Re: Old Hungarian script - progress?

From: Andr Szabolcs Szelp (
Date: Fri Feb 08 2008 - 04:47:57 CST

  • Next message: Sinnathurai Srivas: "Re: minimizing size (was Re: allocation of Georgian letters)"

    Hello, dm,

    I tend to write long posts, I warn you in advance. Also, I'm receiving the messages in digest mode, so cross-posts are possible.

    I'll make a slight detour, also including information you might already know for the sake of other readers. Also, I might be wrong in minor details.

    The 1998 proposal for Old Hungarian Script (OHS) (called rovsrs, literally, 'incision script' in Hungarian; also called "Rovs Script" or "Hungarian Runes" in English literature, seldom also "Szekler Script/Runes" from the extended Hungarian name "szkely rovsrs" referring to the subgroup of Hungarians living in the area where most artefacts have been found) had an ill fate, however quite understandably, I guess.

    It was prepared in good faith by Michael Everson -- in his laudable and great undertaking of promoting minority/little known scripts -- based upon a quasi-encoding created by enthusiasts. This quasi-encoding (using ISO 8859-1/2 as a base) with "fake" glyphs was based on pre-OpenType pre-Unicode technology with no clear notion of graphemes, ligatures, marks, glyph variants, characters vs. glyphs, etc. Also, the creators of the quasi-encoding did some really nice research with respect to attested ligatures and also in terms of character repertoire (though not character in the Unicode sense), but I seem to remember some lacks in the basic set. Also, while the creators boastingly called it "Rovs Szabvny" encoding -- "szabvny" means standard in Hungarian), it has had no official standing anytime.
    Actually, quite some of the little corpus, mostly decorative labels, mottos, etc. is completely corrupted, exactly because of that re-mapping of ISO-8859 family of standards, when a "user" not knowing the script himself simply applied the "Rovs Szabvny" font to Latin Hungarian text, causing completely wrong OHS

    I presume (though I have not been around back then; please correct me), the proposal has not been picked up, as it did not seem priority and the faults of the underlying material were realised: there was no "historic standard" to be incorporated into Unicode, virtually no corpus, and for being encripted anew, the encoding seemed to be ill prepared (not Unicode philosophy; as I understand, glyph/character model not according to Unicode is only encoded in that way if a historic standard endorses it).

    As to your question, I had the plan to bring in a proposal for some years, with the peak of my motivation cca. 3 years ago; however, due to two factors I never came around it. A) Lack of time. This will be something every one of us working will know. B) Being a perfectionist, I would not want to submit an either graphematically, either linguistically incorrect proposal, not until I was sure I did a thorough analysis and nothing has evaded my attention.

    While I have somewhat come out of practice in that matter, I'll pick that thread up, if you're willing to fund me ;-)

    There are however some more controversial issues.

    On the one hand, there is the (extinct(?)) historic script. On the other, there is the script as used by a very limited modern community with a somehow arbitrarily extended character set (so that to map to modern latin-based Hungarian alphabet) and different rules as compared to the historic script, as defined by revivalists (starting with Sndor Forrai) beginning from the 70-ies.

    This revivalists' OHS arguably has similarities to actually merely invented scripts, not unlike Klingon, Cirth or Tengwar (made up by an individual, adopted by a small fan-community, though truly has a very strong historical basis).

    So the question arises: will you encode those characters, which were made up by a single individual in the 70ies and are used by a small enthusiast community? If yes, why would you not encode Cirth? If no, will you not address the actual needs of the living community? However: what defines the needs of "a living community" (c.f. Tengwar users).

    Don't misunderstand: I'm not questioning the legitimacy of encoding OHS _per_se_. On the contrary, I avidly support it! There was the historic script, it's of scholarly interest (though admittedly, barely any academic research has been done since the 50ies; most academic research took place in the first part of the 20th century), it appears in publications. Oh, and I support it, because the script is of particular (personal) interest to me.

    However, the question in my mind occurs about the "modern" revivalists additions to the script. Books have undoubtedly been published in that version (but so have been books in Klingon). Where to delineate? What to encode?

    Actually most additions (if not all, which I suspect, but I'd have to check) have been long counterparts to vowel signs. (Modern Latin-based Hungarian marks them), so even a hybrid (consensus) solution with a VOWEL LENGTH MARKER (kindof a VARIANT SELECTOR) comes to my mind (balancing between the historic fact of a script and modern additions*), where the sequence <(historic) OHS LETTER O><OHS VOWEL LENGTH MARK> is rendered by smart fonts supporting it as the modern long variant, while other fonts might disregard that codepoint.

    This is actually just a question of encoding philosophy/eligibility, as even keyboards can create multiple codepoints on single keystrokes.


    *) How do you define "modern"? Just 20th century? Or do you consider the addition of the equivalent of "" in the early 18th century (if I remember well) which is attested only by a single sentence from a single user (speaking of the 18th century, clearly an early "enthusiast" user, the script being extinct from actual usage by the very latest in the 16th century, probably up to two centuries earlier), used inconsistently even within that one single sentence as "modern" as well?

    -- Original message --
    Date: Thu, 7 Feb 2008 14:55:48 +0100
    From: "=?UTF-8?Q?Jo=C3=B3_=C3=81d=C3=A1m?=" <>
    Subject: =?UTF-8?Q?Old_Hungarian_script_=E2=80=93_progression?=


    I just want to ask if there was any progression regarding the Old
    Hungarian script[1] since 1998.

    Thanks in advance,


    GMX FreeMail: 1 GB Postfach, 5 E-Mail-Adressen, 10 Free SMS.
    Alle Infos und kostenlose Anmeldung:

    This archive was generated by hypermail 2.1.5 : Fri Feb 08 2008 - 04:51:49 CST