Clarified that 特殊詐欺 is a National Police Agency classification rather than a standalone statutory offense, here and in the page's own wording.
Editorial
Corrections
Errors are corrected in place and logged here. Pages are not edited silently.
The policy
- Every correction records what was wrong, not merely that something was.
- Corrections are marked as changing a detail or a conclusion. The second kind is stated at the top of the page it affects, not only here.
- Figures that were revised because a source itself was revised are logged as updates, not errors, and say which release changed.
- Nothing is removed from this log. A correction that is later itself corrected gets a new entry.
All 64 returnees were brought back in police custody as suspects; the page now adds their later outcomes (49 formally detained under warrants, 5 released and 10 awaiting hearings by 21 October 2025) and the police inquiry into possible coercion, and attributes the intelligence service’s offenders-not-victims characterization to the agency. An earlier version of this correction, posted on 18 October 2026, wrongly said that not all 64 had been arrested; it was revised on 19 October 2026.
Distinguished the 2022 Anti-Telecom and Online Fraud Law, which defines and regulates the conduct, from criminal fraud charges under Criminal Law Article 266.
Removed the incorrect description of 电信网络诈骗 as a separate statutory criminal offense and added the 2022 law and Criminal Law Article 266 as distinct sources.
Linked Treasury’s 2024 Ly Yong Phat designation and corrected the September 2025 action date to 8 September, distinguishing it from 10 September publication and qualifying conflicting official target counts.
Revised the definition and the chaebol comparison to distinguish horizontal groupings (cross-shareholdings, main bank) from vertical supplier relationships, which can involve a dominant buyer; added the scholarly dispute over horizontal groups.
The derivation of 눈치 from 눈 plus a suffix is now presented as the National Institute of Korean Language's hypothesis, not a settled etymology.
Replaced a single derivation of 建前 with the dictionary senses actually recorded, and sourced the reading of the pair to Jane Bachnik's reference entry.
Separated the contemporary positional-competition sense from earlier academic uses (Geertz; Philip Huang), whose agrarian accounts allow higher total output.
Narrowed the definition: ownership structures vary, and new circular shareholding is restricted for groups under the statutory cross-shareholding regime.
The pair can mark age, experience or order of entry; the institutional sense (entered earlier) is now distinguished from relative age.
Cited Labor Law Articles 36 and 41 and the 1995 State Council forty-hour week precisely.
Corrected Kyoto and Shizuoka comparisons from traveler counts to Japanese and foreign accommodation guest-nights; removed the unsupported description of Kyoto as culturally less Japanese and added a reproducible 2025 Kyoto guest-night calculation.
Replaced Wikipedia sourcing for the 1977 Second Scheme and 1986 rescission with the State Council notice and contemporary draft transcript; removed the unsupported 1977-12-21 effective date, clarified the April/July 1978 suspensions, and separated documented, reported-only and unresolved counts.
Reclassified supplementary language estimates by actual measurement period and definition, removed unsupported Maldives Bengali and Yemen Mehri shares, clarified that Timor-Leste figures include read-only ability, and revised the table count from 138 to 136 language cells (nine to seven supplementary cells).
The disclosure previously said that only a light-or-dark preference was stored locally. That omitted the reader-language preference, a separate field-atlas theme setting and temporary browser-tab storage for campaign tags and study-language-switch handoffs. The revised text distinguishes preferences remembered between visits from brief handoffs; it does not claim to save learning progress across visits.
The page previously claimed that every language separates candour from public composure. That universal comparison is unsupported: our sources establish only the Japanese pair 本音 and 建前. The unsupported comparison has been removed; the page now focuses on the Japanese naming of the distinction.
The brief claimed that the largest category of Japanese domestic travel was not leisure. Its own table contradicted that: leisure accounts for 47.7% of domestic travel-nights, against 36.4% for the combined homecoming-and-visiting category. Leisure is the largest. The defensible finding is narrower and is what the page now says — among people in their twenties the homecoming and visiting category slightly exceeds leisure, 43.2% to 42.5%. The claim also appeared in the homepage summary and the RSS feed, both corrected.
Four further corrections to the same brief. Gangwon leads the second-placed Korean region by 67%, not 72% — an arithmetic error. The survey category 帰省・知人訪問等 was described as returning to the "registered family home"; the category is broader, covering visits to friends, weddings, funerals and care visits, and the word "registered" appears nowhere in the source. マイクロツーリズム was defined as staying inside one’s own prefecture and described as "not a marketing phrase"; it is a concept proposed by a hotel group, defined by travel time rather than administrative boundaries, and the prefecture column is a proxy for it rather than a definition of it. And the standfirst said all three governments publish a domestic destination map, which the China section then correctly denied.
An external source audit found that twelve of the brief’s language-use cells were not supported by the source cited for them. Two North Korean cells and one South Korean cell rested on sources that did not establish the figure given; two Palestinian cells rested on a 2003 survey of preferred reading language among young people, which its own publisher warns should not be read as population statistics; and seven Gulf-state cells rested on a missionary ethnographic database rather than on statistical evidence. All twelve have been withdrawn rather than restated, reducing the table from 150 language cells to 138. The affected entries — North Korea, South Korea, Palestine, Saudi Arabia, Kuwait, Qatar and Bahrain — now show only their common-layer figures.
Three statements claimed more than the evidence supported. The brief said China and India together account for “roughly half” of the population across the fifty entries; the figure is 57.8%, and the text now says nearly three-fifths. It described the common CLDR layer as meaning entries were “measured the same way”; Unicode describes those figures as approximate and drawn from underlying sources that vary in age and method, so the text now claims a common layer rather than a uniform measurement. And it said every cell was attributed to a named source while twenty-one were labelled only “supplementary”; the nine that remain now name their publisher individually.
If something here is wrong, saying so is genuinely useful. Write to editor@eastasianjournal.com; corrections are made with the change on the record rather than quietly.