Author Archives: Charles

The bend of the leaf

In the process of preparing manuscripts for digital reproduction, I often become more sensitive to the aspects which cannot be reproduced. High-quality, instantly-accessible images are an incredible boon to researchers around the world, but, inevitably, a representation of a 2-dimensional surface will not be able to capture a 3-dimensional object persisting in and changing over time. Moreover, as a human artefact, it is embedded in a rich cultural and historical context, which is only hinted at through imagery.


Two consecutive folios from Indien 54. All images courtesy of the BnF.

Two consecutive folios from Indien 54. All images courtesy of the BnF.


The interaction between materiality and human culture is captured beautifully in palm leaf manuscripts such as Indien 54, pictured here. Continue reading

Punctuation poetry in MS Sanscrit 616

In India, the 19th century was a period in which manuscripts, lithographs, and printed books were all being produced at the same time. We often imagine a linear sequence in which manuscripts are copied as lithographs, which are then replaced by books, but reality is usually not so straightforward. In fact, manuscripts were sometimes produced by copying lithographs and printed books by hand, for a variety of reasons — for example, see Francis 2017, “The Other Way Round: From Print to Manuscript”.


Sanscrit 616. All manuscript images are courtesy of the BnF.

Sanscrit 616. All manuscript images are courtesy of the BnF.


Sanscrit 616 is a particularly interesting example of this phenomenon. It is a palm-leaf manuscript of the Amarakośaviveka, copied from a lithograph made in 1850 in Pune. Continue reading

A wedding invitation around MS Sanscrit 1868

Pieces of scrap paper are often used as sleeves or envelopes for manuscripts. The re-use of discarded material was also prevalent in bookbinding, where they were employed to line or repair a binding, or as endpapers — for example, see “The Fix: Binding Waste” (Smith 2014). The result is that material that was once considered ephemera — such as newspapers, catalogs, or even medieval manuscripts — are unintentionally preserved.

Paper sleeve around MS Sanscrit 1868. All images courtesy of the BNF.

Paper sleeve around MS Sanscrit 1868. All images courtesy of the BNF.

The sleeve around MS Sanscrit 1868 is a slip of paper folded in half. Continue reading

An annotation positioning system in MS Sanscrit 612

In Sanskrit manuscripts, marginal annotations are often followed by a number, indicating the line where that annotation applies (counting either from the top or from the bottom). But in the Amarakośa manuscript that forms part of Sanscrit 612, the scribe takes a different approach.


Sanscrit 612, folio 1v

Sanscrit 612, folio 1v. All manuscript images are courtesy of the BnF.


This paper manuscript in Bengali script takes a familiar form, mimicking the shape and size of palm leaves. The text of the Amarakośa, a popular Sanskrit lexicon, is written four lines to a page, surrounded by a generous margin on four sides. In this manuscript, the scribe has taken the opportunity to fill the margins with a commentary on the text. However, as in this folio, the commentary is much too dense for the line-numbering system usually employed for marginal annotations. Continue reading

Quick palaeographer — a quick app for palaeography

As part of my research with the TST project, I need to study manuscripts in many different scripts from different periods. This can be especially difficult when trying to read annotations — the tiny, scribbled text on the margins of a manuscript. To help myself become familiar with how a particular scribe writes, I’ve created a quick tool for doing palaeography.


You can load an image of a manuscript, and then use SHIFT-click to draw boxes around akṣaras. Then, you can tag those akṣaras. They can have more than one tag; for example, in the demo above, I’ve tagged the ji akṣara as both “ji” and “i_diacritic”, so that it appears under both categories.

The manuscript in the demo is Sanscrit 611, a heavily-annotated Amarakośa in old Bengali script. The photo is courtesy of the BnF.

You can use the tool online at, or you can download the source code at The GitHub source also includes an example of a labeled image. I hope you’ll find it useful!

Back to squinting at annotations…