候Mantic arts dossier · Celestial omens and signs
风角、云气与物候杂占
Fēngjiǎo, Yúnqì yǔ Wùhòu Zázhān·Wind, cloud, phenological and animal omens
Reading winds, cloud forms and colors, rainbows, thunder, animal behavior and unusual phenology for war, weather and affairs, at the boundary between practical observation and symbolic omenology.
- Aliases
- wind divination, cloud omenology, animal and weather signs
- Formation
- Pre-imperial precedents; extensive manuals from Qin-Han through Tang
- Evidence status
Study notes & citation
风角、云气与物候杂占. Institute of Mantic Arts. https://www.manticarts.com/en/arts/wind-cloud-omens
On this page
01Core principles
Core principles
Translate formulas into structure: what enters, how the model is arranged, and which relations drive interpretation.
Time, direction, color, sound and duration were recorded and matched to omen statements through phases, directions, offices, armies or topics. Some rules may preserve weather experience; others are symbolic analogies.
Key concepts
- wind-angle divination
- cloud qi
- phenology
- miscellaneous omens
02Origins and historical development
Origins and historical development
Keep traditional accounts, surviving evidence and later transformations distinct.
Pre-imperial narratives, Qin-Han daybooks, Mawangdui and Dunhuang materials preserve many examples. Tang works such as Yisi Zhan and Kaiyuan Zhanjing compiled earlier omen lore.
TEXT / METHOD / EVIDENCE
Cloud imagery and conditional omen statements
Colour, shape, direction, timing and interpretation form a conditional statement. Separating them opens questions about classification, compilation and the evidence needed for empirical claims.
Matching images to text within a scroll
IDP's learning guide distinguishes the cloud-divination portion from the star atlas later in the scroll. Record the adjacency of images, labels and passages. A proposed image-text match needs positional or labelling evidence; resemblance of subject matter alone is insufficient.
An image-text inventory can record image ID, face or sequence, adjacent text and confidence in the match. Include damage, blank spaces and joins. A gap can affect paragraph boundaries. Modern crops that omit it may make an interpretation appear more certain than the physical arrangement supports. Preserve a wider contextual view when permitted.
Conditions, outcomes and attributed voices
Represent a rule as a described phenomenon under stated circumstances followed by an interpretation. Code colour, shape, direction and duration separately. Distinguish a condition absent from the statement from one lost or illegible: the first concerns what the rule expresses, the second what the present witness allows us to read.
Where a passage attributes a statement to someone, record an internal attribution field. A text's attribution is evidence about its presentation of authority; establishing actual authorship, observation or transmission requires further investigation. Multiple attributed voices can be studied as an editorial arrangement without treating every name as a verified historical contributor.
From historical rules to testable questions
Historical analysis can examine classification, transmission and interpretation. Testing a predictive claim additionally requires definitions of occurrence and outcome set before observation. Retaining only apparent successes removes failures and the denominator. Retrospective narratives and prospective records are different data sources and should not be pooled without explaining their selection mechanisms.
A contemporary dataset needs location, obstruction, lighting, equipment, observation duration and classifier identity. Two independent coders can classify the same image set before inspecting outcomes. Measure agreement first. If the input category cannot be assigned reproducibly, later correspondence estimates rest on an unstable measurement process. Record disagreements rather than silently resolving them after outcomes are known.
A four-step research record
Fix the textual object
Assign image and passage IDs and state the evidence for their association.
Extract conditions
Record phenomenon, time, location and duration; identify loss separately.
Distinguish voices
Distinguish copyist, internal attribution, modern translator and your interpretation.
Choose the research target
Specify textual structure, historical practice or a separately designed empirical test.
WORKED EXAMPLE
From material to claim
Original exercise, not an ancient quotation: rule A requires white, eastward and three time units; rule B requires only white. An observation says white, direction unknown, with no duration.
Within this teaching model, B's condition is supported; A has two unknown conditions and remains indeterminate. Missing values are neither false nor retrospectively fillable from the outcome.
- Are unspecified and missing values distinct?
- Did classifiers see outcomes beforehand?
- Are failures and indeterminate cases both retained?
Review questions
Can a text's assertion of testing become a modern accuracy rate?
A denominator, failures, selection rules and outcome definitions are still needed. The assertion has historical significance; an accuracy estimate requires auditable data.
Does visual resemblance establish identical classification?
Define categories, observational scale and image processing, then examine independent agreement. Resemblance supplies a candidate match.
STUDY / Fēngjiǎo, Yúnqì yǔ Wùhòu Zázhān
Observation protocols for wind and clouds
- Materials and inputs
- Direction, time, weather description and traditional judgment
Research question
Can different observers describe the same phenomenon consistently?
Practical exercise
Have two observers independently describe shared public weather images before comparing traditional terms. Separate descriptive agreement from prognostication.
Close readingDunhuang Star Chart: building an evidence fileStudy pathSpace: orientation, environment and selection03Evidence limits and research debates
Evidence limits and research debates
Distinguish traditional interpretations, documented material and claims requiring testing.
Practical meteorology and auspicious interpretation must be analyzed separately. Animal behavior cannot be treated as portent without ecology, season and observational bias.
This dossier supports historical, textual and methodological study. Internal traditional explanations are not modern empirical validation and do not replace medical, legal, financial or other professional advice.
04Core texts and materials
Core texts and materials
- 01Yisi Zhan
- 02Kaiyuan Zhanjing
- 03Dunhuang cloud and bird-call manuals
- 04Qin-Han daybooks