# Historical-text download research techniques

## Institutional repository verification

### Internet Archive

Use the documented endpoints to avoid guessing filenames:

- Search: `https://archive.org/advancedsearch.php?q=<query>&fl[]=identifier&fl[]=title&fl[]=creator&fl[]=date&rows=20&output=json`
- Metadata/files: `https://archive.org/metadata/<identifier>`
- Stable item page: `https://archive.org/details/<identifier>`
- Direct file: `https://archive.org/download/<identifier>/<filename-from-metadata>`

Inspect metadata fields such as `creator`, `date`, `publisher`, `contributor`, `source`, `rights`, and `licenseurl`. A missing `rights` field is not a public-domain determination. Use the file list to locate the exact PDF. IA OCR is commonly available as `<identifier>_djvu.txt` and is useful for checking title pages, tables of contents, translator attributions, and section boundaries.

### Direct institutional PDFs

Download and inspect rather than trusting search snippets:

- `pdfinfo file.pdf` for title, author, pages, encryption, and producer metadata.
- `pdftotext file.pdf -` or selected page ranges to verify language, edition statement, translator, and completeness.
- Check HTTP status and content type; a URL ending in `.pdf` may return HTML, a bot challenge, login page, or preview.

### Search result hygiene

- Library records, Deutsche Digitale Bibliothek entries, WorldCat, Google Books, and Open Library are discovery records unless a tested download is attached.
- Wikimedia Commons can supply explicit provenance/public-domain templates, but still verify the edition and creators.
- Scribd, personal mirrors, and unattributed “new translations” are not acceptable as authoritative legal sources without a clear rights statement and translator identity.

## German/EU translation-rights checklist

1. Identify the translator, not just the ancient author or volume editor.
2. Confirm death year from a reliable authority record.
3. Under the normal German rule, protection ends 70 years after death at the end of that calendar year; the work becomes public domain on 1 January of the following year.
4. For multiple contributors, determine whether the translation is a joint work and check the last-surviving relevant author.
5. Evaluate introductions, critical notes, reconstructed readings, and modern typography separately.
6. State the date and jurisdiction of the conclusion; availability may differ between the US and Germany/EU.

## Example lessons from pseudepigrapha/Qumran research

- A 1900 anthology can mix public-domain and still-protected translations because each translator has a separate death date. Do not assign one rights status to the whole volume.
- A nineteenth-century edition titled like a complete work may contain only translated fragments plus commentary. Verify the table of contents and body.
- Continuations published across two journal years may be combined into one institutional PDF; record both original page ranges and dates.
- Qumran works discovered after 1947 often have no public-domain German translation. In that case, identify the standard protected German edition, calculate its likely protection term, and distinguish manuscript images or English transcriptions from the requested German full text.
- When reporting exact URLs, include both the repository page and direct file link so provenance remains auditable even if the download path changes.