Scholia Goes Live

After four months of intense work, Scholia opens to the world with six monumental texts structured sentence by sentence with a framework to select, annotate and write (and receive feedback) upon the world's great texts. My ambition is make the study, contemplation and discussion of these foundational texts enhanced, if not reinvigorated, by digital technology.

When I first encountered Aristotle during my bachelor studies it was soon followed by several other, equally fascinating, encounters. These were the great commentators Al-Farabi, Avicenna, Averroes, Boethius, Thomas Aquinas and many others. It was ecstatic to study these thinkers who thought alongside the ideas from the mental supernova of antiquity. Of course, it was later that I realized such conceptual traditions are not unique to the scholiasts, but the norm in philosophy, intellectual traditions and theory-building at large. However, their connections are rarely made explicit but remain implicit and muddled (if not intentionally ignored). Some of this is owed to the practicalities of the matter and the available techonology, other to the necessity of ingenuity (or the attempt thereof), and some to plain laziness.

I do not know what Scholia will become, if it becomes anything, but I do know that it was a sublime experience to read Aristotle, and then find his very text in Averroes, enlarged, expanded and driven to new heights, to then find them both in Aquinas to still more glorious mental combat. I hope that this platform becomes an—if not itself then a stepping stone to—encouragement to partake in these discussions around the ideas and concerns that are ever as pressing to us now as the span of their age.

Want to help us decide which text to get into the platform? Head to this survey!

• https://www.surveymonkey.com/r/2BPSWZY

What We Start Out With

Now, for those still here, here are some numbers. Currently our database has 6 distinct works (the Bible, 2 Kant Critiques, Ibsen's Emperor and Galilean, Milton's Paradise Lost and Shakespeare's Sonnets). The selection is meant to test the robustness of the system in tackling works of a wide variety (poetry, prose and drama). Much time has been spent in trying to figure out a good way, technically, to bring these varieties into a common framework. This work is by no means done, nor is it free of errors and bugs, but it is enough to get started with.

We have, then:

• ~5.23M words across all works.
  • ~4.70M of words in the reading layer (approximately).
  • ~533k of words in orginal-orthography layer (approximately).
  • 28,779,217 characters in total.
• 262,157 sentences across all works (including original orthography).
• 7 block types in use: paragraph, heading, verse, speaker, stage, figure, separator.
• 12 reference systems.

And I hope that this is just the beginning!

Current Feature-Set

Reader.

• Sentence-addressable text. Every sentence is a first-class database row with a stable UUID, individually clickable and citable. This is the primitive everything else is built on.
• Dual-orthography display. Original spelling and modern reading text sit on the same sentence; toggle between them without losing your place (5 editions, 100% coverage).
• Interleaved facing translation. Source and translation locked 1:1 at sentence level, so German/Norwegian and English line up exactly, sentence for sentence.
• Genre-aware rendering- 7 block types (paragraph, heading, verse, speaker, stage direction, figure, separator). Drama renders as drama; verse renders as verse with line numbers; prose as prose.
• Multi-system pagination. A book can carry several reference systems at once (Kant shows both B-edition and Akademie-Ausgabe page numbers), with per-book citation defaults.
• Footnote popovers. Author footnotes resolve inline without leaving the page.
• Panel-based side view: TOC, "About this text", commentary, notes, resources, and article references as switchable panels alongside the text.
• Infinite-scroll paging with prefetch. Cursor-paged node loading, tuned buffer so TOC navigation doesn't stutter (admittedly, this is still not perfect, but it is really hard to tune right).
• Range selection across sentences. Drag to select a passage spanning multiple sentences, blocks, or nodes.
• Guided tour. First-run walkthrough of the reader UI.

Quotations and notes.

• Anchored quotations. A saved quotation binds to sentence UUIDs, not character offsets, so it survives editorial corrections to the text.
• Cross-translation projection. A quotation made in one Bible translation resolves onto the equivalent passage in another, following content alignment rather than verse numbers (839 alignment rows).
• Personal notes on quotations. Free-text notes attached to any saved passage, with tags.
• Auto-generated citations. Each book's cite_template produces the correct form (B 132, AA V 217, John 3:16, Book IX · 1002, Keiser Julian, Tredje handling · p. 214).
• Personal library views. Dedicated pages for all your quotations, all your notes, and your sources.

Writing and publishing.

• Long-form articles with an editor, draft → publish → archive lifecycle.
• Quotations embedded in articles. Pick from your saved quotations; they render as live, citable blocks that link back into the reader.
• Passage references. Articles register against the passages they discuss, so the reader can show "what's been written about this passage" from the text side.
• Peer review workflow. Request review on a draft, threaded comments with replies, review messages, activity feed, assignee management, withdraw.
• Editorial labels. Admin-applied labels on published articles.
• Series. Ordered, curated multi-article collections with drag-reorder.
• Topics and tags for classification.
• Public author profiles at /users/{handle}.

Bibliography.

• Sources and persons as structured records — full CRUD, source types (book, chapter, journal, web), parent/child sources, translation-of relationships, ISBN/DOI/publisher/place/year.
• Person–source roles (author, translator, editor…) as an explicit many-to-many.
• Resources attached to passages — secondary literature linked to specific places in a text, with a public submission queue and admin moderation.
• Cross-edition canonical passages — a resource attached to a passage surfaces on every edition of that passage, in any language.

What Is Not There Yet

A lot of things are on the "todo" list. Full-text search, facsimile page images, cross-references, and many, many more texts we would like to get in.

Known bugs include: interleaved facing source+translation break on anything other than the Kant works (these just were not extensively tested enough), links not working in articles (like this one).