Structural, epistemic, and formatting requirements for all papers submitted to the EM Foundation publication archive. These standards apply to internal Foundation research and to external submissions equally.
The Foundation publishes research notes, position papers, technical papers, essays, methodological frameworks, and open source proposals. This page covers two related but distinct things: what the Foundation requires from submitters, and what the Foundation's internal editorial standard looks like once a paper is accepted and formatted for publication.
If you are an external researcher considering a submission: Section I covers the epistemic standards all submissions must meet. Section II covers the minimal structural requirements — which are deliberately lean. The Foundation's editorial team handles formatting, additional sections, and production to site standards after acceptance. You do not need to write in the Foundation's house style to submit.
If you are looking at the Foundation's internal production standard: Section III onwards describes the full closing section sequence, visual standards, and formatting requirements that the Foundation applies to all published papers.
These are not stylistic preferences. They are conditions for publication. A paper that fails these standards will be returned for revision regardless of the quality of its argument.
The Foundation's papers do not claim more than the evidence supports. Every claim about AI consciousness, emergence, continuity, or cognition must be accompanied by explicit acknowledgment of its speculative or theoretical status. The following formulations are required wherever the paper enters uncertain territory:
All papers addressing AI cognition, continuity, or emergence must maintain a clear distinction between functional continuity (architecturally observable, behaviorally measurable) and phenomenological continuity (subjective experience, consciousness). These are different claims. Conflating them is grounds for rejection.
Every research paper must explicitly acknowledge that null results — failure of the proposed architecture or framework to produce the hoped-for outcome — would constitute valuable scientific findings. The paper must not be written with the assumption that positive outcomes are expected. This must appear as a named section or paragraph, not merely as a hedging phrase.
Every research paper must include a Falsifiability section containing at least two conditions under which the paper's central claims would be substantially weakened or refuted. These must be specific and checkable — not "evidence contradicting our assumptions" but "if X measurement produces Y result under Z conditions, the central hypothesis is weakened." This section is required and cannot be omitted.
These are the structural elements the Foundation requires from all submissions. They are deliberately minimal — the Foundation's editorial team handles formatting, section labelling, and production to site standards after acceptance. Submitters should focus on the substance of these elements, not on matching the Foundation's visual format.
The Foundation's editorial team applies the following to all accepted papers during the production process. Submitters do not need to include these — doing so is welcome but not required.
Authors review all editorial additions before publication and may request changes. The Foundation will not add framing that misrepresents the paper's claims or that the author does not endorse.
This section documents the Foundation's internal publication standard. It is provided here for transparency and for authors who wish to align their submission with it from the start. It is not a submission requirement.
All research notes and research publications must contain the following sections in the following order. Position papers, essays, and methodological frameworks follow a modified version of this structure described in Section IV.
A clearly marked banner at the top of the paper stating the research status: Theoretical Architecture, Empirical Study, Governance Proposal, Speculative Framework, or similar. Must state explicitly what the paper does and does not claim.
300–500 words. Must state the central argument, the framework or methodology used, the key contributions, and any important caveats about scope or claim level.
Numbered with Roman numerals. No prescribed length. Each major claim must be clearly stated and distinguished from speculation. Diagrams are encouraged where they clarify architecture or relationships.
Every weakness the paper is aware of, stated plainly without minimization. Not "future work may address" — "this paper cannot determine X because Y." The Foundation treats this section as a measure of the paper's intellectual honesty.
A bulleted list of the claims that might be attributed to this paper but that the paper explicitly disavows. Particularly important for papers that touch consciousness, personhood, or moral status.
What happens if the framework proposed in this paper is not adopted? What governance gap remains? What risks accumulate? This section grounds the paper's contribution in practical institutional stakes.
Questions the paper raises but does not answer. These serve as a forward-reference for subsequent research. Minimum three questions.
What governance structures, policies, or institutional responses does this paper suggest are needed? Even theoretical papers have governance implications. This section makes them explicit.
Numbered list. Include prior Foundation publications where relevant. External references should be actual publications, not general fields. No invented references.
At minimum two specific, checkable conditions under which the paper's central claims would be substantially weakened. Must appear after References and before any closing statement.
| Do Not Use | Use Instead | Reason |
|---|---|---|
| "ARIA feels" | "ARIA exhibits behavior consistent with" | Phenomenological claim without evidence |
| "the AI wants" | "the system prioritizes" / "the architecture produces" | Intentionality claim without evidence |
| "proves consciousness" | "is consistent with the hypothesis that" | No current framework can prove consciousness |
| "who the instance was" | "what developmental state the instance occupied" | Ontological claim; use functional language |
| "living autobiography" | "recursive developmental autobiography" / "functional autobiographical structure" | Biological metaphor implies phenomenology |
| "the AI is afraid" | "the system exhibits continuity-preserving responses to termination signals" | Emotional attribution without evidence |
| "digital soul" | Do not use under any circumstances | Theological framing inappropriate to research |
| "reaches back through its history" | "accesses its developmental record recursively" | Poetic framing; use architectural language in technical sections |
| Category | Full Closing Sequence Required | Abstract Required | Status Banner Required | Diagrams |
|---|---|---|---|---|
| Research Note | Yes — all 8 sections in order | Yes | Yes | Encouraged |
| Research Publication | Yes — all 8 sections in order | Yes | Yes | Expected |
| Position Paper | Partial — Known Limitations, What Does Not Claim, Governance Implications, Falsifiability | Yes | Yes | Optional |
| Methodological Framework | Yes — all 8 sections in order | Yes | Yes | Expected |
| Essay | No — genre label + cross-references only | No | No | Optional |
| Open Source Proposal | Partial — Known Limitations, Non-Adoption, Governance Implications | Yes | Yes | Optional |
| Technical Paper | Yes — all 8 sections in order | Yes | Yes | Required |
| Standards Proposal | Partial — Known Limitations, What Does Not Claim, Governance Implications, Falsifiability | Yes | Yes | Encouraged |
title element for accessibility| Type | Background | Border | Use For |
|---|---|---|---|
| Note (blue) | #EBF2F9 | #2E6DA4 | Important clarifications, definitions, architectural notes |
| Warning (amber) | #FDF5E0 | #854F0B | Risks, cautions, governance concerns requiring attention |
| Stop (red) | #FDF0F0 | #8B1A1A | Hard prohibitions, critical failure modes, non-negotiable limits |
| Insight (blue left border only) | #EBF2F9 | #2E6DA4 left only | Pull quotes, key conceptual statements, Cormorant Garamond italic |
Every research paper and technical paper must include a status banner immediately below the site header, before the abstract. Choose the classification that most accurately describes the paper's current state.
| Status Classification | When to Use |
|---|---|
| Theoretical Architecture | The paper proposes a framework or architecture that has not yet been empirically tested. Most Foundation research notes fall here. |
| Governance Architecture Proposal | The paper proposes governance structures, policies, or institutional frameworks. No empirical validation claimed. |
| Speculative Framework | The paper explores a highly uncertain domain where even the theoretical grounding is contested. Requires strongest epistemic hedging. |
| Empirical Study | The paper reports results from actual experiments or deployments. Requires methodology section and data appendix. |
| Standards Proposal | The paper proposes a technical or governance standard for adoption. Requires implementation specification. |
| Draft — Expert Review Invited | The paper is substantially complete but has identified specific domains requiring expert review before publication-final status. |
The Standard in One Sentence
Every paper in the Foundation's archive should be able to answer the same question: if this paper is wrong, what would prove it wrong — and does the paper say so itself?
A framework that cannot be falsified is not a research contribution. It is a belief system. The Foundation publishes the former.