Reading edition

Size

Collections, works and characters. Characters are counted separately for the main text and for notes (original notes, collation and marginal notes).

Works per collection

637,203 works in 1,249 collections; the median collection has 339. The largest are the Songja daejeon (9,434), the Yeoyudang jeonseo (8,100) and the Myeonu jip (5,688). 95.7% of all works are by the collection's own author.

Characters per collection

177.4 million characters of text and 5.5 million in notes (3.0%); the median collection has about 80,000. The Yeoyudang jeonseo is the longest at 4.1 million, followed by the Songja daejeon (3.7 million) and the Myeonu jip (2.9 million).

Collections by century

By the author's birth century the collections cluster in the 16th (382), 17th (353) and 18th (272); authors born up to the 14th century account for 48 collections and the 19th century for 82. Of the 586 supplement collections, 48 are by authors born in the 19th century. The 8 authors without a birth year sit in a separate row of the century tables.

Characters per work

The median work has 56 characters (one seven-syllable regulated verse) in the 13th–16th centuries, then 75 in the 17th, 118 in the 18th and 157 in the 19th; the 9th century (Choe Chi-won, 168) and the 12th (80) are exceptions. The direction is the same as the falling share of poetry and the rising share of prose.

Share of characters in notes

Notes make up 3.0% of all characters, between 1.9% (19th century) and 4.1% (16th) in every century except the 9th (Choe Chi-won, 11.6%).

Share of own works

95.7% of works are by the author; the rest are prefaces, postscripts, biographies, elegies and appendices by others. In eight collections (the Manghyeon yugo 忘軒遺稿, Chupa jip 秋坡集, Changju yugo 滄洲遺稿, Neungheogwan mango 凌虛關漫稿, Yongiwa jip 龍耳窩集, Naeheon jip 耐軒集, Eodang jip 峿堂集 and Sosong yugo 小松遺稿) no writer tag matches the author at all — possibly a tagging convention (a hypothesis, flagged for sample checking).

Fascicles and works per fascicle

Of the 851 collections whose introduction states 'n fascicles, n volumes', 691 (81%) have exactly n '卷n' headings in the table of contents. The causes of the 160 mismatches were not coded.

Genre composition

Counts the genre classification the ITKC attached to every work (7 major, 28 minor classes). Headline shares use the author's own works only — works whose writer tag is the collection's author, as opposed to prefaces, postscripts, biographies and elegies by others that are printed in the same collection.

Share of poetry

62.0% of the authors' own 609,494 works are poems. By century the share falls from 88% (14th–15th) to 75% (16th), 62% (17th), 51% (18th) and 42% (19th). By characters poetry is 18.8%: many works, few characters. The direction matches Kwak (2026): poetry largest, declining.

Seven-class genre mix

Of own works: poetry 62.0%, letters 13.4%, miscellany 9.1%, court documents 5.6%, biographies 3.7%, elegies 3.2%, prefaces 3.0%. By characters the order changes: miscellany 25.2%, poetry 18.8%, letters 17.8%, biographies 16.9%. Among minor classes poems (374,153), letters (81,442), 'other' within miscellany (26,505) and memorials (26,386) lead.

Share of letters

Letters are 13.4% of own works: 0.3–2.5% in the 12th–15th centuries, then 5.7% (16th), 15.0% (17th), 15.1% (18th) and 30.2% (19th); the 9th century (Choe Chi-won) stands at 32.9%. The direction matches the rising trend Kwak (2026) reports.

Share of court documents

Court documents (memorials, examination answers, edicts) are 5.6% of own works: 42% in the 9th century (Choe Chi-won), 9.1% and 8.7% in the 12th and 13th, 5.8% and 7.7% in the 16th and 17th, 2.4% in the 19th.

Share of elegies, epitaphs and biographies

Works commemorating others (elegies, epitaphs, biographies) are 6.4% of own works, rising from about 2% in the 14th–15th centuries to 6.3% (17th), 8.3% (18th) and 9.8% (19th).

Share of prefaces and postscripts

Prefaces and postscripts are 3.0% of own works: under 2% through the 16th century except the 13th (3.3%), then 4.6% in the 18th and 5.6% in the 19th.

Residual 'other' share

6.4% of own works sit in an 'other' minor class, between 2.8% (15th century) and 12.1% (12th), with the 9th century at 30%. The largest block is 'other' within miscellany (26,505 works).

Characters per poem

The median poem has 56 characters in every century but the 9th (40) — the length of one seven-syllable regulated verse. Quartiles lie at 28 or 40 (quatrains have 20 or 28 characters, a five-syllable regulated verse 40) and 56–109; the mean (49–93; 48.7 in the 9th century, 93.0 in the 12th) exceeds the median — the distribution has a long right tail.

Relationship genres

Addressees of letters and subjects of epitaphs and biographies, parsed from titles as strings. Version 1.0 does not identify persons: the same person called by courtesy name, pen name or office counts as different strings.

Letter addressees and concentration

In the 895 collections with letters, 93.4% of the 81,842 letters carry an addressee string in the title. Summed over collections there are 34,890 distinct addressees (25,303 distinct strings corpus-wide); a collection's median is 14, and among letters with an addressee the top three take a median of 41%. The Ganjae jip (1,287), Myeonu jip (1,033) and Songja daejeon (798) have the most.

Subjects of epitaphs and biographies

880 collections contain 19,789 epitaphs and biographies by the author; a subject string was parsed from 93.6% (17,972 distinct summed over collections, 15,440 corpus-wide). The Songsa jip (968), Songja daejeon (576) and Myeonam jip (428) have the most.

Other writers

1,120 collections carry 27,709 works by writers other than the author; summed over collections that is 17,675 writers, a median of 8 per collection. The Seongjae jip (173), Hanju jip (143) and Cheogam jip (138) have the most.

Authors and publication

Author dates, linkage to the civil-examination roster, the lag between death and publication, and translation and punctuation progress.

Date completeness and lifespan

Of 1,248 authors, 1,240 have a birth year, 1,231 a death year and 1,228 (98.4%) both. Median lifespan is 65.5 years, from 57.5 (15th century) to 70 (18th). Of the 1,077 authors matched to the AKS person database (by name and birth year, or name and pen name) with a comparable death year, 1,035 (96.1%) agree.

Examination linkage and age at passing

49.1% of the 1,248 authors (613) link to a passer in the civil-examination roster (name and birth year 546, first pass of a re-examined person 65, courtesy or pen name 2): 83% of those born in the 15th century, 62% (16th), 52% (17th), 30% (18th), 22% (19th). An unlinked author is not necessarily a non-passer; variant-character and namesake candidates are reviewed separately. Median age at passing is 30 (27 in the 15th century, 35 in the 18th).

Publication lag

1,038 collections have a printing year and 1,026 also a death year to measure the lag. The median lag from death to printing is 66 years (quartiles 24–177): 175 years for authors born in the 15th century, 128 (16th), 59 (17th), 39 (18th), 15 (19th).

Linked-translation progress

16.6% of works have a linked translation; 1,105 collections have none (585 of the 586 supplement collections). Only 143 first-series collections carry a link, fewer than the 372 Choi (2020) counted as translated — linkage measures database work, not translation.

Collation and punctuation progress

6.5% of works have a linked collated and punctuated edition, concentrated in 58 collections. The share by century ranges from 0% (12th) to 26% (13th).