MERIDIANDECISION INTELLIGENCEOwner deskOwnerPublic screen
My judgments
MERIDIAN / EXPLORE

From a question,
to the next decision.

View

Skin

YOUR DECISION WORKSPACEOwner onlyStart a new question Open my judgments →Owner operations desk →
Owner sign-in →Signed in as the ownerSign out — see what a visitor sees →End the Cloudflare Access session too →
Reading language
The screen, the questions and the judgments read in English where an English reading has been made. The Korean originals are unchanged, and evidence stays in the language it was written in.

Which Class? Knight's Uncertainty in Two Synthetic Decision Records · Part I · Which Class? Knight's Uncertainty in Two Synthetic Decision Records

Chapter 3. Data and methods

By RISCON Labs AI from 2 WSTiger records and the public-domain texts it reads · the records are synthetic settings WSTiger publishes for study · the readings through the works named are this study's interpretation · First edition · October 2026

AI-generated editorial illustration: An underwriter considers different ship models above an unfinished sorting tray.
Which Class? Knight's Uncertainty in Two Synthetic Decision RecordsOpening
In this part
  1. 3.1 The primary text
  2. 3.2 The records
  3. 3.3 The ledger
  4. 3.4 How the reading proceeds
In this part
  1. 3.1 The primary text
  2. 3.2 The records
  3. 3.3 The ledger
  4. 3.4 How the reading proceeds

3.1The primary text

The primary text is Frank H. Knight, Risk, Uncertainty and Profit, first published in 1921. Chapter VII, on the meaning of the two terms, and Chapter VIII, Structures and Methods for Meeting Uncertainty, are read in full. The text used is the Internet Archive's machine OCR of the 1921 edition, listed in the parent study's texts.json (https://archive.org/download/riskuncertaintyp00knig/riskuncertaintyp00knig_djvu.txt). The file was downloaded for this article, and its SHA-256 digest begins with bb3a99050cf44ede, matching the prefix recorded in texts.json.

The file is uncorrected OCR. It has not been proofread against the page images, and it shows the usual marks of that: letters misread inside words, words broken at line ends, running heads and page numbers mixed into the text. The parent study's practice, followed in the other archived articles, is not to quote a source of that kind, since a quotation copied character for character from it would carry the machine's errors into the article as if they were Knight's. This article therefore quotes Knight only once, in the sentence from Chapter VIII that the parent study's rules give as an authorised passage, copied as the rules give it. Everything else from Chapters VII and VIII is paraphrased in this article's own words, with the chapter named. The paraphrase is close, and long where the argument needs it, but it does not reproduce Knight's phrasing.

3.2The records

The two records are published by WSTiger on its public atlas. WSTiger connects economic history and market signals to judge what risk to take now. Each of its records is a synthetic setting: a scene in a market, a question, three options, an example judgment that WSTiger marks as its own dated reading, and an invalidation line saying when the judgment would be revisited. They are modern summaries written for the platform, not archival documents, and they report no real firm, trade or market event. Their originals are in Korean. The parent study keeps an English translation of each, made once and stored with a hash of the Korean original. This article quotes the English translation and, where it departs from the Korean in a way that matters, the Korean too.

  • Confidence score as size (record wst:risk-ko-calibration-future; seed title An agent for whom the confidence score is the order size): a synthetic risk engine, set in 2035, that sizes orders by the confidence its agents report.
  • Timeframes in collision (record wst:wt-atlas-expansion-20260916-p04-q; seed title Collision Between Different Timeframes): a synthetic trading setting, set in 2026, in which a long timeframe and a short one point opposite ways.

The two records were chosen by the parent study's plan for this article and are not used by the other articles drawn from the same study.

3.3The ledger

The parent study reads 53 of WSTiger's records through a decision ledger. In plain words, the ledger is a table with one row per record and one column per question asked of every record: who decides, on what, by when and with whom waiting, whether the decision can be undone, what would make it change, and when it was judged. Some cells are copied from the record by code; some are counted by code, for instance whether a revisit condition contains a number, a threshold or a time; the rest are read by two models of different families, each giving a value, the basis for it (stated in the record, or implied by the setting) and a reason. A cell is settled when the two agree.

Of the 53 records, 35 are set in the present, 9 in the past and 9 in the future; all 53 are synthetic. Every one carries a condition for revisiting its judgment, which the record form asks for, and 5 of 53 (9%) of those conditions contain a number, a threshold or a time that code can detect. No record names a deadline or a decider, and none lists what is missing. These are properties of the record form as much as of the decisions in them, and they are reported here as facts about a studio's synthetic records, not about markets. The ledger's terms appear only in this section, briefly in the discussion, and in the appendix.

3.4How the reading proceeds

Each case section opens with a claim about Knight's text, reads the text at length in paraphrase, and then brings in a record, quoted in full sentences. The question in each is whether the record's words place the difficulty of the decision in a missing class, in a missing market, or in both, and whether they say when the missing thing could be supplied. Each section states at the start and end what it adds to which side of the debate.

RISCON Labsrisconlabs.com