A comment verification platform
말조심 — watch your words. Words are weighed before they are spoken. Each comment is read, measured, and returned with the counsel of careful language. This is a quiet instrument for people who believe that what we say in public ought to carry its proper weight.
§ I. The Reading Room
maljosim accepts a comment, an email, a public reply, a draft tweet — any short passage of language intended for an audience — and returns a structured reading of it. Tone is identified. Ambiguity is highlighted. Statements that may be received as accusations, as condescension, or as careless cruelty are named and explained.
The platform does not censor. It does not rewrite without asking. It returns the comment to the author with its weight made visible, and trusts the author to decide.
We measure four registers: tone, clarity, impact, and civility. Each is reported on a five-point scale, with annotated reasoning. No score is the final word; every score is an argument.
§ II. Specimens
"Honestly, if you'd actually read the thread before replying, you'd know this was already addressed. Maybe slow down next time."
tone condescending
clarity adequate
impact likely to wound
civility low
¶ The phrase "honestly, if you'd actually read" carries a presumption of the reader's failure. Consider whether the point can be made without the assumption.
"그건 좀 아닌 것 같은데요. 제가 보기엔 처음부터 잘못된 접근입니다."
tone measured
clarity clear
impact moderate
civility acceptable
¶ 직접적인 반대 의사를 정중하게 표현하고 있습니다. "처음부터 잘못된" 표현은 상대의 노력을 부정하는 인상을 줄 수 있으니, 구체적 근거를 덧붙이시길 권합니다.
"Thanks for raising this — I hadn't considered the point about timezones, and I think you're right that the schedule needs revisiting. Let me draft an alternative this week and circulate it."
tone warm
clarity precise
impact unlikely to wound
civility high
¶ Acknowledges the other party, names what was learned, and commits to a concrete next step. This passage may be sent without revision.
§ III. Methodology
Each submission moves through four readings. The first is lexical: a careful inventory of the words actually present, with attention to terms that historically carry presumption or contempt. The second is structural: how the sentence arranges blame, agency, and concession. The third is contextual: who the reader is likely to be and what they are likely to bring to the page. The fourth is summary: a single paragraph of counsel, written in the voice of an editor.
Lexical reading
The vocabulary is examined word by word. Loaded terms are flagged with their historical weight noted.
Structural reading
Subject, object, and modal verbs are mapped. Sentences that assign blame asymmetrically are identified.
Contextual reading
Audience and channel are taken into account. A draft tweet and a draft letter are not weighed by the same balance.
Editor's counsel
A single short paragraph, plainly written, offered as advice rather than judgment.
§ IV. Index