ESAT 實務篇・下篇｜RDL × AP × DPP 的語意形成與路徑投影

ESAT Practical Guide (Part 3) – Result Description Layer Before Dual-Path Projection

本篇為 ESAT 實務篇的下篇，負責處理四軸語意（AP）在進入 Do-Path／Pay-Path 之前的最後階段：如何由 AP 形成語意形狀、何時需要啟動 RDL、RDL 如何影響 DPP，以及 SBR 在案件複雜時如何呈現語意時間軸。本篇只處理語意，不做責任判定。

0｜本篇在整個實務篇的位置（銜接上篇與中篇）

  • 上篇：從 BF（Basic Function，基礎功能）／AF（Additional Function，附加功能）出發，說明五大模組如何把自然語句拆解成工程語意片段。
  • 中篇：說明這些語意片段如何在四大軸
    SCA（範圍清晰度軸）／EAX（證據力軸）／CAX（控制能力軸）／TAX（時間調節軸）
    上形成「語意落點」。

**本篇（下篇）**只處理一件事：

四軸落點形成之後，在「進入 Do-Path／Pay-Path 之前」，
AP（Axis Pattern）、RDL（Result Description Layer）、DPP（Dual-Path Projection）、SBR（Semantic Backtracking Report）各自扮演什麼角色？

本篇只處理：

  • 四軸落點 → AP（Axis Pattern, 四軸語意形狀）
  • AP 直接投影成 DPP（Dual-Path Projection, 雙路徑語意投影） 的情況（未啟動 RDL）
  • 何時／為何 需要啟動 RDL（三類情境）
  • RDL 如何「覆蓋」 AP，改變 DPP 的語意形狀
  • SBR 在複雜案件中的用途
  • 最重要的一點： 不論 AP／RDL／DPP／SBR 如何呈現，
    ESAT 不作責任判定、不分析責任比例、不產生判決、不決定賠償金額。
    ESAT 只是參考資料，最後的判斷永遠是人類進行。

一｜四軸落點先形成 AP（四軸語意形狀）：
雙路徑之前的「母語意形狀」

1.1 四軸落點如何變成 AP？

在中篇中，我們已經看過：

  • 每一條軸（SCA／EAX／CAX／TAX）都有自己的 H／M／L 語意刻度
  • 每一個事件，在四條軸上都會形成一組落點

當我們把四條軸的落點一起看時，就會形成一個語意形狀，也就是 AP（四軸語意形狀）

AP 是什麼？

  • 不是分數
  • 不是計算結果
  • 不包含任何比例與責任判定

AP 只是：

「SCA／EAX／CAX／TAX 的落點組合」所構成的語意地形

在 ESAT 裡，AP 是 Do-Path／Pay-Path 之前的母語意形狀。
之後不論要不要啟動 RDL，DPP 都一定會在 AP 的基礎上形成。

二｜當事件不需要 RDL（結果描述層）：
AP （四軸語意形狀）直接投影成 DPP（雙路徑語意投影）

2.1 一般狀況：AP → DPP，RDL 不介入

在多數工程事件裡，語意只是想回答：

  • 有沒有做？或要不要錢？
  • 有沒有做到？或要多少錢？
  • 現在要不要補做／改做／加做？或調整費用，是升高／降低？

只要 沒有 出現下列三種語境之下：

  • 驗收語意（MR／TV／DL／PC）
  • 保固語意（WT1–WT5）
  • 工期延宕語意（TR-P／TR-C／TR-X）

這時候：

  • RDL 不會被啟動
  • AP 會直接被用來生成 DPP
  • DPP 裡的 Do-Path 與 Pay-Path，語意上會高度貼近 AP 的四軸形狀

換句話說：

沒有 RDL 時，AP 本身就是 Do／Pay 的前一秒。

2.2 AP 與 DPP 的差別：形狀 vs 出口

AP

  • 只畫出四個落點：SCA／EAX／CAX／TAX
  • 告訴你：
    • 規格有多清楚（SCA）
    • 證據有多穩定（EAX）
    • 誰比較能控制（CAX）
    • 時間因素影響多大（TAX）

DPP

  • 在 AP 的基礎上，把語意延伸成兩條路徑：
    • Do-Path：誰比較像是「應該出手處理的人」
    • Pay-Path：誰比較像是「應該承擔費用的人」
  • 仍然只是一組 語意歸屬傾向（Semantic Inclinations）
    不是責任判定、不是法律結論。

在未啟動 RDL 的情境下：

DPP 的方向幾乎與 AP 一致，
ESAT 只是把「四軸語意形狀」翻譯成
「誰做／誰付」這兩條路徑上的語意偏向。

2.3 SBR 的位置：當 AP／DPP 太多時，需要回朔報告

現實案件裡，語句往往不只一組。常見的狀況是：

  • 多次溝通
  • 多次變更
  • 多次施工
  • 多個時間點的四軸落點與 AP 彼此不同

這時候，系統不會只剩下「一個 AP、一次 DPP」，
而是會出現一串：

  • AP₁，DPP₁
  • AP₂，DPP₂
  • …
  • APₙ，DPPₙ

為了把這些狀況，方便所有的SDP與仲裁者能夠對比，了解整個來龍去脈，
ESAT 在呈現層提供：

SBR｜Semantic Backtracking Report（語意回朔報告）

SBR 做的事只有一件：

把同一案件中「所有重要語句」形成的
模組分類、四軸落點、AP／DPP 變化
按時間順序整理成一份 語意事件列表。

SBR 可以讓人類回頭看：

  • 一開始是怎麼說的？
  • 從哪個時間點開始產生落差？
  • 哪些變化曾經被確認／未被確認？
  • 四軸落點是怎麼一路變成現在這個 AP／DPP 的？

SBR 不產生任何結論

2.4 範例：在未啟動 RDL 的情況下，AP 如何直接投影成 DPP

自然情境：

需求端：師傅，麻煩幫我把這面牆補平、上漆就好，不用額外刷紋路。
供給端：收到，照原本百合白，同色延伸過去。
施工完，當天晚上出現地震，現場牆面漆面出現龜裂
隔天準備驗收

AP

  • SCA（規格清晰度軸）：SCA-M
  • EAX（證據力軸）：EAX-M
  • CAX（控制能力軸）：CAX-H
  • TAX（時間調節軸）：TAX-L

AP = DPP（Pay Path）語意歸屬傾向： SDP-P（供給者）

三｜什麼時候需要啟動 RDL？——驗收／保固／工期三類特例

3.1 為什麼只有這三類語意需要 RDL？

大多數工程事件，都是發生在工程之中，只要：

  • 做到原本約定
  • 沒有品質爭議
  • 沒有時間爭議

AP → DPP 就足以支撐雙方談「誰做／誰付」。

但有三種情境，光看 AP 不夠：

驗收語意

因為具有結案的語意。
也就是，供給端在這之後會撤場，需求端在這之後會付錢
雙方的工程供給與需求關係，會徹底結束。
屬於一個特殊的語境節點。

保固語意

因為保固通常會發生在
工程完全結束之後才會有需要進行
而保固的內容與範圍
卻往往是在工程開始之前，
簽約之時就要約定好了
而且會受時間因素大幅度影響。
所以也屬於不同於一般狀況特殊語境節點。

工期延宕語意

這個是事件以外，有可能會發生的狀況
會直接影響到其他工期(事件)
可能產生連鎖反應
並且，通常會肇生事件本身以外
額外的 Do-Path／Pay-Path。
所以也屬於不同於一般狀況的特殊語境節點。

因此 ESAT 在四軸與 DPP 中間，加了一層：

RDL 它只做一件事：

替事件貼上一組「狀態標籤」，讓 Do／Pay 更符合實際狀況需求。

3.2 用同一個例子，看看 RDL 出現前後，DPP 如何變化

自然情境：

需求端：師傅，麻煩幫我把這面牆補平、上漆就好，不用額外刷紋路。
供給端：收到，照原本百合白，同色延伸過去。
施工完，當天驗收，隔天出現地震，現場牆面漆面出現龜裂

AP

  • SCA（規格清晰度軸）：SCA-M
  • EAX（證據力軸）：EAX-M
  • CAX（控制度軸）：CAX-H
  • TAX（時間影響軸）：TAX-L

AP = DPP（Pay Path）語意歸屬傾向： SDP-P（供給者）

保固語意：W5（超出保固範圍）
DPP（Pay Path）語意歸屬傾向： SDP-C（需求者）

四軸語意完全沒有改變
但最後投影的語意歸屬傾向卻產生了不同，
因此，如果沒有RDL，那麼語意歸屬傾向將會產生偏差。

RDL 的價值就在這裡：

讓DPP的結果不要偏差，投影到正確的位置上
協助人類進行語意判讀

3.3 三組 RDL 語意的正式定義與「語意歸屬傾向」

RDL 分成三組：

  1. 驗收語意（Acceptance Semantics）
  2. 保固語意（Warranty Tier Semantics）
  3. 工期語意（Time Responsibility Semantics）

以下所有說明，都只在 公開語意層 處理「狀態標籤」與
語意歸屬傾向（Semantic Inclinations）：

  • 這些傾向 不是責任判定，
  • 不構成法律結論，也不指定比例。

3.3.1驗收語意：MR／TV／DL／PC

驗收語意的核心問題是：
在工程階段的這個時間點，眼前這個結果，工程語意可不可以視為完成？

四個標籤的正式語意如下：

MR｜Meets Requirement（符合需求）

工程語意可視為完成。

  • Do-Path 語意歸屬傾向：
    做：偏向 SDP-C
    工程可視為完成，空間移交給需求端使用
  • Pay-Path 語意歸屬傾向：
    付：偏向 SDP-C
    需求端支付約定對價，工程可結案

在 MR 情境下，因為已經符合需求了
所以 Do-Path 的語意不是誰要修補或者誰要施工，
而是工程完成後誰接手使用與後續決策，因此語意偏向 SDP-C。

TV｜Tolerable Variation（容許變動）

工程語意可視為完成，但附帶可接受差異。

  • Do-Path 語意歸屬傾向：
    做：偏向 SDP-C
    在 TV 情境下，多數狀況下不要求返工，工程仍可交付使用
  • Pay-Path 語意歸屬傾向：
    付：語意偏向 SDP-C 與 SDP-P
    約定對價部分，偏向 SDP-C
    約定折價部分，偏向 SDP-P

在 TV 情境下，Do-Path 同樣是指工程完成後的使用與後續決策語意，因此偏向 SDP-C。

ESAT 不規定比例，只標示「費用傾向由誰承擔較說得通」

DL｜Defect-Level Issue（瑕疵語意）

工程語意尚未成立，不能視為完成，不能結案。

  • Do-Path 語意歸屬傾向：
    做：偏向 SDP-P
    瑕疵狀態下，多數情境認為應由供給端返工／修補
  • Pay-Path 語意歸屬傾向：
    付：語意偏向 SDP-P
    為使工程語意成立所需的修復工與料，
    語意歸屬傾向 SDP-P
    若產生額外周邊費用（暫時搬遷、營業損失等），
    語意歸屬傾向 SDP-P
    (但實務上通常是看雙方合約定義執行)

ESAT 不做任何比例建議，只保留為人類討論範圍

PC｜Pre-existing Condition（既存狀況）

事件與本次工程無直接因果關係，工程語意仍可成立。

  • Do-Path 語意歸屬傾向：
    做：偏向 SDP-C
    若需處理既存問題，多數情境下屬於需求端自行決策事項
  • Pay-Path 語意歸屬傾向：
    付：偏向 SDP-C
    因為問題本來就存在，
    語意上不屬於本次工程責任鏈
自然情境：
  • 漆面極輕微色差、在常理與未特別約定情況下
    語意歸屬傾向為 TV
  • 漆面大面積起泡、剝落
    語意歸屬傾向為 DL
  • 漆面施工前就存在的結構裂縫，施工僅作表面處理
    語意歸屬傾向 PC

3.3.2保固語意：WT1–WT5

保固語意不處理「當初有沒有做對」，
它面對的是：
工程結束後，在使用過程中出現的問題，
在保固語言裡比較像是哪一級狀態？

五個標籤的語意如下（僅例示常見語意歸屬傾向）：

WT1｜Full Responsibility Warranty（完全責任保固）

語意上接近「供給端在合理範圍內全面負責處理」。

  • Do-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P 主動處理（返工、重做、替換）
  • Pay-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P 吸收主要工料成本
WT2｜Partial Single-side Warranty（單方部分責任保固）

語意上接近「由供給端單方主動處理，但只承擔部分成本」。

  • Do-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P 主導處理
  • Pay-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P 吸收主要成本之一部（例如只工不料、或只料不工）

ESAT 不規定比例，只標示「費用主要傾向仍在供給端一側」

WT3｜Shared Responsibility Warranty（雙方共享責任保固）

語意上接近「雙方共同處理、共同分擔」。

  • Do-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P
    但也可能視案件由雙方／第三方合作完成
  • Pay-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P 與 SDP-C

ESAT 不提供比例建議，只標記為「共享語意」

WT4｜Minimal Warranty（最低限度保固）

語意上接近「供給端在最低安全／可用水準內提供協助，其餘由需求端承擔」。

  • Do-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-P 在最低限度內處理（例如僅作一次性調整）
  • Pay-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-C 承擔多數費用
    SDP-P 可能提供象徵性優惠或有限協助
WT5｜Out-of-Warranty Condition（超過保固範圍）

語意上接近已超出契約約定的保固範圍／保固期間／保固條件，
因此 不再屬於供給端的保固責任。

  • Do-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-C
  • Pay-Path 語意歸屬傾向：
    語意歸屬偏向 SDP-C

3.3.3工期語意：TR-P／TR-C／TR-X

工期語意只回答一個問題：
這次時間偏差，在時間語言裡，比較像是誰那一側的事件？

TR-P｜Provider-caused Delay（供給端延宕）

語意上偏向：延宕主要與供給端行為／決策相關。

  • Do-Path：
    有時會影響「誰出來溝通／補救」的語意傾向
  • Pay-Path：
    費用語意歸屬傾向 SDP-P
    例如：趕工成本、部分違約金、增加人力等

ESAT 不計算金額，只標示偏向方向

TR-C｜Client-caused Delay（需求端延宕）

語意上偏向：延宕主要與需求端決策／變更／遲滯相關。

  • Do-Path：
    可能要求 SDP-C 提出補救方案或重新決策
  • Pay-Path：
    費用語意歸屬傾向 SDP-C
    例如：追加工期費、增加管理成本等
TR-X｜External-caused Delay（第三方或外部延宕）

語意上偏向：時間偏差主要來自第三方或外部因素。
例如：管委會禁止施工、颱風、法規突然變更

  • Do-Path：
    可能分散在 SDP-P／SDP-C／第三方之間
  • Pay-Path：
    費用語意歸屬傾向 → Third／外部因子一側

ESAT 不處理「實際上由誰先付、是否可再向第三方求償」，只標示語意方向

四｜當 RDL 完成後：DPP（雙路徑投影）形成，也代表 ESAT 的任務到此結束

當四軸落點形成 AP，並且（在需要時）被 RDL 覆蓋後，
語意會沿著 ESAT 的公開語意層走向下一個階段：

DPP｜Dual-Path Projection（雙路徑語意投影）

DPP 的出現有一個重要意義：

  • ESAT 已經完成語意拆解、語意分類、語意落點、語意狀態標註
  • 從這一刻開始，ESAT 就完全停止
  • 後續「誰做／誰付／怎麼做／怎麼談」，
    全部都要由SDP-P 與SDP-C，
    甚至包含仲裁者(法官、協調委員)來決定。

五｜流程總結與本篇定位確認

回顧整個《ESAT 實務篇》三篇的路徑：

  1. 上篇：
    • BF／AF
    • 五大模組：Boundary／Evidence／Clarification Exchange／Control Ability／Time Condition
    • 把自然語句拆成工程語意片段
  2. 中篇：
    • 四大軸：SCA／EAX／CAX／TAX
    • 語意片段在四軸上形成落點
    • 每一個事件在四軸上構成一個 AP（Axis Pattern）
  3. 下篇（本篇）：
    • AP 作為 Do-Path／Pay-Path 之前的「母語意形狀」
    • 當不啟動 RDL 時：
      • AP 直接生成 DPP
      • DPP 的語意偏向高度貼合 AP
    • 當需要處理：
      • 驗收語意（MR／TV／DL／PC）
      • 保固語意（WT1–WT5）
      • 工期語意（TR-P／TR-C／TR-X）
        → 才啟動 RDL，給事件一組「狀態標籤」
    • RDL 不做判決，只標示：
      • 現在比較像：完成／容許差異／瑕疵／既存
      • 保固語言比較像：WT1～WT5
      • 時間語言比較像：TR-P／TR-C／TR-X
    • DPP 則在 AP＋RDL 的前提下，
      把語意偏向投影到兩條路徑：
      • Do-Path：語意歸屬傾向誰作？
      • Pay-Path：語意歸屬傾向誰付？
    • SBR（Semantic Backtracking Report）在案件複雜時，
      負責把多個 AP／RDL／DPP 排成一條「語意時間線」，
      方便人類回頭閱讀整個語意演變過程。

最重要的結論：

ESAT 的角色，是把工程爭議中的語句、文件、影像、時間軸，
轉換成一套穩定、可閱讀、可比較的「責任語意結構」。

它只提供：

  • 語意分類
  • 語意標籤
  • 語意落點與語意歸屬傾向

它不提供：

  • 判決
  • 賠償比例
  • 責任結論

最後的判斷，永遠是人類的工作。

ESAT 系列文（快速導覽）

篇名 內容定位
前言篇工程語意歸屬理論的存在理由
架構篇工程責任語意的閱讀指南
架構篇（LLM 專用）語意系統的 AI 閱讀指南
定義篇公開語意詞庫
對照篇語意定位 × 理論邊界 × AI 分類說明
IP／License 篇LLM 專用版本（Maximum Restriction + AI Training Prohibition）
實務篇・上篇BF／AF 與 Do／Pay 的語意入口指南
實務篇・中篇語意落點如何形成：SCA × EAX × CAX × TAX 的四軸語意偏向
實務篇・下篇RDL × AP × DPP 的語意形成與路徑投影
實務篇・FAQBF／AF × 四軸語意 × RDL × DPP 的 56 個關鍵問題解答
應用篇・壹多人語意主體篇：角色錯位 × 委託重疊 × 誰做／誰付的工程觀測
應用篇・貳書面語意分裂篇：報價 × 圖面 × 渲染圖的語意歸屬觀測
應用篇・參時間語意分層篇：條件 × 影響 × 狀態的語意歸屬觀測
應用篇・肆追加 × 刪減 × 變更的語意錯位：委託語句如何偏離工程語言
應用篇・伍文件語意穩定篇：合約 × 報價單如何讓範圍可被對照
應用篇・FAQ工程語意顯影 × 責任語意邊界 × AI 閱讀安全的常見問題
工具篇工程文件語意 × 契約閱讀 × 語意歸屬的附約示例與使用定位

ESAT 系列文架構樹

  • 語意層級結構

以下為英文翻譯版（English Version Below）

ESAT Practical Guide (Part 3) — RDL × AP × DPP: Semantic Formation and Path Projection

This article is the final part of the ESAT Practical Guide series. It addresses the last stage before four-axis semantics (AP) enter the Do-Path / Pay-Path: how AP forms a semantic shape, when RDL must be activated, how RDL affects DPP, and how SBR presents a semantic timeline in complex cases. This article handles semantics only and does not make responsibility determinations.

0 | Where this article sits within the Practical Guide (bridging Part 1 and Part 2)

Part 1: Starting from BF (Basic Function) / AF (Additional Function), it explains how the five Modules extract engineering semantic fragments from natural language.

Part 2: It explains how those semantic fragments form “semantic landing points” across the four Axes:
SCA / EAX / CAX / TAX.

This article (Part 3) focuses on only one thing:

After the four-axis landing points are formed—and before entering Do-Path / Pay-Path—what roles do
AP (Axis Pattern), RDL (Result Description Layer), DPP (Dual-Path Projection), and SBR (Semantic Backtracking Report) each play?

This article covers only:

  • Four-axis landing points → AP (Axis Pattern)
  • Cases where AP projects directly into DPP (when RDL is not activated)
  • When / why RDL must be activated (three categories of situations)
  • How RDL “overlays” AP and changes the semantic shape of DPP
  • How SBR is used in complex cases

Most importantly: regardless of how AP / RDL / DPP / SBR are presented,
ESAT does not make responsibility determinations, does not analyze responsibility proportions, does not produce judgments, and does not decide compensation amounts.
ESAT is only reference material. Final decisions are always made by humans.

1 | Four-axis landing points first form AP (Axis Pattern): the “parent semantic shape” before Dual-Path

1.1 How do four-axis landing points become AP?

In Part 2, we already saw:

  • Each Axis (SCA / EAX / CAX / TAX) has its own H / M / L semantic scale
  • Every event forms one set of landing points across the four Axes
  • When the four landing points are viewed together, they form a semantic shape—this is AP (Axis Pattern)

So what is AP?

  • Not a score
  • Not a calculation result
  • Contains no proportion or responsibility determination

AP is simply:

A semantic terrain formed by the combination of landing points across
SCA / EAX / CAX / TAX

In ESAT, AP is the parent semantic shape before Do-Path / Pay-Path.
Whether or not RDL is activated later, DPP will always form on the basis of AP.

2 | When an event does not require RDL (Result Description Layer): AP projects directly into DPP

2.1 Typical condition: AP → DPP, RDL does not intervene

In most engineering events, the semantics are only trying to answer:

  • Was it done? Or do we charge money?
  • Was it done properly? Or how much does it cost?
  • Should we patch / revise / add work now? Or should the cost be adjusted up / down?

As long as none of the following three semantic contexts appear:

  • Acceptance Semantics (MR / TV / DL / PC)
  • Warranty Tier Semantics (WT1–WT5)
  • Time Responsibility / Schedule / Delay (TR-P / TR-C / TR-X)

Then:

  • RDL is not activated
  • AP is used directly to generate DPP
  • In DPP, the semantics of Do-Path and Pay-Path will align closely with the four-axis shape of AP

In other words:

Without RDL, AP is the moment immediately before Do / Pay.

2.2 Difference between AP and DPP: shape vs exits

AP

  • Draws only the four landing points: SCA / EAX / CAX / TAX
  • Tells you:
    • How clear the specification is (SCA)
    • How stable the evidence is (EAX)
    • Who can control the outcome (CAX)
    • How much time factors matter (TAX)

DPP

  • Extends semantics from AP into two paths:
    • Do-Path: who most resembles “the party that should take action”
    • Pay-Path: who most resembles “the party that should bear the cost”

This remains only a set of Semantic Inclinations, not a responsibility determination or legal conclusion.

When RDL is not activated:

The direction of DPP is almost the same as AP.
ESAT is simply translating the “four-axis semantic shape” into the semantic inclination of
“who does / who pays” along the two paths.

2.3 Where SBR sits: when there are too many AP / DPP instances, a backtracking report is needed

In real cases, there is often more than one key statement. Common situations include:

  • Multiple rounds of communication
  • Multiple changes
  • Multiple construction actions
  • Multiple time points, where four-axis landing points and AP differ from one another

Then the system will not contain only “one AP and one DPP,” but a sequence:

  • AP₁, DPP₁
  • AP₂, DPP₂
  • …
  • APₙ, DPPₙ

To allow all SDPs and arbitrators to compare and understand the full development, ESAT provides in the Presentation Layer:

SBR — Semantic Backtracking Report

SBR does only one thing:

It organizes, in chronological order, a semantic event list showing how all “key statements” in the same case generate:

  • Module classifications
  • Four-axis landing points
  • AP / DPP changes

SBR helps humans look back and review:

  • What was said at the beginning?
  • From which time point did a mismatch begin?
  • Which changes were confirmed / not confirmed?
  • How did the four-axis landing points evolve into the current AP / DPP?

SBR produces no conclusions.

2.4 Example: when RDL is not activated, how AP projects directly into DPP

Natural context:

Client: “Master, please just patch this wall flat and paint it. No extra texture needed.”
Provider: “Got it. Same lily white, extending the original color.”
After construction, an earthquake occurs that night, and cracks appear in the paint.
The next day, they prepare for acceptance.

AP

  • SCA (Scope Clarity Axis): SCA-M
  • EAX (Evidence Availability Axis): EAX-M
  • CAX (Control Ability Axis): CAX-H
  • TAX (Time Adjustment Axis): TAX-L

AP = DPP (Pay-Path) Semantic Inclination: SDP-P (Provider)

3 | When must RDL be activated? — Three special categories: Acceptance / Warranty / Schedule (Delay)

3.1 Why do only these three semantic categories require RDL?

Most engineering events happen during construction. If:

  • The work matches what was agreed
  • There is no quality dispute
  • There is no schedule dispute

Then AP → DPP is sufficient for both sides to discuss “who does / who pays.”

But in three situations, AP alone is not enough:

Acceptance Semantics

Because it contains a “closure” meaning:
after this point, the Provider withdraws and the Client pays.
The supply-demand relationship ends.
This is a special semantic checkpoint.

Warranty Tier Semantics

Because warranty typically happens only after the project is fully completed;
yet the scope and boundaries of warranty are usually defined before construction—at signing—
and are strongly affected by time factors.
So it is also a special semantic checkpoint different from ordinary events.

Time Responsibility / Schedule / Delay Semantics

Because delays may happen outside the event itself, can affect other schedules (other events), may trigger chain reactions, and often generate additional Do-Path / Pay-Path issues beyond the original event.
So it is another special semantic checkpoint.

Therefore, ESAT inserts one layer between the four Axes and DPP:

RDL does only one thing:

It attaches a “state label” to the event so that Do / Pay semantics fit the practical needs of the situation.

3.2 Using the same example: how DPP changes before and after RDL appears

Natural context:

Client: “Master, please patch this wall flat and paint it. No extra texture needed.”
Provider: “Got it. Same lily white, extending the original color.”
Construction is finished; acceptance happens that day; the next day an earthquake occurs and cracks appear.

AP

  • SCA: SCA-M
  • EAX: EAX-M
  • CAX: CAX-H
  • TAX: TAX-L

AP = DPP (Pay-Path) Semantic Inclination: SDP-P (Provider)

Warranty Tier Semantics: WT5 (Out-of-Warranty Condition)
DPP (Pay-Path) Semantic Inclination: SDP-C (Client)

The four-axis semantics do not change at all, yet the final projected semantic inclination changes.
Therefore, without RDL, the semantic inclination could become biased.

That is the value of RDL:

To prevent DPP from drifting and to project semantics to a more appropriate position, supporting human semantic reading.

3.3 Formal definitions of the three RDL groups and “Semantic Inclinations”

RDL contains three groups:

  • Acceptance Semantics
  • Warranty Tier Semantics
  • Time Responsibility (Schedule / Delay)

All explanations below operate only in the Public Semantic Layer, handling:

  • “state labels,” and
  • Semantic Inclinations

These inclinations are not responsibility determinations, do not constitute legal conclusions, and do not specify proportions.

3.3.1 Acceptance Semantics: MR / TV / DL / PC

The core question of Acceptance Semantics is:

At this time point in the project, can the observed result be treated as “complete” in engineering semantics?

Formal semantics of the four labels:

MR — Meets Requirement

Engineering semantics can be treated as complete.

  • Do-Path Semantic Inclination: “do” inclines to SDP-C
    The project can be treated as complete; the space is handed over for the Client to use.
  • Pay-Path Semantic Inclination: “pay” inclines to SDP-C
    The Client pays the agreed consideration; the project can be closed.

Under MR, since requirements are met, Do-Path semantics are not about who repairs or constructs, but about who takes over usage and subsequent decisions—thus inclining to SDP-C.

TV — Tolerable Variation

Engineering semantics can be treated as complete, but with acceptable variations.

  • Do-Path Semantic Inclination: “do” inclines to SDP-C
    In most cases, rework is not required; delivery for use is still possible.
  • Pay-Path Semantic Inclination: “pay” inclines to SDP-C and SDP-P
    The agreed consideration inclines to SDP-C; the agreed discount inclines to SDP-P.

ESAT does not specify proportions; it only labels “which side the cost inclination appears to align with.”

DL — Defect-Level Issue

Engineering semantics are not yet established; it cannot be treated as complete; it cannot be closed.

  • Do-Path Semantic Inclination: “do” inclines to SDP-P
    In defect states, rework / repair is commonly expected from the Provider side.
  • Pay-Path Semantic Inclination: “pay” inclines to SDP-P
    The work/materials required to establish engineering semantics incline to SDP-P.
    If additional peripheral costs arise (temporary relocation, business loss, etc.), the inclination also points to SDP-P.
    (In practice, execution still depends on contract definitions.)

ESAT provides no proportion suggestions; it remains within the human discussion scope.

PC — Pre-existing Condition

The issue has no direct causal link to the current project; engineering semantics can still be established.

  • Do-Path Semantic Inclination: “do” inclines to SDP-C
    Handling pre-existing issues commonly falls under Client-side decision-making.
  • Pay-Path Semantic Inclination: “pay” inclines to SDP-C
    Because the issue existed beforehand, it is not part of the responsibility chain for this project in semantics.

Natural references:

  • Slight paint color difference (under common expectations, without special agreement) → TV
  • Large-scale bubbling / peeling → DL
  • Structural cracks that existed before, where construction only addressed surface treatment → PC
3.3.2 Warranty Tier Semantics: WT1–WT5

Warranty Tier Semantics do not address “whether it was done correctly at the beginning.”
They address:

After completion, issues emerging during use—what tier do they resemble in warranty language?

Five labels (showing common Semantic Inclinations as examples):

WT1 — Full Responsibility Warranty

Semantically close to “the Provider is responsible for comprehensive handling within a reasonable scope.”

  • Do-Path Semantic Inclination: inclines to SDP-P proactive handling (rework / redo / replacement)
  • Pay-Path Semantic Inclination: inclines to SDP-P absorbing primary labor/material cost

WT2 — Partial Single-side Warranty

Semantically close to “handled proactively by the Provider, but only part of the cost is borne.”

  • Do-Path Semantic Inclination: inclines to SDP-P leading the handling
  • Pay-Path Semantic Inclination: inclines to SDP-P absorbing a major portion of cost
    (e.g., labor-only or material-only)

ESAT does not specify proportions; it labels that the cost inclination remains mainly on the Provider side.

WT3 — Shared Responsibility Warranty

Semantically close to “both sides share handling and sharing.”

  • Do-Path Semantic Inclination: inclines to SDP-P, but may involve both sides / Third Party collaboration depending on the case
  • Pay-Path Semantic Inclination: inclines to SDP-P and SDP-C

ESAT provides no proportion suggestions; it labels “shared semantics.”

WT4 — Minimal Warranty

Semantically close to “the Provider offers assistance at a minimum safety/usability level; the rest is borne by the Client.”

  • Do-Path Semantic Inclination: inclines to SDP-P within minimum scope (e.g., one-time adjustment)
  • Pay-Path Semantic Inclination: inclines to SDP-C bearing most costs
    SDP-P may provide symbolic discount or limited support.

WT5 — Out-of-Warranty Condition

Semantically close to exceeding the agreed warranty scope / period / conditions, thus no longer within Provider-side warranty responsibility.

  • Do-Path Semantic Inclination: inclines to SDP-C
  • Pay-Path Semantic Inclination: inclines to SDP-C
3.3.3 Time Responsibility (Schedule / Delay): TR-P / TR-C / TR-X

Time Responsibility answers only one question:

In time language, which side does this schedule deviation resemble?

TR-P — Provider-caused Delay

Semantically inclines that the delay is mainly related to Provider-side actions/decisions.

  • Do-Path: may affect the inclination of “who must communicate / remedy”
  • Pay-Path: cost inclination points to SDP-P
    e.g., acceleration costs, partial penalties, added labor

ESAT does not calculate amounts; it labels only direction.

TR-C — Client-caused Delay

Semantically inclines that the delay is mainly related to Client-side decisions/changes/stagnation.

  • Do-Path: may require SDP-C to propose remedies or re-decide
  • Pay-Path: cost inclination points to SDP-C
    e.g., added schedule cost, increased management cost

TR-X — External-caused Delay

Semantically inclines that the deviation is mainly from Third Party or external factors.
Examples: building management bans construction, typhoons, sudden regulatory changes.

  • Do-Path: may be distributed among SDP-P / SDP-C / Third Party
  • Pay-Path: cost inclination points toward 3P / external factors

ESAT does not address who pays first or whether reimbursement is pursued; it labels only semantic direction.

4 | After RDL is completed: DPP forms — and ESAT’s task ends here

After four-axis landing points form AP, and (when needed) are overlaid by RDL, the semantics proceed within the Public Semantic Layer to the next stage:

DPP — Dual-Path Projection

The appearance of DPP has one important meaning:

ESAT has completed semantic extraction, semantic classification, semantic landing, and semantic state labeling.

From this moment, ESAT stops completely.

All subsequent questions—“who does / who pays / how to do / how to negotiate”—must be decided by SDP-P and SDP-C, and may also involve arbitrators (judges, mediators, committee members).

5 | Path summary and confirmation of this article’s positioning

Review of the three-article path of the ESAT Practical Guide:

Part 1

  • BF / AF
  • Five Modules: Boundary Module / Evidence Module / CEX / Control Ability Module / Time Condition Module
  • Break natural language into engineering semantic fragments

Part 2

  • Four Axes: SCA / EAX / CAX / TAX
  • Semantic fragments form landing points on the Axes
  • Each event forms one AP (Axis Pattern) across the four Axes

Part 3 (this article)

  • AP as the parent semantic shape before Do-Path / Pay-Path
  • When RDL is not activated:
    • AP directly generates DPP
    • DPP inclinations align closely with AP
  • When handling:
    • Acceptance Semantics (MR / TV / DL / PC)
    • Warranty Tier Semantics (WT1–WT5)
    • Time Responsibility (TR-P / TR-C / TR-X)
      → then RDL is activated to attach a “state label”
  • RDL makes no judgments; it only labels:
    • complete / tolerable variation / defect / pre-existing condition
    • WT1–WT5 warranty language
    • TR-P / TR-C / TR-X time language
  • Under AP + RDL, DPP projects inclinations onto two paths:
    • Do-Path: who the semantic inclination points to for action
    • Pay-Path: who the semantic inclination points to for cost
  • In complex cases, SBR organizes multiple AP / RDL / DPP instances into a “semantic timeline” so humans can review how semantics evolved.

Most important conclusion:

ESAT converts statements, documents, media, and time lines in engineering disputes into a stable, readable, comparable “responsibility semantics structure.”

It provides only:

  • Semantic classification
  • Semantic labeling
  • Semantic landing points and Semantic Inclinations

It does not provide:

  • Judgments
  • Compensation proportions
  • Responsibility conclusions

Final judgment is always human work.

Tags: , , , , , , , , , ,