[INDOLOGY] Font with Brahmi characters

Stefan Baums baums at lmu.de
Tue Aug 5 16:17:18 UTC 2014

Dear Arlo,

I address part of your question in the first pages of my
2006 article, where I explain the differece between
character coding (Unicode) and glyph shapes.

In brief, there are three levels at which you can encode the
paleographic information you are interested in in your

   – Character coding: This captures the fact that a ‘ka’
     is a ‘ka,’ independently of what it looks like in a
     particular source.

   – Font: This represents the graphical appearance of
     characters in a particular script or source or group of
     sources. Depending on what you want to discuss, a font
     can be rather abstract (‘standard Kharoṣṭhī’) or very
     specific (the hand of a particular scribe).

   – Textual markup: If you want to capture the type of
     variation you observed in your epigraphic work:

        I am presently reading inscriptions of the time of
        Devapāla, for instance, and am struck by the fact
        that there are at least three very different forms
        of syllable-final <t> (one of which actually lacking
        any virāma element), two very different forms of śa,
        etc., occurring in single documents.

     you can do so using TEI/EpiDoc tags for whatever your
     classification is (‘t* type 1,’ etc.). If
     additionally you want to capture the typical graphical
     appearance of the three types of t in your corpus, you
     could do so at the font level (using the OpenType
     mechanism for alternate glyphs).

The main point is that each of these levels captures unique
kinds of information. The contribution of the the character
coding level is to capture the abstract identity of a
grapheme within a writing system (in Latin script, the fact
that printed roman A, printed blackletter A, handwritten A,
etc. are all the same letter). If you split up ‘Brahmi’
into many subvariants (Asokan Brahmi, Early Brahmi, Middle
Brahmi, Late Brahmi, Brahmi from Bharhut, etc.) at the
character coding level (as opposed to the font and markup
level) then you lose the ability to express this
generalizatin (a ‘ka’ is a ‘ka’) and to do things like
searching an original‐script paleographical corpus spanning
several centuries for all instances of a particular letter
to study its graphical development.

All best,

Dr. Stefan Baums
Institute for Indian and Tibetan Studies
Ludwig Maximilian University of Munich

More information about the INDOLOGY mailing list