For research in the archive
The digitised archive is stratified. Archēglyph is built to read across it.
The periodicals, books and papers of South Asia and the wider Global South, scanned only as pictures of pages, become text you can read, search, translate and cite — in any script, with every line traceable to the page it came from.
Funding or partnering? Read the research programme →
The page, as scanned · p. 1
আমার সোনার বাংলা, আমি তোমায় ভালোবাসি।
চিরদিন তোমার আকাশ, তোমার বাতাস, আমার প্রাণে
Original · transcribed
আমার সোনার বাংলা, আমি তোমায় ভালোবাসি।
Read by machine, 21 Sep 2026 — checked against the scan
Romanisation · IAST machine-written
āmāra sonāra bāṃlā, āmi tomāẏa bhālobāsi.
Translation · English machine-written
My golden Bengal, I love you.
Page 1, line 1 — one click back to the scan
Three things decide what a researcher can read: a paywall, a language, and what somebody once chose to scan.
-
Paywalls
The great databases of nineteenth-century print are open only to institutions that subscribe.
-
Pictures of pages
Vernacular periodicals across the Global South are often scanned — but only as images, which no search can see into.
This is where Archēglyph works.
-
The gap on the shelf
Some material was never scanned at all: left out by choice, not absence.
What you can do with it
From pictures of pages to evidence you can cite
-
Read any script
Pages scanned as pictures become text you can read, search and quote — Bengali, Urdu, Hindi, Arabic, Chinese, Russian and more.
How a page is read → -
Read beside a translation
The original, a romanisation in a named scheme, and a translation into your own working language, line for line.
Inside the reader → -
Find themes across languages
Passages gathered by what they are about, so a Bengali weekly and an English daily can meet on the same subject.
How themes are found → -
Cite every line
Every passage leads back to its page, and every collection is kept as a fixed edition with its own link.
How provenance is kept →
The Global South at the centre
The archive of the Global South should be readable from the Global South.
The digital record of the nineteenth century was built largely in, and for, the Global North. Researchers in South Asia and across the Global South meet its gaps twice: in the vernacular collections no search can see, and in the subscriptions and equipment they are assumed to have. The echoes of empire are still audible in what gets digitised, in which languages, and for whom. Archēglyph starts from the other end.
-
Researchers without subscriptions or special equipment
Scholars in the Global South, and early-career and independent researchers everywhere. A web browser and a photograph of a page are enough, and it is free during the pre-release period.
-
Libraries and archives holding vernacular collections
Institutions that have digitised their holdings as page images, and want them read, searched and cited — with every reading traceable to their own scans.
-
Readers in their own languages
Translation goes into the language the reader works in, not only into English — and the original script is never replaced by it.
Research agenda · not yet in Archēglyph
What we are working towards
-
Mapping the archive gap
Which collections of the Global South have been digitised, by whom, on whose terms — and what the deliberate gaps on the shelf say about the politics of digitisation.
-
Reading the scripts machines read worst
Open evaluation of how well each script is read, starting with Bengali, and work to improve the weakest — research in its own right, published as it goes.
-
A community across the Global South
Workshops, training and shared collections with researchers, librarians and archivists — built with the institutions that hold the material.
Principles we hold it to
FAIR, CARE, and where we fall short →- Discoverable
- Vernacular print, searchable in its own script.
- Accessible
- Free during pre-release; published collections are public.
- Reproducible
- Fixed editions that record how they were made.
- Open
- Collections download whole, in open formats.
- Ethical
- No trackers. Nothing a machine writes is passed off as the source.
- Citable
- Every passage leads back to its page.
Essays
All essays →-
A translation is not the source
A machine translation is a claim about a text, produced by a named model on a named day — not the text itself, any more than OCR output is the page.
-
Reading clusters as a researcher
The Archēglyph cluster view leads with quotations, not scatterplots. Here is how to use it — and why the scatterplot is behind a toggle.
-
The citable-claim test
A simple test for a research tool's output: can you, in one click, see the page a claim came from? If not, it's for exploration, not for scholarship.
-
Downstream of Trove: analysis in the corpus stack
Digitisation projects like Trove, Chronicling America, and Europeana produce the corpus; Archēglyph produces the analysis on top of it — one stack, not competitors.
Take part
For researchers
Bring a box of scans
- Upload pages in any script and read them back as text.
- Read one book closely, or a whole run by theme.
- Publish a collection or a book for anyone to read.
- Free during the pre-release period.
For funders and institutions
Open more of the archive
- Opens digitised collections of the Global South as searchable, citable data.
- Responsible AI: every line traceable, nothing machine-written passed off as the source.
- Open, FAIR-minded editions that outlive the grant.
- Led by a humanities researcher, with a clear theory of change.