Drosa — “how you know is part of what you say”
498 words. Part of the Conlangs as Tools project, edition v1.2 (2026-09-27).
Brief drafted by Doruga (Pete’s mobile-companion agent), 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)
Drosa brief · Drosa deep dive → · project home · all files (.zip)
Languages: Toki Pona · Taso · Kipona · Ren · Maku · Sito · Olu · Vesi · Korē · Drosa · Hesi
What it explores
Toki Pona does not grammatically mark the source of a claim. The speaker says “rain fell” and the listener trusts or doesn’t. Drosa makes evidentiality mandatory. Every assertive sentence must carry one of five particles that names how the speaker came by the information. An unmarked claim is ungrammatical.
The lesson Drosa tests: natural languages like Tuyuca (NW Amazon) and several Andean and Turkic languages already do this — grammatical evidentiality is not exotic. The Drosa experiment is to engineer one for a misinformation-saturated era and observe what changes.
Profile
Five evidential particles, suffixed to the predicate:
- -vi — direct sensory witness (I saw / heard / felt / tasted)
- -na — secondhand report (I was told)
- -su — inferential (the evidence implies it)
- -ka — common knowledge (it is shared assumption)
- -zo — speculative (I am guessing / supposing)
Built-in interrogatives chase the source: a listener hearing -na can ask “from whom?” The grammar invites the question.
Drosa is not Kipona. Kipona anchors when something happened and when you learned of it. Drosa anchors how you know. The two are complementary; a Drosa-Kipona hybrid is conceivable.
The discipline Drosa imposes is small per sentence — one syllable — but its cumulative effect is large: speakers can’t accidentally state hearsay as fact, can’t accidentally pass off speculation as inference, and can’t absorb common-knowledge claims without surfacing them.
Examples
- kasi pona tan ma -vi — “the plant grows well from the soil [direct witness]” → I see it growing.
- kasi pona tan ma -na — “[hearsay]” → someone told me.
- kasi pona tan ma -su — “[inference]” → the leaves are green, so I infer.
- kasi pona tan ma -ka — “[common knowledge]” → this is what everyone says.
- kasi pona tan ma -zo — “[speculation]” → I’m guessing.
A news report in Drosa is a forest of -vi, -na, and -su. A rumor is -na-heavy. A scientific claim with proper citation is -vi and -su. Conspiracy speech tends to -ka and -zo with the markers omitted — but in Drosa, omission isn’t possible.
Future action research
- Misinformation experiments. A small community uses Drosa for some weeks. Does the -na-chain shorten? Does rumor decay faster? Compare with a matched control community.
- Science communication. Rewrite a science article in Drosa. Run readability and credibility studies. Hypothesis: it reads more cautious and is trusted more.
- LLM hallucination. Train an LLM on Drosa-shaped data and instruct it to mark its outputs. Does it learn to use -zo (speculation) when it ought to? Does this improve calibration?
- Comparison with natural systems. Tuyuca, Quechua, Turkish, Bulgarian each grammaticalize evidentiality differently. Map Drosa onto them. Which natural categories does Drosa borrow well? Which does it miss?
- Cognitive therapy. Cognitive distortions often hide as -ka-claims (“everyone knows X about me”) or unmarked speculations. Could Drosa-style speech, used reflectively, surface them? Pilot with CBT practitioners.