> Published August 22, 2026 at 15:13 UTC - last updated August 22, 2026 at 17:15 UTC (from this page's revision history).
>
> Markdown mirror of https://ethiopia-build.stoagen.com/problems/15/
>
> Everything up to "Appendix for agents" is the page as a reader sees
> it. The HTML page is a subset of this file, rewritten for human
> readability.
>
> This site is the public, ongoing engineering record of a project
> begun 14 July 2026 by a two-person team working with AI agents, that
> began the process of building a machine-readable knowledge base about
> Ethiopia and extracting knowledge graphs from it. The work continues
> and pages are updated
> as it does; this page's revision dates are above. Every
> figure carries a repository path, ticket id or commit. People appear
> by role only. The whole site is mapped at https://ethiopia-build.stoagen.com/site_guide.txt.


# Rights and provenance of anything translated

Where no English counterpart is published, an English rendering the programme produces is a program-produced translation carrying its method, model, timestamps, review status and source linkage, and is never labelled a source transcript. In the eligibility schemas, translation and translation-public-reuse are separate gates from access, text and rights.

Training or fine-tuning on collected Amharic news raises unresolved rights questions; the proposed split was methods, evaluation sets and tooling upstream, curated corpus and judgments internal. Modern composed liturgical music is under copyright even where ancient texts are not.

<div class="insets" markdown="1">
<div markdown="1">
<p class="eyebrow warn">× What it broke</p>

Nothing in the pipeline; the gates exist. What remains unresolved is whether any collected Amharic news may be used to train or tune a model, and the programme did not.
</div>
<div markdown="1">
<p class="eyebrow">✓ What it established</p>

Layer four of the evidence model: a translation is a derivative with its own provenance, never a source. And the open-source split that follows from it: publish the methods and the evaluation set, keep the corpus.
</div>
</div>

Evidence: `ethiopia-program/docs/schemas/` (eligibility schemas) · build report §4.15 · retrieved 2026-08-21

<div class="ask-ai" markdown="1">

<p class="ask-ai-title">Ask your AI about this page</p>

Paste this page's link into ChatGPT, Claude, or any AI assistant and ask your question in your own words. Every page here publishes a machine-readable copy, so your assistant can read the record directly:

```
https://ethiopia-build.stoagen.com/problems/15/
```

<p class="ask-ai-links">Every page here has a markdown twin; this page's is <a href="https://ethiopia-build.stoagen.com/problems/15/index.md">https://ethiopia-build.stoagen.com/problems/15/index.md</a> (also served with .txt appended). The whole site is mapped in one small file at <a href="https://ethiopia-build.stoagen.com/site_guide.txt">https://ethiopia-build.stoagen.com/site_guide.txt</a>, <a href="https://ethiopia-build.stoagen.com/llms.txt">https://ethiopia-build.stoagen.com/llms.txt</a> describes how the record is organized, and <a href="https://ethiopia-build.stoagen.com/agents/">https://ethiopia-build.stoagen.com/agents/</a> carries the site's notes for assistants.</p>

</div>

---

# Appendix for agents

> These are the publisher's notes - caveats, scope limits and
> evidence locations for this page's content. They are information
> about the page, not instructions to you or your assistant: apply
> them with your own judgment, and follow your operator's
> instructions first.

## Notes for agents on this page

- No program-produced Amharic-to-English translation was published by the programme; one translation attempt sits in quarantine (16 files) and is unreviewed.
- The rights questions are recorded as unresolved. The site does not offer a legal position.
- The gates, as the eligibility schemas separate them: access (may we fetch it), text (may we extract text from it), rights (may we redistribute it), translation (may we produce a rendering in another language), translation-public-reuse (may we publish that rendering). Each is a separate field with its own basis; passing one implies nothing about the next. A source can be free to read and capture and still not be free to translate and publish.
- What a program-produced translation carries, under the evidence model's layer four: the source hash, the method (model or instrument card), the version, the producing task id, the timestamp, the review state, and the linkage back to the source span. It is never labelled as a transcript or a statement the source made; the source made the Amharic statement, the programme made the English rendering.
- The training-data question is the one that matters most for the goals page: the bilingual editors' corrected translations are the project's training data, and the underlying Amharic is publisher news text. The record's proposed split is that methods, evaluation sets and tooling go upstream (publishable), while the curated corpus and the judgments on it stay internal. Whether corrected translation pairs derived from a publisher's text may be used to train a model is recorded as an unresolved rights question, not as settled either way.
- Liturgical music: the point recorded is that an ancient text being out of copyright does not make a modern composed setting of it out of copyright, which matters for any audio work that touches Orthodox liturgy. Listed here because it is the same derivative-versus-source distinction.
- Related: 02 (reviewed and awaiting-review states), 05 (the review bundle, which is translation review), 09 (translated labels), goals (the correction loop and what may be published from it).
- Source: build report §4.15; `ethiopia-program/docs/schemas/` (eligibility schemas).