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

Japanese Mode?

Expand Messages
  • hisilome
    Just out of curiosity: Has anybody ever tried to devise a tengwar mode for Japanese? The sound system is pretty straightforward, but surely one would have to
    Message 1 of 9 , Mar 18, 2006
    View Source
    • 0 Attachment
      Just out of curiosity: Has anybody ever tried to devise a tengwar
      mode for Japanese?

      The sound system is pretty straightforward, but surely one would have
      to "break up" the traditional Japanese representation in the kana
      charts (syllable alphabets) and rather deal with the consonants and
      vowels separately, as is usually done in tengwar? That is, syllables
      (e.g. "ka" or "se" or "tsu") should probably best be broken up
      into "k + a" etc, with "tsu" being one of the more interesting
      syllables, since here we're not dealing with a consonant plus pure u-
      vowel anymore, but more of a syllabised "ts" (rather similar to
      Mandarin _c(i)_).

      Other features that would have to be dealt with are the voiced
      counterparts of "k, s, t, b" (daku-on) and the handaku-on "p".
      In Japanese writing, "g, z, d" are represented by the same signs used
      in the "ta, sa, ta"-series with a diacritic called daku-ten (aka
      nigori-ten) added at the top right of the kana, while "b, p" are
      written with the same signs used in the "ha"-series (! what's the
      relation between a glottal fricative and the bilabials?) with the
      daku-ten ("b") or a maru (aka handaku-ten) ("p") added at the top
      right.

      The daku-ten used for the daku-on sounds ("g, z, d, b") looks a bit
      like English double quotation marks (but quite small), slanting
      diagonally from left to right, while the maru/handaku-ten used for
      the handaku-on "p" is a small circle (not dot).

      The palatalized sounds (mya, myu, myo, hya, hyu, hyo, etc) should
      be "easy", as should the doubled consonants...

      Hisilome
    • Arden R. Smith
      ... You didn t really mean to include _b_ here, did you? It s already voiced, and written as the voiced counterpart of _h_ (see below). ... And here you meant
      Message 2 of 9 , Mar 19, 2006
      View Source
      • 0 Attachment
        On Mar 18, 2006, at 9:00 PM, hisilome wrote:

        > Other features that would have to be dealt with are the voiced
        > counterparts of "k, s, t, b" (daku-on)

        You didn't really mean to include _b_ here, did you? It's already
        voiced, and written as the voiced counterpart of _h_ (see below).

        > and the handaku-on "p".
        > In Japanese writing, "g, z, d" are represented by the same signs used
        > in the "ta, sa, ta"-series with a diacritic called daku-ten (aka
        > nigori-ten) added at the top right of the kana,

        And here you meant "ka, sa, ta", right?

        > while "b, p" are
        > written with the same signs used in the "ha"-series (! what's the
        > relation between a glottal fricative and the bilabials?)

        Whatever it is, it's a very real relation in Japanese. Just look at
        _Nihon_ and _Nippon_ as variant pronunciations of the name of Japan.
        Additionally, initial _h_ often becomes _b_ in compound words, e.g.
        _hashi_ 'bridge' in the place-name Nihonbashi 'Japan bridge', and
        _hayashi_ 'forest, grove' in the family-name Kobayashi 'small forest'.
        The alternation one normally finds in this environment is voiceless >
        voiced, e.g. _sushi_ in _nigirizushi_.

        These seemingly dissimilar sounds can also be found in etymological
        relationships outside of Japanese. An example that springs to mind is
        Armenian _hayr_ 'father', in which the initial consonant derives from
        Indo-European _*p_.

        To answer the main question: I've occasionally thought about working
        out a Japanese mode, but I've never actually bothered to do it. It
        would be an interesting exercise, though.

        ***************************************************
        Arden R. Smith erilaz@...

        Perilme metto aimaktur perperienta.
        --Elvish proverb

        ***************************************************
      • hisilome
        ... [ You re right, of course. Typing too quickly there... What I meant was the voiced counterparts of k, s, t = g, z, d, and in addition to that: b, as I
        Message 3 of 9 , Mar 19, 2006
        View Source
        • 0 Attachment
          > On Mar 18, 2006, at 9:00 PM, hisilome wrote:
          >
          > > Other features that would have to be dealt with are the voiced
          > > counterparts of "k, s, t, b" (daku-on)
          >
          > You didn't really mean to include _b_ here, did you? It's already
          > voiced, and written as the voiced counterpart of _h_ (see below).

          [ You're right, of course. Typing too quickly there... What I meant
          was "the voiced counterparts of "k, s, t" = g, z, d, and in addition
          to that: b, as I wrote further down: "The daku-ten used for the daku-
          on sounds ("g, z, d, b") looks a bit like English double quotation
          marks..."

          Even "p" seems not to be a completely "unvoiced" sound in Japanese
          (though I probably pronounce it so, used to the voiced-unvoiced
          dichotomies in most Germanic languages--my "trick" is to say it a
          bit "softer" than, say, the "p" in "pen"), since it's described
          as "semi-voiced" ("handaku-on" translating as something like "half-
          turbid[=voiced]-sound").

          The interesting thing is, since, as I wrote, "b, p" are
          written with the same signs used in the "ha"-series...with the
          daku-ten (for "b" sounds) or a maru (aka handaku-ten) (for "p"
          sounds) added at the top right, this, as you also point out,
          makes "b", at least in writing, the voiced counterpart of "h" in
          Japanese (not of "p"), so we have: k/g, s/z, t/d, h/b--where does
          that leave "p", the "semi-voiced" sound, and how would one fit it
          into the tengwar system?
          Or maybe one shouldn't worry about this, and simply use hyarmen,
          umbar and parma, anyway... ]


          > > and the handaku-on "p". In Japanese writing, "g, z, d" are
          > represented by the same signs used
          > > in the "ta, sa, ta"-series with a diacritic called daku-ten (aka
          > > nigori-ten) added at the top right of the kana,
          >
          > And here you meant "ka, sa, ta", right?

          [ Naturally! Too many typos in this one, sorry. ]


          > > while "b, p" are
          > > written with the same signs used in the "ha"-series (! what's the
          > > relation between a glottal fricative and the bilabials?)
          >
          > Whatever it is, it's a very real relation in Japanese. Just look at
          > _Nihon_ and _Nippon_ as variant pronunciations of the name of
          >Japan.
          > Additionally, initial _h_ often becomes _b_ in compound words, e.g.
          > _hashi_ 'bridge' in the place-name Nihonbashi 'Japan bridge', and
          > _hayashi_ 'forest, grove' in the family-name Kobayashi 'small
          forest'.

          [ Yeah, that sound change's ubiquitous in Japanese, see
          also "combinations" like "ten minutes" jippun, (minute = fun) etc.
          (don't think this is strictly speaking a compound). ]

          > The alternation one normally finds in this environment is voiceless
          > voiced, e.g. _sushi_ in _nigirizushi_.

          [ Now you get my mouth watering. :) ]


          >
          > These seemingly dissimilar sounds can also be found in etymological
          > relationships outside of Japanese. An example that springs to mind
          is Armenian _hayr_ 'father', in which the initial consonant derives
          from Indo-European _*p_.

          [ I guess this is one of the things you can only observe, but not
          necessarily explain then... ]


          > To answer the main question: I've occasionally thought about
          >working
          > out a Japanese mode, but I've never actually bothered to do it. It
          > would be an interesting exercise, though.

          [ Well, if you ever come up with a proposal, let us know! ]

          Hisilome
        • j_mach_wust
          Hisilome wrote: ... ... That s what I d say. There is already a relation between the signs for /t/ and /d/ if we represent them with tinco and ando. You could
          Message 4 of 9 , Mar 19, 2006
          View Source
          • 0 Attachment
            Hisilome wrote:
            ...
            > The interesting thing is, since, as I wrote, "b, p" are
            > written with the same signs used in the "ha"-series...with the
            > daku-ten (for "b" sounds) or a maru (aka handaku-ten) (for "p"
            > sounds) added at the top right, this, as you also point out,
            > makes "b", at least in writing, the voiced counterpart of "h" in
            > Japanese (not of "p"), so we have: k/g, s/z, t/d, h/b--where does
            > that leave "p", the "semi-voiced" sound, and how would one fit it
            > into the tengwar system?
            > Or maybe one shouldn't worry about this, and simply use hyarmen,
            > umbar and parma, anyway... ]
            ...

            That's what I'd say. There is already a relation between the signs for
            /t/ and /d/ if we represent them with tinco and ando. You could
            consider the doubling of the bow a kind of daku-ten (I hope I don't
            mix up things), though I don't see a straightforward representation of
            the relation between /p/, /b/ and /h/.

            Anyway, I think the following two problems are even more important:

            What to do with loanword syllables such as "ti" or "tsi"? If we use
            tinco for all of "ta chi tsu te to", then these are problematic. If we
            use tinco only for "ta ti tu te to", then the natural sound system of
            Japanese is broken.

            How to represent the Japanese pitch accent (if at all)?

            ---------------------------
            j. 'mach' wust
            http://machhezan.tripod.com
            ---------------------------
          • Melroch 'Aestan
            ... I agree. No need to mess up the principles of Tengwar where they are analogous, just because they are not identical. ... Historically _h_ is derived from
            Message 5 of 9 , Mar 19, 2006
            View Source
            • 0 Attachment
              j_mach_wust skrev:
              > Hisilome wrote:
              > ...
              >
              >>The interesting thing is, since, as I wrote, "b, p" are
              >>written with the same signs used in the "ha"-series...with the
              >>daku-ten (for "b" sounds) or a maru (aka handaku-ten) (for "p"
              >>sounds) added at the top right, this, as you also point out,
              >>makes "b", at least in writing, the voiced counterpart of "h" in
              >>Japanese (not of "p"), so we have: k/g, s/z, t/d, h/b--where does
              >>that leave "p", the "semi-voiced" sound, and how would one fit it
              >>into the tengwar system?
              >>Or maybe one shouldn't worry about this, and simply use hyarmen,
              >>umbar and parma, anyway... ]
              >
              > ...
              >
              > That's what I'd say. There is already a relation between the signs for
              > /t/ and /d/ if we represent them with tinco and ando. You could
              > consider the doubling of the bow a kind of daku-ten (I hope I don't

              I agree. No need to mess up the principles of Tengwar where
              they are analogous, just because they are not identical.

              > mix up things), though I don't see a straightforward representation of
              > the relation between /p/, /b/ and /h/.

              Historically _h_ is derived from *f (actually *[p\])
              which in turn is derived from *p rather than the other
              way around, so the most straightforward way would be to
              use Formen for _f_ -- with w-tehta as a diacritic when
              _f_ occurs before vowels other than _u_ and under-dot when _h_
              occurs before _u_. C.f. _huzi/fuji_ which is said to be
              cognate to Korean _phulson_!

              > Anyway, I think the following two problems are even more important:
              >
              > What to do with loanword syllables such as "ti" or "tsi"? If we use
              > tinco for all of "ta chi tsu te to", then these are problematic. If we
              > use tinco only for "ta ti tu te to", then the natural sound system of
              > Japanese is broken.

              I suggest using the underdot as a general diacritic for
              "foreign" pronunciation, thus having an underdot on _ti_
              to distinguish it from _chi_. The Calmatéma should of
              course not be used at all...

              > How to represent the Japanese pitch accent (if at all)?

              Well, since the Japanese, as well as common Romanization,
              usually don't indicate it it should probably be dispensed
              with. Imagine having to look up every word in a pronunciation
              dictionary in order to write something in Tengwar! I've actually
              done that in the course of writing English phonemically...

              --

              /BP 8^)>
              --
              Benct Philip Jonsson -- melroch at melroch dot se
              ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~__
              A h-ammen ledin i phith! \ \
              __ ____ ____ _____________ ____ __ __ __ / /
              \ \/___ \\__ \ /___ _____/\ \\__ \\ \ \ \\ \ / /
              / / / / / \ / /Melroch\ \_/ // / / // / / /
              / /___/ /_ / /\ \ / /Roccondil\_ // /__/ // /__/ /
              /_________//_/ \_\/ /Eowine __ / / \___/\_\\___/\_\
              Gwaedhvenn Angeliniel\ \______/ /a/ /_h-adar Merthol naun
              ~~~~~~~~~Kuinondil~~~\________/~~\__/~~~Noolendur~~~~~~
              || Lenda lenda pellalenda pellatellenda kuivie aiya! ||
              "A coincidence, as we say in Middle-Earth" (JRR Tolkien)
            • hisilome
              ... [ Yes, for t/d and k/g, things would be quite neat . So maybe: k = quesse g = ungwe s = silme (?) z = aaze (?) t = tinco d = ando n = nuumen h = hyarmen p
              Message 6 of 9 , Mar 19, 2006
              View Source
              • 0 Attachment
                --- In elfscript@yahoogroups.com, "j_mach_wust" <j_mach_wust@...>
                wrote:

                > Hisilome wrote:

                >...We have: k/g, s/z, t/d, h/b--where does
                > > that leave "p", the "semi-voiced" sound, and how would one fit
                >it into the tengwar system?
                > > Or maybe one shouldn't worry about this, and simply use hyarmen,
                > > umbar and parma, anyway... ]
                > ...
                >
                > That's what I'd say. There is already a relation between the signs
                >for /t/ and /d/ if we represent them with tinco and ando. You could
                > consider the doubling of the bow a kind of daku-ten (I hope I don't
                > mix up things), though I don't see a straightforward
                >representation of the relation between /p/, /b/ and /h/.

                [ Yes, for t/d and k/g, things would be quite "neat".
                So maybe:

                k = quesse
                g = ungwe
                s = silme (?)
                z = aaze (?)
                t = tinco
                d = ando
                n = nuumen
                h = hyarmen
                p = parma
                b = umbar
                m = nuumen
                y = anna (?)
                r = roomen (Japanese "r" is really a flap I think [akin to
                Spanish "r"], at the same time bearing some similarity to the
                English "l" [though the latter is of course a continuant]), so
                roomen should be better than oore I guess)
                w = vala (basically like English "w", yet a bit "softer" I'd say)

                For palatalization, one could probably use the double under-dots
                (like in Quenya or Mandarin). Japanese here, quite befittingly, uses
                the "Consonant + i" kana from the series in question, followed by a
                half-sized "ya", "yu", or "yo" kana, e.g. "mya" = "mi" + "small "ya".

                For doubled consonants (soku-on, represented in Romanization by
                doubling the letter), the under-bar. (Interestingly, Japanese kana
                script here inserts a half-sized "tsu" kana before the sound in
                question.) ]


                > Anyway, I think the following two problems are even more important:
                >
                > What to do with loanword syllables such as "ti" or "tsi"? If we use
                > tinco for all of "ta chi tsu te to", then these are problematic.
                >If we use tinco only for "ta ti tu te to", then the natural sound
                >system of Japanese is broken.

                [ Well, I was actually focussing on the "traditional" sound system
                first. ;) But you're already thinking ahead, and I agree
                that "imported" sounds are potentially quite troublesome.

                I'm not so sure about "shi/si", "chi/ti", "tsu/tu": as you say, if
                we write, say, "ta chi tsu te to", as "tinco plus A, I, U, E, O"
                respectively (regardless of the fact that "ti, tu" are really "chi,
                tsu"--did I understand correctly that that's what you were thinking
                about?), then some loanwords (and there many of them in Japanase)
                might be problematic.
                I'm not even certain whether this would be a good solution, though.
                In a phonemic mode, at least, shouldn't one represent the actual
                sounds of "chi, tsu", as is even reflected in Romanized versions of
                Japanese. Same for "shi".

                Then there are also sounds like "ji/zi, ja/zya, ju/zyu, jo/zyo", in
                kana spelt as "shi/si kana and daku-ten" (for ji/zi), and "shi/si
                kana and daku-ten + half-sized ya/yu/yo", respectively. All very
                neat in kana, but how would one represent this in tengwar? If, for
                arguments sake, we use the consonant chart I outlined above, would
                we spell these sounds with aaze? Hm, sounds feasible...?

                Analogically for "sha, shu, sho, cha, chu, cho" spelled as "sya,
                syu, syo, tya, tyu, tyo"...? ]


                > How to represent the Japanese pitch accent (if at all)?

                [ This I don't consider to be a problem. First of all, Japanese
                pitch, while essential to speaking "natural-sounding", fluent
                Japanese, is by no means as important as, say, the tones in Chinese.
                I mean, Japanese also has a fair amount of homophones, to be sure,
                and pitch can often help to discern them. Still I have found that
                even when one messes up the "word melody" quite a bit, one is
                usually understood (provided one gets the grammar and vocabulary
                right, that is!)--whereas Chinese mostly tends to become quite
                unintelligible, or at least very hard to understand, if you
                consistently get the tones wrong (and Cantonese or "Fukienese" are
                even "worse" in this respect than Mandarin). I also find that in
                Japanese the "melody" of the entire sentence is often more important
                than the pitch of individual words, though to some degree they are
                of course connected.

                Another thing is that existing transcription systems for Japanese,
                such as the the hebonshiki roomaji or the kunreishiki roomaji, don't
                bother to reflect pitch, either--which is probably also testimony to
                the fact that pitch is not quite as crucial as the tones are in
                Chinese (all transcription systems of Chinese, or its dialects, have
                some way of indicating the tones). ]

                Hisilome
              • Arden R. Smith
                ... I thought of an even better example after I had turned off the computer and gone to bed last night, at the beginning of the series of numbers for counting
                Message 7 of 9 , Mar 19, 2006
                View Source
                • 0 Attachment
                  On Mar 19, 2006, at 2:15 AM, hisilome wrote:

                  >> Whatever it is, it's a very real relation in Japanese. Just look at
                  >> _Nihon_ and _Nippon_ as variant pronunciations of the name of
                  >> Japan.
                  >> Additionally, initial _h_ often becomes _b_ in compound words, e.g.
                  >> _hashi_ 'bridge' in the place-name Nihonbashi 'Japan bridge', and
                  >> _hayashi_ 'forest, grove' in the family-name Kobayashi 'small
                  > forest'.
                  >
                  > [ Yeah, that sound change's ubiquitous in Japanese, see
                  > also "combinations" like "ten minutes" jippun, (minute = fun) etc.
                  > (don't think this is strictly speaking a compound). ]

                  I thought of an even better example after I had turned off the computer
                  and gone to bed last night, at the beginning of the series of numbers
                  for counting long, thin objects: _ippon_, _nihon_, _sanbon_ (or
                  _sambon_, depending on your preferred Romanization). There you get all
                  three right in a row -- bam! bam! bam!

                  I'm afraid that I won't have the time to participate as fully as I
                  would like in the present discussion. Right now I'm trying to finish
                  up a bunch of things in preparation for a trip (coincidentally enough)
                  to Japan. But I do personally favor the option of using the same
                  tengwa to represent the same kana series, in a fashion similar to
                  kunreishiki rômaji (e.g., ta, ti, tu, te, to), even though it doesn't
                  capture the phonetic realizations reflected in Hepburn rômaji (ta, chi,
                  tsu, te, to). I just really like the systematicity of it.

                  Benct's suggestions re f/h are interesting; I'll have to give them some
                  thought.

                  As for /s/ and /z/, I would go with thúle and anto; there's no need to
                  resort to the additional letters if there are perfectly good primary
                  letters lying fallow. I would use óre for /r/ for the same reason.

                  While númen would be good for /nV/ combinations, I'm not too keen on
                  the idea of also using it for the moraic /n/ (or whatever you want to
                  call it). For that, the preceding-nasal tehta would work just fine
                  preconsonantally, but I'm still undecided about how to deal with it in
                  final position, maybe by means of a tehta. My thinking here is
                  undoubtedly colored by the treatment of final /n/ in Qenya alphabets
                  from the 1920s; see e.g. _Parma Eldalamberon_ 14, p. 110. Another
                  possibility would be to use óre for final /n/, which would in turn
                  necessitate the use of rómen for /r/. Or maybe númen for final /n/
                  wouldn't be a bad thing after all, especially if it had a _putta_
                  (_unutikse_). See what I mean by "undecided"?

                  ***************************************************
                  Arden R. Smith erilaz@...

                  Perilme metto aimaktur perperienta.
                  --Elvish proverb

                  ***************************************************
                • hisilome
                  ... [ That certainly makes sense: hu , at least, is often represented as fu in Romanization (and sounds a lot like it, too--it s actually quite hard to
                  Message 8 of 9 , Mar 19, 2006
                  View Source
                  • 0 Attachment
                    --- In elfscript@yahoogroups.com, Melroch 'Aestan <melroch@...> wrote:


                    > Historically _h_ is derived from *f (actually *[p\])
                    > which in turn is derived from *p rather than the other
                    > way around, so the most straightforward way would be to
                    > use Formen for _f_ -- with w-tehta as a diacritic when
                    > _f_ occurs before vowels other than _u_ and under-dot when _h_
                    > occurs before _u_. C.f. _huzi/fuji_ which is said to be
                    > cognate to Korean _phulson_!

                    [ That certainly makes sense: "hu", at least, is often represented
                    as "fu" in Romanization (and sounds a lot like it, too--it's actually
                    quite hard to pronounce "hu" without turning it into "fu", since for
                    the "u" you have to purse the lips so much, and then "h" almost
                    automatically assumes an "f"-like quality).
                    And the "connection" between "h" and "f" also becomes obvious in
                    other environments, for example when native speakers of Minnanyu
                    don't discern them clearly when speaking Mandarin, pronouncing
                    e.g. "hen" as "fen", etc. ]


                    J. 'Mach' Wust wrote:
                    > > Anyway, I think the following two problems are even more
                    >important:
                    > > What to do with loanword syllables such as "ti" or "tsi"? If we
                    >use tinco for all of "ta chi tsu te to", then these are problematic.
                    >If we use tinco only for "ta ti tu te to", then the natural sound
                    >system of Japanese is broken.

                    Melroch replied:
                    > I suggest using the underdot as a general diacritic for
                    > "foreign" pronunciation, thus having an underdot on _ti_
                    > to distinguish it from _chi_. The Calmatéma should of
                    > course not be used at all...

                    [ That's an interesting suggestion, and quite feasible if we don't
                    assign the under-dot to any other function. The under-dot would then
                    be used in lieu of an entire separate kana alphabet (the
                    katakana) which the Japanese employ to make foreign words "stick out"
                    in writing... ]

                    Hisilome
                  • hisilome
                    ... [ Fair enough! ] ... on the idea of also using it for the moraic /n/ (or whatever you want to call it). For that, the preceding-nasal tehta would work
                    Message 9 of 9 , Mar 19, 2006
                    View Source
                    • 0 Attachment
                      --- In elfscript@yahoogroups.com, Arden R. Smith <erilaz@...> wrote:


                      > As for /s/ and /z/, I would go with thúle and anto; there's no need
                      >to resort to the additional letters if there are perfectly good
                      >primary letters lying fallow. I would use óre for /r/ for the same
                      >reason.

                      [ Fair enough! ]


                      > While númen would be good for /nV/ combinations, I'm not too keen
                      on the idea of also using it for the moraic /n/ (or whatever you
                      want to call it). For that, the preceding-nasal tehta would work
                      just fine preconsonantally, but I'm still undecided about how to deal
                      with it in final position, maybe by means of a tehta. My thinking
                      here is undoubtedly colored by the treatment of final /n/ in Qenya
                      alphabets from the 1920s; see e.g. _Parma Eldalamberon_ 14, p. 110.
                      Another possibility would be to use óre for final /n/, which would in
                      turn necessitate the use of rómen for /r/.
                      Or maybe númen for final /n/ wouldn't be a bad thing after all,
                      especially if it had a _putta_ (_unutikse_).

                      [ Maybe I'd go for that: using the over-bar/tilde for
                      preconsonantal /n/, and nuumen, even without a _putta_, for
                      final /n/. After all, it is basically the same sound as initial /n/,
                      just in a different position/function.

                      Then again, Japanese has an independent kana for moraic /n/--I seem
                      to recall that it was a relatively late addition to the syllable
                      alphabets (I think the syllable-final sound only developed after the
                      original kana charts were drawn up). This might speak for the
                      solution with oore, maybe both preconsonantally and word-final? ]

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