PS : regarding e and o : if the examples do refer to long e and o, then the line should perhaps be interpreted as a correction over an existing dot.

On Thu, 16 Mar 2023 23:18:25 +0100
Satyanad Kichenassamy <satyanad.kichenassamy@univ-reims.fr> wrote:

>
> Dear Jean-Luc, Dear Manu (if I may),
>
> Indeed, I was talking about all of the forms both of you mentioned, namely
>
> - long ra with a vinculum
> - short ra with a macron. The macron may even be linked to the ra if one
> writes very fast.
> - line covering a puḷḷi (to correct a mistake)
>
> The first two enable disambiguation. It is not clear to me from a
> historical viewpoint whether they were introduced for the purpose, or
> whether they were natural evolutions (people do tend to link characters
> together when they write fast). The aesthetic dimension may also be
> relevant.
>
> These are forms in use in my family. I am attaching a photograph with
> examples in my own handwriting. My parents used the same and, as I said,
> I have encountered them in manuscripts but didn't make a special note of
> them since these forms were familiar to me.
>
> The vinculum is a line to "link together" several characters;
> mathematicians, especially British, used to write this over expressions
> as a substitute for a parenthesis. Thus, in (self-explanatory) LaTeX
> code, \overline{x+y} means (x+y). (This is consistent with the Latin
> etymology: a /vinculum/ is a link, /in vincula/ means in fetters etc.) 
> The first example seemed to me to be a vinculum rather than a macron
> since it links the two characters.
>
> About the examples with e and o, it may be that the additional marks are
> sloppily written puḷḷi-s and not bars. I am referring to தொல்காப்பியம்,
> எழுத்ததிகாரம், நூன்மரபு 15-16 that read :
>
> மெய்யி னியற்கை புள்ளியொடு நிலையல்
> எகர​ ஒகரத் தியற்கையு மற்றே.
>
> (For people who do not read Tamil, this means that consonants
> intrinsically stand with a dot, and that short e and short o have the
> same intrinsic feature. Tolkaappiyam is the first grammar of classical
> Tamil; it actually includes quite a bit of literary analysis in addition
> to grammar in the narrow sense and, as we see here, treats language as
> written as well as spoken.)
>
> Kind regards to both,
>
>          Satyanad K.
>
> Le 16/03/2023 à 20:12, Jean-Luc Chevillard a écrit :
> > Dear Satyanad,
> >
> > as a clarification,
> > are you talking about the special forms seen in this image,
> > taken from a 17th cent. Goa MS
> > in the words which would nowadays be printed as
> > ஆரோபிக்கிறது, ஆரோகணம், ஆரோக்கியம், ஆரோசை?
> >
> > (I obtained those images from Cristina Muru, many years ago)
> >
> > Best wishes
> >
> > -- Jean-Luc
> >
> >
> > On 16/03/2023 16:33, Satyanad KICHENASSAMY wrote:
> >>
> >> Dear Charles,
> >>
> >> Nice project. Please let us know when the revised font will be
> >> available. Regarding the long ra, I assume you are planning to create
> >> a glyph for the character with a vinculum? (Or is it already there?)
> >> Similarly, the short ra is often written with a macron to avoid
> >> confusion.
> >>
> >> One may want to include also the lines that are used to cover a
> >> pu.l.li (and thus, restore the vocalization).
> >>
> >> Best,
> >>
> >>       Satyanad K.
> >>
> >> Le 16/03/2023 à 11:42, Charles Li via INDOLOGY a écrit :
> >>>
> >>> Hello,
> >>>
> >>> At the TST Project, where we're cataloguing Tamil manuscripts, we've
> >>> forked Noto Tamil and started adding old ligatures, like pre-reform
> >>> ṇā, ṟā, etc. as well as some ligatures that don't seem to have
> >>> appeared before in print, such as the below-base "ma" ligatures. See
> >>> this page for examples:
> >>>
> >>> https://tst-project.github.io/editor/entities.html
> >>>
> >>> It's still a work in progress!
> >>>
> >>> Best,
> >>>
> >>> Charles
> >>>
> >>> On 2023-03-16 11:08, Satyanad KICHENASSAMY wrote:
> >>>> Dear All,
> >>>>
> >>>> To follow up on Harry Spier's query, the typesetting of the older
> >>>> Tamil characters (as well as Tamil Grantha) is sometimes
> >>>> problematic. I use Akshar Unicode for contemporary Tamil, which is
> >>>> very close to the standard printed characters, but insert some
> >>>> characters from Vaigai for the classical characters -- that were
> >>>> actually the standard characters when I grew up. For Grantha, the
> >>>> e-Grantamil font is nice even though less close to the characters
> >>>> in print, but the ligatures are sometimes undone automatically, for
> >>>> reasons that I do not understand. Also, I gather it is encoded in
> >>>> the same segment as Bengali, which is another source of confusion.
> >>>> The final output can be fine, though, see examples in the following
> >>>> paper:
> >>>>
> >>>> https://www.persee.fr/doc/crai_0065-0536_2018_num_162_4_96658
> >>>>
> >>>> This being said, if there is a better solution, I would be interested.
> >>>>
> >>>> For a diplomatic edition, it would be nice to have fonts that
> >>>> contain as many variants as possible. Similarly, Southern Sanskrit
> >>>> manuscripts should be reproduced in their original script if
> >>>> possible, especially in diplomatic editions. For instance, va and
> >>>> ba in printed Grantha are easier to disinguish than in Nagari (this
> >>>> is also true in those palm-leaf mss that I have used).
> >>>>
> >>>> I remember seeing proposals arguing that some characters usually
> >>>> encoded in Unicode as ligatures in Indic language fonts should be
> >>>> treated as stand-alone glyphs, at least in Tamil. The reason is
> >>>> that you sometimes see letters such as "mo" rendered as
> >>>> "kompu-(blank in a dotted circle)-lengthening mark-ma" which is of
> >>>> course nonsense. The placement of diacritics is also misleading at
> >>>> best, as was pointed out on this list a few days ago. This is in
> >>>> addition to the issues raised by Jean-Luc Chevillard (for instance,
> >>>> the ர் cannot be written without the lower diagonal stroke on some
> >>>> fonts).
> >>>>
> >>>> Of course, whether one decides to overlook the differences in
> >>>> variants of one character always involves judgment. An extreme
> >>>> example would be the different versions of the character 之 in the
> >>>> famous calligraphy 蘭亭集序 Lántíngjí Xù by 王羲之 Wáng Xīzhī. For
> >>>> India, the விநாயகர் சுழி vinaayakar cu_li has slightly different
> >>>> forms depending on writers, some of which may be worth recording
> >>>> (recall that this symbol is a form of the pra.nava; the same issue
> >>>> could be raised about the pra.nava in other scripts).
> >>>>
> >>>> Best,
> >>>>
> >>>>        Satyanad Kichenassamy
> >>>>
> >>>> On Wed, 15 Mar 2023 13:33:42 -0400
> >>>> Harry Spier via INDOLOGY<indology@list.indology.info> wrote:
> >>>>
> >>>>> Received thanks to Victor Davella
> >>>>> Harry Spier
> >>>>>
> >>>>>
> >>>>> On Wed, Mar 15, 2023 at 1:21 PM Harry
> >>>>> Spier<vasishtha.spier@gmail.com>
> >>>>> wrote:
> >>>>>
> >>>>>> Can someone recommend a good free unicode font for modern Tamil.
> >>>>>> I.e.
> >>>>>> provide a link to download this.
> >>>>>> Thanks,
> >>>>>> Harry Spier
> >>>>>>
> >>>
> >>> _______________________________________________
> >>> INDOLOGY mailing list
> >>> INDOLOGY@list.indology.info
> >>> https://list.indology.info/mailman/listinfo/indology
> >>
> >> --
> >> **********************************************
> >> Satyanad KICHENASSAMY
> >> Professeur des Universités
> >> LMR (CNRS, UMR9008)
> >> Université de Reims Champagne-Ardenne
> >> F-51687 Reims Cedex 2
> >> France
> >> Web:http://phare.normalesup.org/~kichenassamy
> >> *********************************************
> >>
> >>
> >>
> >> _______________________________________________
> >> INDOLOGY mailing list
> >> INDOLOGY@list.indology.info
> >> https://list.indology.info/mailman/listinfo/indology
>
> --
> **********************************************
> Satyanad KICHENASSAMY
> Professeur des Universités
> Laboratoire de Mathématiques (LMR, CNRS, UMR9008)
> Université de Reims Champagne-Ardenne
> F-51687 Reims Cedex 2
> France
> Web:http://phare.normalesup.org/~kichenassamy
> **********************************************


--
Satyanad KICHENASSAMY <satyanad.kichenassamy@univ-reims.fr>