Research and impact
Widening who can read the archive, and in which languages
Archēglyph is a tool, and it is also a research programme: to make the print of South Asia and the wider Global South readable where it has been digitised only as pictures, to ask why so much of it was left out, and to open that work to the researchers who have been shut out of it.
Much of this page is research agenda — aims, not features. Whatever already works in Archēglyph is marked today; the rest is marked as next, or as an aim.
The problem, and why now
The digital record of the nineteenth century was built largely in, and for, the Global North. Subscription databases hold much of the English-language press. Libraries in South Asia hold millions of pages of vernacular periodicals and books, many digitised and freely available — but only as pictures of pages, which cannot be searched, quoted or translated. Some material was never scanned at all. The result is a record stratified by money, by language and by choice, and a transnational history that is very hard to write from it.
Researchers in the Global South meet those gaps twice: in what they cannot find, and in the subscriptions and equipment they are assumed to have. Machines have only recently begun to read non-Latin print well enough to change this. The question now is whether that reading will be done carelessly — with the source lost behind a fluent paraphrase — or with its provenance intact, and with the Global South at the centre rather than the margin. Archēglyph is built for the second.
Behind a paywall
Searchable, if your institution subscribes
Large English-language newspaper and periodical collections, digitised and fully searchable — for those who can pay.
Outside Archēglyph's reach: no software opens a paywall.
Scanned as pictures
Open to look at, closed to search
Vernacular periodicals and books from South Asia and across the Global South, freely digitised — but only as page images. They cannot be searched, quoted or translated.
Where Archēglyph works
- readable
- searchable
- translatable
- citable
The scan stays beside every line of text drawn from it.
The gap on the shelf
Never digitised at all
The deliberate gap: material left out by choice, not absence — and so absent from every search.
Outside Archēglyph's reach: no software reads what was never scanned. Naming the gap is part of the work.
Theory of change
From the gap in the archive to what changes for the field — with each step marked for what it is today.
-
The problem
The print of the Global South — vernacular and colonial-era — that exists only as pictures of pages, or behind paywalls.
the situation
-
What Archēglyph does
Reads it in any script, keeps its provenance, translates it, and finds its themes.
in use today
-
What it produces
Readable, searchable, citable collections, fixed as editions and downloadable whole.
in use today
-
What changes for researchers
Work across languages and archives — including by researchers in the Global South without subscriptions or specialist equipment.
beginning
-
What changes for the field
A less anglophone, more equitable archive; a community of researchers across the Global South and beyond; the public brought closer to the sources.
our aim
The priorities this work meets
Research councils and foundations in the UK, Europe, North America and Australasia are asking the same questions of digital humanities: is the AI responsible, are collections usable as data, is the work open, who holds authority over it, and does it reach beyond the institutions that already have the most? Here is where Archēglyph stands on each.
-
Responsible AI in the humanities
today Every machine-made line is signed and kept beside its source; nothing is summarised or answered in a machine’s own words.
aim Published evaluations of how well each script is read.
-
Collections as data
today Collections digitised only as images become searchable text, every passage tied to its page.
aim Reading open repositories’ collections where they already live.
-
Open research and FAIR data
today Fixed editions with permanent links, public when published, downloadable whole with a licence.
aim A persistent identifier (DOI) for each edition.
-
Community data governance
today Owners decide when a collection is published, and can unpublish or delete it.
aim Shared stewardship with the communities whose print is being read, in the spirit of the CARE principles.
-
Multilingual and under-served languages
today South Asian, Arabic-script and other scripts; romanisation in named schemes; translation into any working language.
aim Better reading for the scripts machines read worst.
-
Equitable partnerships with the Global South
today Free during pre-release, in a web browser, from a photograph of a page — no specialist equipment.
aim Work designed with the libraries and researchers who hold the material.
-
Skills and capacity building
today Plain-language guides and a glossary written for researchers, not engineers.
aim Workshops and training for researchers, librarians and archivists across the Global South.
-
Public engagement
today Published collections anyone can read; essays on method in plain language.
aim Public-facing work that brings readers closer to the sources, in their own languages.
-
Sustainable research infrastructure
today Open formats, and editions that download whole — so the work outlives the platform.
aim A long-term model that keeps it free for those without institutional support.
Where things stand
The full project roadmap →In use today
- Reading scanned pages in any script, part by part, with every transcription signed.
- A reader that sets the original beside a named romanisation and a translation.
- Themes found across languages, read as passages that lead back to their pages.
- Collections saved as fixed editions, each with its own link, downloadable whole.
Next
- Following people, places and organisations through a whole collection.
- Finding reprinted and near-identical passages.
- Seeing how themes rise and fall over time.
- A persistent identifier (DOI) for each edition.
Research agenda
Research agenda
The research agenda
Archēglyph is not only an engineering project. It began as a humanities question, and its direction is set by research — with technology research in service of it.
Mapping the archive gap
What does one do with the expurgations happening in the here and now — the deliberate gap on the shelf, driven by choice, not absence? This strand asks which collections of the Global South have been digitised, where, by whom and on whose terms, how the echoes of empire inform the way archives are digitised today, and how those choices bear on researchers without sophisticated equipment or institutional support.
Multilingual print, read across borders
The press of the colonial world was multilingual, and its history cannot be written from one language. Reading vernacular and English-language periodicals side by side makes cross-border connections — and a transnational understanding of the key terms of a field — possible where the digital archive has made them hard. It begins with Bengali periodicals, the subject of Maitrayee Roychoudhury's ongoing research.
Reading the scripts machines read worst
Text recognition was built for English and for Latin script, and it still shows. This technology-research strand measures, openly, how well each script is read — starting with Bengali and Urdu — and works to improve the weakest, publishing what it finds so that others can build on it.
Open repositories, read where they live
A great deal of multilingual periodical material has already been digitised and made openly available. We aim to work with open-access repositories — such as Heidelberg University Library's specialised information service for South Asia (FID4SA), the archives of the Centre for Studies in Social Sciences, Calcutta, and collections at the British Library — so that what they hold becomes searchable, translatable into a researcher's language of choice, and citable, without obscuring where it came from.
These are aims. No partnership with any of these institutions is implied.
Objectives
- Build on what exists
- Work downstream of the libraries and digitisation projects already doing the hard work of scanning — alongside them, not in competition.
- Confront bias
- Show how well each script is read, openly, and improve the ones machines read worst.
- Decolonise the data
- Make the vernacular print of the Global South as findable as English print, in its own script and on its own terms.
- FAIR and CARE
- Hold every collection to both sets of principles, and say plainly where we fall short.
- Reach peers and the public
- Discoverable, accessible, reproducible, open, ethical and citable — for scholars and for everyone else.
- Usable and free
- A tool a researcher can use without technical training or institutional money.
- An ecosystem of collaborators
- An international community that shares collections, methods and findings.
Public engagement and knowledge exchange
The intent is an international community of researchers — in the Global South and beyond — who use Archēglyph as a springboard: to forge connections, conduct new research, and share resources. Collections published with Archēglyph can already be read by anyone, and we write about method in plain language in our essays.
What we are working towards: workshops and training for researchers, librarians and archivists, held with institutions that hold vernacular collections; shared collections that other scholars can build on; and public-facing work that brings readers closer to the sources, in their own languages.
- today Public collections anyone can read
- today Essays on method, in plain language
- aim Workshops and training across the Global South
- aim A research community and shared collections
What support would make possible
Funding, a collection to read, or a community to convene — each would move part of the agenda above into use.
-
Opening open repositories
Making already-digitised multilingual collections readable, searchable and citable where they live.
-
Reading more scripts well
Evaluating and improving how Bengali, Urdu and other scripts of the Global South are read — the ones machines read worst today.
-
Mapping the archive gap
A study of what has been digitised, where, by whom and on whose terms — and what has been left out.
-
A community across the Global South
Workshops and training for researchers, librarians and archivists, and a network that builds on one another’s collections.
-
Keeping it free
For the researchers without institutional support that the project exists to serve.
Frequently asked questions
- What is Archēglyph?
- A digital humanities project that makes the scanned archives of South Asia and the wider Global South readable, searchable, translatable and citable, in any script. It is both a research tool, in use today, and a research programme on how archives are digitised and whom they are digitised for.
- Why focus on the Global South?
- Because that is where the digital record is thinnest and hardest to reach. Vernacular periodicals are often digitised only as page images that no search can see into; the large databases of nineteenth-century print sit behind paywalls and lean heavily towards English; and researchers in the Global South are often assumed to have subscriptions and equipment they do not have.
- How does Archēglyph use AI responsibly?
- Machines read the page, translate it and group passages into themes, but every line they produce is signed with what produced it and kept beside the source, never in place of it. Archēglyph does not write summaries or answer questions in its own words, and every passage can be walked back to the scan it came from.
- Is the work open, and does it follow the FAIR principles?
- Collections are saved as fixed editions with their own permanent links. Published collections are public, and every edition downloads whole in open, documented formats, with a record of how it was made and a licence chosen by its owner. A persistent identifier (DOI) for each edition is next on the roadmap.
- How does it approach the CARE principles?
- Collections stay private until their owners publish them, and owners can download, unpublish or delete them at any time. Sharing the stewardship of a collection with the community it comes from is part of the research agenda — an aim, not yet a feature.
- What exists today, and what would funding make possible?
- Reading pages in any script, the side-by-side reader, themes across languages and citable editions are in use today. Support would open repositories of digitised multilingual material, improve how poorly served scripts are read, map the gaps in the digitised archive, build a community of researchers across the Global South, and keep the tool free for those without institutional support.
- Who is behind Archēglyph?
- Maitrayee Roychoudhury, a researcher of Victorian and Bengali periodicals who directs the research, and Dipankar Sarkar, who builds the tool.
- How can a funder or an institution get in touch?
- Write to hello@archeglyph.com about funding, partnership, or a collection you would like read.
Who is behind it
Archēglyph is led by Maitrayee Roychoudhury, who directs the research, and Dipankar Sarkar, who builds the tool. Maitrayee's own account of why it exists, and both of their backgrounds, are on the about page.
Talk to us about funding or partnership
We would be glad to hear from funders, libraries, archives and research groups — whether you hold a collection, support this kind of work, or want to build on it.