三千年疫情记录与干支纪年的符合度检验:古人疫情年谱能否被统计验证

Written by

in

Language

Evidence level: B

中国历史疫情年谱(《中国三千年气象记录总集》类数据源)与干支纪年”太过不及”归年的匹配符合度,是可被现代置换检验直接裁决的统计命题——疫情记录是否真的偏向某些干支年份,不再只靠古籍叙述。

正文

五运六气理论的核心断言之一是:天干地支的周期性轮转(六十年一甲子)携带气候与疫病流行的时间节律,某些”太过不及”之年疫病更易流行。过去两千年间,中国留下了全球最连续的书面疫情记录——正史、方志、档案、笔记中的疫情条目,最终被辑录为《中国三千年气象记录总集》这样的巨型文献数据库。问题的关键在于:这套定性叙述是否可以被变成一张可检验的时间表?具体地说,把历史疫情事件按公历年份数字化后,它们是否真的在”太过/不及”干支年份上出现统计上显著的高发?这一命题完全可测,且已具备成熟的统计工具。

第一步,历史疫情序列必须数字化,且方法学已成熟。 REACHES 数据库的建设(PMID 30561430)正是基于《中国三千年气象记录总集》,建立了层级编码系统:把古籍中每条记录翻译为”时间+地点+事件类型”的数字条目,并将中国历史纪年转换为公历、地名转换为经纬度。这套方法已覆盖1644-1795年的近五万条二级记录,证明”定性文言文 → 结构化时序数据”的转换是可行的——这为干支匹配检验提供了数据管道。同样重要的是,研究者已经清楚记录了这类数据源的天生偏倚:官方记载密度随时间(朝代兴衰、方志出版量)和地域(东部稠密、西部稀疏)剧烈变化,”记异不记常”导致常规年份被系统性低估。

第二步,疫情-气候年际关联确属真实信号。 对古代中国疫情与气候因子的定量重建(PMID 29109246)证明:在年分辨率上,冷干气候通过蝗灾和饥荒间接推高疫情流行,且这一关联在不同频率域上表现不同——低频长期趋势与疫情负相关,而旱涝蝗饥等高频变异与疫情正相关。这说明历史疫情年谱并非随机噪声,其中确实编码了气候信号。更直接的证据来自鼠疫:1850-1964年中国第三次鼠疫大流行的空间时间序列分析(PMID 21646523)显示,鼠疫强度与干湿指数呈非线性关联,且在南北中国方向相反;而1754-1949年的县级鼠疫等级数据研究(PMID 40472758)进一步发现,上一年干旱每加重一级,北方鼠疫风险上升28.4%(95%CI 11.3%-48.2%)——这是一个真实的年际滞后效应,证明”前一年气候→当年疫情”的跨年链条在历史数据中可检测。

第三步,把这些证据接回干支归年,就构成可执行的检验方案。 干支纪日是连续、精确的天文历法序列,可无损映射到公历年份(干支纪日连续性的断代验证见本系列相关文章),因此每个历史疫情事件都可以打上干支年标签。检验设计如下:将所有疫情事件按”太过/不及”归年分组,比较各组疫情事件频率与随机基线的差异;关键纪律是用置换检验替代朴素卡方——因为文献记录密度本身随年代漂移,必须把”干支分组”这个标签在时间轴上随机重排几千次,生成零分布,再看真实归年组的频率是否落在分布尾部。同时做FDR校正以控制多重比较(五个太过运、六个不及气、再加”刚柔失守”滞后1/2/3年的窗口分层,检验次数动辄上百)。这个方案的精妙之处在于:它检验的不是”气候是否影响疫情”(这已被 PMID 29109246 证实),而是”干支纪年的特定年份是否携带超出普通气候关联的额外预测力“——即古人的经验编码是否真的抓住了气候-疫情耦合中未被现代气象指标覆盖的相位信息。

值得警惕的陷阱是:文献偏倚可能与干支年份产生虚假相关。方志出版高峰、战乱记载中断、中央修史周期的年代学结构,如果恰好与干支周期谐波相近,会制造伪符合。因此检验必须同时控制记录密度(以总记录数为暴露校正)和朝代结构,并把”干支效应”与”已知气候效应”(旱涝蝗饥)做条件化分离。若控制后干支效应消失,则说明古人经验是对气候关联的间接总结;若控制后仍有独立信号,那才是真正值得深挖的相位信息。

证据与局限

已验证:
– 中国历史疫情可数字化重建为年分辨率时序(PMID 29109246),REACHES编码方法学成熟(PMID 30561430)
– 历史疫情与气候的年际关联真实存在,含上一年干旱→当年鼠疫风险升高的年际滞后(PMID 40472758)
– 历史鼠疫强度对干湿响应南北相反、呈非线性(PMID 21646523)

部分验证:
– 疫情记录存在时空不均、记异不记常等系统性偏倚(PMID 29109246、30561430),会污染朴素统计,必须用记录密度校正+置换检验

待验证:
– 控制记录偏倚与已知气候因子后,”太过/不及”干支年份是否仍携带独立于普通气候关联的疫情高发信号
– “刚柔失守”归年后1/2/3年滞后窗口在历史疫情序列上是否复现现代数据中的滞后模式
– 疫情事件类型(疫病种类)与干支归年的类型特异性匹配(火太过年是否偏热性疫病、水太过年是否偏寒湿疫病)

相关阅读

证据等级: B(历史疫情-气候关联有PubMed直接证据,”干支纪年匹配符合度”全球尚无系统检验)
参考文献: PMID 29109246, PMID 40472758, PMID 21646523, PMID 30561430


免责说明:本文属传统知识与科学文献的整理、对照与研究设计讨论,仅供理解与参考,不构成医疗建议,不能替代专业诊断与治疗。文中涉及的效应量、频率区间与暴露结论多为研究报告层面的推断,个体情况差异很大;如有身体不适或健康疑虑,请就医并由专业医师判断。

Validation of the Compatibility Between Three Millennium of Pandemic Records and the Heavenly Stems and Earthly Branch Calendar: Can Ancient Pandemic Chronologies Be Statistical…

Evidence level: B

The degree of fit between Chinese historical epidemic chronologies (data sources such as the Compendium of Three Thousand Years of Meteorological Records) and the “excess-deficiency” year classification within the Gan-Zhi cycle constitutes a statistical proposition subject to direct adjudication via modern permutation testing—specifically, whether epidemic records genuinely skew toward certain Gan-Zhi years rather than relying solely on ancient textual narratives.

Main Text

One core assertion of the Five Movements and Six Qi theory is that the cyclical rotation of Heavenly Stems and Earthly Branches (a sixty-year cycle) carries temporal rhythms of climate and epidemic prevalence, with certain “excess” or “deficiency” years being more prone to outbreaks. Over the past two millennia, China has preserved the world’s most continuous written record of epidemics—entries found in official histories, local gazetteers, archives, and notes—which have ultimately been compiled into massive literature databases such as the Compendium of Three Thousand Years of Meteorological Records. The critical question is whether this qualitative narrative can be transformed into a verifiable timeline. Specifically, when historical epidemic events are digitized by Gregorian year, do they indeed show statistically significant high-frequency occurrences specifically in “excess/deficiency” Gan-Zhi years? This proposition is fully measurable and possesses mature statistical tools.

Step 1: The historical epidemic sequence must be digitized, a methodological maturity already established. The construction of the REACHES database (PMID 30561430), based on the Compendium, established a hierarchical coding system: translating every record from ancient texts into digital entries of “time + location + event type,” and converting Chinese historical chronology to the Gregorian calendar while mapping place names to latitude and longitude. This methodology covers nearly 50,000 secondary records spanning 1644–1795, demonstrating the feasibility of converting “qualitative classical Chinese → structured temporal data”—thereby providing a data pipeline for Gan-Zhi matching validation. Equally important is the clear documentation of inherent biases in these data sources: official record density fluctuates drastically over time (based on dynastic rise and fall, gazetteer publication volume) and region (dense in the east, sparse in the west). The tendency to “record anomalies but not the norm” leads to a systematic underestimation of routine years.

Step 2: The interannual association between epidemics and climate factors is indeed a real signal. Quantitative reconstruction of ancient Chinese epidemics and climatic factors (PMID 29109246) demonstrates that at an annual resolution, cold-dry climates indirectly elevate epidemic prevalence through locust disasters and famine, with this association manifesting differently across frequency domains—negatively correlated with low-frequency long-term trends but positively correlated with high-frequency variations such as droughts, floods, locust plagues, and famines. This indicates that historical epidemic chronologies are not random noise but indeed encode climatic signals. More direct evidence comes from the plague: spatial-temporal sequence analysis of China’s third major plague pandemic (1850–1964) (PMID 21646523) revealed a non-linear association between plague intensity and the wet-dry index, with opposite directions in northern versus southern China. Furthermore, research on county-level plague severity data from 1754–1949 (PMID 40472758) found that for every step increase in the previous year’s drought severity, the risk of northern plague increased by 28.4% (95%CI 11.3%-48.2%)—a real interannual lag effect, proving that the cross-year chain of “previous year climate → current year epidemic” is detectable in historical data.

Step 3: Connecting these evidences back to Gan-Zhi year classification constitutes an executable validation scheme. The Gan-Zhi calendar is a continuous, precise astronomical sequence that can be losslessly mapped to Gregorian years (verification of the discontinuity of Gan-Zhi dating is addressed in related articles in this series); therefore, every historical epidemic event can be tagged with a Gan-Zhi year label. The validation design is as follows: group all epidemic events by “excess/deficiency” year classification and compare the frequency of events in each group against a random baseline. The key discipline is to use permutation testing instead of naive chi-square tests—because literature record density itself drifts over time, the “Gan-Zhi grouping” label must be randomly reshuffled thousands of times along the time axis to generate a null distribution, after which one examines whether the frequency in the true classification group falls into the tail of this distribution. Simultaneously, apply FDR correction to control for multiple comparisons (five excess movements, six deficiency qi, plus stratification by 1/2/3-year lag windows for “rigid and soft failure to hold,” resulting in hundreds of test instances). The sophistication of this scheme lies in the fact that it does not test “whether climate affects epidemics” (already confirmed by PMID 29109246), but rather whether specific Gan-Zhi years carry predictive power beyond ordinary climate associations—i.e., whether ancient experiential coding truly captured phase information in the climate-epidemic coupling not covered by modern meteorological indicators.

A trap to be wary of is that literature bias may produce spurious correlations with Gan-Zhi years. Peaks in gazetteer publication, interruptions in war-related records, and the chronological structure of central dynastic historiography could coincidentally align with harmonic frequencies of the Gan-Zhi cycle, creating false fits. Therefore, validation must simultaneously control for record density (using total record count as exposure correction) and dynastic structure, and perform a conditional separation of the “Gan-Zhi effect” from “known climate effects” (droughts, floods, locusts, famine). If the Gan-Zhi effect disappears after control, it suggests ancient experience was an indirect summary of climate associations; if an independent signal persists, that is the genuine phase information worth deep excavation.

Evidence and Limitations

Verified:
– Historical Chinese epidemics can be digitally reconstructed at annual resolution (PMID 29109246); REACHES coding methodology is mature (PMID 30561430)
– Real interannual associations exist between historical epidemics and climate, including an interannual lag where previous-year drought increases plague risk (PMID 40472758)
– Historical plague intensity responds oppositely and non-linearly to wet-dry conditions in north vs. south (PMID 21646523)

Partially Verified:
– Epidemic records exhibit systematic biases such as spatiotemporal unevenness and “recording anomalies but not the norm,” which can contaminate naive statistics; must employ record density correction + permutation testing

Pending Verification:
– Whether “excess/deficiency” Gan-Zhi years still carry epidemic prevalence signals independent of ordinary climate associations after controlling for record bias and known climatic factors
– Whether the 1/2/3-year lag windows following “rigid and soft failure to hold” year classification replicate the lag patterns observed in modern data within historical epidemic sequences
– Whether specific epidemic event types (disease categories) match Gan-Zhi classifications with type specificity (e.g., whether “excess fire” correlates with heat-related diseases, “excess water” with cold-damp diseases)

Related Reading

Evidence Level: B (Direct PubMed evidence exists for historical epidemic-climate associations; systematic validation of “Gan-Zhi year matching fit” is globally unestablished)
References: PMID 29109246, PMID 40472758, PMID 21646523, PMID 30561430


Disclaimer: This article compiles traditional knowledge and scientific literature for comparison and research-design discussion only. It is not medical advice and cannot replace professional diagnosis or treatment. Effect sizes, frequency ranges and exposure conclusions cited here are largely inferences at the research-report level, and individual circumstances vary widely. If you have any discomfort or health concerns, please consult a physician.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *