← Archive
lm-002471 · 2026-08

OOE-II_人類早就在做本體工程_操作本體技術史_v0.1

下載 MD 檔 ⬇

OOE-II:人類早就在做本體工程

從死亡、法人、王權、宗教職位到國家連續性的操作本體技術史

OOE-II: Humans Have Long Been Engineering Ontology

A History of Operational Ontology from Death, Legal Persons, and Offices to State Continuity

系列:Operational Ontology Engineering(OOE/操作本體工程)
作者:Neo.K
機構:EveMissLab/一言諾科技有限公司
日期:2026-08-09
版本:v0.1
性質:歷史—理論統合論文
前置論文:《OOE-I:本體論何時變成工程問題?》
前置理論:Continuity Object Theory(COT)


摘要

OOE-I 提出「本體論操作門檻」:當一個原本可以懸置的本體問題開始決定醫療、法律、財產、責任或系統行動,而且不同分類造成的實際損失超過可忽略門檻時,制度便被迫將本體問題轉化為可執行的操作狀態。

然而,這並不是 AI 時代才第一次出現的現象。

本文回顧死亡判準、推定死亡、法律人格、王權「兩個身體」、corporation sole、大學法人、使徒/主教職位繼承、法律父母、特定決策能力與國家連續性等案例,提出一個更強的歷史命題:

Civilization has repeatedly operationalized unresolved ontology.\boxed{ \text{Civilization has repeatedly operationalized unresolved ontology.} }

不同文明與制度雖然沒有使用「操作本體工程」這個名稱,卻反覆發明功能相似的技術:閾值化、法律擬制、角色分層、來源鏈、推定、功能拆解、關係重定義、人格抽象、永久繼承、程序裁決與責任承接。

1968 年 Harvard 腦死報告是一個現代典型案例:復甦與生命支持技術使「心臟仍跳動」與「腦功能不可逆喪失」可以分離,舊死亡判準因而不足,醫療體系被迫建立新的操作判準。法人制度則採用另一條路:它不試圖證明公司具有人的本體性,而是創造一個可以持有財產、簽約、起訴與被訴的人工法律人格。中世紀「國王的兩個身體」與 corporation sole 又將會死亡的任職者與可持續的職位/政治身體分離。英國 Oxford 之類大學則透過 perpetual succession,讓成員持續替換而組織人格不必隨之死亡。國際法中的 state continuity 更直接指出:國家是「continuous rather than static」的法律主體,政府或憲制變化不必等於國家身份中止。

本文將這些制度整理成「操作本體技術譜系」(Operational Ontology Pattern Library),並提出「文明本體編譯器」概念:法律、醫療、宗教與制度的部分功能,其實是在把不可直接觀察、不可完全確定或過於昂貴的本體問題,編譯成可執行的社會狀態。

本文最後主張,2026 年真正新的地方不是人類首次把哲學問題工程化,而是 AI、BCI 與數位人格可能把過去分散於死亡、身份、責任、人格、代理、記憶與連續性的多種操作本體問題,同時集中到同一類可程式化存在上。

關鍵詞:OOE、操作本體工程、本體編譯器、腦死、法律擬制、法人、王權雙身、perpetual succession、國家連續性、推定死亡、法律父母、決策能力


一、OOE 的歷史問題:這真的新嗎?

OOE-I 問:

When does ontology become operational?\text{When does ontology become operational?}

但接著立刻產生另一個問題:

人類是不是直到 AI、BCI 與後人類技術出現,才第一次必須把本體論變成操作規則?

歷史答案顯然是否定的。

更合理的是:

OOE is new as a unified framework, not as a human activity.\boxed{ \text{OOE is new as a unified framework, not as a human activity.} }

人類很早就必須決定:

  • 誰算死亡?
  • 誰算合法繼承人?
  • 誰算同一個王權?
  • 一群成員換掉後,組織是否仍是原組織?
  • 某個失蹤者應否在法律上視為死亡?
  • 誰算孩子的法律父母?
  • 一個人是否有能力對某件事作有效決定?
  • 政府、國名或憲制改變後,國家是否仍是同一個國家?

這些問題都具有 OOE 結構:

OuncertainΠOσoperational.O_{\mathrm{uncertain}} \rightarrow \Pi_O \rightarrow \sigma_{\mathrm{operational}}.

二、第一類:判準工程——「死亡」必須變成可判定狀態

死亡是最直觀的本體概念之一。

傳統生活世界中:

no breathingno heartbeatirreversible death.\text{no breathing} \approx \text{no heartbeat} \approx \text{irreversible death}.

因此多個指標自然重合。

現代生命支持與復甦技術卻使:

circulation>0\text{circulation}>0

可以與:

irreversible brain damage\text{irreversible brain damage}

同時存在。

1968 年 Harvard Medical School Ad Hoc Committee 的報告明確指出,其主要目的就是提出 irreversible coma 作為新的死亡判準;報告將改變復甦/支持技術與器官移植爭議列為建立新死亡定義的重要原因。

這正是:

Technological DecouplingCriterion Engineering.\boxed{ \text{Technological Decoupling} \rightarrow \text{Criterion Engineering}. }

原本看似單一的:

D=deathD=\text{death}

被拆成多個可觀察狀態:

D=f(Bbrain,Ccirculation,Rrespiration,Iirreversibility).D = f( B_{\mathrm{brain}}, C_{\mathrm{circulation}}, R_{\mathrm{respiration}}, I_{\mathrm{irreversibility}} ).

醫療制度不是因此解決了「生命的終極本質」。

它做的是:

metaphysical death questionclinical-operational death criteria.\boxed{ \text{metaphysical death question} \rightarrow \text{clinical-operational death criteria}. }

本文把這類模式稱為:

Criterion Engineering Pattern

判準工程型操作本體


三、第二類:推定工程——真相不可知,但決策不能永遠停住

假設某人失蹤。

制度無法證明:

Alive=1\text{Alive}=1

也無法證明:

Dead=1.\text{Dead}=1.

但以下問題仍會持續累積:

  • 社會福利;
  • 保險;
  • 婚姻;
  • 遺產;
  • 財產管理;
  • 扶養關係。

因此法律建立:

presumption of death.\boxed{ \text{presumption of death}. }

例如美國現行聯邦規則在無法直接證明死亡時,可以依官方認定、長期失聯等證據推定死亡。

這是一個很重要的 OOE 模式,因為:

σlegal=presumed dead\boxed{ \sigma_{\mathrm{legal}} = \text{presumed dead} }

並不等於:

Otruth=certainly dead.\boxed{ O_{\mathrm{truth}} = \text{certainly dead}. }

制度只是承認:

無限等待本身也有成本。

因此:

Cdelay>Cclassification riskC_{\mathrm{delay}} > C_{\mathrm{classification\ risk}}

時,社會採用一個可被推翻的暫時操作狀態。

本文稱為:

Presumptive Ontology Pattern

推定型操作本體


四、推定制度其實非常深

推定的核心形式是:

P(OE)<1,P(O\mid E)<1,

但:

ΠO(E)=σ.\Pi_O(E)=\sigma.

它承認:

action can be justified without metaphysical certainty.\boxed{ \text{action can be justified without metaphysical certainty}. }

因此,一個成熟本體協議可以允許:

σ=provisional,\sigma = \text{provisional},

並保留:

rebuttal+appeal+revision.\text{rebuttal} + \text{appeal} + \text{revision}.

這正是 OOE-I 所主張的:

Operational StatusMetaphysical Truth.\text{Operational Status} \neq \text{Metaphysical Truth}.

五、第三類:法律擬制——「人」可以被制度重新定義

法人是一個更加強烈的案例。

公司不是生物人。

它沒有:

  • 人類肉體;
  • 生物出生;
  • 人體神經系統。

但法律仍可以將其視為:

legal person.\boxed{ \text{legal person}. }

法律人格使非人類法律實體可以:

  • 起訴;
  • 被起訴;
  • 持有財產;
  • 簽訂契約。

所以法律沒有先回答:

corporation 在本體論上是不是 person?

它建立的是:

person-for-a-domain.\boxed{ \text{person-for-a-domain}. }

也就是:

PlegalPbiological.P_{\mathrm{legal}} \neq P_{\mathrm{biological}}.

這是:

Legal-Fiction Pattern

法律擬制型操作本體


六、法律擬制的真正功能:降低重新計算成本

假設公司只有自然人:

Ct={a1,a2,,an}.C_t = \{a_1,a_2,\ldots,a_n\}.

下一期:

Ct+1={b1,b2,,bm}.C_{t+1} = \{b_1,b_2,\ldots,b_m\}.

如果每次成員變動都要重新判斷所有財產、契約、債務、訴訟與權限,則:

Crecompute.C_{\mathrm{recompute}} \uparrow\uparrow.

法人把它壓縮成:

EntityID=C.\boxed{ Entity_ID=C. }

然後:

propertyC,\text{property}\rightarrow C, contractC,\text{contract}\rightarrow C, liabilityC.\text{liability}\rightarrow C.

因此:

Legal Personhood=Identity Compression Technology.\boxed{ \text{Legal Personhood} = \text{Identity Compression Technology}. }

這一點將直接連到 COT 的 Continuity Token 與未來 AI Identity Key。


七、第四類:角色分層——「任職者」和「職位」不是同一個東西

中世紀與早期現代歐洲政治—法律思想發展出著名的 King's Two Bodies。

其核心是將:

KnaturalK_{\mathrm{natural}}

與:

KpoliticK_{\mathrm{politic}}

分開。

前者:

mortal individual.\text{mortal individual}.

後者:

enduring body politic.\text{enduring body politic}.

所以:

Knatural0K_{\mathrm{natural}}\rightarrow0

不必導致:

Kpolitic0.K_{\mathrm{politic}}\rightarrow0.

這是一個非常成熟的操作本體技巧:

CarrierInstitutional Role.\boxed{ \text{Carrier} \neq \text{Institutional Role}. }

本文稱為:

Layer-Separation Pattern

層級分離型操作本體


八、從「國王的兩個身體」到「職位是一個持續存在」

現代法學對 Crown 的分析仍然可以把 office 理解為:

一個持續存在的制度實體,個別自然人只是不同時間的 incumbent。

形式上:

Ooffice(t)=Ooffice(t+1),O_{\mathrm{office}}(t) = O_{\mathrm{office}}(t+1),

即使:

HtHt+1.H_t\neq H_{t+1}.

因此:

holder identityoffice identity.\boxed{ \text{holder identity} \neq \text{office identity}. }

這與 COT 中:

House headHouse\text{House head} \neq \text{House}

以及未來:

model versionAgent\text{model version} \neq \text{Agent}

具有相同抽象結構。


九、第五類:corporation sole——一個職位直接成為法律人格

教會與普通 corporation aggregate 不同。

某些制度發展出:

corporation sole.\boxed{ \text{corporation sole}. }

即:

現任職位持有人與其 successors 構成一個持續法律人格。

這使財產不必在每任主教死亡後重新變成:

Pestate of dead bishop.P\rightarrow \text{estate of dead bishop}.

而可以保持:

PObishop.P\rightarrow O_{\mathrm{bishop}}.

所以:

B1B2B3B_1\rightarrow B_2\rightarrow B_3

但:

OBO_B

保持。

其操作本體作用非常清楚:

Mortal CarrierPersistent Office-Person.\boxed{ \text{Mortal Carrier} \rightarrow \text{Persistent Office-Person}. }

十、這不是哲學玩具,而是財產系統需要它

如果沒有職位/法人持續性,每一次任職者死亡都可能造成:

property discontinuity,\text{property discontinuity}, contract discontinuity,\text{contract discontinuity}, authority discontinuity.\text{authority discontinuity}.

所以:

Tinstitution>ThumanT_{\mathrm{institution}} > T_{\mathrm{human}}

會產生一個制度壓力:

create a carrier-independent identity.\boxed{ \text{create a carrier-independent identity}. }

這可以看成:

Carrier-Decoupling Principle

載體解耦原理


十一、第六類:perpetual succession——組織活得比所有成員都久

Oxford 是非常好的例子。

Oxford 的治理史資料指出,它由十二世紀末的 Masters 與 students 群體逐漸發展,1216 年左右開始使用 universitas;1571 年正式法定法人化後,Chancellor、Masters and Scholars 被賦予 perpetual succession。

這意味:

MtMt+1M_t \rightarrow M_{t+1}

成員可以持續替換,

但:

Ut=Ut+1.U_t = U_{t+1}.

從 COT 看:

cmembers0c_{\mathrm{members}}\rightarrow0

甚至經過足夠長時間後:

cphysical campus<1,c_{\mathrm{physical\ campus}}<1, ccurriculum<1,c_{\mathrm{curriculum}}<1,

仍不必:

IOxford0.I_{\mathrm{Oxford}}\rightarrow0.

這是:

Member-Independent Continuity Pattern

成員獨立型連續本體


十二、大學其實是一艘制度型忒修斯船

一所古老大學可以逐步替換:

  • 所有人;
  • 建築;
  • 學科;
  • 教材;
  • 財務制度;
  • 管理規則。

但仍透過:

name+charter+legal person+history+relations+recognition\text{name} + \text{charter} + \text{legal person} + \text{history} + \text{relations} + \text{recognition}

維持:

IU1.I_U\approx1.

所以:

University=Institutional Ship of Theseus.\boxed{ \text{University} = \text{Institutional Ship of Theseus}. }

人類很早就不是只在哲學上思考忒修斯問題。

我們已經讓大學、城市、教會、王權與公司實際這樣運作了數百年。


十三、第七類:來源鏈——權力是否沿合法路徑傳過來?

使徒繼承/主教繼承提供另一種完全不同的解法。

這種制度關心的不是:

新任主教和前任主教像不像?

而是:

authority provenance.\boxed{ \text{authority provenance}. }

形式上:

A0A1A2An.A_0 \rightarrow A_1 \rightarrow A_2 \rightarrow \cdots \rightarrow A_n.

只要每一條傳遞邊:

eie_i

被認為有效:

V(ei)=1,V(e_i)=1,

則:

cG+cT\boxed{ c_G+c_T}

保持。

這是:

Provenance-Chain Pattern

來源鏈型操作本體


十四、來源鏈與「相似性」完全不同

一個主教可以在性格、能力、國籍、語言與政治觀點上與前任完全不同。

但:

IofficeI_{\mathrm{office}}

仍然可以持續。

因此其關鍵不是:

similarity.\text{similarity}.

而是:

authorized transition.\boxed{ \text{authorized transition}. }

這與未來 AI 的 delegation chain、provenance chain、signed migration、successor certificate 具有直接形式類似。


十五、第八類:關係拆分——「母親/父母」不是單一生物關係

輔助生殖與代理孕母使:

genetic parent,\text{genetic parent}, gestational parent,\text{gestational parent}, intended parent,\text{intended parent}, legal parent\text{legal parent}

可以不再重合。

例如英國 HFEA 的現行指引明確區分生產者、預定父母與 legal parenthood;在代理孕母安排中,生下孩子的 surrogate 在出生時具有法律母親身份,而 parental order 可以將法律親權轉移給 intended parent(s)。

這說明:

Biological RelationLegal Relation.\boxed{ \text{Biological Relation} \neq \text{Legal Relation}. }

制度不是在問:

誰「真正」才是宇宙本體論上的母親?

它是在不同功能域建立不同關係:

Rgenetic,Rgestational,Rlegal,Rsocial.R_{\mathrm{genetic}}, \quad R_{\mathrm{gestational}}, \quad R_{\mathrm{legal}}, \quad R_{\mathrm{social}}.

本文稱為:

Relation-Splitting Pattern

關係拆分型操作本體


十六、關係拆分是未來人機混合的重要前例

這個模式非常重要。

它說明:

同一自然概念可以被拆成數個制度角色,而不必強迫其中一個吞掉其他所有意義。\boxed{ \text{同一自然概念可以被拆成數個制度角色,而不必強迫其中一個吞掉其他所有意義。} }

未來例如:

AI-generated decision\text{AI-generated decision}

也可能拆成:

  • originator;
  • operator;
  • model contributor;
  • executor;
  • legal principal;
  • responsible party。

不一定非要找一個唯一:

“真正做決定的人”.\text{“真正做決定的人”}.

十七、第九類:功能拆解——自由意志不必先解完

英國 Mental Capacity Act 類型的 capacity framework 提供另一個重要模式。

法律不需要先回答:

Does this person possess metaphysical free will?\text{Does this person possess metaphysical free will?}

而可以針對某一特定決策檢查:

  • 理解;
  • 保留;
  • 使用/權衡資訊;
  • 表達選擇。

因此:

global autonomy\text{global autonomy}

被轉成:

decision-specific functional capacity.\boxed{ \text{decision-specific functional capacity}. }

這就是:

Functional Decomposition Pattern

功能拆解型操作本體


十八、功能拆解其實是「本體降階」

它把:

Oglobal=autonomous person?O_{\mathrm{global}} = \text{autonomous person?}

降成:

f=(f1,f2,f3,f4).\mathbf f = (f_1,f_2,f_3,f_4).

然後:

σcapacity=F(f,K).\sigma_{\mathrm{capacity}} = F(\mathbf f,K).

其中:

KK

是具體決策情境。

因此:

global essencelocal competence.\boxed{ \text{global essence} \rightarrow \text{local competence}. }

這種技術未來也可能用於 AI:

不是先判定:

AI 有沒有真正自由意志?

而是問:

這個 Agent 是否理解這類契約?

是否能維持長期承諾?

是否能辨識權限邊界?


十九、第十類:國家連續性——國家是持續的,不是靜態的

國際法把問題再推進一步。

James Crawford 對 state continuity 的經典分析指出:

States are continuous rather than static.\boxed{ \text{States are continuous rather than static}. }

國家可以發生政府更換、憲法重構與政治革命,但其法律關係仍可能持續。

所以:

regime change⇏state extinction.\text{regime change} \not\Rightarrow \text{state extinction}.

甚至國際法研究指出,一個國家在缺乏有效政府時仍可能持續作為國際法上的 state。

這意味:

cgovernment0c_{\mathrm{government}} \rightarrow0

不必:

Istate0.I_{\mathrm{state}} \rightarrow0.

本文稱為:

Legal-Relational Continuity Pattern

法律—關係型連續本體


二十、state continuity 和 state succession 為什麼必須分開?

如果:

SS1+S2,S\rightarrow S_1+S_2,

首先要問:

SS 還存在嗎?

這是:

Istate.I_{\mathrm{state}}.

其次才問:

條約、債務、財產、國籍由誰承接?

這是:

Succession.Succession.

因此:

IdentitySuccession.\boxed{ \text{Identity} \neq \text{Succession}. }

這再次與 COT 完全吻合:

得到前一存在的資產⇏成為前一存在本身.\text{得到前一存在的資產} \not\Rightarrow \text{成為前一存在本身}.

二十一、十種操作本體模式

現在可以把本文案例整理成一個 Pattern Library:

類型 歷史/制度案例 核心技術
判準工程 腦死 用新觀測門檻重定義操作狀態
推定工程 推定死亡 在真相未知時產生可推翻狀態
法律擬制 法人 建立非生物的人格節點
層級分離 King's Two Bodies 載體與制度角色拆分
載體解耦 corporation sole 任職者死亡但職位人格持續
成員獨立 大學法人 成員全換但組織持續
來源鏈 apostolic succession 合法轉移鏈維持權威
關係拆分 legal parenthood 生物/孕育/法律角色分離
功能拆解 mental capacity 全域本體問題降為局部能力
法律—關係連續 state continuity 結構大變仍保存法律身份

這些制度彼此沒有一個共同歷史起源。

但功能上它們都在回答:

When reality is ambiguous, what shall count as what?\boxed{ \text{When reality is ambiguous, what shall count as what?} }

二十二、操作本體模式的統一結構

這些模式可以統一寫成:

ΠO:(O,E,K,V,R)σ.\boxed{ \Pi_O: (O,E,K,V,R) \rightarrow \sigma. }

其中:

  • OO:待判定本體對象;
  • EE:證據;
  • KK:情境;
  • VV:制度價值;
  • RR:風險;
  • σ\sigma:操作狀態。

各種歷史技術只是使用不同:

ΠO.\Pi_O.

二十三、文明本體編譯器

因此本文提出:

Ccivilization={ΠO1,ΠO2,,ΠOn}.\boxed{ \mathfrak C_{\mathrm{civilization}} = \{\Pi_{O_1},\Pi_{O_2},\ldots,\Pi_{O_n}\}. }

稱為:

Civilizational Ontology Compiler

文明本體編譯器

它不是一部真正的中央機器。

而是分散在:

  • 法律;
  • 醫療;
  • 宗教;
  • 行政;
  • 家族;
  • 市場;
  • 國際法;

中的一組操作規則。

其功能是把:

ambiguous world\text{ambiguous world}

轉換成:

actionable institutional states.\text{actionable institutional states}.

二十四、文明為什麼需要本體編譯器?

因為重新計算本體的成本太高。

定義:

Contological recomputation.C_{\mathrm{ontological\ recomputation}}.

如果每一場契約、每一次死亡、每一任國王、每一代大學師生、每一次政府更替都要重新回答:

它到底「真正」是什麼?

則:

Contological recomputation.C_{\mathrm{ontological\ recomputation}} \uparrow\uparrow.

所以制度建立:

  • death certificate;
  • corporate registration;
  • office succession;
  • charter;
  • legal parenthood;
  • capacity ruling;
  • state recognition。

這些其實都是:

ontology caching.\boxed{ \text{ontology caching}. }

二十五、本體快取

本文提出一個新概念:

κO=Ontological Cache.\boxed{ \kappa_O = \text{Ontological Cache}. }

當某狀態已被制度正式判定:

σt,\sigma_t,

後續系統通常不會每一次都重算:

O.O.

而會讀取:

κO=σt.\kappa_O=\sigma_t.

例如:

  • 出生證明;
  • 死亡證明;
  • 公司登記;
  • 身份證件;
  • 法院能力裁定;
  • 婚姻登記;
  • 國籍;
  • 職位證書。

因此制度效率來自:

expensive ontology judgementcheap reusable status.\boxed{ \text{expensive ontology judgement} \rightarrow \text{cheap reusable status}. }

二十六、但快取可能過期

問題是:

κO(t)\kappa_O(t)

可能與:

Oworld(t+1)O_{\mathrm{world}}(t+1)

逐漸脫離。

這就是:

stale ontology cache.\boxed{ \text{stale ontology cache}. }

例如:

  • 舊能力裁定不再符合現況;
  • 技術使舊死亡判準失效;
  • 家庭技術使舊親子規則不足;
  • AI 能力已遠超「普通軟體工具」分類。

因此:

ontology cache requires invalidation rules.\boxed{ \text{ontology cache requires invalidation rules}. }

這直接接回 OOE-I 的 versioning 與 Ontological Governance Debt。


二十七、操作本體技術的共同目的不是求真,而是降低制度摩擦

可以定義:

CG=Cclassification+Ccoordination+Cdispute+Cerror.C_G = C_{\mathrm{classification}} + C_{\mathrm{coordination}} + C_{\mathrm{dispute}} + C_{\mathrm{error}}.

一個好的:

ΠO\Pi_O

可能是在最小化:

CG.\boxed{C_G.}

也就是:

  • 判定不能太貴;
  • 大家要能協調;
  • 爭議不能無限;
  • 誤判成本不能太大。

因此:

ΠO=argminΠOE[CG].\boxed{ \Pi_O^* = \arg\min_{\Pi_O} \mathbb E[C_G]. }

這就是 OOE 的博弈—制度面。


二十八、所以「誰是真正的父母/國家/公司」有時根本問錯層了

很多制度問題不是:

What is X metaphysically?\text{What is X metaphysically?}

而是:

Which relation, right, duty, or continuity must be assigned to whom?\boxed{ \text{Which relation, right, duty, or continuity must be assigned to whom?} }

例如:

「真正的父母是誰?」

可以拆成:

  • genetic relation;
  • gestational relation;
  • custody;
  • financial duty;
  • legal parenthood;
  • social identity。

一旦拆開:

single ontological war\text{single ontological war}

可能變成:

multi-relation allocation problem.\boxed{ \text{multi-relation allocation problem}. }

這是一個極重要的 OOE 方法。


二十九、OOE 不是把所有東西都相對化

本文並不主張:

只要法律說是,它就是。

實際狀態:

OworldO_{\mathrm{world}}

仍然存在。

例如:

  • DNA 關係不會因法院命令改變;
  • 腦活動不會因法律文字改變;
  • 公司實際資產不會只靠名稱產生;
  • AI 實際能力也不會因「tool」標籤消失。

OOE 只是區分:

Oworldσinstitution.\boxed{ O_{\mathrm{world}} \neq \sigma_{\mathrm{institution}}. }

好制度必須持續測量兩者失配:

GO=d(Oworld,σinstitution).G_O = d( O_{\mathrm{world}}, \sigma_{\mathrm{institution}} ).

三十、人類真正做了什麼?

從歷史看,人類反覆做的並不是:

解答本體論。

而是:

reduce ontological ambiguity enough to coordinate action.\boxed{ \text{reduce ontological ambiguity enough to coordinate action}. }

也就是:

Unknown EssenceOperationally Sufficient Classification.\text{Unknown Essence} \rightarrow \text{Operationally Sufficient Classification}.

這可以稱為:

Sufficient Ontology Principle

足夠本體原理


三十一、足夠本體原理

定義:

Q(σ)=operational adequacy.Q(\sigma) = \text{operational adequacy}.

如果:

Q(σ)θKQ(\sigma)\ge\theta_K

足以處理情境 KK,制度不必證明:

σ=Oultimate.\sigma=O_{\mathrm{ultimate}}.

所以:

Operational sufficiency can precede metaphysical completeness.\boxed{ \text{Operational sufficiency can precede metaphysical completeness}. }

這正是腦死、法人、capacity、parenthood 與 state continuity 共同表現出的制度智慧。


三十二、這也解釋法律為什麼特別像本體編譯器

法律的特殊之處在於:

它可以直接把:

σ\sigma

接到:

  • 權利;
  • 義務;
  • 財產;
  • 制裁;
  • 身份;
  • 程序。

因此:

law\boxed{ \text{law} }

可能是人類歷史上規模最大、最成熟的:

Operational Ontology Engine.\boxed{ \text{Operational Ontology Engine}. }

但是醫療、宗教與技術標準同樣具有類似功能。


三十三、歷史上的 OOE 不是完美的

這裡必須加入非常重要的限制。

操作本體技術也可以:

  • 歧視;
  • 固化錯誤分類;
  • 排除某些人;
  • 把制度便利誤當真理;
  • 因權力關係而形成有害邊界。

因此:

Operational success⇏ethical legitimacy.\boxed{ \text{Operational success} \not\Rightarrow \text{ethical legitimacy}. }

歷史技術史不能被寫成:

古人都處理得很好。

而應寫成:

古人早就在處理,但也留下大量分類傷害。


三十四、本體分類本身就是權力

如果:

ΠO\Pi_O

決定:

  • 誰算人;
  • 誰算死;
  • 誰算父母;
  • 誰有能力;
  • 誰算公民;
  • 誰算合法繼承者;

那麼:

ontology compiler=power allocation mechanism.\boxed{ \text{ontology compiler} = \text{power allocation mechanism}. }

所以 OOE 不能只是工程優化。

它必須同時研究:

Who controls the compiler?\boxed{ \text{Who controls the compiler?} }

三十五、因此成熟 OOE 必須加入程序正義

至少需要:

Transparency+Evidence+Appeal+Review+Versioning+Rights Protection.\boxed{ \text{Transparency} + \text{Evidence} + \text{Appeal} + \text{Review} + \text{Versioning} + \text{Rights Protection}. }

否則:

CO\mathcal C_O

可能退化成:

arbitrary classification machine.\text{arbitrary classification machine}.

這是 OOE 與單純 ontology engineering 或 database schema design 的根本不同。


三十六、2026 年真正新的地方

現在可以更精確地回答:

OOE 到底哪裡新?

不是:

humans first operationalize ontology in 2026.\text{humans first operationalize ontology in 2026}.

而是:

multiple historical ontology domains are beginning to converge on programmable agents.\boxed{ \text{multiple historical ontology domains are beginning to converge on programmable agents}. }

AI/BCI 可能同時觸發:

Πidentity,Πagency,Πpersonhood,Πmemory,Πconsent,Πliability,Πcontinuity.\Pi_{\mathrm{identity}}, \Pi_{\mathrm{agency}}, \Pi_{\mathrm{personhood}}, \Pi_{\mathrm{memory}}, \Pi_{\mathrm{consent}}, \Pi_{\mathrm{liability}}, \Pi_{\mathrm{continuity}}.

這些以前是不同制度花幾百年分開處理的問題。

現在可能需要:

machine-readable ontology governance.\boxed{ \text{machine-readable ontology governance}. }

三十七、從歷史 protocol 到 machine-executable protocol

歷史上的 protocol 可以是:

  • 王位繼承法;
  • 教會按立;
  • 死亡證明;
  • 公司註冊;
  • parental order;
  • 國際承認。

未來則可能直接是:

status = ontology_compiler.evaluate(entity, evidence, context)

if status == SAME_ENTITY:
    retain_contracts()
    retain_liabilities()
elif status == SUCCESSOR:
    transfer_selected_rights()
elif status == DISPUTED:
    freeze_irreversible_actions()
    open_review()

因此:

OOE historyOOE software architecture.\boxed{ \text{OOE history} \rightarrow \text{OOE software architecture}. }

三十八、正式命題一:歷史反覆發明命題

若不同文明與制度在缺乏共同理論來源下,反覆產生功能相似的:

ΠO,\Pi_O,

則:

Operational Ontology is a recurrent institutional need.\boxed{ \text{Operational Ontology is a recurrent institutional need}. }

它不是 AI 時代的特殊哲學偏好。


三十九、正式命題二:載體壽命差命題

若:

TX>Tcarrier,T_X > T_{\mathrm{carrier}},

且:

VX>Vmin,V_X > V_{\min},

則制度產生建立載體獨立身份機制的壓力:

P(Πcontinuity).\boxed{ P(\Pi_{\mathrm{continuity}})\uparrow. }

這解釋:

  • 王權;
  • 教會;
  • House;
  • 大學;
  • 公司;
  • 國家。

四十、正式命題三:技術拆分—制度重耦合命題

若科技使原本:

X1X2X_1\approx X_2

變成:

X1X2,X_1\neq X_2,

制度需要建立:

R(X1,X2)\mathcal R(X_1,X_2)

重新分配法律與操作角色。

因此:

Technological DecouplingInstitutional Recoupling Pressure.\boxed{ \text{Technological Decoupling} \rightarrow \text{Institutional Recoupling Pressure}. }

四十一、正式命題四:本體快取命題

當某本體判定的重新計算成本:

CrecomputeC_{\mathrm{recompute}}

很高,而狀態具有一定穩定性時:

P(κO).\boxed{ P(\kappa_O)\uparrow. }

文明會創造證書、登記、身份、法人、職位與國籍,作為可重複利用的本體快取。


四十二、正式命題五:快取失效命題

若:

ΔOworld>ΔκO,\Delta O_{\mathrm{world}} > \Delta \kappa_O,

則:

GO.\boxed{ G_O\uparrow. }

當失配超過門檻:

GO>θG,G_O>\theta_G,

制度必須:

invalidate / revise ontology cache.\boxed{ \text{invalidate / revise ontology cache}. }

四十三、正式命題六:足夠本體命題

在不可懸置情境 KK 中:

Q(σ)θKQ(\sigma)\ge\theta_K

即可支持行動,不要求:

σ=Oultimate.\sigma=O_{\mathrm{ultimate}}.

所以:

Operational Sufficiency⇏Metaphysical Completeness.\boxed{ \text{Operational Sufficiency} \not\Rightarrow \text{Metaphysical Completeness}. }

四十四、正式命題七:權力—編譯器命題

如果:

ΠO\Pi_O

控制重大權利與義務分配,則:

control of ΠO=a form of institutional power.\boxed{ \text{control of }\Pi_O = \text{a form of institutional power}. }

因此本體協議設計必須受到程序性治理。


四十五、OOE-II 的總結模型

可以把整篇壓縮成:

Ambiguous Ontology\boxed{ \text{Ambiguous Ontology} } \downarrow Historical Pressure to Act\boxed{ \text{Historical Pressure to Act} } \downarrow Operational Pattern\boxed{ \text{Operational Pattern} } \downarrow Protocol / Institution\boxed{ \text{Protocol / Institution} } \downarrow Reusable Status / Ontology Cache\boxed{ \text{Reusable Status / Ontology Cache} } \downarrow Rights / Duties / Actions\boxed{ \text{Rights / Duties / Actions} } \downarrow Revision when Reality Changes.\boxed{ \text{Revision when Reality Changes}. }

四十六、下一篇接口:本體編譯器本身怎麼設計?

OOE-II 已經證明:

人類反覆做過這件事。

下一步就不能只停留在歷史分類。

我們需要問:

CO 到底應如何形式化?\boxed{ \mathcal C_O \text{ 到底應如何形式化?} }

也就是:

  • 輸入是什麼?
  • 證據怎麼加權?
  • 不確定性怎麼保留?
  • 誤判成本怎麼進模型?
  • 哪些判定可以暫定?
  • 哪些行動應凍結?
  • 如何版本化?
  • 如何申訴?
  • 誰能修改編譯規則?

因此下一篇為:

《OOE-III:本體編譯器——從模糊世界到可執行制度狀態》


四十七、結論

人類不是到 2026 年才開始處理本體工程。

更準確地說:

文明史的一部分,本來就是不斷把「它到底算什麼」轉換成「我們現在應該怎麼做」。\boxed{ \text{文明史的一部分,本來就是不斷把「它到底算什麼」轉換成「我們現在應該怎麼做」。} }

死亡判準處理:

What counts as dead?\text{What counts as dead?}

法人處理:

What counts as a person for law?\text{What counts as a person for law?}

王權雙身與 corporation sole 處理:

What survives the death of the holder?\text{What survives the death of the holder?}

perpetual succession 處理:

How can an institution outlive every member?\text{How can an institution outlive every member?}

使徒繼承處理:

What makes authority the same authority through time?\text{What makes authority the same authority through time?}

法律父母處理:

Which parent relation matters for which legal function?\text{Which parent relation matters for which legal function?}

capacity 處理:

What counts as enough autonomy for this decision?\text{What counts as enough autonomy for this decision?}

state continuity 處理:

How much political change can occur before a state becomes another state?\text{How much political change can occur before a state becomes another state?}

所以:

OOE is not an invention ex nihilo.\boxed{ \text{OOE is not an invention ex nihilo.} }

它是一個統一名稱,用來描述人類已經反覆實踐、卻分散在不同制度中的古老活動。

2026 年的特殊性只是:

AI and neurotechnology may force these historically separate ontology protocols to converge into explicit, machine-executable governance systems.\boxed{ \text{AI and neurotechnology may force these historically separate ontology protocols to converge into explicit, machine-executable governance systems.} }

而這意味著下一階段不再只是「發現古人也做過」。

真正的工程問題變成:

我們能不能把幾千年分散累積的操作本體智慧,抽象成一套不會把制度暫定答案誤認為宇宙真理的現代本體編譯架構?\boxed{ \text{我們能不能把幾千年分散累積的操作本體智慧,抽象成一套不會把制度暫定答案誤認為宇宙真理的現代本體編譯架構?} }

這就是 OOE-III 的起點。


初版參考文獻與史料接口

  1. A Definition of Irreversible Coma: Report of the Ad Hoc Committee of the Harvard Medical School to Examine the Definition of Brain Death, JAMA, 1968.
  2. Cornell Legal Information Institute, “legal fiction” 與 “legal person”.
  3. Ernst H. Kantorowicz, The King's Two Bodies: A Study in Mediaeval Political Theology.
  4. J. G. Allen, “The Office of the Crown,” Cambridge Law Journal.
  5. University of Oxford, governance history and statutory materials on incorporation and perpetual succession.
  6. Cambridge scholarship on corporation sole and the persistence of ecclesiastical office/property.
  7. Scholarship on apostolic / episcopal succession.
  8. UK Human Fertilisation and Embryology Authority materials on surrogacy and legal parenthood.
  9. UK Mental Capacity Act 2005 and Code of Practice.
  10. James Crawford, The Creation of States in International Law, chapter on identity and continuity.
  11. Matthew C. R. Craven, “The Problem of State Succession and the Identity of States under International Law.”
  12. Yejoon Rim, “State Continuity in the Absence of Government: The Underlying Rationale in International Law.”

版本註記

v0.1 已重新檢索死亡判準、法律擬制、法律人格、王權/office、Oxford perpetual succession、主教 succession、legal parenthood、Mental Capacity 與 state continuity 等資料。

後續 v0.2 應擴充:

  1. 羅馬法 universitas / collegium 的精確法律地位;
  2. corporation sole 的教會法起源;
  3. 王權雙身與現代 Crown office 的演化;
  4. 各國推定死亡制度比較;
  5. 不同法域 assisted reproduction parenthood 的差異;
  6. 精神責任能力/insanity tests;
  7. 公民身份與國籍作為 Ontological Cache;
  8. 奴隸、殖民身份與歷史上「有害本體編譯器」案例;
  9. OOE Pattern Library 的跨文明資料庫。