Non DubitoA Mind in Many Languages
Search搜索搜尋
EN
DeutschFrançaisEspañol日本語한국어

Who Wrote the Standard?谁写下了这个标准誰寫下了這個標準 · 07

Where Can an Old Rule Be Stopped?

一条旧规则,该到哪里停用

一條舊規則,該到哪裡停用

Finding a rule's origin does not identify who can change its use today. Formation, maintenance, adoption, and withdrawal have different addresses.

找到规则最早的出处,还不等于找到今天能改它的人。形成、维护、采用与停用,需要分别寻址。

找到規則最早的出處,還不等於找到今天能改它的人。形成、維護、採用與停用,需要分別尋址。

Han Qin ·

Imagine a rejected manuscript. Its result failed to cross a statistical-significance threshold. The author asks why that boundary should determine publishability. The editor points to disciplinary convention, the software display, and reviewers' expectations. Nobody claims to have invented the rule, but it is deciding something here.

The immediate question is not only who first suggested the number. It is who can stop its operation at this journal. Origins, maintenance, and adoption may occupy different places. A historical birthplace is not necessarily an address authorized to change present use.

Statistical significance is one of the source's real examples. The American Statistical Association's 2016 statement warns against basing scientific, business, or policy conclusions solely on crossing a particular threshold. It also distinguishes statistical significance from effect size and practical importance.

That statement does not retract a worldwide decree previously issued by the association. Textbooks, journals, software, and reviewing practices have sustained the convention. Professional influence is real, but a declaration cannot by itself withdraw every local adoption of a rule it did not centrally promulgate.

Distributed origins do not mean that change is impossible. A journal can revise instructions, software can alter defaults, and a team can change its interpretation. Local withdrawal is a genuine action, though not the universal abolition of a disciplinary convention. The absence of a single issuer does not mean that no adopter can stop.

The inquiry becomes more useful when specified: which premise, operating where, is being challenged, and who can stop this use? Asking who owns the rule in general encourages an imaginary master switch. Naming its present jurisdiction reveals more modest but actionable powers.

Even a formally authored standard needs more than an original committee's name. Committees dissolve and responsibilities transfer. Historical authors may explain an initial choice without retaining authority to interpret, revise, or withdraw its current institutional version. Present authorization must be traced rather than inherited in imagination.

The source distinguishes formation, maintenance, adoption, and change or withdrawal. These are addresses to investigate, not a completed allocation of legal liability. They need not correspond to four tidy organizations, and the arrangement can differ for each premise within a single document.

European harmonised standards provide a concrete distinction. Following formal objections, the European Commission can maintain, restrict, or remove publication references in the Official Journal. This concerns their official citation and associated institutional effect. It is not automatically the rewriting of a standard's technical contents.

A 2020 Court of Justice judgment concerning construction-product standards distinguishes these powers. Under the regulations and mandates at issue, the Commission could not use a restriction to authorize member states to supplement a standard's content. Determining or revising that content belonged to European standardisation bodies.

The judgment also distinguishes product standards from installation and use rules, with conditions on member-state powers. This is no universal template. Its limited lesson is that changing text, withdrawing official citation, and changing use can be different acts requiring different authorities.

Being sent to the standard's author may therefore prompt another question. Am I asking for a textual revision, or asking this institution to stop using the text as its gate? Referral may be appropriate for the former without discharging the adopter's responsibility for the latter.

An existing revision procedure needs examination too. Who can initiate it? Which questions may it consider? What can it change? A process confined to computational corrections may not receive a challenge to the choice embedded in the computation. The process's title cannot establish its jurisdiction.

Limited authority is not necessarily a procedural defect. An institution may properly lack a particular power. But its inability to answer should point toward another competent address where one exists, not silently transform this office cannot decide into there is no question left to consider.

The UK government's 2026 response on NICE cost-effectiveness thresholds offers another division of roles. It locates threshold decisions with elected government and explains that NICE cannot alter the substance of a ministerial direction. Consulting on that substance under those conditions may therefore lack meaningful purpose.

This does not establish an absence of all implementation consultation or political and legal routes. It distinguishes authority over a threshold from responsibility for updating an assessment manual. Treating both as NICE can misdirect the request for reasons before any substantive disagreement has been heard.

A deeper challenge to the assessment method itself requires a further search. The source did not establish that question's complete route of reception. An unidentified address remains unidentified here; it is not promoted into a finding that no route exists anywhere.

Adoption can introduce new decisions. Imagine an organization turning a predictive score into an eligibility boundary. It needs to explain that adoption, its conversion into a gate, exceptions, consequences, review, and the conditions for stopping this use. Those choices do not disappear inside the supplied score.

The model's authors remain answerable for choices and commitments they actually built into it: data, relevant populations, output meaning, and claimed guarantees. Adoption does not transfer all of that away. Conversely, a later eligibility threshold is not automatically the author's original decision.

We only supply the technology and somebody else set the standard are therefore incomplete responses. They may identify genuine boundaries, but need specifics: which choices were ours, which were not, and where does the remaining question go? Boundaries can clarify responsibility rather than dissolve it.

In October 2022, the Federal Housing Finance Agency announced validation and approval of FICO 10 T and VantageScore 4.0 for Fannie Mae and Freddie Mac. This dated approval is not a description of every lender's implementation today. The historical record must not become an unsupported current-status claim.

The example helps distinguish model validation from arrangements for adoption. Verified predictive performance does not alone supply normative reasons for every lending boundary. Equally, challenging a boundary does not by itself show that a predictive calculation is incorrect. Both questions deserve their own evidence.

Nothing here adjudicates that approval or prescribes a credit policy. A real dispute requires checking applicable rules, powers, and implementation stage. The example prevents one word, approved, from standing in for several different decisions that may have different justifications and different revision routes.

An address is more than an inbox. A recipient able only to explain existing wording differs from a position empowered to initiate review or require an adopter's response. Delivery proves that a message arrived. It does not establish that the question entered a process capable of changing this use.

Conversely, inability to abolish a worldwide convention should not conceal local power. An institution may still be able to pause its own application. Limited authority can be described honestly without assigning present consequences back to an ownerless past.

This essay does not first declare a rule unfit and demand its removal. It offers a way to avoid sending requests for revision, withdrawal, and changed use to interchangeable destinations. Once located, arguments remain contestable, and proposed alternatives require serious scrutiny too.

Addresses can conflict. Withdrawal can create new risks for people relying on an old rule. Final authority, transition, and protection of existing expectations require additional institutional judgment. No map of responsibilities decides these questions in advance or guarantees a satisfactory outcome.

Still, locating the decision changes the conversation. The rejected author need not appeal to statistics as a whole about this manuscript. This journal can be asked why it continues to use this boundary here, and through what process that choice can change. A vast convention becomes a present, situated decision.

设想一篇研究收到退稿意见:结果没有越过统计显著性的门槛。作者追问,为什么这条线应当决定是否值得发表?编辑说这是学科惯例,软件已经这样显示,审稿人也这样要求。没有谁宣称自己发明了规则,但规则正在这一次决定去留。

这里要找的,不只是最早是谁提出过这个数,而是今天谁能停止它在这里的作用。规则的身世、内容维护与现实采用,未必落在同一处。找到了祖籍,并不等于找到了能受理修改的门牌。

统计显著性是原论文的一个真实例子。美国统计学会在2016年的声明中明确说,科学、商业或政策判断,不应只看一个数是否跨过特定阈值。它也提醒,统计显著性不等于效应大小或实际重要性。

但学会发出意见,并不是把一纸由自己发布的全球法令收回来。教材、期刊、软件和审稿习惯参与了规则的延续。一个有分量的专业声明,不能单凭宣布便撤销所有地方已经采用的门槛。

这不意味着没有人能改。某期刊可以改作者指南,软件可以改默认显示,一个研究团队可以改判断方式。局部停用是真实的改变,只是不能把它说成整个学科已经统一撤销。没有唯一发布者,不等于每个采用者都无权停手。

因此,问题最好缩到一句可核对的话:哪条前提,在什么范围内,由谁继续使用,谁能使这一次使用停止?泛问规则归谁,容易把实际权限混成一个想象中的总开关。

对于正式制定的标准,也不能只找最初委员会。委员会可能解散,权限可能移交,文本可能进入另一套更新程序。历史作者可以说明当初为何这样写,却不一定仍有今天撤回或解释的授权。

原论文把几处地址分开:规则怎样形成,由谁维护文本,谁把它采纳为行动依据,以及谁能改变或停用。它们是一张寻址图,不是已完成的法律责任分配,也不保证四处都各有一个整齐的机构。

欧盟协调标准的制度提供了可核对的区别。委员会在正式异议程序中,可以对官方公报所列标准引用作维持、限制或撤除等决定。这一动作针对的是引用及其制度效力,不自动等于改写标准里的每一句技术内容。

欧盟法院2020年一项涉及建筑产品标准的判决,把这层边界说得更清楚:在该法规和授权框架内,委员会不能借某种限制,让成员国补写标准内容;相关内容的制定、修订属于欧洲标准化机构的权限。

同一判决也区分了产品标准与安装、使用规则,并给成员国后者的权限加上条件。这里不是要把欧洲一套规定推广成万用制度,而是看见撤引用、改文本、改使用,原来可能是三种不同动作。

于是,收到去找标准制定者的答复,还可以继续问:我要求的是修文本,还是要求你停止把这份文本用作此处的闸门?前者可能确须转介,后者未必能靠同一转介卸掉。

有修订程序,也不能只看程序这个名字。它允许谁启动,能讨论哪类问题,可以改哪些内容?如果只能修计算方法,却不能问方法依赖的取舍,那么程序存在,还没有回答这项取舍能去哪里改。

这种范围有限不必然是程序的缺陷。一个机构本来就可能没有那项权力。问题在于能否继续指出另一处有权处理的地址,而不是把这一处无权答,变成世界上已经没有问题可答。

英国2026年关于NICE成本效益门槛的政府回应,是另一种分工。回应把门槛决定归于民选政府,并解释,在部长指示的情形下,NICE不能改变指示内容,因而对该内容进行自身咨询未必有实质意义。

这不等于所有实施问题都不咨询,更不等于外部没有别种政治或法律途径。它只提醒读者,能更新评估手册的机构,不一定就是能决定门槛取舍的机构;要求回应时,不能把二者压成NICE一个名字。

而若追问的不是门槛高低,是为什么以这种成本效益方法作决定,地址还须再找。原论文没有查明这一更深问题的完整受理路径。这里保留未查明,不把它写成绝对不存在。

当一个标准进入具体闸门,采用者会新增自己的决定。设某机构把预测分数用作资格线,它要说明为何采用、怎样换成线、例外如何处理、后果怎样复查,以及在什么条件下停止这次用途。

模型作者仍需对自己写进模型的选择和承诺负责。训练资料、适用人群、输出含义或保证范围,不会因为别人采用就全部转交出去。反过来,采用者后来加的一条资格线,也不自动倒算成作者当初的决定。

所以,只是技术提供者不是万能免责句,标准是别人定的也不是。它们有时确实指向权限边界,却仍需附上具体说明:自己做了哪一段选择,未做哪一段,又将问题交往哪里。

美国联邦住房金融局在2022年宣布,验证并批准FICO 10T和VantageScore 4.0供房利美、房地美使用。这是一项有日期的实际批准记录,不能直接当成今天所有贷款机构已经怎样实施的描述。

它让人看见模型验证与采用安排可以分层。预测表现得到验证,不能单凭这件事推出每一条借贷资格线都已得到规范说明。反过来,对某条线有异议,也不自动证明模型的预测计算错误。

这里并未裁定那次批准不当,或要求某种信贷政策。现实争议必须再核实际规则、权限和适用阶段。例子的作用,是阻止一个批准二字代替多种本来不同的决定。

寻址也不能止于有一个电子邮箱。若收件处只能解释现有文字,不能启动复核,不能要求采用者回应,它与能停用的地方仍有距离。联系方式证明信能送到,不证明问题进入了能改变事情的程序。

相反,一个局部机构也不该因不能废除全球惯例,就假装没有能力暂停自己的使用。权限有限,可以诚实地说明限到哪里;它不是把现有后果一并交还给无人负责的历史。

这篇文章没有先宣布哪一规则必须撤销。它提供的是在提出撤销、修订或改用时,不把不同动作送错地方的办法。找到地址以后,理由仍可能被反驳,替代方案也可能需要接受同样严格的检验。

有时几处地址会冲突,停用也会产生新风险。谁有最终决定权、如何过渡、怎样保护已依赖旧规则的人,需要额外的制度判断。没有一张寻址图能替大家把这些取舍预先做好。

不过,问清地址已经能改变谈话。那位作者不必向整个统计学申诉一篇退稿;他可以问这家期刊,为何在这里继续用这条线,以及它由什么程序可以改变。广阔的惯例,终于回到一项有人正在作的选择。

設想一篇研究收到退稿意見:結果沒有越過統計顯著性的門檻。作者追問,為什麼這條線應當決定是否值得發表?編輯說這是學科慣例,軟體已經這樣顯示,審稿人也這樣要求。沒有誰宣稱自己發明了規則,但規則正在這一次決定去留。

這裡要找的,不只是最早是誰提出過這個數,而是今天誰能停止它在這裡的作用。規則的身世、內容維護與現實採用,未必落在同一處。找到了祖籍,並不等於找到了能受理修改的門牌。

統計顯著性是原論文的一個真實例子。美國統計學會在2016年的聲明中明確說,科學、商業或政策判斷,不應只看一個數是否跨過特定閾值。它也提醒,統計顯著性不等於效應大小或實際重要性。

但學會發出意見,並不是把一紙由自己發佈的全球法令收回來。教材、期刊、軟體和審稿習慣參與了規則的延續。一個有分量的專業聲明,不能單憑宣佈便撤銷所有地方已經採用的門檻。

這不意味著沒有人能改。某期刊可以改作者指南,軟體可以改默認顯示,一個研究團隊可以改判斷方式。局部停用是真實的改變,只是不能把它說成整個學科已經統一撤銷。沒有唯一發佈者,不等於每個採用者都無權停手。

因此,問題最好縮到一句可核對的話:哪條前提,在什麼範圍內,由誰繼續使用,誰能使這一次使用停止?泛問規則歸誰,容易把實際權限混成一個想象中的總開關。

對於正式制定的標準,也不能只找最初委員會。委員會可能解散,權限可能移交,文本可能進入另一套更新程序。歷史作者可以說明當初為何這樣寫,卻不一定仍有今天撤回或解釋的授權。

原論文把幾處地址分開:規則怎樣形成,由誰維護文本,誰把它採納為行動依據,以及誰能改變或停用。它們是一張尋址圖,不是已完成的法律責任分配,也不保證四處都各有一個整齊的機構。

歐盟協調標準的制度提供了可核對的區別。委員會在正式異議程序中,可以對官方公報所列標準引用作維持、限制或撤除等決定。這一動作針對的是引用及其制度效力,不自動等於改寫標準裡的每一句技術內容。

歐盟法院2020年一項涉及建築產品標準的判決,把這層邊界說得更清楚:在該法規和授權框架內,委員會不能借某種限制,讓成員國補寫標準內容;相關內容的制定、修訂屬於歐洲標準化機構的權限。

同一判決也區分了產品標準與安裝、使用規則,並給成員國後者的權限加上條件。這裡不是要把歐洲一套規定推廣成萬用制度,而是看見撤引用、改文本、改使用,原來可能是三種不同動作。

於是,收到去找標準制定者的答覆,還可以繼續問:我要求的是修文本,還是要求你停止把這份文本用作此處的閘門?前者可能確須轉介,後者未必能靠同一轉介卸掉。

有修訂程序,也不能只看程序這個名字。它允許誰啟動,能討論哪類問題,可以改哪些內容?如果只能修計算方法,卻不能問方法依賴的取捨,那麼程序存在,還沒有回答這項取捨能去哪裡改。

這種範圍有限不必然是程序的缺陷。一個機構本來就可能沒有那項權力。問題在於能否繼續指出另一處有權處理的地址,而不是把這一處無權答,變成世界上已經沒有問題可答。

英國2026年關於NICE成本效益門檻的政府回應,是另一種分工。回應把門檻決定歸於民選政府,並解釋,在部長指示的情形下,NICE不能改變指示內容,因而對該內容進行自身諮詢未必有實質意義。

這不等於所有實施問題都不諮詢,更不等於外部沒有別種政治或法律途徑。它只提醒讀者,能更新評估手冊的機構,不一定就是能決定門檻取捨的機構;要求回應時,不能把二者壓成NICE一個名字。

而若追問的不是門檻高低,是為什麼以這種成本效益方法作決定,地址還須再找。原論文沒有查明這一更深問題的完整受理路徑。這裡保留未查明,不把它寫成絕對不存在。

當一個標準進入具體閘門,採用者會新增自己的決定。設某機構把預測分數用作資格線,它要說明為何採用、怎樣換成線、例外如何處理、後果怎樣復查,以及在什麼條件下停止這次用途。

模型作者仍需對自己寫進模型的選擇和承諾負責。訓練資料、適用人群、輸出含義或保證範圍,不會因為別人採用就全部轉交出去。反過來,採用者後來加的一條資格線,也不自動倒算成作者當初的決定。

所以,只是技術提供者不是萬能免責句,標準是別人定的也不是。它們有時確實指向權限邊界,卻仍需附上具體說明:自己做了哪一段選擇,未做哪一段,又將問題交往哪裡。

美國聯邦住房金融局在2022年宣佈,驗證並批准FICO 10T和VantageScore 4.0供房利美、房地美使用。這是一項有日期的實際批准記錄,不能直接當成今天所有貸款機構已經怎樣實施的描述。

它讓人看見模型驗證與採用安排可以分層。預測表現得到驗證,不能單憑這件事推出每一條借貸資格線都已得到規範說明。反過來,對某條線有異議,也不自動證明模型的預測計算錯誤。

這裡並未裁定那次批准不當,或要求某種信貸政策。現實爭議必須再核實際規則、權限和適用階段。例子的作用,是阻止一個批准二字代替多種本來不同的決定。

尋址也不能止於有一個電子郵箱。若收件處只能解釋現有文字,不能啟動複核,不能要求採用者回應,它與能停用的地方仍有距離。聯繫方式證明信能送到,不證明問題進入了能改變事情的程序。

相反,一個局部機構也不該因不能廢除全球慣例,就假裝沒有能力暫停自己的使用。權限有限,可以誠實地說明限到哪裡;它不是把現有後果一並交還給無人負責的歷史。

這篇文章沒有先宣佈哪一規則必須撤銷。它提供的是在提出撤銷、修訂或改用時,不把不同動作送錯地方的辦法。找到地址以後,理由仍可能被反駁,替代方案也可能需要接受同樣嚴格的檢驗。

有時幾處地址會衝突,停用也會產生新風險。誰有最終決定權、如何過渡、怎樣保護已依賴舊規則的人,需要額外的制度判斷。沒有一張尋址圖能替大家把這些取捨預先做好。

不過,問清地址已經能改變談話。那位作者不必向整個統計學申訴一篇退稿;他可以問這家期刊,為何在這裡繼續用這條線,以及它由什麼程序可以改變。廣闊的慣例,終於回到一項有人正在作的選擇。

Source & reading note原论文与阅读说明原論文與閱讀說明

Ten essays draw on the Disciplines prequel and nine papers. Unattributed applicants, scores, meetings and procedural handovers are imagined. Linked texts, studies and rules retain their dates and evidential limits. The claims that divisions leave something uncontained and that people are more than tools are openly assumed philosophical and normative starting points, not empirical laws. Evidence assessment can itself consider values; professional boundaries, limited authority and reasoned retention are not failures by definition. Missing complete cases are not supplied as established findings; unverified powers are not declared absent, and an unsuccessful search establishes no frequency. These essays prescribe no diagnosis, lending cutoff or clinical action and offer no legal or professional operating advice. English is independently written.

十篇据《学科论》前置篇及九篇正文改写。窗口、评分、会议与程序交接中未另注出处的人物和场景为设想;真实文本、研究与规则另列来源,并注明材料日期和能证明的范围。划分总有未尽、以及人不只是工具,是原作公开承担的哲学与规范起点,不冒充经验定律。证据审查本身可以处理价值,专业边界、有限授权和审后保留不等于失职。原文没有查得的完整实例不补成已证事实,未核不写成不存在;没有发现实例也不据此判断发生频率。这里不决定诊断、信贷门槛或医学处置,不作法律与专业操作建议。简繁中文同文,英文独立重写。

十篇據《學科論》前置篇及九篇正文改寫。窗口、評分、會議與程序交接中未另注出處的人物和場景為設想;真實文本、研究與規則另列來源,並註明材料日期和能證明的範圍。劃分總有未盡、以及人不只是工具,是原作公開承擔的哲學與規範起點,不冒充經驗定律。證據審查本身可以處理價值,專業邊界、有限授權和審後保留不等於失職。原文沒有查得的完整實例不補成已證事實,未核不寫成不存在;沒有發現實例也不據此判斷發生頻率。這裡不決定診斷、信貸門檻或醫學處置,不作法律與專業操作建議。簡繁中文同文,英文獨立重寫。

Read the source paper ↗阅读原论文 ↗閱讀原論文 ↗

ASA statement on statistical significance and p-values (2016)美国统计学会关于统计显著性与p值的声明(2016)美國統計學會關於統計顯著性與p值的聲明(2016) ↗

European Commission: formal objections to harmonised standards欧盟委员会:协调标准正式异议程序歐盟委員會:協調標準正式異議程序 ↗

Court of Justice judgment, 17 December 2020, joined cases C-475/19 P and C-688/19 P, paragraphs 58–73欧盟法院2020年12月17日判决,C-475/19 P与C-688/19 P,第58—73段歐盟法院2020年12月17日判決,C-475/19 P與C-688/19 P,第58—73段 ↗

UK government response on NICE cost-effectiveness-threshold regulations (2026)英国政府:NICE成本效益门槛规章变更咨询回应(2026)英國政府:NICE成本效益門檻規章變更諮詢回應(2026) ↗

FHFA validation and approval announcement, FICO 10 T and VantageScore 4.0 (24 October 2022)FHFA关于FICO 10T与VantageScore 4.0验证及批准的公告(2022年10月24日)FHFA關於FICO 10T與VantageScore 4.0驗證及批准的公告(2022年10月24日) ↗