Abstract
As long as free universal elections have not been achieved in mainland China, no provincial constitutional preparatory body—whether its members are inside the country or in exile abroad—may substitute consultative participation, public reputation or professional credentials for the democratic mandate of the people of the province. This paper distinguishes five powers at different levels (the power of initiative, internal organizational authorization, consultative power, text-drafting power, and the formal power to make and ratify a constitution) and argues that none of the first four can give rise to the fifth. It sets out an institutional division among three types of body (research committee, consultative representative conference, formal provincial constituent assembly) and designs an auditable, revocable and time-limited governance framework for interim consultative mechanisms. The historical comparison uses the 1776 Virginia Convention and the 1921 Guangdong provincial constitution draft as mirrors to show that drafting, deliberation, ratification, promulgation and entry into force are five mutually independent factual levels, and that a normative design must not be written up as an accomplished fact. The legal part draws on the Additional Articles of the Constitution currently in force in the Republic of China (Taiwan), Judicial Yuan interpretations and the Local Government Act to delimit the boundary between private preparatory bodies and statutory provincial government organs; L01–L07 are research drafts without legal force and may not be cited as current law. This paper is a normative institutional design study (Grade D); every historical assertion carries an evidence grade, and unverified ones are recorded as pending verification.
Keywords: constitutional preparation; representation; mandate boundaries; ratification procedure; provincial constitution movement; Additional Articles of the Constitution; Local Government Act
Research Questions and Method
I. To compare the Virginia revolutionary conventions of 1774–1776, the 1921 Guangdong provincial constitution draft and the provincial representative institutions of 1909–1925, testing the limits of comparability of historical precedents; a historical analogy does not by itself generate any present-day legal mandate.
II. To determine what limited procedural authorization exiles and volunteers inside the country can establish when residents inside the country cannot participate freely; the courage, credentials and expertise of participants cannot be converted into a power to govern others.
III. To distinguish research committees, consultative representative conferences and provincial constituent assemblies lawfully elected in the future; the three types of body differ in the source of their power and in the constraints that bind them.
IV. To design safeguards for participant safety, privacy protection and regional and viewpoint diversity; anonymous consultation and verifiable elections must not be conflated.
V. To examine side by side three paths of constitutional continuity—revival of a historical provincial constitution, enactment of a new provincial constitution, and connection to the current constitutional order of the Republic of China (Taiwan)—without prejudging any one of them; each path must independently establish its legal basis and its ratification mechanism.
VI. To compare the legal preconditions and democratic legitimacy of three ratification mechanisms: an elected constituent assembly, ratification by an elected legislature, and a citizens' referendum; which route is lawful depends on the higher-ranking law and the transitional legal basis in force at the time.
Evidence grading: Grade A covers original legal texts, records of proceedings and verifiable archival images; Grade B covers reliable scholarship, official historical materials and attributed translations and annotations; Grade C covers verified research findings of other projects (the original chain of evidence must be preserved); Grade D covers policy proposals, model assumptions and normative inferences (clearly marked as the author's argument). Historical assertions that have not been independently verified are not marked A or B. The institutional design arguments in the chapters that follow belong to Grade D.
Chapter 1 Introduction: The Problem and the Limits of the Argument
Constitutional preparatory work in the pre-transition period can improve the stock of knowledge and the procedural quality of future public constitution-making, but neither the public visibility, the professional reputation nor the exile credentials of a preparatory body can substitute for a mandate from the people. This is the starting point of this paper and a boundary that its argument does not cross.
This paper assesses four criteria separately: legal validity (whether a claim has a legal basis), democratic legitimacy (whether a procedure has been authorized by the governed), participant safety (whether participants inside the country are protected) and feasibility of implementation (whether a design can be put into practice). The four criteria are independent of one another; meeting one cannot make up for failing another. This paper does not present political goals as established legal facts, does not present research proposals as current law, and does not equate the number of participants with the will of the whole province.
Division of labour between this paper and the other studies in the C-T04 series: P01 is responsible for the primary research on the Guangdong–Virginia historical comparison, and P02 for the national database of provincial representative institutions from 1909 to 1925; P03 cites the verified conclusions and evidence numbers of both directly and does not re-examine unverified facts. All historical materials cited from P01/P02 retain their evidence-grade labels.
Chapter 2 Historical Precedents and the Limits of Comparability
The value of historical precedents lies in revealing institutional mechanisms, not in supplying ready-made legitimacy for present-day claims. Two historical cases serve as the empirical anchors of P03's argument; the detailed examination is in Chapter 10, and the limits of comparability are set out here first.
What the 1776 Virginia case shows: a representative convention that adopted a constitution on its own authority, without a popular ratification procedure, did exist historically; but this is no template for modern democratic legitimacy. The franchise at the time was limited to adult white men who owned land, and by present-day standards its representativeness was deficient. Citing this case can only show that "constitution-making by a representative convention" is a historical form; it cannot show that constitution-making without democratic ratification is legitimate today.
What the 1921 Guangdong provincial constitution draft shows: the existence of a text does not mean the institution took effect. The drafting committee's printed edition proves that the proposal existed, and the record of the Provincial Assembly's deliberations proves that a debate took place, but neither proves ratification by the province's citizens or formal promulgation and entry into force. Reading the ratification procedure designed in the draft (the clause on a general vote of citizens) as "a referendum was held at the time" mistakes a normative text for historical fact; this is the core error against which this paper repeatedly guards.
What the provincial representative institutions of 1909–1925 (P02 database) show: an expansion of statutory seats does not mean an expansion of actual participation; the production of institutional texts, the allocation of the ratification power and administrative implementation are three links that frequently break apart. Any claim to "revive" a historical provincial constitution must first answer: what is the legal status of the original text, by what procedure was it terminated or altered, and where is the legal basis for revival. Historical legitimacy does not automatically turn into present validity.
Chapter 3 Distinguishing the Five Powers and the Non-Transferability Boundary
This is the core legal proposition of this paper. The five powers belong to different levels of authorization and cannot be derived from one another:
(1) Power of initiative: any voluntary group may put forward an initiative for constitutional research. This falls within freedom of expression and creates no power over others.
(2) Internal organizational authorization: binds only members who have joined voluntarily and accepted the charter. The effect of a charter, an oath or an internal vote stops at the boundary of the organization and cannot spill over into a power to govern the residents of the whole province.
(3) Consultative power: to collect and publish the views of participants. Consultation results represent only those who actually took part, not residents who did not; the denominator of a participant turnout rate is the eligible registered participants, and it may never be called "the province-wide turnout".
(4) Text-drafting power: to produce non-binding alternative drafts. Drafts may be opened for public comment and may allow alternative texts, but they have no legal force in themselves.
(5) Formal power to make and ratify a constitution: can be acquired only under the law in force at the time and through an appropriate democratic procedure. This is the only power capable of creating generally binding force.
Non-transfer rule: the fifth power may not be derived from any of the first four. Common reasoning errors in practice include: inferring the will of the whole province from the number of consultation participants (substitution of the denominator); passing off an organization's internal election as a statutory election (substitution of nature); claiming present validity from the existence of a historical text (substitution of time); and claiming statutory power from the provisions of a research draft (substitution of legal effect). This paper requires every claim to power to state its direct legal basis; a claim that cannot point to a legal basis is classed as a Grade D argument and may not be written into statements of fact.
Four independent tests of representativeness (for assessing the quality of consultation, not as a substitute for an electoral mandate): first, descriptive coverage—whether participants come from different regions and groups; second, procedural openness—whether there are transparent mechanisms for recruitment, expression and dissent; third, traceability of authorization—whether an individual's consent to take part in consultation has been wrongly interpreted as authorizing someone to exercise public power on his or her behalf; fourth, correctability—whether those affected can withdraw their views, correct misstatements and put forward alternatives. The four indicators can only measure the quality of consultative work; whether or not they are met, they produce no constitution-making mandate.
Chapter 4 Three Types of Body
(1) Research committee. Functions: cataloguing historical materials, proposing alternatives, independently verifying facts. It makes no claim of representativeness, holds no "election" under any name and issues no binding documents. Its legitimacy comes from scholarly honesty and procedural transparency, not from numbers or reputation.
(2) Consultative representative conference. Functions: discussing several sets of proposals and recording majority and minority views, without making binding decisions. If it holds internal elections, it must declare publicly that the delegates chosen represent only the participants who actually registered; the election rules, voter eligibility, counting method and channels for objection are all made public. Regional thematic groups may hold discussions, but a regional group may not hold itself out as the statutory representative of that region.
(3) Formal provincial constituent assembly. It comes into being lawfully only when two conditions hold at the same time: there is an applicable and valid legal basis, and conditions for free and fair elections exist. The scope of its mandate, its term and its powers must have a clear legal basis and are subject to judicial review.
The three types of body are a functional division, not a ladder of promotion. However excellent a research committee's work, it will not "grow" into a constituent assembly; however lively the participation in a consultative representative conference, it will not "turn into" a democratic mandate. Crossing from one type of body to another can only be achieved through a new and independent legal authorization and democratic procedure; there is no automatic transition.
Chapter 5 The Denominator Problem in Representativeness Statistics
Consultation reports must disclose the following denominators side by side and may not publish them selectively: the number actually invited, the number of eligible registrants, the number taking part in the consultation, the number taking part in internal elections, the number of valid votes, geographical coverage, age and other aggregate structures that can safely be made public, recruitment channels, and identified coverage gaps. It is prohibited to use any model to extrapolate a non-existent "province-wide support rate"; it is prohibited to report the results of anonymous questionnaires, open calls, random sampling and formal elections in mixed form; participation results not based on a reliable sampling frame may not be used for statistical inference about the population.
Counterfactual test: if the great majority of a province's residents cannot vote, then even a turnout of one hundred per cent among exile participants may not be called the will of the whole province. The honest course is to publish the denominator of actual participants, the eligibility conditions, the regional blanks, and the unknown quantity of the population that cannot safely take part—without filling the unknown with conjecture.
Voluntary participation in consultation may record provincial origin or regional affiliation, but unverified native places or population proportions may not be used to generate "formal provincial seats"; a quota is not a mandate. Diversity of region, gender, occupation, language and political opinion can only be achieved through open participation and honest disclosure of gaps; it cannot be fabricated by digital techniques.
Chapter 6 A Negative List of Interim Powers
A preparatory body may not claim or exercise the following powers: provincial legislative power, the power to levy revenue, command of police or armed forces, power to dispose of land and property, power to appoint courts, or power to make compulsory determinations of citizenship. It may not confer constitution-making power on itself through a charter signed in advance, an oath taken by members or an online vote.
What it may do: publish drafts, comparative reports, minority opinions, financial audits and procedural recommendations—but each of these must be clearly labelled as non-binding. Where participants or affiliated organizations hold an independent mandate under another legal system (such as political rights attached to the nationality of a particular country), its legal basis must be listed separately, and that mandate may not be merged into the name of the preparatory body.
The point of the negative list is falsifiability: any overreach can be identified and pointed out by outside observers on the basis of the list. A preparatory body that refuses to be bound by a negative list cannot credibly describe itself as "time-limited, open and revocable".
Chapter 7 Safety, Inclusion and Conflicts of Evidence
Participants inside the country face real personal risk. The safety principles of this paper are: voluntariness, informed consent, data minimization, non-disclosure of identity by default, tiered access, independent security assessment, and the right to withdraw. No identifying information that could lead to reprisals is made public; without a security assessment, small cross-tabulated samples of provincial origin, age, occupation and the like that could be used to re-identify individuals may not be published.
There is an inherent conflict between safety and evidence: anonymity limits the ability to verify against duplicate voting, and confidentiality limits the depth of external audit. The honest course is to state the limits of credibility publicly, not to force the disclosure of sensitive identities in order to improve data quality. "Courage" or an experience of persecution must never be made a qualification for participation—that would turn risky behaviour into a condition of entry, distort the sample and endanger personal safety.
Inclusion requirements: ensure that different regions, genders, occupations, languages and political views can be expressed; publish coverage gaps rather than fabricating representativeness; provide dissenters with equal channels for submitting proposals. Sources of funding, organizers' conflicts of interest and links with political forces inside and outside the country are all brought within the funding-disclosure and recusal system.
Chapter 8 Three Competing Paths of Constitutional Continuity
Path one: revival of a historical provincial constitution. It must be shown what the legal status of the original text is, what its scope of application is, and on what basis it was terminated or altered. Historical texts have research value and symbolic significance, but "once existed" does not mean "currently in force". A claim of revival must answer: whether the original ratification procedure was completed, through what legal act the text subsequently lost force, and what new authorization revival itself requires. Present validity cannot be presumed merely because a text has historical legitimacy.
Path two: enactment of a new provincial constitution. It must first be established what the legal basis of provincial constitution-making power is, what the hierarchical relationship is between a provincial constitution and the national constitution, and where the line runs between provincial powers and local self-government. Without authorization from higher-ranking law, "enacting a provincial constitution" is merely the writing of a draft, not an act of legislation.
Path three: connection to the current constitutional order of the Republic of China (Taiwan). This requires separate study of the current Constitution and its Additional Articles, the present structure of state organs, the threshold for constitutional amendment, and the legal relationship between the free area and political change on the mainland. Under the current Additional Articles, Article 9 provides for the appointment of provincial government organs[1], Article 12 requires a constitutional amendment to be proposed by a special majority of the Legislative Yuan and approved by referendum of the electors of the free area[2], Judicial Yuan Interpretation No. 467 confirms that the province is not a local self-governing body[3], and the Local Government Act expressly makes the province an agency of the Executive Yuan[4]. Within this framework it may not be assumed that political transition on the mainland will automatically change the legal powers of the existing organs of the Republic of China (Taiwan); it may not be assumed that a provincial consultative body holds national-level constitution-making power; and a political vision may not be treated as an automatically effective interpretation of current law.
The three paths may be mutually exclusive in legal principle, or they may require different transitional procedures. This paper keeps the scenarios open and does not prejudge which takes precedence; progress on each path must be independently audited by ROC-CONST-01 (national constitutional structure) and ROC-TRANS-02 (transition sequencing and cross-Strait institutional relations). This paper does not replace the conclusions of those two studies.
Chapter 9 Procedural Thresholds from Consultation to Democratic Ratification
Moving from consultation to democratic ratification requires passing six thresholds separately; the six thresholds are independent of one another, and a single vote cannot make up for a legal mandate that was missing earlier:
Threshold one: credible guarantees of freedom of expression, association and the press and of participant safety.
Threshold two: public and contestable rules on voter eligibility and electoral districts.
Threshold three: the manner of establishment, term and powers of the constitution-making body determined by law.
Threshold four: publication of the full text of the draft, of alternative proposals and of fiscal impact.
Threshold five: independent administration of voting, election observation, and complaint and remedy procedures.
Threshold six: ratification completed in accordance with the applicable higher-ranking constitution or a lawful transitional legal basis.
The triggering conditions for a formal handover must be reviewed independently, one by one: whether an applicable and valid legal basis exists; whether association, expression, electoral competition and general voter participation are guaranteed; whether independent election administration, open remedies for disputes and a credible count have been established; whether the division of powers between the provincial and national constitutions is clear; and whether the procedures for adoption, promulgation, entry into force and judicial review are clear. Once the conditions are met, the results of interim consultation may only be submitted as materials to the formal body, which decides on its own whether to adopt, amend or reject them; the original consultative organization may not claim automatic succession to governmental power and may not refuse to dissolve or hand over.
Chapter 10 Positive and Negative Lessons from Historical Comparison: Virginia (1776) and Guangdong (1921)
Virginia, 1776. The Fifth Virginia Convention adopted a constitution on its own authority on 29 June 1776 without holding a separate ratifying referendum; this is a historical instance of "constitution-making by a representative convention"[7]. The franchise at the time was limited to adult white men who owned land, and by modern standards its representativeness was deficient. This case can therefore only show that a form of constitution-making without a referendum existed historically; it cannot serve as a template for modern democratic legitimacy, still less directly prove the legitimacy of any present-day provincial draft. For sources, see the references.
Guangdong, 1921. This paper applies a four-stage evidentiary standard: (1) drafting—the drafting committee's printed edition proves that the proposal existed, not that it was authorized; (2) representative deliberation—the Provincial Assembly version must be supported by minutes of proceedings, a resolution of adoption and a chain of original versions, and cannot be established on the strength of a signature line on a web page; (3) popular ratification—whether the province-wide general vote of citizens provided for in Article 135 of the draft took place must be proved by a notice of the vote, voter eligibility, vote counts and certification records; (4) promulgation and entry into force—the transitional election deadline in Article 132 runs from the actual date of promulgation, which requires a gazette or equivalent formal proof of promulgation.
Collation of versions (at the level of public transcriptions; word-by-word collation against the original prints is still to be completed): the September 1921 drafting committee draft[8] and the Provincial Assembly version dated 19 December 1921[9] contain substantive variant readings in Article 129 (the parties co-signing an amendment: "county assemblies and special municipal assemblies" versus "county assembly members and special municipal assembly members") and Article 134 (the ratio of transitional military expenditure and the implementation timetable); in Article 135 both versions provide that "this Law shall be decided by a general vote of the citizens of the whole province and shall take effect on the day of its promulgation"—this is a condition for entry into force in the draft, not evidence that a vote was held or that the law was promulgated and took effect. Article 45 (on dissolution of the Provincial Assembly) and Article 129 (referendum on amendments) concern different matters and cannot serve as proof for each other. The election deadline in Article 132 runs from "after the promulgation of this Law" and cannot be counted from the date of the draft. For the original printed copy, see [10].
Historical conclusion: on 19 December 1921 the Provincial Assembly completed only the second reading; the draft did not pass a third reading or a citizens' vote and never took effect[9][14]; no evidence has been found that a province-wide citizens' vote was held or that the draft was formally promulgated at the time. No exile consultative vote may pass itself off as the province-wide citizens' ratification envisaged in that draft. For the detailed collation, see Appendix B.
The common lesson of the two cases: normative design (what procedure the draft prescribed) and historical fact (what actually happened) are two different things; reading the former as the latter is the most common and most dangerous error in the study of constitutional history. This paper's requirement for present-day preparatory bodies follows the same logic: what a charter says is not the same as what has been lawfully obtained.
Chapter 11 International Comparison and a Review Framework
Two studies by International IDEA provide an analytical framework for this paper; they do not constitute a specific legal mandate for any region.
Erin C. Houlihan and Sumit Bisarya (2021), Practical Considerations for Public Participation in Constitution-Building: What, When, How and Why? (International IDEA Policy Paper No. 24)[11], distinguish different stages of participation—agenda-setting, selection of bodies, provision of input and final approval—and stress that differences of context determine the design of participation. This supports this paper's layered framework in which "consultation is not ratification".
Kimana Zulueta-Fülscher and Sumit Bisarya (2018), (S)electing Constitution-Making Bodies in Fragile and Conflict-Affected Settings[12], compare the methods of selecting bodies in 37 constitution-making processes between 1991 and 2018 and show that the way a constitution-making body is formed profoundly affects the legitimacy and content of the process. Their conclusion is consistent with this paper: "how" a body is formed matters as much as "what" the text says.
Review framework: the proposals of any provincial preparatory body should undergo a threefold review—historical review (whether precedents are accurately described), legal review (whether claims have a basis in current law) and procedural review (whether the mechanisms for participation, decision-making and error correction are auditable). The three reviews are independent of one another, and none may be omitted.
Chapter 12 Model Charter, Scenarios and Risks
Key points of a model charter. Purpose: non-governmental, non-coercive research and consultation on a provincial constitution. Limits on power: no taxation, law enforcement, establishment of armed forces, appointment to public office or promulgation of a binding constitution (see the negative list in Chapter 6). Governance: public rules of procedure, disclosure of funding, recusal for conflicts of interest, an appendix of minority opinions, external audit. Term: periodic review of the mandate; once formal democratic organs have been established, records are handed over and the special preparatory functions end. Disputes: independent complaints, error-correction mechanisms and a public record of versions. This paper does not regard any model charter as a document of a statutory government body.
Three scenarios. Scenario A: free participation remains impossible for a long time—work is limited to confidential research and non-representative consultation, with limitations honestly labelled. Scenario B: partial opening—public hearings are expanded, but no province-wide mandate is presumed. Scenario C: free elections become feasible—the work is handed over to the formal body in accordance with law, and the original organization dissolves or becomes an ordinary research group.
Main risks: elite capture (the organization is controlled by a few people), identity security (participants inside the country are exposed), imbalance between views inside and outside the country, conflicts of legal basis (claims under different legal systems clash), romanticization of the old provincial constitution (sanctifying the historical text), and perpetuation of interim bodies (failure to dissolve when the term ends). Each risk must have a corresponding institutional safeguard, and the charter must set out the conditions that trigger a review.
Chapter 13 Cross-Project Reuse, Conflict Register and Audit
P01: comparative historical study of constitution-making mandates, selection of representatives and ratification paths. P02: database of provincial elections, provincial consultative assemblies and the provincial constitution movement. ROC-CONST-01: national constitutional structure and amendment procedure. ROC-TRANS-02: transition sequencing and cross-Strait institutional relations. L01–L07: design of the powers of interim bodies, elections, emergency powers and judicial remedies—the official page describes them itself as model legislation drafts with no present legal force[15]; this paper cites them only as objects of comparison, not as a legal basis.
Cross-project conflict register (first edition; every item may be closed only after cross-review against the original texts and the latest version of each project):
C1: Whether a provincial consultative representative conference has been wrongly described as holding a constitution-making mandate—verify against P01/P02 and the seven laws. Position of this paper: no legal basis has been found, and it may not be so described.
C2: Whether a historical provincial constitution has been wrongly described as automatically restorable—verify against the original provincial constitution texts and the changes in the national constitution. Position of this paper: revival must be argued independently and may not be presumed.
C3: Whether a transitional executive body may appoint members of a constituent assembly—verify against the seven laws, judicial review and electoral rules. Position of this paper: appointment is not a democratic mandate; the design intent must be confirmed with the drafters of the seven laws.
C4: Whether transition on the mainland automatically changes the current parliament and amendment procedure of the Republic of China (Taiwan)—verify against ROC-CONST-01. Position of this paper: it does not change them automatically; any change must follow the amendment procedure of Article 12 of the Additional Articles[2].
C5: The basis of mandate for Taiwan's and the mainland's participation in national-level constitutional decision-making—verify against ROC-TRANS-02. Position of this paper: the provision on electors of the free area in current Article 12 does not mean that the question of representation for future residents of the mainland has been resolved[2]; Article 11 leaves the relationship of rights and obligations between the people of the mainland area and the free area to be governed by law, the present position is likewise unsettled, and the future institutional connection requires separate analysis[16].
Every item of evidence must record its source URL or file number, page number, verification status, reusing project and objections; items not yet verified are marked pending verification and may not be upgraded to confirmed. The article-by-article conflict matrix for the full text of the seven laws is still to be completed and falls outside the scope of this publication.
Conclusion
Being able to discuss a future constitutional order lawfully is not the same as having lawfully acquired the right to make a constitution on behalf of the people. This is where the entire argument of this paper leads.
A credible preparatory mechanism must satisfy four requirements at once: restraint in its powers (knowing clearly what it cannot do), openness to scrutiny (every claim can be checked against evidence), participant safety (not at the cost of other people's personal safety), and ultimate democratic mandate (leaving the decision to formal procedures under future conditions of freedom). Without any one of these, preparatory work slides from "paving the way for democratic constitution-making" to "exercising unauthorized power in the name of constitution-making".
What this paper offers is not a ready-made institutional blueprint but a set of constraints: a firewall between the five powers, a functional division among the three types of body, four tests of representativeness, six ratification thresholds and a negative list. Within these constraints, research, consultation and drafting can proceed lawfully; beyond them, any organization—however courageous its members, however deep their credentials, however polished their text—is merely authorizing itself.
History has already given a warning: the 1921 Guangdong provincial constitution draft wrote in a clause for a general vote of citizens, but the vote was never held and the constitution never took effect. The quality of a text cannot substitute for the completion of a procedure. Today's preparers should be more honest than their predecessors and draw a strict line between "we hope the future will be like this" and "we have lawfully obtained this". That is exactly why this paper requires every claim to carry an evidence grade.
Appendix A Evidence Grades and Verification Status (as of 2026-10-15)
- Grade A (original texts verified word for word):
- Additional Articles of the Constitution of the Republic of China, Article 9 (appointment of provincial government organs): official text of the Office of the President verified.
- Additional Articles of the Constitution of the Republic of China, Article 12 (amendment threshold): official text in the Ministry of Justice's Laws & Regulations Database verified; "half of the total number of electors" and "a majority of those actually voting" have been distinguished.
- Judicial Yuan Interpretation No. 467 (the province is not a local self-governing body): official interpretation text of the Constitutional Court verified.
- Local Government Act, Article 2 (the provincial government is an agency of the Executive Yuan; the province is not a local self-governing body), Article 5 (the province has a provincial government and a provincial advisory council) and Article 14 (local self-governing bodies do not include the province): current text in the Laws & Regulations Database verified word for word (version amended 3 December 2025); Articles 1, 8–13, 16 and 30 follow the verification record of 2026-10-14.
- Bibliographic information for the two International IDEA publications verified (2021 Policy Paper No. 24; 2018 comparative study of 37 constitution-making processes).
- Grade B (reliable transcriptions or official library catalogue records; collation against the original prints pending):
- September 1921 drafting committee version of the Draft Provincial Constitution of Guangdong: catalogue record of the 39-page scan held by the National Library confirmed (15 chapters, 135 articles); the Wikisource index marks it "not proofread"; the transcription of the articles is Grade B.
- Provincial Assembly version of 19 December 1921: only online transcriptions are available, and the page itself states that its source has not been checked; the variant readings against the September version (Articles 129 and 134) are observations at the level of transcriptions.
- Procedure of the 1776 Virginia constitutional convention: exhibition page of the Library of Virginia verified—on 29 June 1776 the convention unanimously adopted the first constitution, which was declared in effect without a ratifying referendum; under a 1736 statute the franchise was limited to adult white men who owned or held long-term leases on land.
- Grade C (reuse of verified research findings):
- P01 conclusions of the Guangdong–Virginia comparison and the P02 database of provincial representation: the evidence numbers of their published versions are cited.
- Grade D (normative argument):
- The distinction among five powers, the three types of body, the tests of representativeness, denominator disclosure, the negative list, the six ratification thresholds, the model charter and the scenario analysis: all are institutional design proposals of this paper and are not claimed to be current law or historical fact.
- Pending verification (to continue after publication):
- Locating the original printed copy of the December 1921 Guangdong version and collating it word for word; checking the clause on the cap on military expenditure in Article 116 against the original print.
- Article-by-article conflict matrix for the full text of the seven laws (L01–L07).
- Item-by-item conflict review of the latest formal drafts of ROC-CONST-01 and ROC-TRANS-02.
Appendix B Summary of the Collation of the 1921 Guangdong Versions
Physical copies. Scan one: the 1921 facsimile of the Draft Provincial Constitution of Guangdong (39 pages) held by the National Library of China, whose catalogue record lists the editor as the Guangdong Provincial Constitution Drafting Committee and the whole work as 15 chapters and 135 articles. Scan two: a 40-page copy, likewise credited to the drafting committee as editor, 1921. Neither scan can be identified, on the strength of its catalogue record alone, as the text adopted by the Provincial Assembly on 19 December 1921.
Transcriptions. The September 1921 drafting committee draft has an independent text transcription; the version credited to the Guangdong Provincial Assembly and dated 19 December 1921 also has an online transcription, but the page expressly warns that its original source has not been checked. The 19 December text reproduced by Zhonghua Wenku is identical in content to the Wikisource transcription and cannot be regarded as independent corroboration.
Variant readings (at the level of transcriptions, not a definitive finding from the original prints). Article 129: the September version reads "co-signature by more than half of the county assemblies and special municipal assemblies", the December version "co-signature by more than half of the county assembly members and special municipal assembly members"—"assemblies" versus "assembly members" may affect the body on which the threshold for a proposal is calculated, and must be settled against the original print. Article 134: the September version provides for reducing transitional military expenditure to one third of annual expenditure and for phased arrangements after the national constitution is established; the December version provides for reducing it to one half of annual expenditure before a formal government is established—a substantive difference in terms of fiscal and institutional timetables, and the two may not be spliced together. Article 135: both versions provide that "this Law shall be decided by a general vote of the citizens of the whole province and shall take effect on the day of its promulgation"—a condition for entry into force in the draft, not evidence that a vote was held. Article 132: both versions take "after the promulgation of this Law" as the starting point of the election deadline, which may not be counted from the date of the draft. Article 116 (military expenditure capped at thirty per cent of the province's total annual expenditure; see the unproofread transcription, p. 34) and Article 134 (the transitional ratio) are different norms and must not be confused.
Standard. Drafting, deliberation in the assembly, citizens' ratification, formal promulgation and actual entry into force are five independent evidentiary stages. The 19 December version proves at most that deliberation reached the second reading; there is no evidence of a citizens' vote, formal promulgation or entry into force. This appendix closes the item "identification of variant readings in public transcriptions"; collation against the original prints and the historical facts of any vote remain pending verification.
References and Evidence Entry Points
- [1] Office of the President, Additional Articles of the Constitution of the Republic of China, Article 9 (provincial government organs): the provincial government has nine members, and members of the provincial advisory council are likewise appointed by the President on the nomination of the President of the Executive Yuan; county council members and county magistrates are elected by the people of the county. Official text: Office of the President (English website)
- [2] Ministry of Justice, Laws & Regulations Database, Additional Articles of the Constitution of the Republic of China, Article 12 (amendment procedure): proposed upon the motion of one quarter of the legislators, with three quarters present and a resolution of three quarters of those present; after public announcement for half a year, submitted to a referendum of the electors of the free area, with valid votes in favour exceeding half of the total number of electors. Official text: Laws & Regulations Database of the Republic of China (Taiwan)
- [3] Judicial Yuan Interpretation No. 467 (22 October 1998): the province is a public-law juridical person that is not in the nature of a local self-governing body; within the scope of specific statutory powers it may still have the status of a public-law juridical person. Text of the interpretation: Constitutional Court website
- [4] Ministry of Justice, Laws & Regulations Database, Local Government Act (version amended 3 December 2025): Article 2, the provincial government is an agency of the Executive Yuan and the province is not a local self-governing body; Article 5, the province has a provincial government and a provincial advisory council; Article 14, local self-governing bodies are special municipalities, counties (cities) and townships (towns, cities). Official text: Laws & Regulations Database of the Republic of China (Taiwan)
- [5] Office of the President, explanation of the seventh amendment of 2005 (history of the amendment procedure): Office of the President website
- [6] Office of the President, introduction to the third amendment of 1994 (historical version: the provincial assembly and the governor were once elected by the people of the province): Office of the President website —historical provisions do not automatically revive; the version and date must be stated.
- [7] Library of Virginia, Virginia Constitutions exhibition: procedure of the 1776 constitutional convention: Library of Virginia exhibition page
- [8] Draft Provincial Constitution of Guangdong, transcription of the September 1921 drafting committee version: Chinese Wikisource
- [9] Draft Provincial Constitution of Guangdong, transcription of the version credited to the Provincial Assembly and dated 19 December 1921 (the page itself states that its source has not been checked): Chinese Wikisource
- [10] Catalogue record of the 1921 facsimile (39 pages) of the Draft Provincial Constitution of Guangdong held by the National Library: Wikimedia Commons
- [11] Erin C. Houlihan and Sumit Bisarya, Practical Considerations for Public Participation in Constitution-Building: What, When, How and Why?, International IDEA Policy Paper No. 24, 2021, DOI 10.31752/idea.2021.63: International IDEA publication page
- [12] Kimana Zulueta-Fülscher and Sumit Bisarya, (S)electing Constitution-Making Bodies in Fragile and Conflict-Affected Settings, International IDEA, 2018 (compares 37 constitution-making processes, 1991–2018): International IDEA publication page
- [13] C-T04-P01 Guangdong–Virginia comparative study (published version; evidence numbers carried over).
- [14] C-T04-P02, "From the Late-Qing Provincial Consultative Assemblies to the Republican Provincial Assemblies: Provincial Representative Institutions, Electoral Representativeness and the Provincial Constitution Movement in China, 1909–1925", Research Paper v1.0 (published): FreeChina.live (in Chinese)
- [15] L01–L07 research drafts: the banner at the top of the official FreeChina.live page states expressly "MODEL LEGISLATION · ADOPTION-CANDIDATE TEXT · NOT CURRENT LAW" and "It has no present legal force"—the drafts have no legal force and may not be cited as a source of statutory power: FreeChina.live
- [16] Office of the President, Additional Articles of the Constitution of the Republic of China, Article 11 (rights and obligations between the people of the mainland area and the free area left to be governed by law): same official page as [1].
Editorial Note
This text was independently reviewed and revised directly. Seven legal provisions were independently verified against official sources; the collation of the 1921 Guangdong versions keeps its conclusions at the level of transcriptions, and collation against the original prints is listed as pending verification. This text combines Grade D normative research with Grade B historical transcriptions and does not present itself as Grade A evidence from original prints.
Other language editions:
Traditional Chinese and Spanish editions are in preparation and will be linked here when published.