Eight case files
Seven are admitted transformation or framing cases; one is a negative control. The pilot does not claim a corpus-wide error rate or an exhaustive census.
Symbiain Misrepresentation Case 001
Chinese Classical Thought
in Western Translation.
Source, commentary, translation and the later lives of altered words
A text travels. A commentary joins it. A translation selects its terms. What, precisely, does the next reader inherit?
An archival object, not an imagined reconstruction. [S03]
Research update / 13 September 2026
A clearer record.
A narrower claim.
The expanded Field Manual does not offer an undifferentiated charge of “mistranslation”. It follows eight worked case files and separates an observed textual difference from its classification and any alleged downstream effect.
Commentary is not invention
In the Daxue opening, Heaven is absent from the base clause but appears in Zhu Xi’s Chinese commentary. That distinction changes what the evidence can support.
Draft, not a verdict
The research has not been peer reviewed or independently specialist-signed off. Chinese witnesses, Latin grammar and reception links remain open to focused review.
Chinese witness≠textual observation≠interpretation≠causal claim
This is the governing discipline for the project. A translation difference can be demonstrated locally; claims about intent, prevalence or long historical effects require their own evidence.
01 / Executive intelligence assessment
A documented difference.
An open causal question.
A major printed gateway
Confucius Sinarum Philosophus appeared in Paris in 1687, a collaborative Latin presentation associated with Prospero Intorcetta, Christian Herdtrich, François de Rougemont and Philippe Couplet. [S02]
A specific semantic addition
Oxford historian Ross Moncrieff reports “Heaven” in Nathaniel Vincent’s 1685 rendering of a Chinese line without 天. Vincent’s source was Intorcetta’s 1662 Sapientia Sinica. [S01]
A larger transformation
Systematic relational loss, centuries of intellectual effects and deliberate manipulation require different evidence. This case offers a testable protocol; it does not claim those findings have been demonstrated.
Misrepresentation does not always replace the facts. Sometimes it preserves the facts and changes the relationships between them.
Governing proposition: a Symbiain analytical claim. The historical evidence below tests particular instances; it does not establish a universal contrast between Chinese and Western thought.
02 / Source and commentary
Heaven is absent
from the clause.
It is present in Zhu Xi’s commentary. The question is not whether a European translator simply invented an alien idea, but how distinct textual layers were combined and attributed for later readers.
大學之道，在明明德
Working gloss: the way of great learning lies in making bright virtue manifest.
A modern working gloss for comparison, not a critical translation. There is no 天 in this quoted clause.
Heaven belongs to the Chinese commentary; Vincent’s reported 1685 English passage presents spiritual power “conveyed to us from Heaven”.
The task is to establish the route and attribution of the added relation, rather than collapse commentary, Latin and English into one voice. [S01] [S09]
and commentary
Sapientia Sinica
reported excerpt
related wider object
The parallel marker separates publications. Vincent’s earlier English wording cannot descend from the later 1687 volume.
What this exhibit establishes—and what remains open
The lexical absence in the quoted base clause and the commentary’s 人之所得乎天 are inspectable. The precise Chinese–1662 Latin–1685 English relationship still needs direct collation. The evidence does not establish invention, deception or a single European motive.
The English passage is reported through published scholarship, rather than independently collated here against Vincent’s original printed page. No corresponding 1687 Latin passage is invented or silently substituted.
03 / Pilot research register
Eight files.
Different burdens of proof.
The new draft makes a crucial distinction: a transformation can be a commentary fusion, editorial frame or reception change rather than a literal mistranslation.
| Case | Source point | Current classification |
|---|---|---|
| C01 | Daxue 明明德 | Recoding and attribution compression; commentary explains Heaven. |
| C02 | Daxue 親民 / 新 reading | Negative control: commentary-dependent reading, not a proven literal error. |
| C03 | Daxue 知止 sequence | Amplification and local attribution shift. |
| C04–C05 | Zhongyong opening and emotions | Rational and normative specification; downstream proposition untraced. |
| C06 | Daxue 物格而後知至 | Provisional strong grammatical and semantic mismatch. |
| C07 | Liezi western sage | Extraction and theological framing; uncertainty retained. |
| C08 | Orphan first-act song | Lyrical omission and later cultural generalisation; reception requires separate evidence. |
03 / Element Dice & Slice
Audit the relation,
not just the vocabulary.
Object → representation → structure → relationship → transformation → judgement.
Object
Fix edition, folio, characters, punctuation and surrounding commentary. Keep text and commentary separate.
Elements
Record characters, compounds and attested meanings; do not turn graphic components into a translation.
Relationships
Identify predicates, agents, objects, modifiers, sequence and what remains underdetermined.
Transformation
Align the Chinese with the Latin or English. Mark additions, omissions, narrowed meanings and changed relations.
Alternatives
Compare independent renderings. Test the strongest charitable reading and passages that resist the hypothesis.
Consequence
Identify who read which version and how they used it. A translation difference alone cannot prove an effect on later thought.
| Layer | Observation | Assessment |
|---|---|---|
| Elements | Chinese clause: 大學 / 之道 / 在 / 明明德 | Segmentation is analytical; 明明德 also needs grammatical interpretation. |
| Addition | English excerpt includes “from Heaven”. | Reported textual difference; no matching 天 in this clause. |
| Relationship | The English supplies a source for spiritual power. | Symbiain inference about the relation introduced by this wording. |
| Intent | No evidence of deliberate deception presented. | Do not infer intent from the difference. |
| Consequence | Vincent used Confucian material in an anti-Hobbesian setting. [S01] | Local reception context, not proof of centuries of causation. |
04 / Mandarin Semantic Register
Before the word
was cooked.
“Uncooked” is a metaphor for reopening a compressed translation. It is not a claim to recover a pure meaning hidden inside a character.
The programme retains its working name, Mandarin Semantic Register, but these sources demand historical Literary Chinese, period usage and genre-specific reading. Modern Mandarin glosses cannot supply that evidence by themselves. [S08]
Graphic
Historical forms; semantic and phonetic components. A prompt for investigation, never proof of contextual meaning.
Concrete
Material senses attested in dated sources; distinguish documented usage from an appealing origin story.
Process
Candidate actions or transformations. Mark an analytical reconstruction as such.
Classical
The lexeme and compound in comparable literary contexts, including counterexamples.
Period
The senses supported in this genre and edition; do not substitute modern Mandarin usage for a Kangxi-period text.
Translation
The actual Latin or English wording, its syntax, commentary and interpretive commitments.
Reconstruction
Competing contextual readings, ranked by evidence with unresolved ambiguity retained.
05 / A second kind of representation
Jieziyuan Huazhuan.
Instruction becomes an image.
The Mustard Seed Garden Manual makes a different inquiry possible: how do marks, techniques and arrangements become a teachable representation?
Wang Gai. Woodblock print; ink and colour on paper. The Metropolitan Museum of Art, Rogers Fund, 1924, CP37. Public domain. This sheet provides visual context; it is not the textual page quoted below. [S04]
A plate can be read as an inventory of trees, rocks and water, or examined through placement, scale and relationships. That contrast is a proposed analytical test. The existence of this manual alone does not prove that a relational reading is historically correct.
07 / Semantic choke points
Where a choice
can close an alternative.
These are research prompts, not dictionary entries or verdicts about a language.
Configuration / power
Is an effect attributed to an isolated actor or to an arrangement?
Pattern / principle
Does the rendering preserve contextual organisation or posit a universal rule?
Method / law
Does the local passage concern practice, a model, a norm or legislation?
Quality / resonance
Does the compound require a different reading from either element alone?
Life / movement / change
Is a process attested by the sentence, or supplied by our analysis?
Heart-mind / harmony
Does the translation add a mental division or turn coordination into a fixed state?
08 / Comparisons with boundaries
Mai-mai Sze.
Sze’s translation, first published in 1956, offers a later English comparator. The published facsimile concerns the 1887–1888 Shanghai edition, not an unchanged 1679 witness. [S07]
Align the Chinese passages and edition variants before comparing her choices with alternative readings. Sze is a comparator, not a neutral gold standard or a completed experimental control. No claim of a demonstrated Sze mistranslation is made here.
Return to Qingming.
In a future study, compare object inventories with carefully evidenced spatial and relational readings of the scroll. Record which interpretation changes a judgement, and what visible evidence supports it.
No genealogical claim: similarity would not demonstrate transmission from the later manual to the earlier scroll. This is a comparison across media, not a proposed historical line of descent.
Visit the existing Qingming field study →09 / Sinarum Transformation Hypothesis
Five hypotheses.
Separate burdens of proof.
These are competing or overlapping explanations, not a ladder on which each step proves the next.
Baseline / live alternative
Ordinary translation loss
Differences arise from language, genre, commentary and unavoidable interpretation.
Discriminating test: Compare transformations against ordinary translation variation and genuine preservation.
Plausible / corpus test pending
Systematic conceptual accommodation
Repeated choices favour categories familiar to the intended European readership.
Discriminating test: Use a preselected Chinese–Latin sample, explicit coding and independent readers.
Local evidence / scope unresolved
Missionary or theological framing
Theological interpretation changes particular representations of Chinese philosophy.
Discriminating test: Separate a commentator’s gloss from translation and show recurrent effects, not merely religious vocabulary.
Research hypothesis
Downstream path dependence
Later judgements depend on a transformed representation that readers inherit.
Discriminating test: Join textual transformation to a documented reader, uptake and consequential reuse. Test alternative sources.
Not established / quarantined
Deliberate ideological manipulation
A consequential alteration was deliberately designed to produce a particular ideological effect.
Discriminating test: Require attributable archival evidence of that intent. Coherence and missionary affiliation are insufficient.
10 / Strange Signal firewall
Coherence → agency → intent.
Coherence
Find a reproducible pattern across a defined sample, including negative cases. A compelling exhibit is not a corpus result.
Agency
Identify who selected, translated, edited and transmitted the passage. A pattern can emerge without central direction.
Intent
Require documentary evidence connecting a particular actor to a particular intended alteration and effect.
No evidence may silently cross this firewall.
11 / Representational path dependence
The next reader may
inherit the frame.
Source object
The text in its edition, language and context.
First representation
A translation, selection or commentary.
Inherited representation
A later text or judgement based on R₁.
This is an analytical model, not a demonstrated transmission chain for every example in the case. Its central test is whether a later judgement depends on an earlier transformation and would plausibly change if the source relationship were restored.
Companion film / The case explained
Same elements.
Changed relationships.
When readers depend on a translation, source text and commentary can become difficult to distinguish. Heaven is absent from the quoted clause but present in Zhu Xi’s Chinese commentary. The question is how translation combined those layers and interpreted their meaning. [S09]
2 minutes 41 seconds · Archival plates, original explanatory diagrams and synthetic British English narration. The documented example is separated from provisional interpretations.
Read the film transcript
Today, Symbiain Intelligence brings together AI-assisted comparison and human judgement to examine how meaning changes as it passes from one account to another. By tracing sources, interpretations and inherited assumptions, we investigate how the accounts we receive shape our understanding.
In sixteen eighty-seven, Couplet and his collaborators published Confucian texts in Latin for European readers. We plan to explore that work in a later film.
Our example here comes earlier. In sixteen eighty-five, Nathaniel Vincent published a partial English translation of the Great Learning, drawing on Intorcetta’s Latin work of sixteen sixty-two. His rendering includes the words “from Heaven”.
The Chinese character for Heaven is absent from the short opening clause being translated. Yet Heaven appears in Zhu Xi’s commentary on that clause, written centuries before the European translations. Its absence from the clause does not mean the idea was absent from the Chinese interpretive tradition.
That distinction changes the question. How were source text, commentary and theological interpretation brought together? What meaning did the translation convey? And could a reader distinguish the original wording from the explanations surrounding it? Symbiain’s role is to make those layers visible: what the source says, what commentary explains, how translation frames it, and what a later reader inherits.
The Mustard Seed Garden Manual opens a related inquiry. Its discussion of method describes learning rules before moving beyond dependence on them.
Mai-mai Sze’s English translation provides a comparison through which Symbiain can examine how those relationships are expressed in another language. Our broader interpretation remains provisional.
When these layers become difficult to distinguish, an interpretation may acquire the authority of the original. Inherited assumptions may then shape thoughts, decisions, behaviour and wider cultural understanding.
Establishing those effects requires evidence at each step.
Symbiain Intelligence helps us trace possible biases, investigate their effects, and recognise when the evidence requires us to correct our own explanation.
Evidence / provenance / open questions
Keep the drawer open.
High confidence marks a bibliographic or directly inspectable observation. Documented in scholarship identifies secondary reporting. Provisional marks an interpretation awaiting testing. Not established means this case supplies no adequate proof.
S01The Reception of Confucianism in Early Modern Britain (2023)
Ross Moncrieff / Oxford Centre for Intellectual History
Secondary scholarship: Vincent 1685, Intorcetta 1662 and English reception. The English quotation is reported here through this source; an edition-and-page collation remains open.
Open the source ↗S02Confucius Sinarum philosophus (Paris, 1687)
University of Tsukuba Library
Institutional bibliographic record for the historical object and its named Jesuit collaborators.
Open the source ↗S03Confucius Sinarum Philosophus frontispiece
Wikimedia Commons / Internet Archive scan
Plate A: anonymous, 1687. Public domain / Public Domain Mark 1.0. Reduced-resolution reproduction; no inscriptions altered.
Open the source ↗S04Wang Gai, Page from the Mustard Seed Garden Manual of Painting
The Metropolitan Museum of Art
Plate B: first edition, printed 1679; woodblock print, ink and colour on paper. Rogers Fund, 1924, CP37. Museum image labelled Public Domain; API confirms isPublicDomain.
Open the source ↗S05芥子園畫傳卷一 · 青在堂畫學淺説
Shidian Guji / 識典古籍
Digital transcription of the primary text. Supports the method/no-method sequence. Facsimile collation remains open; transcription has apparent variants/errors, including 氣運 in its Six Principles passage.
Open the source ↗S06繪畫六法 / The Six Criteria of Chinese Painting
University Museum and Art Gallery, University of Hong Kong
Institutional source for Xie He’s formulation 氣韻，生動是也. This is not a diplomatic transcription of a specified Jieziyuan edition.
Open the source ↗S07The Mustard Seed Garden Manual of Painting, translated and edited by Mai-mai Sze
Princeton University Press / JSTOR
Copyright 1956; facsimile of the 1887–1888 Shanghai edition. Bibliographic comparator; the publisher’s excerpt also supports a sequence of learning rules and subsequently modifying practice. No modern translation pages or illustrations reproduced.
Open the source ↗S08Translating and Interpreting Chinese Philosophy
Stanford Encyclopedia of Philosophy
Methodological context: translation requires interpretation; context constrains meaning. Not evidence that all Chinese thought is relational or all European thought categorical.
Open the source ↗S09四書章句集註 · 大學章句
Zhu Xi / Wikisource primary-text transcription
The short opening clause contains no 天, while Zhu Xi’s commentary on 明德 includes 人之所得乎天. This establishes the distinction between base clause and Chinese commentary; it does not establish the precise route into the Latin and English witnesses.
Open the source ↗