Classical Chinese
Enable an AI agent to recognise Classical Chinese usage, assess the evidence and uncertainty behind a reading, and decide whether to transcribe, annotate, translate, compare or request specialist review.
Research draft, second pass
A second pass drafted this model: the structure a model of this thing needs, and what is known about it in the world. The line under this one says how the second half was obtained - researched against sources, or recalled without web access, in which case nothing here was read anywhere and every claim is a lead to verify. Unreviewed either way.
Researched by: Codex + Grok
Purpose and description
Enable an AI agent to recognise Classical Chinese usage, assess the evidence and uncertainty behind a reading, and decide whether to transcribe, annotate, translate, compare or request specialist review.
Classical Chinese is the prestige written register of Sinitic, grammatically and lexically modelled on late Zhou and Western Han prose, used for administration, scholarship, and belles-lettres in China and the wider Sinosphere until vernacular writing displaced it as the default medium in the early twentieth century.
It can be Classify a passage's linguistic register with explicit evidence and boundaries.; Propose punctuation, lexical segmentation and grammatical analyses while preserving alternatives.; Compare witnesses and parallel passages to investigate variant characters or disputed readings.; Produce aligned glosses and translations that identify supplied content and interpretive choices.; Generate explicitly labelled Classical Chinese imitations against a declared period and genre profile.; Route unresolved graph readings, constructions or allusions to targeted research or specialist review..
Distinguishing features
Test connected clauses for combinations of particle functions, word order and contextual lexical usage; isolated characters such as 之 or 也 do not establish Classical Chinese identity. Grammatical orientation: [Stanford Chinese Texts](https://chinesetexts.stanford.edu/1-classical-chinese/).
Distinguish language from script: traditional characters, simplified characters or an unpunctuated presentation cannot independently distinguish Classical Chinese from modern written Chinese.
Distinguish Classical Chinese from reconstructed Old Chinese speech: a written passage does not by itself establish its historical pronunciation or its correspondence to everyday speech.
Distinguish early Classical Chinese from later Literary Chinese by recording composition context and the attested linguistic profile separately; an archaizing construction alone does not date a text.
Distinguish sustained Classical Chinese usage from quotations and idioms embedded in modern prose by testing the surrounding clauses and recording the exact extent of each register.
Scope
+ Evidence for identifying Classical Chinese at passage level, including mixed or imitative usage.
+ Period-sensitive vocabulary, grammatical constructions and discourse conventions.
+ Character readings, textual variants and editorial segmentation where they affect interpretation.
+ Competing interpretations, their supporting attestations and unresolved ambiguities.
+ Conditions for producing traceable annotations, translations and Classical Chinese compositions.
- Physical manuscripts, conservation condition and ownership, which belong to document and collection models.
- The Chinese writing system as a whole, including calligraphic styles and character encoding standards.
- Modern spoken Chinese varieties and general modern Chinese proficiency.
- Reconstructed historical pronunciation as an independent phonological system.
- The truth of philosophical, historical or scientific claims expressed in a text.
- Complete bibliographies, author biographies and literary-work lifecycle records.
Characteristics
- Classical Chinese identification
- unassessed | supported | disputed | unsupported, with passage-level evidence Prevents script appearance or a familiar quotation from determining the classification of an entire text.
- Working chronological scope
- early classical core | later literary continuation | modern imitation | mixed | unresolved; record the adopted boundary Makes the operational meaning of Classical Chinese explicit without silently deciding the registry's unresolved scope.
- Attestation context
- Links to dated or approximately dated passages, genres, witnesses and editions; composition and witness dates distinguished A construction or sense needs contextual evidence before it can support a reading.
- Graph interpretation
- direct reading | variant graph | proposed phonetic loan | suspected corruption | unresolved, with evidence Different explanations for a written form justify different editorial and interpretive actions.
- Segmentation status
- inherited from edition | independently proposed | competing segmentations | unresolved Word and clause boundaries can determine the resulting grammatical analysis.
- Construction analysis
- Passage spans linked to proposed particle functions, constituent roles and comparable attestations Exposes the basis for a parse rather than treating a fluent translation as grammatical evidence.
- Interpretive intervention
- transcription | punctuation | normalization | supplied referent | emendation | translation | imitation; multiple values allowed Keeps observed wording distinguishable from material introduced by an editor or agent.
- Unresolved consequential readings
- Count of unresolved alternatives that change a named passage's proposition, participant roles or requested output Helps decide whether a translation can proceed with alternatives or needs further research.
Also called
Where this came from
wikidata · CC0 1.0
Drafted structure
Bundle to layer to finding to question, as the second pass will find it: 6 bundles · 11 layers · 18 findings · 28 questions.
Classical language boundary Establish what Classical Chinese means for this entry and which stretches of language meet that working definition.
The registry supplies no definition, so neither chronology nor inclusion of later Literary Chinese should be silently assumed.
Classical scope convention
Record the adopted relationship between the early classical core and later literary practice.
Declared classical boundary
Require an explicit operational boundary supported by named references and corpus examples, with disputed extensions retained as unresolved.
- Does Classical Chinese here designate an early written-language core, or also specified later Literary Chinese practices? definition
- Which periods, corpora or usages remain undecided under that definition? boundary
Passage register evidence
Locate Classical Chinese usage within texts that may contain quotations, commentary or other registers.
Bounded register attribution
Record exact spans and linguistic evidence for classification, separating a quoted classical passage from its surrounding prose.
- Which spans show sustained Classical Chinese grammar and vocabulary rather than isolated inherited expressions? boundary
- Which observed constructions support the classification, and what evidence would weaken it? definition
Written evidence and readings Connect interpretive work to the actual characters and textual witnesses used.
A reading may depend on a variant graph, an edition's intervention or a transcription error; the agent must identify which evidence it is interpreting.
Witness and edition
Identify the source of the wording and distinguish transmitted text from editorial presentation.
Traceable passage wording
Retain a locatable source for the passage and identify transcription, punctuation and normalization interventions. Digital comparison capabilities are illustrated by the [Chinese Text Project introduction](https://ctext.org/introduction).
- Which witness, edition, page or stable passage locator supplies the characters being analysed? provenance
- Which characters, punctuation marks or normalized forms were introduced by a transcriber or editor? provenance
Graph and variant resolution
Evaluate written-form alternatives without silently replacing the evidence.
Justified graph reading
Distinguish a documented variant from a proposed loan reading, damaged text, OCR error or conjectural correction, and preserve the original form.
- What supports treating this graph as a variant or loan for another word rather than reading it directly? provenance
- Can the proposed interpretation retain the transmitted graph, or must an explicit emendation be presented for review? action
Lexicon and clause analysis Represent the lexical and grammatical decisions required to read a passage.
An agent needs passage-specific analyses of words, particles and constructions rather than automatic substitutions from modern Chinese.
Contextual lexical sense
Determine candidate word boundaries and senses using relevant attestations.
Attested word and sense
Record candidate lexical units, contextual senses and supporting examples without assuming each graph is a separate word or importing a modern compound reading.
- Which adjacent graphs should be read together here, and what supports that segmentation? definition
- Which period- and genre-relevant attestations support the proposed sense over competing senses? provenance
Particles and constructions
Resolve grammatical functions from the clause and its context.
Supported clause parse
Record how word order and particles support a parse, including alternatives for forms such as 之, 者, 所 and 也. Reference orientation: [Stanford Chinese Texts](https://chinesetexts.stanford.edu/1-classical-chinese/).
- What function does each relevant particle perform in this construction, and which constituent does it govern or relate to? definition
- Which alternative punctuation or constituent analysis changes who acts, what is predicated or what is negated? boundary
Discourse and inherited context Track information recovered from surrounding discourse, genre conventions and earlier texts.
Grammatical plausibility alone may leave participants, speech force or an allusive expression unresolved.
Implicit participants and force
Expose contextual inferences that supply missing participants or determine the utterance's function.
Recoverable discourse inference
Record candidate referents and utterance readings separately from explicit wording; omitted subjects and indicative-versus-imperative readings can be disputed. Discussion: [Stanford Encyclopedia of Philosophy](https://plato.stanford.edu/entries/chinese-translate-interpret/).
- Which subjects, objects or possessors are supplied by the reading, and where are their proposed referents established? provenance
- What supports reading the clause as a statement, command, question or another speech act? definition
Allusions and genre expectations
Test whether parallel wording or genre practice materially constrains a reading.
Evidenced intertextual reading
Separate demonstrated quotation, plausible allusion and coincidental resemblance; identify commentary-derived interpretations without treating them as the passage's explicit wording.
- Which earlier or parallel passage supports the proposed allusion, and is the chronological relationship established? provenance
- Does the proposed genre convention or allusion change the reading enough to require verification before translation? action
Interpretive state and use Determine what outputs the available linguistic evidence supports.
The agent must distinguish a usable provisional reading from one whose unresolved choices materially change the requested result.
Reading readiness
Assess consequential uncertainty for a specific passage and task.
Task-relative reading state
Identify unresolved graphs, clause boundaries, senses and referents that affect the requested output, and state what evidence would resolve them.
- How many unresolved alternatives change the passage's proposition or participant roles, and which spans contain them? measurement
- Can the requested use proceed with labelled alternatives, or does it require another witness, a commentary or specialist review? action
Translation and composition controls
Keep transformations aligned with their source and declared linguistic target.
Declared output contract
Specify whether the output is a gloss, readable translation, editorial reconstruction or new imitation, with supplied content and stylistic targets explicit.
- How will the output identify supplied pronouns, inferred relations, emendations and unresolved readings? action
- For new composition, which period, genre and attested constructions define the target, and how will the result be labelled as newly generated? action
Evidence and external alignment What the world already says about this thing, gathered so the model can be checked against it.
A model that cannot be lined up against existing standards, identifiers and practice cannot be adopted by anyone who already uses them.
Reported evidence
Findings from the breadth pass, kept separate from the structural claims.
Kinds and varieties
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Pre-Qin canonical prose (the grammatical model of the Analects, Mencius, Zuozhuan, Zhuangzi and allied texts)
- Han historical and rhapsodic prose (Shiji, Hanshu, fu)
- Six Dynasties parallel prose (pianwen)
- Tang-Song ancient-style prose (guwen)
- Civil-service examination prose, including the eight-legged essay (baguwen)
- Buddhist and other translated Literary Chinese
- Sinospheric Literary Chinese as a written lingua franca (Japanese kanbun, Korean hanmun, Vietnamese Hán văn)
- Residual modern Classical used for inscriptions, couplets, mottos, seals and ceremonial titles
- Which of these kinds and varieties hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Identifiers and schemes
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- ISO 639-3 - lzh - Registered name: Literary Chinese; this is the code used for the written tradition of which Classical Chinese is the canonical core.
- BCP 47 / IETF language tag - lzh - Language subtag for Literary Chinese.
- Wikidata - Q37041 - Item for Classical Chinese / Literary Chinese as a written language.
- Glottolog - lite1248 - Literary Chinese.
- Which of these identifiers and schemes hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Standards and regulation
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- ISO 639-3 code lzh, issued by the ISO 639-3 Registration Authority (SIL International), identifies the written language for cataloguing and tagging.
- Unicode Standard and ISO/IEC 10646 (Unicode Consortium; ISO/IEC JTC 1/SC 2) govern encoding of the CJK characters in which the register is written.
- PRC Ministry of Education ordinary-high-school Chinese curriculum sets mandatory 文言文 passages and examination requirements.
- Republic of China (Taiwan) Ministry of Education 12-year basic-education Chinese curriculum sets a larger 文言文 share than the PRC school course.
- Historically, Qing and earlier keju (civil-service examination) statutes governed who had to demonstrate competence; the Qing court abolished the examinations in 1905.
- Which of these standards and regulation hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Real-world use
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Read and examined as 文言文 in secondary schools and university-entrance tests in the PRC, Taiwan and Hong Kong.
- Used for calligraphy, temple and ancestral-hall inscriptions, New Year couplets (对联), seals and university or state mottos.
- The working language of dynastic histories, local gazetteers, legal codes, the Confucian canon and much premodern technical writing.
- Read in Japan as 漢文 through kundoku; historically the written medium of elite Korea (한문) and Vietnam (Hán văn).
- Searched and cited in full-text corpora (Chinese Text Project, CBETA, Scripta Sinica) for historical and philological work.
- Which of these real-world use hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Typical measurements
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Length of a single school or gaokao 文言文 reading passage - 300-800 - Chinese characters
- Function-word (虚词) inventory a competent reader is expected to control - 100-250 - item types
- Character types often cited as a threshold for independent reading of unpunctuated texts - 2000-5000 - character types
- Span during which the register was the default written medium of elite communication - c. 500 BCE - c. 1920 - years CE/BCE
- Which of these typical measurements hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Failure modes and hazards
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Reading a graph with its modern Mandarin gloss and thereby misparsing syntax, especially 之, 其, 而, 以, 于, 也, 矣.
- Conflating the written register with reconstructed Old Chinese phonology, or with spoken Middle Chinese.
- Clause-boundary and anaphora errors in originally unpunctuated text (zero anaphora, topic drop).
- 通假 (phonetic-loan) and character-variant mismatch, worsened by simplified/traditional conversion.
- Quoting canonical tags as political or commercial slogans stripped of source syntax and context.
- Pedagogical translation into 白话 that erases register and invents subjects the original leaves implicit.
- Which of these failure modes and hazards hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Regional variation
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- PRC education names it 文言文 and teaches it in simplified characters against 白话文; Taiwan and Hong Kong use traditional characters and generally keep a larger classical share.
- Sinologists often reserve "Classical Chinese" for the pre-Qin/Han model and use "Literary Chinese" (ISO lzh) for the whole later written tradition; Chinese school usage does not make that split.
- Japan reads the same written medium as 漢文 via kundoku, not as spoken Chinese.
- Korea treats 한문/漢文 as a classical written subject; Vietnam historically used Hán văn, distinct in practice from chữ Nôm vernacular writing.
- Local documentary styles (edicts, contracts, epitaphs) diverge from the school canon even where the grammar remains classical.
- Which of these regional variation hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Neighbouring kinds and how to tell them apart
Reported by the breadth pass; each item needs checking against its source before it becomes normative.
- Old Chinese - Old Chinese is a reconstructed spoken language of the Zhou period (rhymes, phonetic series, morphology); Classical Chinese is a written grammatical-lexical register. A phonological reconstruction is not a test of whether a sentence is classical prose.
- Middle Chinese - Middle Chinese is the Qieyun-system phonological stage used to explain readings and regulated verse; it does not define the syntax of 文言文.
- Written vernacular Chinese (白话/白话文) - Vernacular writing uses spoken-based grammar (aspect 了/着/过, copula 是, particle 的). Classical prose uses 也/矣/焉, 之/其, and does not rely on those vernacular markers.
- Modern Standard Chinese - A living spoken-and-written standard intelligible orally; Classical Chinese is not a spoken vernacular and is not mutually intelligible with modern speech without training.
- Literary Chinese of medieval East Asia - Same written medium, but local reading traditions (kundoku, hanmun) and some local graphs and documentary formulae; the test is reading practice and local usages, not a different core grammar.
- Regulated verse (近体诗) - A poem may be classical in lexicon while obeying Tang tonal metre and parallelism; the test is whether tonal-metre constraints, not classical prose grammar, license the line.
- Which of these neighbouring kinds and how to tell them apart hold for the sense of Classical Chinese this model covers, and on what evidence? provenance
Sources
- Outline of Classical Chinese Grammar - Specialist grammatical description of the late Zhou-Han written register that sinologists call Classical Chinese.
- Chinese - Places Classical/Literary Chinese in the history of Sinitic and distinguishes it from spoken Old, Middle and modern varieties.
- ISO 639-3 code lzh (Literary Chinese) - The registered language identifier and the standard name Literary Chinese for this written tradition.
- 古代汉语 (Ancient Chinese) - The standard PRC textbook distinction of 文言文, 虚词, 通假 and the school canon that still structures how the register is taught.
What the second pass must settle
- What definition and chronological boundary should this registry entry adopt, particularly for later Literary Chinese?
- Does an existing Vercy world model already own this linguistic concept and therefore require this entry to link to it?
- How should Literary Sinitic composed outside China, vernacular mixtures and modern classical imitation relate to this entry without creating unauthorized registry splits?
- Which reference corpora, editions and grammatical descriptions provide sufficient coverage across the periods and genres ultimately included?
- Which task-specific evidence thresholds should distinguish a publishable reading from a provisional analysis requiring review?