Loading ...
Sorry, an error occurred while loading the content.

Re: Romanization: digraphs vs. diacritics

Expand Messages
  • David McCann
    On Sat, 19 Jan 2013 18:26:31 +0100 ... In the English-speaking world, the ISO transcription is tending to give way to the Library of Congress system, since
    Message 1 of 12 , Jan 20, 2013
    • 0 Attachment
      On Sat, 19 Jan 2013 18:26:31 +0100
      Jörg Rhiemeier <joerg_rhiemeier@...> wrote:

      > Sure. For Russian, for instance, there is a scholarly
      > transliteration that maps Cyrillic letters on Latin letters,
      > often with diacritics, loosely according to the conventions
      > used for languages such as Czech or Croatian. But in German
      > newspapers, one rather find spellings such as _Gorbatschow_
      > which simply attempt to apply German spelling conventions to
      > Russian names. This has many shortcomings. You never know
      > whether _s_ represents /s/ or /z/, or whether _sch_ represents
      > /ʃ/ or /ʒ/! (Also, it has a *heptagraph*: _schtsch_ for what
      > is a single phoneme in names such as _Chruschtschow_!)

      In the English-speaking world, the ISO transcription is tending to give
      way to the Library of Congress system, since that's the one in the
      library catalogues.

      It's interesting that (as far as I know) the Russians don't
      transliterate, but transcribe. My favourite example is the old Гул
      "Gul" for Hull, but the other day I found Halle Berry as Хэлли
      "Khelli". The perception of /æ/ as /ɛ/ seems common, as in фен "fen" <
      "fan".
    • Eugene Oh
      Sometimes this reveals hypercorrections too, such as Айлингтон /ajliNtOn/ for Islington /IzlINt@n/, the London borough, on Google maps. Eugene Sent
      Message 2 of 12 , Jan 20, 2013
      • 0 Attachment
        Sometimes this reveals hypercorrections too, such as Айлингтон /ajliNtOn/ for Islington /IzlINt@n/, the London borough, on Google maps.

        Eugene

        Sent from my iPhone

        On 20 Jan 2013, at 16:52, David McCann <david@...> wrote:

        > On Sat, 19 Jan 2013 18:26:31 +0100
        > Jörg Rhiemeier <joerg_rhiemeier@...> wrote:
        >
        >> Sure. For Russian, for instance, there is a scholarly
        >> transliteration that maps Cyrillic letters on Latin letters,
        >> often with diacritics, loosely according to the conventions
        >> used for languages such as Czech or Croatian. But in German
        >> newspapers, one rather find spellings such as _Gorbatschow_
        >> which simply attempt to apply German spelling conventions to
        >> Russian names. This has many shortcomings. You never know
        >> whether _s_ represents /s/ or /z/, or whether _sch_ represents
        >> /ʃ/ or /ʒ/! (Also, it has a *heptagraph*: _schtsch_ for what
        >> is a single phoneme in names such as _Chruschtschow_!)
        >
        > In the English-speaking world, the ISO transcription is tending to give
        > way to the Library of Congress system, since that's the one in the
        > library catalogues.
        >
        > It's interesting that (as far as I know) the Russians don't
        > transliterate, but transcribe. My favourite example is the old Гул
        > "Gul" for Hull, but the other day I found Halle Berry as Хэлли
        > "Khelli". The perception of /æ/ as /ɛ/ seems common, as in фен "fen" <
        > "fan".
      • Jörg Rhiemeier
        Hallo conlangers! ... In my romanization of Old Albic, _ng_ is always /ŋ/; the sequence /ŋg/ is transcribed _ngg_, /ŋk/ is _ngc_, and /ng/ does not occur (a
        Message 3 of 12 , Jan 20, 2013
        • 0 Attachment
          Hallo conlangers!

          On Sunday 20 January 2013 00:02:28 Herman Miller wrote:

          > On 1/19/2013 8:44 AM, Melroch wrote:
          > > My own shifts in conventions for transcribing Sohlob. I started out over
          > > 15 years ago with an ASCII-based system using <tj sj dj zj> for
          > > alveopalatals. This was actually sub-phonemic since I decided very early
          > > that [ʑ] was an allophone of /dʑ/, and <j> was used only in those
          > > digraphs. At the same time I used <ny hl hr> for single phonemes and <ng>
          > > ambiguously for /ŋ/ and /ŋg/, justified by the conventions in the
          > > 'native' script which was and is under-specifying to a high degree.
          >
          > The ambiguity with "ng" is admittedly one of the drawbacks of
          > conventions like these including the spelling I use on the map (in names
          > like "Kerngat" and "Nagmingo"). A name like "Nagmingo" could be
          > [naɡminɡo], [naɡmiŋɡo], or [naɡmiŋo] (not counting possible variations
          > in the vowels that aren't distinguished in the romanization).

          In my romanization of Old Albic, _ng_ is always /ŋ/; the sequence
          /ŋg/ is transcribed _ngg_, /ŋk/ is _ngc_, and /ng/ does not occur
          (a nasal preceding a stop always assimilates to the latter's POA).

          Of course, in the native script, there is a letter for /ŋ/ (also
          letters for /ɸ/, /θ/ and /x/, so no digraphs are needed at all).

          --
          ... brought to you by the Weeping Elf
          http://www.joerg-rhiemeier.de/Conlang/index.html
          "Bêsel asa Éam, a Éam atha cvanthal a cvanth atha Éamal." - SiM 1:1
        • Herman Miller
          ... I just noticed a fourth possibility with the name Kerngat, since the n could be part of an rn digraph, which would make it [keɳɡat]. Many of the
          Message 4 of 12 , Jan 20, 2013
          • 0 Attachment
            On 1/20/2013 3:39 PM, Jörg Rhiemeier wrote:
            > Hallo conlangers!
            >
            > On Sunday 20 January 2013 00:02:28 Herman Miller wrote:
            >
            >> The ambiguity with "ng" is admittedly one of the drawbacks of
            >> conventions like these including the spelling I use on the map (in names
            >> like "Kerngat" and "Nagmingo"). A name like "Nagmingo" could be
            >> [naɡminɡo], [naɡmiŋɡo], or [naɡmiŋo] (not counting possible variations
            >> in the vowels that aren't distinguished in the romanization).
            >
            > In my romanization of Old Albic, _ng_ is always /ŋ/; the sequence
            > /ŋg/ is transcribed _ngg_, /ŋk/ is _ngc_, and /ng/ does not occur
            > (a nasal preceding a stop always assimilates to the latter's POA).
            >
            > Of course, in the native script, there is a letter for /ŋ/ (also
            > letters for /ɸ/, /θ/ and /x/, so no digraphs are needed at all).

            I just noticed a fourth possibility with the name Kerngat, since the "n"
            could be part of an "rn" digraph, which would make it [keɳɡat].

            Many of the languages on Sarangia are written in alphabets that could be
            described as featural, although sound changes over time might complicate
            things. So the languages that have /ŋ/ as a phoneme probably all have a
            way to write the sound with a single character (as Tirelat does).

            I've used "ñ" to write the /ŋ/ sound in Tirelat, but currently I just
            use "ŋ". (If I'm going to use an inconvenient character in the first
            place, I might as well use one that won't be misread.)
          • Douglas Koller
            ... Just so in Géarthnuns, too. Bangui (in the news of late) is rendered Banggísars ; Ankara , Angkarasars . ... This is true of syllabic nasals in
            Message 5 of 12 , Jan 20, 2013
            • 0 Attachment
              > Date: Sun, 20 Jan 2013 21:39:40 +0100
              > From: joerg_rhiemeier@...
              > Subject: Re: Romanization: digraphs vs. diacritics
              > To: CONLANG@...

              > In my romanization of Old Albic, _ng_ is always /ŋ/; the sequence
              > /ŋg/ is transcribed _ngg_, /ŋk/ is _ngc_,

              Just so in Géarthnuns, too. "Bangui" (in the news of late) is rendered "Banggísars"; "Ankara", "Angkarasars".

              > (a nasal preceding a stop always assimilates to the latter's POA).

              This is true of syllabic nasals in Géarthuns, certainly, but one can easily imagine morpheme boundaries knocking up against each other à la "input", "painkiller", and "ingrown" (examples in Géarthnuns are not legion, though).

              > so /ng/ does not occur

              For the Miami store "ManGear", you'd have to run with "Man'gírs" using our trusty friend, the apostrophe, to break up digraph romanizations. That pops up most often with a'u, sometimes with o'u and r'h, and only one set of related words with t'h that I can think of off hand. s'h, d'h, z'h, k'h may exist out there in the lexicon, but nothing springs to mind. I don't think n'g has occurred...yet. It's really hard to see z'ç happening any time soon if at all.

              > Of course, in the native script, there is a letter for /ŋ/ (also
              > letters for /ɸ/, /θ/ and /x/, so no digraphs are needed at all).

              Indeed. Just the opposite, in fact. There are fourteen letters which represent two consonant sounds together. (I'm really warming to [ɸ] BTW -- perhaps I can shoehorn it in as an allophone somewhere ;) )

              Kou
            Your message has been successfully submitted and would be delivered to recipients shortly.