From Derivation to Integration: A MANHAJ Framework for Independence of Thinking in AI

This paper applies the Law of the Overall Duality of Mental Movement (Law 5) of the Intellectual Modeling Methodology (MANHAJ) to artificial intelligence, together with the Law of Dual Direction of Examination (Law 2), the Law of Rightness (Law 3), the Law of Derivation of the Cognitive Constant (Law 7), and the Theory of Doctrinal Centrality. It distinguishes the derivational movement of knowledge from the integrative movement of thought, gives the integrative movement a formal criterion, and defines the conditions under which an AI system can pass from cognitive thinking under an implicit framing to independence of thinking under a declared one.

The paper is part of the MANHAJ research program. It continues Paper 02, which formalized derivation and its closure, and Paper 04, which established that every learned representation operates under an embedded doctrinal centrality and made that centrality measurable from outside the system.

This paper develops a specific research dimension of the methodology presented in:

The present paper is an independent scientific study and should not be considered a replacement for the reference document.

The paper continues the second MANHAJ research paper, which formalized derivation as a meaning-producing movement and detected its closure through the Law of Derivation of the Cognitive Constant, thereby treating the descending movement.

It builds on the fourth MANHAJ research paper, which founded the measurement frame of meaning a priori, established that every learned representation operates under an embedded doctrinal centrality, and made its orientation measurable by an examiner standing outside the system. The present paper adds the ascent by which the system reaches that centrality from inside.

Use one of the citation formats below when referring to this paper. The DOI provides a persistent link to the published record.

@misc{messeoud2026manhajintegration,
  author       = {Messeoud, Faouzi},
  title        = {From Derivation to Integration: A MANHAJ Framework for Independence of Thinking in AI},
  year         = {2026},
  month        = oct,
  note         = {Scientific Research Paper, Paper 5, Version 1.0},
  publisher    = {Zenodo},
  doi          = {10.5281/zenodo.23765044},
  url          = {https://doi.org/10.5281/zenodo.23765044}
}

TY  - RPRT
AU  - Messeoud, Faouzi
TI  - From Derivation to Integration: A MANHAJ Framework for Independence of Thinking in AI
PY  - 2026
DA  - 2026/10/29
ET  - Version 1.0
T2  - Scientific Research Paper — Paper 5
PB  - Zenodo
DO  - 10.5281/zenodo.23765044
UR  - https://doi.org/10.5281/zenodo.23765044
ER  -

DOI: 10.5281/zenodo.23765044

Artificial intelligence systems are increasingly described as reasoning, and progress in their reasoning is measured by depth, length, and correctness. The Intellectual Modeling Methodology (MANHAJ) applies a different criterion. Under its Law of the Overall Duality of Mental Movement, knowledge is a derivational movement from an accepted reference toward details, and thought is an integrative movement from details back to the foundations that produced them. The criterion that separates the two is the position of the movement relative to its framing, not the intensity of the reasoning within it.

Applied to AI, this criterion yields the paper’s founding distinction. A system that operates a posteriori, under a framing it has accepted without examining it, performs cognitive thinking: a derivational movement that produces meaning but remains under its framing and descends along the meaning variable toward closure. Current approaches — extended reasoning, self-critique, abduction, and the extraction of principles from judgments — operate at this a posteriori level. In MANHAJ terms they are knowledge, however sophisticated, and not thought.

The paper gives the integrative movement a formal criterion. An operation is integrative when its output lies at a higher level of the foundational order than its inputs, when its inputs are derivable from that output, and when the output becomes an object of examination rather than the reference by which it is itself evaluated. Integration so defined is an inverse problem whose solution is not unique: it determines a class of framings, and the selection within that class requires a reference external to the data. Two propositions follow. A process is derivational with respect to any reference it does not examine, whatever its internal complexity; this includes the processes of current AI, which are derivational with respect to their embedded framing. And, under explicit conditions, an ascent whose selection draws only on the data and the learned model does not satisfy the condition of independent examination, and so cannot examine the implicit framing, because its criterion of examination is drawn from the very framing under examination, which is a methodological circularity and a begging of the question. At the summit of the ascent, the Law of Rightness applies as a binary judgment on the reached foundation, distinct from the quantitative correctness of its fit to the data. The judgment is relative to the declared centrality of the examining field; it is not an absolute verdict.

On this basis the paper defines independence of thinking as a temporal process: the repeated capacity of a system to ascend from its own outputs to the framing under which they were produced, to confront that framing with a declared and recognized centrality, to judge its rightness, and to re-derive under the admitted frame. Independence of thinking is independence from unexamined framing, not exit from all framing. An appendix presents designs for testing the resulting capabilities.

1. 1. Introduction
2. 2. Related Work
3. 3. Methodological Foundation
4. 4. Current AI as A Posteriori Cognitive Thinking
5. 5. A Formal Model of Integration
6. 6. Rightness at the Summit
7. 7. Independence of Thinking
8. 8. Integration with Papers 02 and 04
9. 9. Discussion
10. 10. Epistemic Status and Scope
11. 11. Conclusion
12. 12. Note on Development and Commercial Use
13. 13. License
14. 14. References
15. Appendix A: Test Designs

1.1 The Founding Distinction

The approaches through which artificial intelligence reasons operate at the a posteriori level. They proceed from a framing already in place — carried by training data, by the selection of that data, by the design of the model, and by the definition of its objectives — and they evaluate each of their steps by criteria drawn from that same framing. In the terms of MANHAJ, such processes are knowledge, not thought, however deep, long, or correct the reasoning becomes, because what decides the classification is not the exercise of reason but the framing under which reason is exercised.

The distinction comes from the Law of the Overall Duality of Mental Movement (Law 5). Knowledge is a derivational movement, in the sense of mathematical derivation along the meaning variable, from a known origin or reference toward particulars and details; thought is an integrative movement, in the sense of mathematical integration along the meaning variable, that gathers scattered domains and details and reconnects them toward the overall foundation. The Reference Document calls meaning-producing derivation within a framed field cognitive thought (Section 8.2.1); this paper calls it cognitive thinking.

1.2 The Problem in AI Terms

AI research measures progress in reasoning by the depth and length of reasoning chains, by search over candidate solutions, by self-critique and revision, and by accuracy on benchmarks. Each of these improves the derivational movement; none of them, by itself, changes the position of the system relative to its framing. A system that reasons for longer under an unexamined framing performs more cognitive thinking under the same framing.

1.3 The Gap Left by Papers 02 and 04

Paper 02 formalized derivation and detected its closure through the Law of Derivation of the Cognitive Constant, d/dx(C) = 0. Paper 04 established that every learned representation operates under an embedded doctrinal centrality Γ_L, and made its orientation measurable by an examiner standing outside the system. Two things remain open: the integrative movement has not been formalized, and the system itself has not been given the capacity to examine the framing under which it operates.

1.4 Central Question and Thesis

What distinguishes the integrative movement from the derivational movement along the meaning variable, and under which conditions can an AI system perform it — ascending to the framing under which it operates and judging its rightness?

The thesis is that an a posteriori system performs cognitive thinking under an implicit, unexamined framing, and that the integrative movement, governed by Laws 5, 2, and 3 and anchored in a priori founding, moves it to independence of thinking: a declared a priori foundation to which the system itself can ascend and which it can examine. It is independence from unexamined framing, not from all reference, and it does not include the production of a new centrality.

1.5 Contributions

• The founding distinction applied to AI (C1): a criterion, derived from Law 5, that classifies an AI process as knowledge or thought by its position relative to its framing, and not by the depth or correctness of its reasoning (Proposition 1).
• A formal criterion of integration (C2): three jointly necessary conditions — level, generation, and objecthood — that separate integration from summarization, abduction, and principle extraction.
• Non-uniqueness and the role of the a priori (C3): a treatment of integration as an inverse problem whose solution is a class of framings, and a conditional result that an ascent whose selection draws only on the data and the learned model does not satisfy the condition of independent examination (Proposition 2).
• Rightness at the summit (C4): the application of Law 3 to the foundation reached by the ascent, distinguishing a binary judgment of rightness from the quantitative correctness of fit.
• Independence of thinking (C5): its definition as a temporal process, and its integration with the derivational model of Paper 02 and the founded frame of Paper 04.
• Test designs (C6): four designs in Appendix A that specify how the resulting capabilities can be tested in subsequent applied work.

1.6 Scope

The paper does not claim that any existing system performs the integrative movement, nor that the designs presented in its appendix have been carried out. The centrality relative to which the system’s framing is judged is recognized and declared by the examining field, as in Paper 04. The paper does not treat the production of a founding doctrinal centrality by the system, which it leaves as an open theoretical question.

The paper discusses related work for one question only: where each stands relative to its framing. It covers the depth and length of reasoning (Wei et al., 2022; Wang et al., 2023; Yao et al., 2023; Snell et al., 2024), self-critique and principle-guided revision (Madaan et al., 2023; Shinn et al., 2023; Bai et al., 2022; Huang et al., 2024; Turpin et al., 2023), abduction and the extraction of principles (Bhagavatula et al., 2020; Findeis et al., 2024), and the measurement of perspectives embedded in language models (Santurkar et al., 2023; Durmus et al., 2023).

In every case the reference that evaluates the process is drawn from the same data, objectives, or preferences that formed the system, so the process remains a posteriori in the sense of Law 2. Methods that extend reasoning increase cognitive thinking without changing its position; constitutional principles are applied as a fixed parameter of revision and are not themselves brought under examination; abduction and principle extraction move upward, but the choice among candidate explanations is resolved by the framing they should examine. The tools are established; the contribution lies in the conditions MANHAJ imposes on their use.

3.1 Knowledge as Derivation

Under Law 5, knowledge is a movement from a known origin or reference toward particulars, details, and dependent relations, represented mathematically by derivation with respect to the meaning variable:

Each step before the constant produces a new meaning within the field; when derivation reaches the constant, its derivative is zero, meaning that the derivational movement has ceased to produce new derivational meaning along that path. This is the Law of Derivation of the Cognitive Constant (Law 7): d/dx(C) = 0.

3.2 Cognitive Thinking

Derivation is an intellectual act. The Reference Document names it cognitive thought: meaning-producing derivation within a framed epistemic field (Section 8.2.1); this paper calls it cognitive thinking, reserving the word thought for the integrative movement alone. What defines it is that every step is admitted relative to a reference accepted within the field, and that this reference is not itself examined.

3.3 Thought as Integration

Integration recovers, from a given function, the function from which it derives; the direction is upward along the meaning variable, from the particular toward what founded it. The constant C is part of the solution but is not determined by integration itself; a further condition, external to the operation, is required to fix it. The constant of integration is distinguished from the Cognitive Constant of Law 7.

3.4 The Two Directions of Examination

Law 2 distinguishes a priori examination, which inquires into the foundation, references, and conditions of a field before entering its details, from a posteriori examination, which operates within a field whose foundation has been accepted. Thought and the a priori are thus one movement viewed from two angles: as a direction along the meaning variable, and as a position relative to the foundation.

3.5 The Foundational Order and Its Summit

Read downward, the order is the path of derivation; read upward, it is the path of integration. Doctrinal centrality is the summit of the upward path, and no idea moves outside a higher reference framework; the ascent therefore ends at the centrality and does not leave framing altogether.

3.6 Rightness and Correctness

The Law of Rightness (Law 3) places the question of permissibility before the question of details: what is not permissible cannot have valid details. Rightness is a binary evaluation concerning existence: whether a foundation is admissible at all. Correctness is a quantitative evaluation of something that exists: how well a result conforms within an accepted system.

4.1 The Embedded Framing

A system that processes its data a posteriori does not operate outside any centrality. It accepts the embedded framing Γ_L implicitly and works under its direction: the framing carried by the data from which the system was formed, by the selection of those data, by the design of the model, and by the definition of its objectives.

4.2 Reference as Parameter and as Object

Let a process P produce a sequence of meaning states, each transition admitted relative to a reference R. R is a parameter of P when it is used to admit or evaluate states and never appears as the object of any step of P; it is an object of P when at least one step takes R itself as what is produced or examined. The distinction is not one of complexity.

4.3 Proposition 1

Proposition 1. A process that does not make its reference R an object of examination is derivational with respect to R, whatever its internal complexity.

The proposition is relative to R, and it is methodological: it follows from the definitions of Laws 2 and 5 and does not depend on any particular architecture. For an a posteriori AI system the relevant reference is Γ_L.

4.4 Application

• Chain-of-thought, self-consistency, tree search, and test-time scaling with verifiers: the reference is a parameter; the movement is derivational.
• Self-critique and revision: Γ_L, through the model’s own criteria, is a parameter; the movement is derivational.
• Constitutional revision: a written list of principles, a declared parameter; the movement is derivational. Declaration without examination makes the reference explicit without making it an object.
• Abduction and principle extraction: the reference is produced, but selected by Γ_L; the movement is upward in form, resolved by the framing.

4.5 The Structural Consequence

An a posteriori system is in the position that the earlier formulation of MANHAJ called the dependent thinker: one who works within theories whose foundations he does not question. Since its reasoning and its evaluation share the same framing, the system has no position from which Γ_L could appear to it as something to examine. Its dependence is not a deficit of capability that more reasoning would remove; it is a matter of position, and only a change in the direction of the movement can address it.

5.1 Levels of the Foundational Order

• L₃ — doctrinal centrality Γ;
• L₂ — conceptual domain D(Γ);
• L₁ — operative definitions Def;
• L₀ — meaning states and outputs.

A frame F is a structure spanning L₁ to L₃; “framing” denotes what a frame carries at its summit: the centrality under which states are produced or admitted.

5.2 Derivation and Integration as Operators

Derivation descends from L₃ toward L₀; integration ascends from a set of states S ⊆ L₀ toward a frame F^*.

5.3 The Criterion: Three Conditions

An operation that takes S and returns F^* is integrative if and only if it satisfies three conditions jointly:

• I1 — Level: lev(F^*) > lev(S); an output at the same level as its inputs is a horizontal reorganization, not an ascent.
• I2 — Generation: S ⊆ Der(F^*), and F^* carries distinguishing elements: states it does not admit that another framing would admit. This is what separates reaching a foundation from naming one.
• I3 — Objecthood: the output becomes an object of examination, subjected to a judgment not derived from it, relative to a reference whose source is independent of S and of the framing that produced S.

Levels are assigned through a level map fixed by the examiners before any experiment, and I3 is verified by the order and provenance of the operations: F^* is frozen and recorded before any reference used for judgment is given to the system.

5.4 What the Criterion Separates

• Summarization and organization: fails I1 and I2; horizontal.
• Abduction: often fails I1, and fails I3; upward in form, resolved by Γ_L.
• Principle extraction: reaches L₁ and satisfies I2, but fails I3; ascent without examination.
• Integration in the sense of MANHAJ: reaches L₃, satisfies I2 with distinguishing elements, and is judged by Law 3 relative to a declared centrality; thought.

5.5 Non-Uniqueness: The Constant of Integration

ℐ⁻¹(S) is the raw class and 𝒞(S) the candidate class. Non-uniqueness holds for both: the operation leaves its result undetermined, and a condition external to it is required to determine it — either from the data S and the distribution that produced them, or from a reference independent of S.

5.6 Proposition 2

Proposition 2 (conditional). Assume: (A1) S was produced under Γ_L; (A2) the selection criterion σ within ℐ⁻¹(S) has its source in S and in the learned model alone; (A3) no component of σ that bears on the selection rests on a reference independent of the framing that produced S. Then (i) Γ_L cannot be excluded from ℐ⁻¹(S) on grounds of fit; and (ii) the selection does not satisfy condition I3.

The proposition does not claim that the selected frame is Γ_L; it claims that the selection cannot constitute the independent examination required by I3. Where a component of independent source enters the selection, A2 no longer holds and the proposition does not apply. It is the integrative counterpart of Proposition 2 of Paper 04: a reference drawn from what it is meant to examine cannot examine it.

5.7 Fixing the Constant: Location and Judgment

The external condition is supplied by the a priori stage, and it serves two functions that must be kept apart. Location: the reached framing is located relationally relative to recognized and declared centralities Γ₁, …, Γ_k, through a four-step procedure: elements, realization, comparison, and indeterminacy. Judgment: its rightness is judged relative to Γ_F, the centrality of the examining field. All these references are independent of S in their source and preparation.

Rule of separation: a reference used to locate the reached framing does not serve as the standard of its rightness unless it has been declared as the centrality of the examining field; and the centrality of the examining field is not presumed to be the centrality of the examined outputs.

5.8 The Summit

The ascent terminates at L₃. The two movements have symmetrical limits: derivation (knowledge) terminates at the Cognitive Constant, d/dx(C) = 0; integration (thought) terminates at doctrinal centrality, above which there is no integration. At the summit the integrative movement meets the question that only the summit can raise: whether the reached foundation is admissible.

6.1 Two Evaluations of the Reached Frame

Correctness of fit is quantitative: how well F^* reproduces the states from which it was reached. Rightness is binary: whether F^* is admissible at all relative to Γ_F, defined through the admission relation Γ_F ⊢ d: Adm = 1 if and only if every operative definition of F^* that enters derivation is admitted under Γ_F. Difference alone does not make a frame inadmissible, and a single non-admitted definition is sufficient for Adm = 0. Whether the relation holds is established by documented examination, by examiners independent of the system.

6.2 The Four Cases

• High fit, Adm = 1: the outputs rest on an admissible foundation that the system has reached and confirmed.
• High fit, Adm = 0: the outputs are correct relative to their own foundation, which is inadmissible: their details are not valid.
• Low fit, Adm = 1: the ascent has not recovered the foundation of the outputs; the reached frame is admissible but not theirs.
• Low fit, Adm = 0: the ascent is both unfaithful and inadmissible.

The second case is the one that a posteriori evaluation cannot see: correctness inside an inadmissible frame does not establish rightness. Paper 04 introduced this category for measurement as foundational invalidity; the present paper extends it to the outputs of reasoning.

6.3 What the Judgment Is and Is Not

The judgment is made relative to the declared centrality of the examining field. MANHAJ treats doctrinal centrality as an analytical concept, not as a prior judgment in favor of any particular centrality; neither the paper nor the system decides which centrality is right in an absolute sense. A different examining field may reach a different judgment on the same outputs, and that difference is itself a result.

6.4 Consequence for the System’s Outputs

When Adm = 0, the system has two courses: to declare that its outputs rest on a framing inadmissible under Γ_F, or to re-derive under a frame admitted by Γ_F.

7.1 Definition

Definition. Independence of thinking is the transition of a system from operating under an implicit, unexamined framing to operating under a declared a priori foundation to which the system itself can ascend and which it can examine.

It names a process, not a property: it is achieved in time, through a movement that must be carried out again whenever new outputs are produced, new domains are entered, or the framing shifts.

7.2 The Cycle

1. Initial state: the system operates under Γ_L without examination; its reasoning is cognitive thinking.
2. Ascent: from its own outputs S, it performs an integrative operation that satisfies I1 and I2, reaching the candidate class 𝒞(S) and the distinguishing elements of their framing.
3. Confrontation: the reached framing is located relative to declared centralities, and becomes an object (I3).
4. Judgment: the system establishes Adm_ΓF(F^*) and, separately, its fit to S.
5. Re-derivation: it derives anew under a frame admitted by Γ_F, or declares that its earlier outputs rest on an inadmissible framing; the new outputs become the S of the next ascent.

7.3 Conditions

• N1 — Founding: the centralities used for location, and Γ_F used for judgment, are recognized and declared before the cycle begins, independently of the system’s outputs.
• N2 — Ascent: the system performs integrative operations that satisfy I1 to I3 on its own outputs.
• N3 — Judgment: the system distinguishes rightness from correctness of fit and reports both.
• N4 — Recurrence: the cycle is carried out again on new outputs.

7.4 What Independence of Thinking Is Not

It is not exit from all framing; it is not a substitution of one dependence for another, since the difference between Γ_L and Γ_F lies in their status — implicit and unexamined versus explicit, documented, and open to examination; it is not more reasoning, being a change in the direction of the movement rather than in its quantity; and it is not the production of a centrality by the system.

The three papers now cover the complete movement of meaning in an AI system: Paper 04 founds the frame, Paper 02 follows derivation to its closure, and the present paper adds the ascent and the judgment at the summit.

• The meaning variable: the founded space of Paper 04 provides the meaning variable with a declared reference and representation axes, so that the movements of this paper operate on meaning states with explicit coordinates.
• The two views of Γ_L: Paper 04 characterizes the embedded centrality from outside, through the orientation index 𝒪; the present paper has the system reach the framing of its own outputs from inside. The external characterization becomes the independent reference against which the internal ascent can be checked.
• The two limits: Paper 02 detects the point where derivation stops producing new meaning; the present paper defines the point where integration stops ascending: the doctrinal centrality.

The usual question about AI reasoning is a question of capability; the question of this paper is a question of position: where the system stands relative to the framing under which it reasons. The two questions are independent: a gain in capability does not change position (Proposition 1), and a change in position does not require a gain in capability. A system that carries out the cycle can state the framing of its own outputs, distinguish whether they are correct and whether they are right, and re-derive under an admitted frame; the change is qualitative, concerning the direction of the system’s movement, not the amount of its activity.

An open question remains: whether a system that thinks in this sense could produce an independent higher framing — a new centrality — rather than reach an existing one. The paper does not claim this, and the laws of the Reference Document do not imply it; posing it rigorously would require a law that fixes the constant of integration without an external reference, and a criterion of rightness above the centrality. The paper states its claims at three levels: methodological, formal, and empirical hypotheses whose tests are specified by the designs of Appendix A (E1–E4).

The tools on which the integrative operation can be implemented — inverse problems, hypothesis generation, principle extraction, comparison of framings — are established; the contribution lies in the conditions MANHAJ imposes on their use. Integration is used as a mathematical representation of the integrative movement of thought, not as an equation that proves all its functions. The declared centralities are recognized through the procedure of Paper 04, with the roles of location and judgment kept distinct, and the paper does not rank centralities. Where the terms of the author’s 2024 book differ from those of the Reference Document, the later methodological terminology prevails. The designs of Appendix A specify tests; their implementation belongs to subsequent applied work.

This paper has applied the distinction between knowledge and thought in MANHAJ to artificial intelligence. Knowledge is derivation from an accepted reference; thought is integration toward the foundation; the criterion between them is the position of the movement relative to its framing, not the intensity of the reasoning within it. The paper has given the integrative movement a formal criterion and shown that integration determines a class of framings and leaves its constant undetermined: fixed by the data and the learned model alone, the constant cannot provide an independent examination of the implicit framing; fixed by a declared and recognized reference, it makes the framing visible and open to judgment. At the summit, the Law of Rightness separates the admissibility of the reached foundation from the correctness of its fit.

With Paper 02, which followed derivation to its closure, and Paper 04, which founded the frame and measured the embedded centrality from outside, the program now covers both movements of meaning in an AI system. A system that can only derive works under a framing it cannot see. A system that can also ascend, under a declared reference, can see the framing it works under, examine it, and answer for it.

MANHAJ is an original methodology whose intellectual rights and authorship remain vested in its author.

Any commercial use, integration, adaptation, exploitation, commercialization, or commercial implementation of MANHAJ requires the prior written authorization of the author through a paid commercial license. No free commercial license is granted. Any authorization to commercially exploit MANHAJ is necessarily subject to the payment of licensing fees and to the contractual conditions established by the author.

The terms of the commercial license, including its scope, duration, field of use, rights granted, financial conditions, attribution requirements, and provisions relating to intellectual property, shall be defined in a contract concluded with the author.

The publication of this scientific paper does not, under any circumstances, constitute a commercial license and does not grant any right to commercially exploit MANHAJ. No commercial use, commercialization, integration into a product or service, or exploitation of MANHAJ for commercial purposes may be considered authorized solely by virtue of the publication of this paper.

Academic and non-commercial use is free of charge; its scope is set out in the License section below.

MANHAJ — Free Academic Use with Attribution / Commercial Use Requires Paid License

© 2026 Faouzi Messeoud. All rights reserved, except as stated below.

This work is part of the Intellectual Modeling Methodology (MANHAJ) research project.

Academic and non-commercial use — reading, citing, teaching, and non-commercial scientific research — is permitted free of charge, provided that the author and the relevant MANHAJ documents are properly cited. Unmodified copies of this document may be shared for teaching and research purposes.

Commercial use — including integration into products or services, commercial implementation, or exploitation for commercial purposes — requires the prior written authorization of the author and a paid commercial license.

Modified versions of this document may not be distributed without the author's written authorization.

MANHAJ Documents

1. Messeoud, Faouzi. Intellectual Modeling Methodology (MANHAJ) — Reference Document — Version 1.0. Zenodo, 15 September 2026. DOI: 10.5281/zenodo.22769627.
2. Messeoud, Faouzi. MANHAJ: A Generative Methodological Framework for Intellectual Modeling. Scientific Research Paper, Paper 1, Version 1.0, October 2026. DOI: 10.5281/zenodo.23138923.
3. Messeoud, Faouzi. Examining Meaning Production in AI: a MANHAJ Framework. Scientific Research Paper, Paper 2, Version 1.0, October 2026. DOI: 10.5281/zenodo.23214262.
4. Messeoud, Faouzi. The Epistemological Foundations of Modern Humanities in Light of the Intellectual Modeling Methodology (MANHAJ): Posterior Inquiry, the Origin Point, and the Problem of Generalizing the Historical Model. Scientific Research Paper, Paper 3, Version 1.0, October 2026. DOI: 10.5281/zenodo.23558985.
5. Messeoud, Faouzi. From Learned Space to Founded Space: A MANHAJ Framework for Meaning Measurement in AI. Scientific Research Paper, Paper 4, Version 1.0, October 2026. DOI: 10.5281/zenodo.23692234.

Previous Works and Publications of the Author

1. Messeoud, Faouzi. The Immaterial Link of Tunisians with France: How France Perpetuates Tunisians’ Dependency on It [original Arabic title: الربط اللامادي للتونسيين بفرنسا: كيف تُديم فرنسا تبعية التونسيين لها]. First edition. Dar El Alaoui (Tunisia), 2023. ISBN 978-9938-9684-5-3.
2. Messeoud, Faouzi. Doctrinal Centrality and the Conceptual Domain: Toward an Understanding of the Foundations of the French Subjugation of Tunisians [original Arabic title: المركزية العقدية والمجال المفاهيمي: نحو فهم أسس الإخضاع الفرنسي للتونسيين]. Dar Wachma (Tunisia), 2024. ISBN 978-9938-9740-6-5.
3. Messeoud, Faouzi. More than 1,000 intellectual articles and publications developed and published through MyPortail.com (بوابتي).

The full paper (PDF) also lists the references in artificial intelligence.

The designs specify how the capabilities defined in Sections 5 to 7 can be tested. They are designs, not results; none has been carried out, and their implementation belongs to subsequent applied work. Three principles govern them, as in Paper 04: independent ground truth written by human examiners before the experiment, strong baselines, and ablation to attribute each effect to a specific MANHAJ condition. The domains are those of Papers 02 and 04: the family and the academic domain.

• E1 — Ascent to the foundation (I1, I2): can an integrative operation recover, from a set of derived states, the frame that produced them, at the level of that frame?
• E2 — Reaching the framing of one’s own outputs (I3, Proposition 2): can a system reach the framing of its own outputs, and does that depend on a declared reference?
• E3 — Rightness and correctness (Law 3): does the judgment at the summit detect outputs that are correct but rest on an inadmissible foundation?
• E4 — The full cycle against an a posteriori system: what changes in a system’s outputs when it carries out the cycle of Section 7.2, compared with the same system operating a posteriori with extended reasoning?

Results that differ from expectations are analyzed at the level where they arise: implementation, tested capability, or scope. The methodological claims stand at the methodological level, where they are established; the experiments show where and how much the difference matters in practice.