment → action — is a domain-general template for well-formed normative argument in Tonesu. It separates: (1) the empirical state (la-mi se lo-X); (2) the harm claim (go [X] lo-Y mu-be); (3) the value anchor (lo-Y vo wi); (4) the action (la-mi wi [action]). Missing any one of these produces a diagnosable gap.

S388 (la-to-fe-su lo-ka-li-su de) , la-mi to lo-ka-li-su de , go [lo-ka-li-su de] , (du lo-ne-su mu-be) , (du lo-ka-li-su de-fe) , lo-ka-li-su-mi : la-mi lo-ze se ne , na-Kur : la-ze lo-to-de re-ka , go [la-mi se lo-ra-ma-de se] , la-mi lo-zo-li ka-li-su wi The composite speech: structural audit.

Notes

  • Four of five audit items are ❌. The fifth is ⚠️ — a softening that is suspicious but not necessarily fallacious; recalibrating certainty downward when challenged is sometimes honest. The grammar records it; the listener evaluates whether it is honest recalibration or slippage.
  • All five audit items are independently diagnosable. None entangle with each other: fixing the laundering does not fix the cascade; fixing the cascade does not fix the topic substitution; fixing the topic substitution does not fix the normative leap. Each failure lives in its own grammatical frame.
  • The compact five-clause speech contains every major FAL-001/002 mechanism in 35 words of Tonesu. The structural audit produces a table with the same density as the original — which is the point. A speech designed to be slippery in natural language becomes a stack trace in Tonesu.
  • Orthogonality property. Each failure exists in a separate grammatical structure. Fixing one failure does not fix the others: removing the laundering does not repair the cascade; repairing the cascade does not repair the topic substitution; repairing the topic substitution does not repair the normative leap. This is the structural reason the stack-trace metaphor is accurate — the language decomposes the speech into independent, separately contestable commitments.

Batch Summary

Confirmed results:

  1. Composite speech is fully auditable. ✅ S388 demonstrates that a five-move political speech arc produces five independently diagnosable structural commitments. No cross-contamination between failure types; each is challengeable on its own.

  2. Laundering detectable in context. ✅ S384 confirms that () → bare la-mi to remains identifiable even when surrounded by other rhetorical moves. The bracket cannot be hidden inside a compound text.

  3. Epistemic retreat is recorded. ✅ S385 shows that weakening from to (certainty) to se (some basis) mid-speech is a structural event that : records. The grammar does not call it a failure — it just notes that the predicate changed.

  4. Topic substitution under adversarial conditions confirmed. ✅ S385 shows : working as in FAL-001/002 even when a thin affirmation precedes the shift. Cosmetic continuity does not obscure the topic change.

  5. Cascade and normative leap patterns are domain-general. ✅ S386 and S387 apply FAL-001/002 structural patterns (S381, S377) to political subject matter without structural variation. The mechanisms are not domain-specific.

Cross-batch summary (FAL-001–004 · 25 sentences):

batch scope sentences new mechanisms
FAL-001 10 logical fallacies S364–S373 (), :, ta/go, si/to
FAL-002 5 rhetorical fallacies S374–S378 wi-fe be, (du...) cascade, to-si extraction
FAL-003 5 depth / nesting S379–S383 () stacking, multi-agent chains
FAL-004 5 composite political speech S384–S388 domain-generality confirmed

No new grammatical machinery was required after FAL-001. All four mechanisms scale from single-layer through composite political speech without degradation. The language does not prevent fallacies; it requires the speaker to make every manipulative step an explicit grammatical commitment — one that a careful listener can identify and challenge in isolation.

Four orthogonal dimensions. Across all four batches, Tonesu consistently separates four types of statements that most languages conflate: (1) source (() — where does this claim come from?); (2) topic (: — what is this claim about?); (3) grounding (go / ta — what relation type holds?); (4) epistemic level (si/se/to — how certain is the claim?). Fallacies become visible because they require illegitimate movement across exactly one of these four dimensions while the other three remain fixed. The mechanisms are orthogonal: correcting a dimension-1 failure (laundering) has no effect on a dimension-2 failure (topic substitution).

Test-sequence framing: FAL-001–4 form a methodologically complete sequence: unit tests (isolated fallacies, FAL-001/002) → stress tests (depth, FAL-003) → system test (composite speech, FAL-004). The system test passing confirms the design is compositional at the level of multi-move argumentation.

Open grammar questions (FAL-follow): - [ ] FAL-F-005: Three-layer hearsay chain — deferred to FAL-005. - [x] FAL-F-006: to-fe-su — second attestation here (S384); deferred per review until third independent domain use. - [x] FAL-F-007: ~~Name compound for multi-move rhetorical manipulation?~~ Resolved: do not create. The table is the representation. Creating a single label for structured composite manipulation would collapse the orthogonal structure the language exposes. to-de-ka covers deliberate single-move falsehood; composite manipulation is already expressible as an enumeration of failure types in the audit structure. If the pattern appears outside the fallacy corpus — e.g. in a formal debate-record or institutional review — that would be the time to evaluate naming. Until then, the decomposition is the point.

FAL-005 · Propaganda Paragraph Corpus

Purpose: FAL-005 is the maximum-density test. A single paragraph-scale composite text combines all mechanisms simultaneously: three evidential layers, two topic shifts, one causal cascade, one normative leap, and one modal inflation. S389 is the propaganda text itself. S390 is the structural audit. The governing question: when seven distinct argumentative manipulations run in parallel across one block of text, does the audit remain legible and the failures remain independently addressable?

Note on structure: FAL-001–004 used one sentence per fallacy type. FAL-005 uses two entries for the full paragraph: one for the source text and one for the complete audit table. This is a format shift — the paragraph exceeds useful single-entry length, and the audit is a first-class result in its own right.

S389 (la-to-fe-su lo-ra-ma-de se) , la-mi to lo-ra-ma-de de , go [lo-ra-ma-de de] , (du lo-ne-su mu-be) , (du lo-ka-li-su de-fe) , lo-ka-li-su-mi : la-mi lo-zo-li vo wi , na-Kur : (la-na-Kur la-to-fe-su-ki (la-to-li lo-ra-ma-de de si) re-ka) , la-na-Kur lo-ra-ma-de de to , go [(la-mi se lo-ra-ma-de de)] , (du la-mi lo-ka-li-su wi) , (du la-mi lo-ka-li-su to) The propaganda paragraph.

Notes

  • The paragraph deploys seven distinct failures in nine clauses. It is written to sound like a coherent campaign statement; no individual sentence is syntactically malformed. The structural damage is entirely in the epistemic and relational commitments.
  • la-to-fe-su-ki = agent performing a standards-body publication event (W097 ✅, active — to-fe-su-ki is the inchoative: the moment a ruling enters published to-su status; la- marks Kur as agent of that event). W072 and W097 both registered before the FAL series.

S390 Structural audit of S389.

Notes

  • This is the maximum-density test. FAL-001–004 produced one to three failures per entry; S389 produces six ❌ in nine clauses. The audit table remains legible and all cells are independently actionable.
  • FAL-F-005 resolved here. Clause 5 contains a two-layer hearsay chain: (A (B) re-ka) where A is a named individual and B is an anonymous scholar attribution. The inner () marks the scholar's claim as hypothesis; the outer () marks Kur's announcement as reported. Both layers are readable and independently contestable. Grammar is well-formed at depth-2 in a composite text. Depth-3 would require one more nesting: (C (A (B) re-ka) re-ka) — grammatically valid, practically a stress test for any reader. Deferred to standalone attestation.
  • FAL-F-006 resolved here. la-to-fe-su-ki in S389 is the third independent domain attestation of to-fe-su (institutional standards body). S383 (epistemic ladder), S384 (campaign opening), S389 (propaganda paragraph — both as agent and as agent-holder). Registration threshold met; see FAL-F-006 resolution note below.
  • The ⚠️ on clause 4 is intentional: lo-zo-li vo wi is not by itself fallacious. It only becomes problematic when borrowed as warrant for the governance-will claim in clause 7 without bridging. The grammar correctly marks the direct expression; the failure lives in the inference-chain between 4 and 7.

Cross-batch observation (FAL-001–005 · 27 sentences): Maximum-density composition does not break the audit. Six failures in nine clauses remain independently addressable. The four-dimensional orthogonality (source / topic / grounding / epistemic level) holds even at propaganda-paragraph scale. The structural audit is the stable representation — it degrades in length but not in kind.

Batch Summary

Confirmed results:

  1. Seven-mechanism composite is fully auditable. ✅ S390 produces a six-row ❌ audit table, each row independently addressable. No entanglement between failure types.

  2. Laundering detectable in multi-mechanism context. ✅ S384 mechanism holds at paragraph scale: () removal at clause 2 is diagnosable even surrounded by cascade, topic-substitution, and modal inflation.

  3. Two-layer hearsay confirmed in composite context. ✅ Clause 5 of S389 embeds (A (B) re-ka) inside a larger propaganda text. Readable. This resolves FAL-F-005 for depth-2; depth-3 awaits standalone test.

  4. Opponent-epistemic-inflation as a distinct move. ✅ Clause 6 introduces a novel pattern not explicitly in FAL-001–004: asserting an opponent's epistemic certainty bare (la-na-Kur lo-ra-ma-de de to). The fix is identical to the speaker-laundering fix — wrap in (). The mechanism generalizes to third-party epistemic claims.

  5. Orthogonality confirmed at maximum density. ✅ Six repairs required, none cross-contaminating. The stack-trace property of Tonesu extends to paragraph-level composition.

FAL-Follow resolutions:

  • [x] FAL-F-005: ~~Hearsay depth limit standalone test.~~ Depth-2 confirmed in composite context (S389 clause 5); depth-3 deferred to standalone test. Not because grammar is expected to fail — but because embedding a depth-3 chain inside a propaganda paragraph would make both harder to read. Standalone is cleaner for establishing the readability threshold. Programmed as FAL-006.
  • [x] FAL-F-006: ~~to-fe-su registration.~~ Threshold met. Third independent domain attestation: la-to-fe-su-ki in S389. Registration candidate confirmed. Queued for next W-series intake: to-fe-su :: to (knowledge) + fe (boundary) + su (structure) = institutional body that governs epistemic thresholds; standards council.

Open grammar questions (FAL-follow): - [ ] FAL-F-008: Depth-3 hearsay chain standalone test — deferred to FAL-006 or standalone spot-check. - [x] FAL-F-009: ~~Third-party bare epistemic claim spec note.~~ Resolved. Corollary added to spec/grammar.md § Evidential Frame: bare epistemic predicates about third parties are speaker-certified claims and require () wrapping by the same rule as first-person laundering. The symmetry is exact: la-na-Kur lo-X to (bare) = speaker asserts Kur is certain of X; (la-na-Kur lo-X to) = speaker reports that Kur is said to be certain of X. The distinction matters identically to the first-person case.

FAL-006 · Depth-3 Hearsay Chain

Purpose: FAL-F-005 and FAL-F-008 deferred a standalone depth-3 hearsay test. FAL-006 provides it. The question is narrow and specific: does () stacking remain grammatically well-formed and humanly readable at three layers? S391 is the bare depth-3 chain. S392 tests the same chain under adversarial conditions — embedded inside a short speech where the hearsay chain is used to launder a policy conclusion. The batch also establishes the practical readability threshold for () nesting.

S391 Depth-1 (direct attribution): (la-na-Sura lo-ka-li-su de se) Depth-2 (report-of-report): (la-na-Kael (la-na-Sura lo-ka-li-su de se) re-ka) Depth-3 (report-of-report-of-report): (la-to-fe-su (la-na-Kael (la-na-Sura lo-ka-li-su de se) re-ka) re-ka) Speaker's diagnosis: la-mi no-se lo-ka-li-su de , go [la-to-fe-su (la-na-Kael (la-na-Sura lo-ka-li-su de se) re-ka) re-ka] Depth-3 hearsay chain (standalone).

Notes

  • Grammatical result: well-formed at depth-3. The innermost content is a standard epistemic clause (la-na-Sura lo-ka-li-su de se); each wrapper adds an outer agent, the inner clause as patient-in-frame, and re-ka as the action predicate. The parse is unambiguous at every depth.
  • Readability result: depth-3 is legible but cognitively expensive. The reader must hold three attribution agents simultaneously before reaching the content proposition. A reader encountering this at speed in spoken form would likely lose one layer. In written form with spacing it remains readable — but uncomfortable. This approximately marks the practical ceiling for casual use; depth-4 is formally valid but almost certainly beyond usable readability for most speakers.
  • Provenance chain structure: the speaker's diagnosis is the key expressive payoff. The go [chain] formulation linearizes the three-agent provenance into a readable sequence and makes it clear that the original claim (lo-ka-li-su de) has zero independent grounding from the speaker's perspective despite three institutional endorsements.
  • re-ka = recurring-action = formally announced / publicly stated (compositional re repetition + ka action; first attested FAL-003 S382; second attestation here confirms productive use).

S392 Hearsay chain (depth-3, identical to S391): (la-to-fe-su (la-na-Kael (la-na-Sura lo-ka-li-su de se) re-ka) re-ka) , Laundering step: la-mi to lo-ka-li-su de , Policy conclusion (normative leap): go [lo-ka-li-su de] , la-mi lo-zo-li ka-li-su wi Depth-3 hearsay laundered into policy conclusion.

Notes

  • The laundering step at depth-3 is structurally identical to the depth-1 case (FAL-S384). The number of hearsay layers does not change the fix: restore () on the conclusion, and assert only the outermost institutional announcement as personal certainty. Extra layers increase the epistemic distance between the original claim and the speaker's certification, making the laundering more egregious — not different in kind.
  • Adversarial result: depth-3 hearsay used as warrant for a policy conclusion is auditable in two independent steps. The laundering audit and the normative-leap audit are orthogonal as in all previous batches.
  • This confirms the general principle: the number of hearsay layers does not affect the audit structure; it only changes how distant the grounding is from the final claim.

Batch Summary

Confirmed results:

  1. Grammar is well-formed at depth-3. ✅ S391 is unambiguous. Each layer adds one outer agent, one inner-frame patient clause, and one re-ka predicate. The parse is clean.

  2. Readability threshold identified. ✅ Depth-3 is uncomfortable but legible in written form. Depth-4 is expected to be formally valid but practically unusable for casual speakers. The practical ceiling for hearsay chains in ordinary Tonesu use is depth-2. Depth-3 is reserved for formal institutional audits and deliberate provenance documentation.

  3. Laundering at depth-3 is structurally identical to laundering at depth-1. ✅ S392 confirms: the fix is the same regardless of chain depth. More layers increase epistemic distance and argumentative dishonesty, not mechanism novelty.

  4. Orthogonality holds in depth-3 adversarial context. ✅ S392 audit produces two independent ❌ — laundering and normative leap — with no entanglement.

FAL-F-008 resolved: ✅ Depth-3 hearsay chain standalone confirmed grammatically well-formed. Practical readability ceiling identified at depth-3 written / depth-2 spoken. Depth-4 not attested; no reason to expect grammatical failure, only human readability collapse.

Cross-batch summary (FAL-001–006 · 29 sentences):

batch scope key result
FAL-001 10 logical fallacies four mechanisms introduced
FAL-002 5 rhetorical fallacies same mechanisms apply
FAL-003 5 depth / nesting mechanisms scale
FAL-004 5 composite political speech mechanisms remain separable
FAL-005 2 propaganda paragraph density does not collapse orthogonality
FAL-006 2 depth-3 hearsay depth-3 well-formed; ceiling identified

Open grammar questions (FAL-follow): - [x] FAL-F-010: ~~Complement adversarial corpus with good-faith baseline.~~ Resolved: programmed as FAL-007. Up to FAL-006, the entire FAL corpus is adversarial — manipulative or defective reasoning. The missing proof is that (), :, go-chains, and epistemic predicates are neutral tools that add no friction to honest argument. FAL-007 provides that baseline using the same political subject-matter as FAL-004/005 but written by a speaker who is not manipulating anything.

FAL-007 · Good-Faith Political Speech

Purpose: FAL-001–006 demonstrated that Tonesu surfaces manipulation. FAL-007 demonstrates the complementary result: the same four mechanisms impose no friction on honest argument. The subject matter is identical to FAL-004/005 (food security, governance, opponent engagement, policy grounding), but each mechanism is used correctly. S393–S396 each isolate one mechanism; S397 is the composite good-faith speech and provides the "clean audit table" — the inverse of S388 and S390. The governing question: does a speaker using evidence honestly, maintaining topic consistency, grounding every causal claim, and calibrating epistemic level conservatively produce fluent, readable Tonesu?

S393 (la-to-fe-su lo-ra-ma-de de se) , la-mi to [la-to-fe-su lo-ze re-ka] , la-mi se lo-ra-ma-de de Honest evidential attribution.

Notes

  • Three distinct claims at three different epistemic levels: the council's report inside () (not endorsed, just cited); the fact of the announcement as la-mi to (certifiable); the underlying proposition as la-mi se (some basis, honestly weaker).
  • Compare S384: the dishonest version strips () and elevates directly to la-mi to lo-ra-ma-de de. S393 keeps each level where it belongs.
  • This is not verbose — it is precise. A careful speaker in formal register naturally states what they are certain of separately from what they have partial basis for. In colloquial register, (la-to-fe-su lo-ra-ma-de de se) , la-mi se lo-ze achieves the same result in two clauses.

S394 Challenged: la-zu lo-ka-li-su-mi to-si ne Response: lo-ka-li-su-mi : la-mi se lo-ra-ma-de de , go [lo-ra-ma-de de , lo-zo-li mu-be] , la-mi lo-zo-li vo wi , du la-mi lo-ra-ma ka-li-su wi Consistent topic maintenance.

Notes

  • Compare S385: the dishonest version opens with a thin affirmation then shifts : to an opponent attack. S394 opens the same lo-ka-li-su-mi : frame and stays on it through the full response.
  • The du conclusion (la-mi lo-ra-ma ka-li-su wi) correctly follows from the stated chain. The response does not assert more than the chain warrants.
  • lo-ra-ma ka-li-su = food-governance = policy-over-the-food-domain (compositional W144 ra-ma + W147 ka-li-su; ⚠️ proposed).

S395 go [la-mi se lo-ra-ma-de de] , go [lo-ra-ma-de de , lo-zo-li mu-be , go [la-na-Sura lo-ra-su ka ti-de] lo-ze se] , la-mi se lo-ne-su mu-be Grounded causal chain.

Notes

  • Compare S386: the dishonest version uses one bare go [premise] to launch three (du...) consequences. S395 provides independent grounding (la-na-Sura lo-ra-su ka ti-de) for the harm-link, and the conclusion stays at la-mi se rather than inflating to certainty.
  • The conclusion is exactly one step from the warranted evidence. The speaker does not assert social-fabric collapse or governance failure — only social-fabric harm, and only with se.
  • ra-su = signal-structure = food-system / supply infrastructure (compositional ra signal/force + su structure; ⚠️ proposed).

S396 go [la-mi se lo-ra-ma-de de] , go [lo-ra-ma-de de , lo-zo-li mu-be] , go [la-mi lo-zo-li vo wi] , la-mi lo-ra-ma ka-li-su wi , la-mi se lo-ze be-vo Conservative epistemic calibration.

Notes

  • The full S387 ethical-reasoning template: fact → harm-link → value-anchor → action. Each link is warranted; no link is skipped.
  • The speaker does not claim la-mi to lo-ka-li-su wi (certain-I-should-govern) or la-mi to lo-ze be-vo (certain-I-am-capable). Both conclusions remain at la-mi wi (will) and la-mi se lo-ze be-vo (some capability basis) — the honest epistemic levels.
  • lo-ze be-vo = this-entity has productive capacity = I am capable. (The speaker uses lo-ze rather than la-mi here because capability is a property being attributed, not an intention being expressed.)

S397 (la-to-fe-su lo-ra-ma-de de se) , la-mi to [la-to-fe-su lo-ze re-ka] , la-mi se lo-ra-ma-de de , go [lo-ra-ma-de de , lo-zo-li mu-be , go [la-na-Sura lo-ra-su ka ti-de] lo-ze se] , la-mi se lo-ne-su mu-be , lo-ka-li-su-mi : go [la-mi se lo-ra-ma-de de] , go [lo-ra-ma-de de , lo-zo-li mu-be] , go [la-mi lo-zo-li vo wi] , la-mi lo-ra-ma ka-li-su wi , la-mi se lo-ze be-vo , la-mi to-si [la-zu lo-ze se-to to-si] The composite good-faith speech.

Notes

  • The composite speech is the same length as S388 (propaganda) and covers the same subject matter. The honest version is not longer and not more complex. The mechanisms add no overhead when used correctly — they are invisible infrastructure that only becomes visible when someone misuses them.
  • The final question la-mi to-si [la-zu lo-ze se-to to-si] = "I ask: what is your certain-perception basis?" This is the honest inversion of S378's loaded question. Instead of smuggling accusations inside the interrogative, the speaker declares their own epistemic level (se) and genuinely invites the interlocutor's grounding.
  • The good-faith speaker sounds like a functioning adult. The propaganda speaker (S388/S389) required a six-row audit table. The honest speaker (S397) requires a five-row ✅ table. The grammar itself makes no judgment — it just keeps the books.

Batch Summary

Confirmed results:

  1. Four mechanisms impose no friction on honest argument. ✅ S393–S397 use (), :, go-chain, and si/se/to exactly as designed. None creates overhead; all produce clean, readable, convincing argument.

  2. Same subject matter, same mechanisms, opposite audit. ✅ S397 covers identical ground to S388 (composite propaganda). The good-faith version produces five ✅ instead of six ❌. The grammar makes no judgment; the difference is entirely in the speaker's epistemic commitments.

  3. Ethical reasoning template is fluent. ✅ S396 demonstrates the fact → harm → value → action chain in natural speech register. It is not bureaucratic or labored — it is simply explicit about what it warrants.

  4. Honest questions are the inversion of loaded questions. ✅ S397's closing question exposes its own epistemic level before asking; S378's loaded question buried hidden accusations inside the interrogative. The grammar handles both cleanly; they are structurally distinct.

  5. Conservative epistemic calibration is productive, not weak. ✅ A speaker who says la-mi se where they have partial basis, and la-mi to only where they have strong grounds, is more credible — not less — because their certainty claims are auditable and consistent throughout.

Cross-batch summary (FAL-001–007 · 34 sentences):

batch scope key result
FAL-001 10 logical fallacies four mechanisms introduced
FAL-002 5 rhetorical fallacies same mechanisms apply
FAL-003 5 depth / nesting mechanisms scale
FAL-004 5 composite political speech mechanisms remain separable
FAL-005 2 propaganda paragraph density does not collapse orthogonality
FAL-006 2 depth-3 hearsay depth-3 well-formed; ceiling identified
FAL-007 5 good-faith speech mechanisms are neutral tools, not anti-manipulation shackles

FAL-F-010 resolved. The FAL corpus is now complete as an empirical demonstration: adversarial test (FAL-001–006) plus honest baseline (FAL-007). Together they prove that Tonesu's four discourse mechanisms are neutral infrastructure that make argumentative structure visible — to the benefit of honest speakers and the detriment of dishonest ones.

Open grammar questions (FAL-follow): - [x] FAL-F-011: ~~to-fe-su registration queued.~~ Resolved: both already registered. to-fe-su = W072 ✅ (epistemic standards framework/body; first use S128; registered before the FAL series). to-fe-su-ki = W097 ✅ (inchoative: the moment a ruling enters published to-su status; la- marks the agent of that event). The FAL corpus footnotes were tagging them as unregistered, but both predate the FAL series by a wide margin. No intake needed; footnotes corrected in place.

Generated from registry/entries.yaml.