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

Tengwar for Chinese: a few more considerations ;)

Expand Messages
  • Dave
    A few more considerations occurred to me. When talking about affricates, I should have mentioned that j(i) and q(i) are in some respects special. First lets
    Message 1 of 23 , Nov 9, 2004
    • 0 Attachment
      A few more considerations occurred to me.

      When talking about affricates, I should have mentioned that j(i) and
      q(i) are in some respects special.

      First lets recap in a more clarified (I think) manner the relevant part
      of my notes on Chinese syllabic affricates/fricatives in general:

      One consideration with these sounds is how to spell them when they do
      _not_ occur in combination with other sounds (i.e. when they are not
      part of more complex syllables such as zhuang, chuang, jian, qian, where
      they are combined with "uang" or "ian", for example). When they do occur
      on their own, they are syllabic consonants, if you like (that's why in
      Pinyin an "i" is added, which I've given in brackets. This "i" is only
      used in romanization when no other sounds follow (so "zhuang" instead of
      "zhiuang": in "zhuang" you have the following "ua" vowel that makes the
      syllable "sound", so you don't need the "i" sound, and since it's not
      there in the pronunciation in such cases, it's not found in the spelling
      either; but when there's only "zh", it has to be spelled "zhi", and
      analogically "chi", ji, qi, etc.). In other words, this "i" represent
      the "sound" you have to produce to make the consonant "sound" at all
      when it appears "on its own":)). Like all other syllables, they
      represent one syllable/character.

      (This "i" is thus different from the one seen in syllables such as "ti,
      di, mi, ni, liao, tiao, diao, miao, lian, liang, qian, jian, qiao, jiao
      etc etc.)

      I would propose to simply use an under-dot, even though in the
      orthographic English tehta mode (which I'm generally basing this mode
      on) the under-dot indicates a silent "e" rather than a syllabic consonant.

      Still, I think we can "borrow" this feature from the "Bombadil mode",
      since it a) doesn't conflict with any other sounds/spellings and b)
      comes in really handy here.

      So far, so good. The problem with q(i) and j(i) is that they do occur in
      combinations such as "jian" or "qian". Here, the "a"-sound is pronounced
      different than usually. In many words, the "a" in Hanyu Pinyin signifies
      a pretty straightforward "a"-sound similar to the sound in German "matt"
      or the Quenya short "a" (examples would be the "a"-sounds ta, da, ma,
      na, cha, zha, fa, a, tiao, jiao, qiao, liao, guai, kuai, zhang, chang,
      kang, gang, liang, han, gan, man, nan, kan etc.).

      But in "jian" or "qian" (and also in combinations such as tian, dian,
      lian, mian, nian, i.e. when we are dealing with syllables ending in "ia
      + n" (_not_ "ia + ng": here the "a" is pronounced "normal"), the "a"
      sound much more like the "e"-sound in the unstressed English "can" (as
      in "I can go now"), i.e. the whole sound "ian" sounds pretty much like
      the English (or Scottish :)) name "Ian".
      Therefore, in these syllables I would propose to add a phonetic touch to
      this proposed orthographic mode by using the E-tehta in the relevant
      place. Thus, the syllable "jian" (notice that this "i" is not the same
      that occurs in "ji", where the affricate plus "i"-sound forms the entire
      syllabe; in "jian" the "i" rather forms a complex vowel with the "a") I
      would propose to spell with anca + I-tehta on top of yanta (not stemless
      calma) + nuumen. Analog for "qian", with harma (aha) initially for "q".
      And of course also use the E-tehta for "a" in the other syllables
      mentioned above: tian, dian, lian, mian, nian, to which we might add
      bian, pian, xian.

      But this use of a phonetic feature is open to discussion: it may in fact
      be unnecessary, because even when spelled orthographically, someone who
      knows Chinese will not be confused anyway :).
      Also, this would probably make it "necessary" to also discern in
      spelling the different pronuncation of "e" in syllables such as "hen,
      leng, men, nen etc" vs. "feng, peng, beng etc.", and personally I think
      this would be going a bit too far and make this mode too complicated.
      Conclusion: Maybe skip the E-tehta in syllables in "ian", and rather use
      purely the orthographic A-tehta.
      (But in a phonetic Chinese mode, one might consider representing the
      sounds differently.)

      I'd stick to the other proposed "phonetic" feature though, the under-dot
      under the syllabic affricates/fricatives described earlier. Here, I
      would still rather use the under-dot than an I-tengwa on top, (e.g. si =
      silme with under-dot, zhi = anga with under-dot, qi = harma with
      under-dot, ci = thuule with under-dot etc.).

      Oh well, I also noticed that there is one more trisyllabic vowel sound
      apart from "iao" (and I don't think there are any more than these two,
      having a look at the syllable chart----hopefully I didn't miss anything
      this time), which is "uai" as in shuai, kuai, guai.
      In accordance with my descriptions so far, I'd spell these as follows:

      shuai: hwesta with U-tehta on top + anna with A-tehta on top.
      kuai: quesse with U-tehta on top + anna with A-tehta on top.
      guai: ungwe with U-tehta on top + anna with A-tehta on top.

      A reminder:
      In "kuai" and "guai" we see the previously mentioned dichotomy again
      between aspirated vs. unaspirated (as also in Hanyu Pinyin p and b, t
      and d. This was probably represented "more accurately" in older
      transcription systems such as Wade Giles, where these pairs looked like
      this: p' and p, t' and t, k' and k, the apostrophe marking aspiration
      (so people would less easily confuse aspirated vs unaspirated with
      unvoiced vs. voiced).

      The same dichotomy is BTW found in other Chinese "consonantal pairs",
      such as ch(i) and zh(i), q(i) and j(i) etc. (Wade Giles ch' and ch etc.).
      As I said before, I think that it is ok to represent this dichotomy by
      tengwar "pairs" that Tolkien used to distinguish unvoiced/voiced pairs
      (of plosives/fricatives, and, in English but not Elvish modes, one
      affricate pair: church and jazz), since Chinese has no such distinction,
      but features aspirated/unaspirated (plosives and fricatives/affricates)
      prominently in its stead.

      Notice how the retroflex affricates "ch" and "zh" that I propose to
      spell with calma and anga do in fact sound quite similar to the English
      post-alveolar affricates as in church and jazz, which in English modes
      are usually represented by----calma and anga. They are so similar in
      fact, that many non-native speakers (and quite a few Chinese speakers,
      especially those in southern and other parts of China where Mandarin is
      not the native tongue but "only" the official language people have to
      learn, while most still speak their own dialect of Chinese as first
      language) pronounce the retroflex affricates rather similar to English
      post-alveolar ones. The reason is simple: People are lazy, and rolling
      (or flexing upward) your tongue (which is what "retroflex" really means)
      requires an extra effort, so the post-alveolar sounds, while still not
      the "easiest to pronounce", are less demanding.

      The same phenomenon can be seen for the other retroflex sounds, the
      fricatives "sh(i), r(i)". This being so, I do have a mind to maybe let
      them "switch places" with q(i) and j(i): this way not only would we have
      the retroflex sounds all in the same teema (the third), but as with the
      affricates ch(i), zh(i), these sounds are by many Mandarin speakers
      pronounced rather similar to their post-alveolar English "counterparts"
      (shy and beige) anyway, so it would be most befitting for them to occupy
      the same place in the chart (i.e. be represented with the same tengwa).

      Notice again, though, that the English fricatives/affricate pair(s)
      distinguish between unvoiced and voiced, while the Chinese ones normally
      discern aspirated and unaspirated. The only exception are Chinese
      retroflex sh(i) and r(i): here the distinction is between unvoiced and
      voiced, respectively, just as it is between English post-alveolar shy
      and beige.

      Notice also that q(i) and j(i) would not be so ideally placed (being
      alveola-palatal, and the third series [in English modes] usually being
      palatal), but it's hard to get everything right I guess. After all,
      tengwar weren't really made for Chinese :). I think consistently using
      the third series for the retroflexes is fine, even though we have to
      accept the fact that the fourth teema become a "mixed" velar/palatal
      one. Distinct languages called for different solutions: Surely that's
      something Tolkien would have agreed with, as his adaptations of Tengwar
      to different languages show?

      <>One (hopefully) final thing occurred to me: Chinese (Hanyu Pinyin)
      "x(i)" (alveolo-palatal fricative) should probably be represented by
      wilya (vilya) to ensure it is in the fourth teema together with the
      alveolo-palatal affricates j(i) and ch(i). It's not ideal, since the
      sixth tyelle is usually reserved for approximants (semi-vowels), but
      there would at least be a certain orderliness in the fourth series:
      tyeller 1, 2 and 5 for velars, tyeller 3,4 and 6 for alveolo-palatals.
      The tengwar noldo would not be used.

      Another problem is that I previously assigned wilya to Chinese "w", but
      I think there is an easy solution to this: we can use hwesta sindarinwa
      for "w" instead, since this is in many orthographic English samples uses
      for English "w", which is basically the same as Chinese "w".

      <>Thus:

      ch(i): calma
      zh(i): anga
      sh(i): harma (aha)
      r(i): anca
      q(i): hwesta
      j(i): unque
      x(i): wilya (vilya)
      c(i): thuule (suule)
      z(i): anto
      s(i): silme
      w: hwesta sindarinwa

      <>Other consonants as described in this and my previous two mails:

      t (t'): tinco
      d (t): ando
      p (p'): parma
      b (p): umbar
      k (k'): quesse
      g (k): ungwe
      ng (as in "king"): nwalme
      f: formen
      n: nuumen
      m: malta
      y (consonant, also featuring as vowel tehta "i"): anna
      h: hyarmen
      l: lambe

      Now it should be possible to write every of the 400+ Chinese syllables
      (and thus every of the thousands of Chinese characters as pronounced in
      modern Mandarin) in Tengwar (wouldn't take _too_ long), or if one's too
      lazy for that, just fill in the values I've described in a more
      traditional Tengwar chart. We can start writing Chinese (Mandarin, to be
      correct) in Tengwar :).

      If we accept, where "necessary", the substitution of the Chinese
      aspirated/unaspirated distinction for the English unvoiced/voiced one,
      and the fact that we need to create a teema for retroflexes (though in
      practice often pronounced more like post-alveolars!), the only
      consonants in the 24 value teemar/tyeller grid absolutely _not_ being in
      an adequate spot (in terms of point or method of articulation, or both)
      would be consonantal y (as anna, with the retroflexes), and of course
      q(i), j(i), x(i) (hwesta, unque, wilya in teema four, with the velars k,
      g (k'), and ng). Personally, I would be willing to accept this "mixed"
      teema.

      <>Hísilómë

      Thanks again, Mach, for clarifying a lot of my questions!
      <>







      [Non-text portions of this message have been removed]
    • j_mach_wust
      I don t know Chinese, so I can t help you much. ... The tengwar have two extra to avoid the confusion with diphthongs: yanta and úre, and since you re already
      Message 2 of 23 , Nov 10, 2004
      • 0 Attachment
        I don't know Chinese, so I can't help you much.

        Dave wrote:

        > In syllables such as tui, dui, rui, sui, mao, tao, dao, rao etc. I'd
        > use the relevant consonant tengwa and treat the complex vowel as the
        > diphthong that it is (e.g. U-tehta on anna for "ui", a-tehta on úre
        > for ao).
        >
        > A syllable like "tang" would be spelled tinco with A-tehta on top
        > plus nwalme (ng).
        >
        > Where Y represent an initial palatal sound (consonantal Y), I'd use
        > a long carrier (not anna to avoid confusion with diphthongs), e.g.
        > "yang" = long carrier with a-tehta on top plus nwalme. Using anna
        > plus E-tehta on top plus nwalme would read "aing"--though maybe this
        > is no problem, since this sound doesn't exist in Chinese. Hmm. So,
        > on second thought, I might stick to this spelling, since it couldn't
        > really be misread by someone who understands Chinese.

        The tengwar have two extra to avoid the confusion with diphthongs:
        yanta and úre, and since you're already using úre, you should use
        yanta as well.

        > In "kuai" and "guai" we see the previously mentioned dichotomy again
        > between aspirated vs. unaspirated (as also in Hanyu Pinyin p and b,
        > t and d. This was probably represented "more accurately" in older
        > transcription systems such as Wade Giles, where these pairs looked
        > like this: p' and p, t' and t, k' and k, the apostrophe marking
        > aspiration (so people would less easily confuse aspirated vs
        > unaspirated with unvoiced vs. voiced).
        >
        > The same dichotomy is BTW found in other Chinese "consonantal
        > pairs", such as ch(i) and zh(i), q(i) and j(i) etc. (Wade Giles ch'
        > and ch etc.).As I said before, I think that it is ok to represent
        > this dichotomy by tengwar "pairs" that Tolkien used to distinguish
        > unvoiced/voiced pairs (of plosives/fricatives, and, in English but
        > not Elvish modes, one affricate pair: church and jazz), since
        > Chinese has no such distinction, but features aspirated/unaspirated
        > (plosives and fricatives/affricates) prominently in its stead.

        I don't see any problem with giving the opposition of single bow vs.
        double bow a new meaning, as long as this is consistently done.

        However, I think that the consistency of the témar and tyeller is
        something very important to preserve. There are only two attested
        violations, but both can be justified more or less: The use of vilya
        for /w/ in the mode of Beleriand, even though the óretyelle is used
        for short nasals, but this might be a relatively recent development;
        and the use of calma and anga for /tS, dZ/, but there are reasons in
        the sound structure of English to treat these affricates as stops.

        Secondary users of tengwar, like us, have to be more careful about the
        consistency of our tengwar use if we want it to be approved by the
        critical, because we don't have any authority to justify specific
        inconsistencies.

        > If we accept, where "necessary", the substitution of the Chinese
        > aspirated/unaspirated distinction for the English unvoiced/voiced
        > one, and the fact that we need to create a teema for retroflexes
        > (though in practice often pronounced more like post-alveolars!), the
        > only consonants in the 24 value teemar/tyeller grid absolutely _not_
        > being in an adequate spot (in terms of point or method of
        > articulation, or both) would be consonantal y (as anna, with the
        > retroflexes), and of course q(i), j(i), x(i) (hwesta, unque, wilya
        > in teema four, with the velars k, g (k'), and ng). Personally, I
        > would be willing to accept this "mixed" teema.

        I don't know whether these back consonants could be considered to
        belong to the same "point of articulation".

        If there's a need for more than four points of articulation, then I
        prefer to differentiate them by tehtar, as in the tyelpetéma of
        Quenya. Here's a proposition of an Arabic tengwar mode that uses the
        s-hook to mark a témar difference:

        http://groups.yahoo.com/group/elfscript/message/1501

        ---------------------------
        j. 'mach' wust
        http://machhezan.tripod.com
        ---------------------------
      • Benct Philip Jonsson
        ... Neither do I, but I know some things *about* Mandarin phonology, so I have taken the liberty to upload my suggestion for a transcription. I admit that my
        Message 3 of 23 , Nov 11, 2004
        • 0 Attachment
          j_mach_wust wrote:

          >
          > I don't know Chinese, so I can't help you much.
          >

          Neither do I, but I know some things *about*
          Mandarin phonology, so I have taken the liberty to upload
          my suggestion for a transcription. I admit that my use
          of Kalma, Anga, Harma, Anka for _q, j, ch, zh_ but I hope
          it can be justified with optimalization in the use of the
          available signs. Also Mandarin speakers are said to feel
          that _q, j_ are related to _k, g_, while _ch, zh_ are related
          to _c, z_, so at least they end up next to their "relatives".
          The tone transcription is a real kludge, but still better than
          nothing...


          /BP 8^)
          --
          Benct Philip Jonsson -- melroch at melroch dot se

          Solitudinem faciunt pacem appellant!
          (Tacitus)
        • hisilome
          ... [What exactly does that mean?] ... [Who says that? How would they be related? I ve never read about this, or heard a Chinese person say this :)] ... next
          Message 4 of 23 , Nov 11, 2004
          • 0 Attachment
            --- In elfscript@yahoogroups.com, Benct Philip Jonsson <melroch@m...>
            wrote:
            >
            > Neither do I, but I know some things *about*
            > Mandarin phonology, so I have taken the liberty to upload
            > my suggestion for a transcription. I admit that my use
            > of Kalma, Anga, Harma, Anka for _q, j, ch, zh_ but I hope
            > it can be justified with optimalization in the use of the
            > available signs.

            [What exactly does that mean?]

            > Also Mandarin speakers are said to feel
            > that _q, j_ are related to _k, g_,

            [Who says that? How would they be related? I've never read about this,
            or heard a Chinese person say this :)]

            >while _ch, zh_ are >related> to _c, z_, so at least they end >up
            next to their >"relatives".

            [Well, in as far as the retroflex affricates ch(i) and zh(i) are often
            pronounced "lazily" (see my mail on "lazy pronunciation"), i.e. they
            "deteriorate" to c(i) and z(i), maybe here is some relationship :)
            For q(i), j(i), x(i), I propose using palatalized versions (with two
            underdots) of the dental/alveolar affricates (i.e. of thuule, anto and
            silme).

            Your chart is very interesting, but I do have some questions:
            1) How would you propose to represent complex vowels both efficiently
            and elegantly? Looks like you would need to use an awful lot of
            carriers. Besides yanta and uure, you have proposed no signs for
            tengwar-vowels that could be used in dipththongs and triphthongs.

            2) What's with the wilya for initial "u" (u-)? There is no initial "u"
            in Mandarin :).

            3) No need for roomen and oore! There you've really made me aware of a
            mistake I made in my previous posts, thanks!
            The only "r" in Chinese is the voiced retroflex fricative, and that
            (at least in my mode) is represented by anca: I am much more
            comfortable with putting all the retroflexes in one teema.]

            > The tone transcription is a real kludge, but still better than nothing.
            >
            [Hm. Doesn't look too bad :)! But somehow, I prefer to stick with
            macron, acute, breve and grave.]

            BTW, I'll soon post a fine-tuned version of my Mandarin mode :)

            Hisilome
          • Dave
            j_mach_wust wrote: [First, thanks for your suggestions! I ve incorporated some of them into my fine-tuned sound chart (see post), particularly I have found a
            Message 5 of 23 , Nov 11, 2004
            • 0 Attachment
              j_mach_wust wrote:

              [First, thanks for your suggestions! I've incorporated some of them into
              my fine-tuned sound chart (see post), particularly I have found a better
              solution for alveolo-palatal affricates/fricative (see below), since I
              agree that each teema should basically represent sounds with the same
              point of articulation].

              > I don't know Chinese, so I can't help you much.
              >
              > Dave wrote:
              >
              > > In syllables such as tui, dui, rui, sui, mao, tao, dao, rao etc. I'd
              > > use the relevant consonant tengwa and treat the complex vowel as the
              > > diphthong that it is (e.g. U-tehta on anna for "ui", a-tehta on úre
              > > for ao).
              > >
              > > A syllable like "tang" would be spelled tinco with A-tehta on top
              > > plus nwalme (ng).
              > >
              > > Where Y represent an initial palatal sound (consonantal Y), I'd use
              > > a long carrier (not anna to avoid confusion with diphthongs), e.g.
              > > "yang" = long carrier with a-tehta on top plus nwalme. Using anna
              > > plus E-tehta on top plus nwalme would read "aing"--though maybe this
              > > is no problem, since this sound doesn't exist in Chinese. Hmm. So,
              > > on second thought, I might stick to this spelling, since it couldn't
              > > really be misread by someone who understands Chinese.
              >
              > The tengwar have two extra to avoid the confusion with diphthongs:
              > yanta and úre, and since you're already using úre,

              [I use úre only as a vowel-tehta, not consonantal!----which is what you
              seem to suggest when you say "to avoid confusion with diphthongs"
              (vowels), and suggest, as I think you do (?), that yanta should be used
              for consonantal "y" (which in fact I have decided to do, but for a
              different reason, not because I'm "using úre already" :). Anyway, in
              your orthographic English tehta mode, yanta, úre and anna are _all_ used
              as the second sound in various dipthongs, and anna is also used for
              consonantal "y" (in accordance with DTS 5, 10, 62), while the other two
              _have no_ consonantal values----so here yanta and úre are _not_ used to
              "avoid confusion with diphthongs, I think, but as part of diphthongs.
              Well, I admit I didn't quite get you there, maybe you can clarify this :)).

              > you should use yanta as well.

              [I see. Thanks for this suggestion. As I pointed out, it wouldn't really
              matter, since a Chinese speaker would not usually misread "yang" even
              when spelled anna with A tehta on top + nwalme (i.e. he/she would know
              that this can't be a diphthong but a simple consonant + vowel
              combination). However, yanta with A-tehta on top plus nwalme on top is
              better to avoid confusion in the few cases were it might occur (see below).

              As for úre, as I said, it is used _only_ in diphthongs in the
              orthographic English tehta mode (proposed by you in accordance with DTS
              5, 10, 62) I'm basing this on, and even there its use is hypothetical
              (though I like it, and have incorporated this use into my Chinese mode).

              And úre is not used as a consonant in the known samples (by Tolkien) of
              the English orthographic tehta mode neither, right?

              Well I guess you mean yanta and úre could be used for "y" and "w(u)" in
              English modes to avoid confusion with anna/vilya, but which attested
              mode uses these values? For "w", English modes have vala (not vilya)
              anyway, and anna as second part in diphthongs is again hypothetical (but
              useful and also incorporated into my Mandarin mode).
              (In Quenya, anna with two under-dots is used for consonantal "y".)

              Since I use úre only as a vowel-tengwa (not for any consonant), there
              won't be any confusion.

              Anyway, analysed carefully, initial "y" in the Hanyu Romanization script
              can represent three sound values (this is the one single letter that's
              used most inconsistently in Pinyin I find):

              1) consonantal (palatal) y: as in ya, yan, yang, yao, yo, yong, you and
              yuan.

              2) vowel (really written "yi") (sound of "e" in English "me"): as in yi,
              yin and ying. "Yi" in all these three syllables is really just the
              "e"-sound on its own, not e + i: this should really better be spelled
              just with "y", but I guess the inventors of Pinyin though that wouldn't
              look good: you would get y, yn and yng:). Very confusing to the
              learner... Another special case is the syllable "ye": here we have y
              (e-sound in English me) + schwah, but when you pronounce it, the "y" is
              somewhat palatalized----but not as strongly as in the syllables listed
              under 1). It's something in between.

              3) no value of its own, serves as a vowel qualifier, as in yu, yuan, yue
              and yun. Here, "y" indicates that the following "u" is not read as "u"
              (sound in English "too" or Quenya "u"), but as the German Umlaut "ü" (or
              French u-sound in lune, Sindarin vocalic "y").

              This means, "y" serves a variety of functions in Pinyin, and in a
              phonetic mode you might have to worry about different representation.
              Since I'm proposing a consonantal mode, though, I think the safest and
              most convenient way is too simiply represent "y" consistently by "yanta".
              Thanks again for suggesting that!

              This also solves the problem that there _could_ be confusion otherwise
              (i.e. if you use anna for "y") in case of a few syllables, e.g.

              ai and ya would be spelled identically (anna plus A-tehta on top).

              Actually, this is the only example I can think of, but even so, this
              settles it in favor of "yanta" for Pinyin "y" _in all its functions_
              (the three described above).

              y = yanta.

              No confusion possible with yanta as second part of diphthongs, since
              initially, no sounds like "ae, oe, ie, ee, ue" can occur.

              Now we can use anna as second part of diphthongs/vowel combinations
              (those in "-i") without any possible misreadings.

              >
              > > In "kuai" and "guai" we see the previously mentioned dichotomy again
              > > between aspirated vs. unaspirated (as also in Hanyu Pinyin p and b,
              > > t and d. This was probably represented "more accurately" in older
              > > transcription systems such as Wade Giles, where these pairs looked
              > > like this: p' and p, t' and t, k' and k, the apostrophe marking
              > > aspiration (so people would less easily confuse aspirated vs
              > > unaspirated with unvoiced vs. voiced).
              > >
              > > The same dichotomy is BTW found in other Chinese "consonantal
              > > pairs", such as ch(i) and zh(i), q(i) and j(i) etc. (Wade Giles ch'
              > > and ch etc.).As I said before, I think that it is ok to represent
              > > this dichotomy by tengwar "pairs" that Tolkien used to distinguish
              > > unvoiced/voiced pairs (of plosives/fricatives, and, in English but
              > > not Elvish modes, one affricate pair: church and jazz), since
              > > Chinese has no such distinction, but features aspirated/unaspirated
              > > (plosives and fricatives/affricates) prominently in its stead.
              >
              > I don't see any problem with giving the opposition of single bow vs.
              > double bow a new meaning, as long as this is consistently done.

              [It is, with the one and only mentioned exception of sh(i) and r(i),
              where we have to revert to the original English distinction between
              unvoiced and voiced. This shouldn't be a problem though.]

              >
              > However, I think that the consistency of the témar and tyeller is
              > something very important to preserve. There are only two attested
              > violations, but both can be justified more or less: The use of vilya
              > for /w/ in the mode of Beleriand, even though the óretyelle is used
              > for short nasals, but this might be a relatively recent development;
              > and the use of calma and anga for /tS, dZ/, but there are reasons in
              > the sound structure of English to treat these affricates as stops.
              >
              > Secondary users of tengwar, like us, have to be more careful about the
              > consistency of our tengwar use if we want it to be approved by the
              > critical, because we don't have any authority to justify specific
              > inconsistencies.


              [I guess this is a good point. See below.]

              >
              > > If we accept, where "necessary", the substitution of the Chinese
              > > aspirated/unaspirated distinction for the English unvoiced/voiced
              > > one, and the fact that we need to create a teema for retroflexes
              > > (though in practice often pronounced more like post-alveolars!), the
              > > only consonants in the 24 value teemar/tyeller grid absolutely _not_
              > > being in an adequate spot (in terms of point or method of
              > > articulation, or both) would be consonantal y (as anna, with the
              > > retroflexes), and of course q(i), j(i), x(i) (hwesta, unque, wilya
              > > in teema four, with the velars k, g (k'), and ng). Personally, I
              > > would be willing to accept this "mixed" teema.
              >
              > I don't know whether these back consonants could be considered to
              > belong to the same "point of articulation".
              >
              > If there's a need for more than four points of articulation, then I
              > prefer to differentiate them by tehtar, as in the tyelpetéma of
              > Quenya. Here's a proposition of an Arabic tengwar mode that uses the
              > s-hook to mark a témar difference:

              [Yes, I was thinking to use only quesse (k), ungwe (g) and nwalme (ng)
              from the quessetema, and find a better solution than hwesta, unque and
              wilya for the alveolo-palatal sounds q(i), j(i), x(i), namely to spell
              them with thuule, anto and silme with two under-dots to indicate
              palatalization (alveolo-palatal sounds are also palatal in a broader
              sense :)).
              This would be all the more befitting since these are in fact
              "palatalized versions" of the sounds represented by thuule, anto and
              silme _without_ under-dots in my mode, the dental/alveolar sounds c(i),
              z(i) and s(i).
              Also, the affricates would all be represented by tengwar in the third
              and fourth tyeller, while the fricatives in both groups of sounds, x(i)
              and s(i) would both be represented by silme (with/without under-dots).

              The only slight "problem" would be that the affricates ch(i) and zh(i)
              would be in the first and second (plosive) tyller, but as you've pointed
              out, English modes also make an "exception" here by placing the
              palato-alveolar affricates as in English church/jazz in the first and
              second tyelle of the third teema, so this should be ok.

              The third and fourth tyelle of the third teema wouls also be affricates
              [sh(i) and r(i)], which would be in sync with the affricates in grade 3
              and 4 of teema 1 and 4. Parmateema would have the third Chinese
              fricative (after s[i] and x[i], "f".

              It has also occured to me that one could still do without yanta for
              consonantal "y", BTW, by simply putting those under-dots under anna when
              it's used for "y" (no under-dots of course when it stands for "i" in
              diphthongs). After all, even in the examples yu, yuan, yue, yun, where
              the "y" qualifies the "u" to become a German Umlaut "ü", this is also
              kind of a "palatalized" vowel, which is probably why in the Sindarin
              Mode of Gondor it's spelled with two dots (though over-dots) and
              romanized as "y".

              You said since I'm useing uure, I should use anna, but I _am_ using both
              for vowels. The only difference it would make to use yanta for "y" would
              be that then uure and anna would _both_ be used _only_ for vowels, while
              not using yanta for "y" means that anna (but with two under-dots to mark
              the difference) would be used as both vowel and consonant tehta.

              I prefer anna with under-dots for consonantal y, actually, since the
              palatal y _is_ of course a palatalized version of the vowel "i" (which
              it can still represent without the dots in vowels/diphthongs).

              So:

              y: yanta with under-dots
              i as part of i-glide diphthongs: yanta without under-dots (as in the
              Mode of Gondor)

              Hísilómë








              [Non-text portions of this message have been removed]
            • Benct Philip Jonsson
              ... I meant the medial -u- in e.g. _guo_. -- /BP 8^) -- Benct Philip Jonsson -- melroch at melroch dot se Solitudinem faciunt pacem appellant! (Tacitus)
              Message 6 of 23 , Nov 12, 2004
              • 0 Attachment
                hisilome wrote:
                > 2) What's with the wilya for initial "u" (u-)? There is no initial "u"
                > in Mandarin :) .

                I meant the medial -u- in e.g. _guo_.

                --

                /BP 8^)
                --
                Benct Philip Jonsson -- melroch at melroch dot se

                Solitudinem faciunt pacem appellant!
                (Tacitus)
              • j_mach_wust
                The following proposals are based on http://www.wfu.edu/~moran/Cathay_Cafe/IPA_NPA_4.htm and on http://en.wikipedia.org/wiki/Pinyin . ... (I suppose you meant
                Message 7 of 23 , Nov 16, 2004
                • 0 Attachment
                  The following proposals are based on
                  http://www.wfu.edu/~moran/Cathay_Cafe/IPA_NPA_4.htm and on
                  http://en.wikipedia.org/wiki/Pinyin .

                  Dave wrote:

                  > I use úre only as a vowel-tehta, not consonantal!

                  (I suppose you meant vowel-tengwa.) I think talking of vowel tengwar
                  in a tehtar mode is not helpful at all. I'd rather say that in a
                  tehtar mode, all vowels are represented by tehtar, and tengwar
                  represent only consonants. All diphthongs are analyzed into sequences
                  of vowels (represented with tehtar) and consonants (represented with
                  tengwar). The diphthong _au_ (as in _house_), for instance, is
                  analyzed into the vowel _a_ and the consonant _w_ (haws); the
                  diphthong _iu_ (as in cute), for instance, is analyzed into the
                  consonant _y_ and the vowel _u:_ (kyu:t).

                  Luckily, the Latin alphabet provides us with some special letters to
                  mark the consonanticity: consonant _u_ and _i_ are represented with
                  the letters _w_ and _j_ (or, more anglocentric, with _y_).
                  Unfortunately, there are no special letters for consonant _e/o_ in the
                  Latin script. Only in tengwar, the distinction between vowel _e/o_ and
                  consonant _e/o_ can be marked. The same applies to consonant _y_.

                  As for the Pinyin diphthongs, I'd suggest to analyze them into
                  preceding consonant j/w/y + vowel + following consonant j/w. According
                  to http://en.wikipedia.org/wiki/Pinyin , this analyzis covers all
                  possible diphthongs if only we consider _ao_ to be _aw_.

                  Consonant _j/w_ are no problem in tengwar, since they're well attested
                  in many different modes: Preceding consonant _j/w_ is represented
                  either with anna*/vala if there's no preceding consonant or with two
                  dots below/a modified u-curl above a preceding consonant; following
                  _j/w_ is written with yanta/úre that bear the preceding vowel tehta.

                  The only problem is the preceding consonant _y_. I suggest to
                  represent it by a combination of the signs for consonant _j_ and
                  consonant _w_: either by a modified u-curl on anna* if there's no
                  preceding consonant or by two dots below and a modified u-curl above a
                  preceding consonant. Actually, this procedure can also be applied for
                  the transcription of _ü_ (u-umlaut), if only we consider _ü_ to be _yu_.

                  *side-note on anna:

                  > (In Quenya, anna with two under-dots is used for consonantal "y".)

                  That's true, but I think it's most important that in that particular
                  mode, anna is the weakest sound (óretyelle) of the velar series
                  (calmatéma), a sound tending to disappear and thus taking the function
                  of a carrier for special purposes. I think the tyeller consistency is
                  very important for this use. So if there's no palatal series, the
                  tengwa for consonantal j should be the óretyelle tengwa of the velar
                  series with two dots below.

                  > > I don't know whether these back consonants could be considered to
                  > > belong to the same "point of articulation".
                  > >
                  > > If there's a need for more than four points of articulation, then
                  > > I prefer to differentiate them by tehtar, as in the tyelpetéma of
                  > > Quenya. Here's a proposition of an Arabic tengwar mode that uses
                  > > the s-hook to mark a témar difference:
                  >
                  > Yes, I was thinking to use only quesse (k), ungwe (g) and nwalme
                  > (ng) from the quessetema, and find a better solution than hwesta,
                  > unque and wilya for the alveolo-palatal sounds q(i), j(i), x(i),

                  If Benct's remark that these sounds feel related to the velar sounds
                  for Chinese speakers is true, then it'd be the most elegant solution
                  to use the calmatéma as a palatal-velar series.

                  > Also, the affricates would all be represented by tengwar in the
                  > third and fourth tyeller, while the fricatives in both groups of
                  > sounds, x(i) and s(i) would both be represented by silme
                  > (with/without under-dots).
                  >
                  > The only slight "problem" would be that the affricates ch(i) and
                  > zh(i) would be in the first and second (plosive) tyller, but as
                  > you've pointed out, English modes also make an "exception" here by
                  > placing the palato-alveolar affricates as in English church/jazz in
                  > the first and second tyelle of the third teema, so this should be
                  > ok.

                  In English, this may be justified since /tS/ and /dZ/ are the only
                  English affricates. Chinese, however, has much more of them, and I
                  think it's an unnecessary inconsistency that some affricates are
                  represented by tinco-andotyelle signs, but other affricates by
                  thúle-antotyelle signs (the latter tyeller, by the way, aren't pure
                  affricate tyeller either since there's at least one fricative: f, and
                  there might be a second: h).

                  I'd rather suggest to use the tyeller with extended stems (above and
                  below the lúva) for affricates, since they're said to be used for
                  required combinations.

                  As for the fifth téma: If the problem isn't solved in Benct's way,
                  then I'd rather mark the retroflex sounds as variants of the alveolar
                  sounds by adding an s-hook to the latter, since the two dots below are
                  required for the diphthongs and can't be used for distinguishing
                  another téma.

                  ---------------------------
                  j. 'mach' wust
                  http://machhezan.tripod.com
                  ---------------------------
                • Dave
                  j_mach_wust wrote: [Thanks a lot for your suggestions! I m still trying to figure out how to incorporate some of them into the mode... I got the impression
                  Message 8 of 23 , Nov 17, 2004
                  • 0 Attachment
                    j_mach_wust wrote:

                    [Thanks a lot for your suggestions! I'm still trying to figure out how
                    to incorporate some of them into the mode...
                    I got the impression from your suggestions that you are not very happy
                    anymore with all your proposed transcriptions of diphthongs in your
                    proposed English orthographic tehta mode?]

                    > The following proposals are based on
                    > http://www.wfu.edu/~moran/Cathay_Cafe/IPA_NPA_4.htm
                    > <http://www.wfu.edu/%7Emoran/Cathay_Cafe/IPA_NPA_4.htm> and on
                    > http://en.wikipedia.org/wiki/Pinyin .
                    >
                    > Dave wrote:
                    >
                    > > I use úre only as a vowel-tehta, not consonantal!
                    >
                    > (I suppose you meant vowel-tengwa.)

                    [Right.]

                    > I think talking of vowel tengwar
                    > in a tehtar mode is not helpful at all. I'd rather say that in a
                    > tehtar mode, all vowels are represented by tehtar, and tengwar
                    > represent only consonants. All diphthongs are analyzed into sequences
                    > of vowels (represented with tehtar) and consonants (represented with
                    > tengwar). The diphthong _au_ (as in _house_), for instance, is
                    > analyzed into the vowel _a_ and the consonant _w_ (haws); the
                    > diphthong _iu_ (as in cute), for instance, is analyzed into the
                    > consonant _y_ and the vowel _u:_ (kyu:t).
                    >
                    > Luckily, the Latin alphabet provides us with some special letters to
                    > mark the consonanticity: consonant _u_ and _i_ are represented with
                    > the letters _w_ and _j_ (or, more anglocentric, with _y_).
                    > Unfortunately, there are no special letters for consonant _e/o_ in the
                    > Latin script. Only in tengwar, the distinction between vowel _e/o_ and
                    > consonant _e/o_ can be marked. The same applies to consonant _y_.

                    [This sounds very good. But as e/o are marked with yanta and and uure in
                    my Chinese mode, as in the English orthographic tehta mode (EOTM) and
                    the the Sindarin modes, how can you propose to also use yanta for
                    following i(j)(see below)?]

                    >
                    > As for the Pinyin diphthongs, I'd suggest to analyze them into
                    > preceding consonant j/w/y + vowel + following consonant j/w. According
                    > to http://en.wikipedia.org/wiki/Pinyin , this analyzis covers all
                    > possible diphthongs if only we consider _ao_ to be _aw_.

                    [I think this could be problematic. Mandarin (-)ao is not the same as
                    the _au_ in Quenya or German (and thus not quite _aw_), even though the
                    difference may not be a great as between -o and -uo (see below).]

                    >
                    > Consonant _j/w_ are no problem in tengwar, since they're well attested
                    > in many different modes: Preceding consonant _j/w_ is represented
                    > either with anna*/vala if there's no preceding consonant or with two
                    > dots below/a modified u-curl above a preceding consonant;

                    [Hm. Were is the problem with representing preceding j(i)/w(u) with
                    I-tehta and U-tehta?]

                    > following
                    > _j/w_ is written with yanta/úre that bear the preceding vowel tehta.

                    [I hinted above that this creates a problem: the frequent Mandarin
                    combinations -ie and -ue are in my mode already expressed by
                    I-tehta/U-tehta plus yanta (cf. spelling of "Michael" in Endorion
                    Dedication), if we also spell following -i(j) with yanta it would be
                    quite confusing, because combinations like -ui and -ai would incorporate
                    the same vowel tengwa, and -ui and -ai in particular would become
                    identical. Why not stick to anna for following i/j as you also proposed
                    in your EOTM?
                    I also see no problem with vala for u-glide diphthongs such as -u. Also,
                    if you use uure for following -u(w), what about following -o (for which
                    I propose uure), which is a distinct sound?
                    And BTW, I don't think the Wikipedia article is quite correct when it
                    states that "'o' is written after b p m f, "uo" everywhere else". The -o
                    in bo, po, mo, fo is quite distinct from the o-glide diphthong seen in
                    syllables such as duo, tuo or luo, therefore we cannot simply treat uo
                    as o, and that means that we need a distinct representation for (-)ao
                    and -uo.]

                    >
                    > The only problem is the preceding consonant _y_.

                    [I don't see why it would be. As you note further above ("consonant _u_
                    and _i_ are represented with the letters _w_ and _j_ [or, more
                    anglocentric, with _y_]"), i,j,y in diphthongs all represent the same
                    sound, i.e. consonantal y (as in "yes"), so why not use anna with two
                    underdots (as in Quenya, in my proposed mode the calmateema is not
                    _strictly_ speaking a palatal series)?]

                    > I suggest to
                    > represent it by a combination of the signs for consonant _j_ and
                    > consonant _w_: either by a modified u-curl on anna* if there's no
                    > preceding consonant or by two dots below and a modified u-curl above a
                    > preceding consonant.

                    [As I mentioned, Pinyin y occurs only initially.]

                    > Actually, this procedure can also be applied for
                    > the transcription of _ü_ (u-umlaut), if only we consider _ü_ to be _yu_.

                    [Not sure what you mean here. _yu_ is really nothing but _ü_, and y is
                    only used in my mode as an orthographical spelling of the sound,
                    corresponding to Pinyin "yu" (also "yuan").]

                    >
                    > *side-note on anna:
                    >
                    > > (In Quenya, anna with two under-dots is used for consonantal "y".)
                    >
                    > That's true, but I think it's most important that in that particular
                    > mode, anna is the weakest sound (óretyelle) of the velar series
                    > (calmatéma), a sound tending to disappear and thus taking the function
                    > of a carrier for special purposes. I think the tyeller consistency is
                    > very important for this use. So if there's no palatal series, the
                    > tengwa for consonantal j should be the óretyelle tengwa of the velar
                    > series with two dots below.
                    >
                    > > > I don't know whether these back consonants could be considered to
                    > > > belong to the same "point of articulation".
                    > > >
                    > > > If there's a need for more than four points of articulation, then
                    > > > I prefer to differentiate them by tehtar, as in the tyelpetéma of
                    > > > Quenya. Here's a proposition of an Arabic tengwar mode that uses
                    > > > the s-hook to mark a témar difference:
                    > >
                    > > Yes, I was thinking to use only quesse (k), ungwe (g) and nwalme
                    > > (ng) from the quessetema, and find a better solution than hwesta,
                    > > unque and wilya for the alveolo-palatal sounds q(i), j(i), x(i),
                    >
                    > If Benct's remark that these sounds feel related to the velar sounds
                    > for Chinese speakers is true, then it'd be the most elegant solution
                    > to use the calmatéma as a palatal-velar series.

                    [Hm. But then you would have a mixed teema again with quite distinct
                    sounds lumped together. (BTW, this is exactly what I suggested in an
                    earlier version of my mode: q, j, x plus k and g (maybe h) in one teema,
                    but I changed that following your suggestion to "preserve the
                    consistency of the teemar and tyeller".) Also, I still haven't found a
                    Chinese person to confirm the theory that velar sounds and
                    alveolo-palatal ones "feel related" :). Certainly don't to me.]

                    >
                    > > Also, the affricates would all be represented by tengwar in the
                    > > third and fourth tyeller, while the fricatives in both groups of
                    > > sounds, x(i) and s(i) would both be represented by silme
                    > > (with/without under-dots).
                    > >
                    > > The only slight "problem" would be that the affricates ch(i) and
                    > > zh(i) would be in the first and second (plosive) tyller, but as
                    > > you've pointed out, English modes also make an "exception" here by
                    > > placing the palato-alveolar affricates as in English church/jazz in
                    > > the first and second tyelle of the third teema, so this should be
                    > > ok.
                    >
                    > In English, this may be justified since /tS/ and /dZ/ are the only
                    > English affricates. Chinese, however, has much more of them, and I
                    > think it's an unnecessary inconsistency that some affricates are
                    > represented by tinco-andotyelle signs, but other affricates by
                    > thúle-antotyelle signs (the latter tyeller, by the way, aren't pure
                    > affricate tyeller either since there's at least one fricative: f, and
                    > there might be a second: h).

                    [This is a valid point. Thinking about it.]

                    >
                    > I'd rather suggest to use the tyeller with extended stems (above and
                    > below the lúva) for affricates, since they're said to be used for
                    > required combinations.

                    [Another interesting suggestion. Never liked the extended stems much,
                    but I'll give it some thought :).
                    I still feel though that only two affricates and one fricative (if we
                    use hyarmen for h) "out of line" are not too bad a price to pay, if it
                    allows us to make use of almost all the basic 24 tengwar. But it is not
                    perfect, so maybe I'll experiment with those extended stems.]

                    >
                    > As for the fifth téma: If the problem isn't solved in Benct's way,

                    [I don't think it is. In his chart, the retroflex sounds are scattered
                    over three different teemar (tinco, parma, calma).]

                    > then I'd rather mark the retroflex sounds as variants of the alveolar
                    > sounds by adding an s-hook to the latter, since the two dots below are
                    > required for the diphthongs and can't be used for distinguishing
                    > another téma.

                    [So you'd use the calmateema for velar/palatal sounds, and mark the
                    retroflexes as variant of the dental/alveolar sounds----and hten what
                    about the quesseteema? Largely empty? That doesn't strike me as economic...
                    As for the under-dots: As I pointed out, Mandarin q, j, x (the
                    alveolo-palatal affricates and fricative) can very well be considered to
                    be palatalized versions of c, z, s (their dental/alveolar counterparts),
                    therefore I do still think using two underdots a good solution,
                    especially when one uses your suggestions from the EOTM for representing
                    _i_(j) in dipthongs (I-tehta for preceding/anna for following).
                    Confusion with preceding "y" (anna plus underdots) is not possible.]

                    Hísilómë




                    [Non-text portions of this message have been removed]
                  • Benct Philip Jonsson
                    ... Well it is. The palatal initials developed historically out of the velars. In fact they are still in complementary distribution. See
                    Message 9 of 23 , Nov 17, 2004
                    • 0 Attachment
                      j_mach_wust wrote:

                      >
                      > If Benct's remark that these sounds feel related to the velar sounds
                      > for Chinese speakers is true, then it'd be the most elegant solution
                      > to use the calmatéma as a palatal-velar series.
                      >
                      >

                      Well it is. The palatal initials developed historically out of
                      the velars. In fact they are still in complementary distribution.
                      See <http://www.wfu.edu/~moran/Cathay_Cafe/IPA_NPA_4.htm>
                      There have also been romanization systems which exploited this fact,
                      so that e.g. _jing_ was written _ging_ or _king_.


                      --

                      /BP 8^)
                      --
                      Benct Philip Jonsson -- melroch at melroch dot se

                      Solitudinem faciunt pacem appellant!
                      (Tacitus)
                    • hisilome
                      ... [I never denied this phonological development. What I do doubt is that an awful lot of average native speakers are aware of it, and actually consider these
                      Message 10 of 23 , Nov 17, 2004
                      • 0 Attachment
                        --- In elfscript@yahoogroups.com, Benct Philip Jonsson <melroch@m...>
                        wrote:
                        > j_mach_wust wrote:
                        >
                        > >
                        > > If Benct's remark that these sounds feel related to the velar sounds
                        > > for Chinese speakers is true, then it'd be the most elegant solution
                        > > to use the calmatéma as a palatal-velar series.
                        > >
                        > >
                        >
                        > Well it is. The palatal initials developed historically out of
                        > the velars. In fact they are still in complementary distribution.
                        > See <http://www.wfu.edu/~moran/Cathay_Cafe/IPA_NPA_4.htm>
                        > There have also been romanization systems which exploited this fact,
                        > so that e.g. _jing_ was written _ging_ or _king_.
                        >
                        >
                        [I never denied this phonological development. What I do doubt is that
                        an awful lot of average native speakers are aware of it, and actually
                        consider these sounsds as "relatives", which is quite a different
                        matter. The original wording was "Mandarin speakers are said to feel
                        that _q, j_ are related to _k, g_, while _ch, zh_ are related to _c,
                        z_,(...)".

                        What speakers of a language _feel_ may be quite different from what
                        they actually _know_ :). For example, I doubt that many English
                        speakers know the phonological/etymological/historical causes for,
                        say, the many "irregularities" in English spelling----in fact, most
                        native speakers are probably not even acutely _aware_ of them :).

                        Consider also that Romanization systems _before_ Pinyin were virtually
                        all exclusively designed by Western linguists/scholars who looked at
                        the language very differently (with academic rigor and distance)
                        than the average speaker.

                        So, if you say that for historical reasons (i.e. from a diachronic
                        point of view) these sounds should somehow be put in each other's
                        vicinity, I'd have much less of a problem with that. And since the
                        Elves are said to have devised their writing system(s) with a keen
                        awareness of linguistic/phonological/historical details and
                        devolopments, this is certainly a valid consideration.]

                        Hisilome
                      • j_mach_wust
                        ... What made you think that? I m very hesitating about úre, since it s not attested at all, but that s all. ... There s an important difference between the
                        Message 11 of 23 , Nov 17, 2004
                        • 0 Attachment
                          Dave 'Hísilómë' wrote:

                          > I got the impression from your suggestions that you are
                          > not very happy anymore with all your proposed
                          > transcriptions of diphthongs in your proposed English
                          > orthographic tehta mode?]

                          What made you think that? I'm very hesitating about úre, since it's
                          not attested at all, but that's all.

                          > But as e/o are marked with yanta and and uure in my
                          > Chinese mode, as in the English orthographic tehta mode
                          > (EOTM) and the the Sindarin modes, how can you propose to
                          > also use yanta for following i(j)(see below)?]

                          There's an important difference between the Chinese mode and the
                          English mode: The English mode places the vowel tehtar on the
                          following tengwar, so that yanta/úre are free to represent other
                          sounds than j/w, whereas the Chinese mode places the vowel tehtar on
                          the preceding tengwar, so that yanta/úre are mere variants for the
                          representations of j/w (at least they are in all attested modes, and
                          there's no need to break with this in a Chinese mode).

                          > > As for the Pinyin diphthongs, I'd suggest to analyze
                          > > them into preceding consonant j/w/y + vowel + following
                          > > consonant j/w. According to
                          > > http://en.wikipedia.org/wiki/Pinyin , this analyzis
                          > > covers all possible diphthongs if only we consider _ao_
                          > > to be _aw_.
                          >
                          > [I think this could be problematic. Mandarin (-)ao is not
                          > the same as the _au_ in Quenya or German (and thus not
                          > quite _aw_), even though the difference may not be a great
                          > as between -o and -uo (see below).]

                          The pronunciation of /aU/ isn't the same in German and in English
                          either. Actually, the German sound is often understood to be [ao]
                          rather than [aU], but still it's written with the letters <au> and
                          nobody has a problem with that. There are also Chinese romanization
                          systems that use the letters <au>. It a similar case like the English
                          tengwar modes where the same tengwar transcription is used for both
                          <au> and <aw>, no the English modes are even much worse, since in the
                          Chinese modes, the transcription of Pinyin <ao> as if it were <aw>
                          can't create any ambiguity at all, whereas in English, there might
                          possibly be a pair of words that is distinguished only by the letters
                          <au> vs. <aw>.

                          To make it explicit: I suggest not to use yanta and úre for the
                          representation of the letters <e/o>. See below.

                          > > Consonant _j/w_ are no problem in tengwar, since they're
                          > > well attested in many different modes: Preceding
                          > > consonant _j/w_ is represented either with anna*/vala if
                          > > there's no preceding consonant or with two dots below/a
                          > > modified u-curl above a preceding consonant;
                          >
                          > [Hm. Were is the problem with representing preceding
                          > j(i)/w(u) with I-tehta and U-tehta?]

                          It's not attested at all. Of course you can choose whatever strange
                          unattested spellings you wish, but the chance that a mode of your
                          creation will be accepted by others will increase significantly if you
                          base it on the attested samples by J. R. R. Tolkien, since they are
                          the only samples that everybody accepts.

                          > > following _j/w_ is written with yanta/úre that bear the
                          > > preceding vowel tehta.
                          >
                          > [I hinted above that this creates a problem: the frequent
                          > Mandarin combinations -ie and -ue are in my mode already
                          > expressed by I-tehta/U-tehta plus yanta (cf. spelling of
                          > "Michael" in Endorion Dedication), if we also spell
                          > following -i(j) with yanta it would be quite confusing,
                          > because combinations like -ui and -ai would incorporate
                          > the same vowel tengwa, and -ui and -ai in particular would
                          > become identical. Why not stick to anna for following i/j
                          > as you also proposed in your EOTM?

                          If we assume that the calmatéma really is the palatal series, then the
                          reason is that I'd like to distinguish <ai> from <ya>. The óretyelle
                          sign of the palatal series is in the vast majority of the attested
                          modes the normal sign for the sound of j (consonantal i/anglocentric
                          consonantal y). In the modes where the tengwar bear the preceding
                          vowel tehtar, anna bears the preceding vowel tehta, in the modes where
                          the tengwar bear the following vowel tehtar, anna bears the following
                          vowel tehta (and yanta is used for the same sound when required to
                          bear the preceding vowel tehta).

                          > I also see no problem with vala for u-glide diphthongs
                          > such as -u. Also, if you use uure for following -u(w),
                          > what about following -o (for which I propose uure), which
                          > is a distinct sound?

                          Analogous to above.

                          > And BTW, I don't think the Wikipedia article is quite
                          > correct when it states that "'o' is written after b p m f,
                          > "uo" everywhere else". The -o in bo, po, mo, fo is quite
                          > distinct from the o-glide diphthong seen in syllables such
                          > as duo, tuo or luo, therefore we cannot simply treat uo as
                          > o,

                          Did anybody ever suggest such a thing?

                          > and that means that we need a distinct representation for
                          > (-)ao and -uo.]

                          No. The Pinyin transcription <ao> is analyzed as <aw> (as I've written
                          in the previous mail and above), and Pinyin <uo> is analyzed as <wo>.

                          > > The only problem is the preceding consonant _y_.
                          >
                          > [I don't see why it would be.

                          I wasn't precise enough. I meant preceding consonant version of the
                          soun [y], that is the preceding sound [H], that is preceding Pinyin
                          letters <yu> or <ü>.

                          > > I suggest to represent it by a combination of the signs
                          > > for consonant _j_ and consonant _w_: either by a
                          > > modified u-curl on anna* if there's no preceding
                          > > consonant or by two dots below and a modified u-curl
                          > > above a preceding consonant.
                          >
                          > [As I mentioned, Pinyin y occurs only initially.]
                          >
                          > > Actually, this procedure can also be applied for the
                          > > transcription of _ü_ (u-umlaut), if only we consider _ü_
                          > > to be _yu_.
                          >
                          > [Not sure what you mean here. _yu_ is really nothing but
                          > _ü_, and y is only used in my mode as an orthographical
                          > spelling of the sound, corresponding to Pinyin "yu" (also
                          > "yuan").]

                          I meant: Transcribe all instances of Pinyin <ü> as if they were Pinyin
                          <yu> (if another vowel follows as jw, if not, as ju).

                          > > If Benct's remark that these sounds feel related to the
                          > > velar sounds for Chinese speakers is true, then it'd be
                          > > the most elegant solution to use the calmatéma as a
                          > > palatal-velar series.
                          >
                          > [Hm. But then you would have a mixed teema again with
                          > quite distinct sounds lumped together.

                          No, Benct's trick is to consider these sounds not to be distinct, but
                          to belong to one and the same consonantal series (like the bilabials
                          and the labiodentals which are neither considered to be distinct, but
                          to belong to one and the same consonantal series).

                          > > As for the fifth téma: If the problem isn't solved in
                          > > Benct's way,
                          >
                          > [I don't think it is. In his chart, the retroflex sounds
                          > are scattered over three different teemar (tinco, parma,
                          > calma).]

                          I only meant his way of considering the palatalveolars and the velars
                          to be identical.

                          > > then I'd rather mark the retroflex sounds as variants of
                          > > the alveolar sounds by adding an s-hook to the latter,
                          > > since the two dots below are required for the diphthongs
                          > > and can't be used for distinguishing another téma.
                          >
                          > [So you'd use the calmateema for velar/palatal sounds, and
                          > mark the retroflexes as variant of the dental/alveolar
                          > sounds----and then what about the quesseteema? Largely
                          > empty? That doesn't strike me as economic...

                          No, this'd be quessetéma for velars and calmatéma for palatalveolars.
                          Both largely empty (only three/four signs used), but I rate
                          consistency with the attested modes much higher than economy (see above).

                          > As for the under-dots: As I pointed out, Mandarin q, j, x
                          > (the alveolo-palatal affricates and fricative) can very
                          > well be considered to be palatalized versions of c, z, s
                          > (their dental/alveolar counterparts),

                          Sure.

                          > therefore I do still think using two underdots a good
                          > solution, especially when one uses your suggestions from
                          > the EOTM for representing _i_(j) in dipthongs (I-tehta for
                          > preceding/anna for following). Confusion with preceding
                          > "y" (anna plus underdots) is not possible.]

                          In Tolkien's English orthographic tehtar modes as I understand them,
                          preceding j is not used at all, and the only reason why following j is
                          written with anna is because the tehtar are placed on the following
                          tengwar.

                          ---------------------------
                          j. 'mach' wust
                          http://machhezan.tripod.com
                          ---------------------------
                        • Dave
                          ... whereas the Chinese mode places the vowel tehtar on ... , and there s no need to break with this in a Chinese mode). ... [So following -e would be spelled
                          Message 12 of 23 , Nov 17, 2004
                          • 0 Attachment
                            j_mach_wust wrote:

                            >
                            > Dave 'Hísilómë' wrote:
                            >
                            > > I got the impression from your suggestions that you are
                            > > not very happy anymore with all your proposed
                            > > transcriptions of diphthongs in your proposed English
                            > > orthographic tehta mode?]
                            >
                            > What made you think that? I'm very hesitating about úre, since it's
                            > not attested at all, but that's all.
                            >
                            > > But as e/o are marked with yanta and and uure in my
                            > > Chinese mode, as in the English orthographic tehta mode
                            > > (EOTM) and the the Sindarin modes, how can you propose to
                            > > also use yanta for following i(j)(see below)?]
                            >
                            > There's an important difference between the Chinese mode and the
                            > English mode: The English mode places the vowel tehtar on the
                            > following tengwar,

                            > so that yanta/úre are free to represent other
                            > sounds than j/w,

                            whereas the Chinese mode places the vowel tehtar on

                            > the preceding tengwar, so that yanta/úre are mere variants for the
                            > representations of j/w (at least they are in all attested modes

                            , and there's no need to break with this in a Chinese mode).

                            >
                            > > > As for the Pinyin diphthongs, I'd suggest to analyze
                            > > > them into preceding consonant j/w/y + vowel + following
                            > > > consonant j/w. According to
                            > > > http://en.wikipedia.org/wiki/Pinyin , this analyzis
                            > > > covers all possible diphthongs if only we consider _ao_
                            > > > to be _aw_.
                            > >
                            > > [I think this could be problematic. Mandarin (-)ao is not
                            > > the same as the _au_ in Quenya or German (and thus not
                            > > quite _aw_), even though the difference may not be a great
                            > > as between -o and -uo (see below).]
                            >
                            > The pronunciation of /aU/ isn't the same in German and in English
                            > either. Actually, the German sound is often understood to be [ao]
                            > rather than [aU], but still it's written with the letters <au> and
                            > nobody has a problem with that. There are also Chinese romanization
                            > systems that use the letters <au>. It a similar case like the English
                            > tengwar modes where the same tengwar transcription is used for both
                            > <au> and <aw>, no the English modes are even much worse, since in the
                            > Chinese modes, the transcription of Pinyin <ao> as if it were <aw>
                            > can't create any ambiguity at all, whereas in English, there might
                            > possibly be a pair of words that is distinguished only by the letters
                            > <au> vs. <aw>.
                            >
                            > To make it explicit: I suggest not to use yanta and úre for the
                            > representation of the letters <e/o>. See below.

                            [So following -e would be spelled how?]

                            >
                            > > > Consonant _j/w_ are no problem in tengwar, since they're
                            > > > well attested in many different modes: Preceding
                            > > > consonant _j/w_ is represented either with anna*/vala if
                            > > > there's no preceding consonant or with two dots below/a
                            > > > modified u-curl above a preceding consonant;
                            > >
                            > > [Hm. Were is the problem with representing preceding
                            > > j(i)/w(u) with I-tehta and U-tehta?]
                            >
                            > It's not attested at all. Of course you can choose whatever strange
                            > unattested spellings you wish, but the chance that a mode of your
                            > creation will be accepted by others will increase significantly if you
                            > base it on the attested samples by J. R. R. Tolkien, since they are
                            > the only samples that everybody accepts.
                            >
                            > > > following _j/w_ is written with yanta/úre that bear the
                            > > > preceding vowel tehta.
                            > >
                            > > [I hinted above that this creates a problem: the frequent
                            > > Mandarin combinations -ie and -ue are in my mode already
                            > > expressed by I-tehta/U-tehta plus yanta (cf. spelling of
                            > > "Michael" in Endorion Dedication), if we also spell
                            > > following -i(j) with yanta it would be quite confusing,
                            > > because combinations like -ui and -ai would incorporate
                            > > the same vowel tengwa, and -ui and -ai in particular would
                            > > become identical. Why not stick to anna for following i/j
                            > > as you also proposed in your EOTM?
                            >
                            > If we assume that the calmatéma really is the palatal series, then the
                            > reason is that I'd like to distinguish <ai> from <ya>.

                            [Not necessary if you spell <ya> with anna plus two underdots.]

                            > The óretyelle
                            > sign of the palatal series is in the vast majority of the attested
                            > modes the normal sign for the sound of j (consonantal i/anglocentric
                            > consonantal y). In the modes where the tengwar bear the preceding
                            > vowel tehtar, anna bears the preceding vowel tehta, in the modes where
                            > the tengwar bear the following vowel tehtar, anna bears the following
                            > vowel tehta (and yanta is used for the same sound when required to
                            > bear the preceding vowel tehta).
                            >
                            > > I also see no problem with vala for u-glide diphthongs
                            > > such as -u. Also, if you use uure for following -u(w),
                            > > what about following -o (for which I propose uure), which
                            > > is a distinct sound?
                            >
                            > Analogous to above.

                            [So how do you propose to spell -o (second/third part in o-glide
                            diphthongs/triphthongs then? If I understand you correctly, -o and -u
                            are simply all lumped together as -w? I would have a problem with that,
                            especially in an orthographic mode:), and particularly in -uo which is
                            _not_ <uw> or <uu>.]

                            >
                            > > And BTW, I don't think the Wikipedia article is quite
                            > > correct when it states that "'o' is written after b p m f,
                            > > "uo" everywhere else". The -o in bo, po, mo, fo is quite
                            > > distinct from the o-glide diphthong seen in syllables such
                            > > as duo, tuo or luo, therefore we cannot simply treat uo as
                            > > o,
                            >
                            > Did anybody ever suggest such a thing?

                            [Yes, the Wikipedia Chart.]

                            >
                            > > and that means that we need a distinct representation for
                            > > (-)ao and -uo.]
                            >
                            > No. The Pinyin transcription <ao> is analyzed as <aw> (as I've written
                            > in the previous mail and above), and Pinyin <uo> is analyzed as <wo>.
                            >
                            > > > The only problem is the preceding consonant _y_.
                            > >
                            > > [I don't see why it would be.
                            >
                            > I wasn't precise enough. I meant preceding consonant version of the
                            > soun [y], that is the preceding sound [H],

                            [Which sound is [H] representing?]

                            > that is preceding Pinyin
                            > letters <yu> or <ü>.

                            >
                            > > > I suggest to represent it by a combination of the signs
                            > > > for consonant _j_ and consonant _w_: either by a
                            > > > modified u-curl on anna* if there's no preceding
                            > > > consonant or by two dots below and a modified u-curl
                            > > > above a preceding consonant.
                            > >
                            > > [As I mentioned, Pinyin y occurs only initially.]
                            > >
                            > > > Actually, this procedure can also be applied for the
                            > > > transcription of _ü_ (u-umlaut), if only we consider _ü_
                            > > > to be _yu_.
                            > >
                            > > [Not sure what you mean here. _yu_ is really nothing but
                            > > _ü_, and y is only used in my mode as an orthographical
                            > > spelling of the sound, corresponding to Pinyin "yu" (also
                            > > "yuan").]
                            >
                            > I meant: Transcribe all instances of Pinyin <ü> as if they were Pinyin
                            > <yu> (if another vowel follows as jw, if not, as ju).
                            >
                            > > > If Benct's remark that these sounds feel related to the
                            > > > velar sounds for Chinese speakers is true, then it'd be
                            > > > the most elegant solution to use the calmatéma as a
                            > > > palatal-velar series.
                            > >
                            > > [Hm. But then you would have a mixed teema again with
                            > > quite distinct sounds lumped together.
                            >
                            > No, Benct's trick is to consider these sounds not to be distinct, but
                            > to belong to one and the same consonantal series (like the bilabials
                            > and the labiodentals which are neither considered to be distinct, but
                            > to belong to one and the same consonantal series).

                            [True. But this sounds as if you could arbitrarily decide to "not
                            consider distinct" any two different sets of sounds (with a different
                            point of articulation) you like, doesn't it? :)]

                            >
                            > > > As for the fifth téma: If the problem isn't solved in
                            > > > Benct's way,
                            > >
                            > > [I don't think it is. In his chart, the retroflex sounds
                            > > are scattered over three different teemar (tinco, parma,
                            > > calma).]
                            >
                            > I only meant his way of considering the palatalveolars and the velars
                            > to be identical.
                            >
                            > > > then I'd rather mark the retroflex sounds as variants of
                            > > > the alveolar sounds by adding an s-hook to the latter,
                            > > > since the two dots below are required for the diphthongs
                            > > > and can't be used for distinguishing another téma.
                            > >
                            > > [So you'd use the calmateema for velar/palatal sounds, and
                            > > mark the retroflexes as variant of the dental/alveolar
                            > > sounds----and then what about the quesseteema? Largely
                            > > empty? That doesn't strike me as economic...
                            >
                            > No, this'd be quessetéma for velars and calmatéma for palatalveolars.
                            > Both largely empty (only three/four signs used), but I rate
                            > consistency with the attested modes much higher than economy (see above).
                            >
                            > > As for the under-dots: As I pointed out, Mandarin q, j, x
                            > > (the alveolo-palatal affricates and fricative) can very
                            > > well be considered to be palatalized versions of c, z, s
                            > > (their dental/alveolar counterparts),
                            >
                            > Sure.
                            >
                            > > therefore I do still think using two underdots a good
                            > > solution, especially when one uses your suggestions from
                            > > the EOTM for representing _i_(j) in dipthongs (I-tehta for
                            > > preceding/anna for following). Confusion with preceding
                            > > "y" (anna plus underdots) is not possible.]
                            >
                            > In Tolkien's English orthographic tehtar modes as I understand them,
                            > preceding j is not used at all, and the only reason why following j is
                            > written with anna is because the tehtar are placed on the following
                            > tengwar.

                            [How about "you" spelled with anna + O-tehta on top of vala (DTS10)?]



                            [Non-text portions of this message have been removed]
                          • j_mach_wust
                            CONSONANTS t: tinco d: ando s: thúle c: extended tinco z: extended ando n: númen p: parma b: umbar f: formen m: malta w: vala (1) sh: aha ch: extended calma
                            Message 13 of 23 , Nov 17, 2004
                            • 0 Attachment
                              CONSONANTS

                              t: tinco
                              d: ando
                              s: thúle
                              c: extended tinco
                              z: extended ando
                              n: númen

                              p: parma
                              b: umbar
                              f: formen
                              m: malta
                              w: vala (1)

                              sh: aha
                              ch: extended calma
                              zh: extended anga
                              r: anna

                              k: quesse
                              g: ungwe
                              x: hwesta
                              q: extended quesse
                              j: extended ungwe
                              y: vilya (2)

                              l: lambe
                              h: hyarmen


                              NOTES TO THE CONSONANTS

                              (1): vala is only used when the <w> is the first sound of a syllable.
                              If it follows a consonant, then the <w> is represented by a left-curl
                              above that consonant. If it follows a vowel, then it's written with an
                              úre that bears that preceding vowel.

                              (2): vilya is only used when the <y> is the first sound of a syllable.
                              If it follows a consonant, then the <y> is represented by two dots
                              below that consonant. If it follows a vowel, then it's written with a
                              yanta that bears that preceding vowel.


                              SIMPLE VOWELS

                              The vowels are placed on the preceding consonant sign. If there's no
                              preceding consonant sign, then they're placed on a short carrier. For
                              exceptions see the notes to the consonants.

                              a: three dots
                              e: acute
                              i: single dot
                              o: right curl
                              u: left curl


                              COMPLEX VOWELS AND Ü

                              All complex vowels and ü are written by combinations of the following
                              kind:

                              y/w/yw + vowel + y/w

                              Like this:

                              ua = w + a
                              uo = w + o
                              uai = w + a + y
                              ao = a + w
                              iu = i + w (y + o + w?)
                              ou = o + w
                              iao = y + a + w

                              ia = y + a
                              ie = y + e
                              io = y + o
                              iao = y + a + w
                              ai = a + y
                              ei = e + y
                              ui = u + y (w + e + y?)
                              uai = w + a + y

                              ü = y + u
                              üe = yw + e
                              üa = yw + a


                              TONES

                              I suggest to use the same tone signs as in the pinyin romanizations
                              (like Dave 'Hísilómë' does). However, I suggest to place them below
                              the same sign where the vowel tehta is.

                              ---------------------------
                              j. 'mach' wust
                              http://machhezan.tripod.com
                              ---------------------------
                            • j_mach_wust
                              I m sorry, I hit the wrong key and posted before I could complete the subject. I wanted it to be Chinese tengwar mode with Tolkienian vowel representation
                              Message 14 of 23 , Nov 17, 2004
                              • 0 Attachment
                                I'm sorry, I hit the wrong key and posted before I could complete the
                                subject. I wanted it to be "Chinese tengwar mode with Tolkienian vowel
                                representation"
                              • Benct Philip Jonsson
                                ... Sure, but noone less than YR Chao claimed that many speakers are aware, or at least have a hunch about, the complementary distribution of velars and
                                Message 15 of 23 , Nov 17, 2004
                                • 0 Attachment
                                  hisilome wrote:

                                  > [I never denied this phonological development. What I do doubt is that
                                  > an awful lot of average native speakers are aware of it, and actually
                                  > consider these sounsds as "relatives", which is quite a different
                                  > matter. The original wording was "Mandarin speakers are said to feel
                                  > that _q, j_ are related to _k, g_, while _ch, zh_ are related to _c,
                                  > z_,(...)".
                                  >
                                  > What speakers of a language _feel_ may be quite different from what
                                  > they actually _know_ :) .

                                  Sure, but noone less than YR Chao claimed that many speakers are
                                  aware, or at least have a hunch about, the complementary distribution
                                  of velars and palatals. After all complementary distribution is
                                  discoverable by inspection! As for retroflexes and dentals they are
                                  known to be related by the fact that some dialects have dentals
                                  in place of retroflexes. That is allegedly why they selected the
                                  pinyin romanization as dental+h.

                                  --

                                  /BP 8^)
                                  --
                                  Benct Philip Jonsson -- melroch at melroch dot se

                                  Solitudinem faciunt pacem appellant!
                                  (Tacitus)
                                • hisilome
                                  ... [Right, that s why I pointed out that in many parts of China, even the Mandarin retroflexes are pronounced lazily, as dental/alveolars. If you don t roll
                                  Message 16 of 23 , Nov 17, 2004
                                  • 0 Attachment
                                    --- In elfscript@yahoogroups.com, Benct Philip Jonsson <melroch@m...>
                                    wrote:
                                    > hisilome wrote:
                                    >
                                    > > [I never denied this phonological development. What I do doubt is that
                                    > > an awful lot of average native speakers are aware of it, and actually
                                    > > consider these sounsds as "relatives", which is quite a different
                                    > > matter. The original wording was "Mandarin speakers are said to feel
                                    > > that _q, j_ are related to _k, g_, while _ch, zh_ are related to _c,
                                    > > z_,(...)".
                                    > >
                                    > > What speakers of a language _feel_ may be quite different from what
                                    > > they actually _know_ :) .
                                    >
                                    > Sure, but noone less than YR Chao claimed that many speakers are
                                    > aware, or at least have a hunch about, the complementary distribution
                                    > of velars and palatals. After all complementary distribution is
                                    > discoverable by inspection! As for retroflexes and dentals they are
                                    > known to be related by the fact that some dialects have dentals
                                    > in place of retroflexes. That is allegedly why they selected the
                                    > pinyin romanization as dental+h.
                                    >
                                    [Right, that's why I pointed out that in many parts of China, even the
                                    Mandarin retroflexes are pronounced lazily, as dental/alveolars. If
                                    you don't roll your tongue, <zh, ch, shi> come out quite "naturally"
                                    as <z, c, s>. As I also wrote, the dental/alveolar counterpart for
                                    retroflex <r> is not part of the Chinese sound system, so that's where
                                    this phenomenon stops.
                                    As for the implications of the complementary distribution of velars
                                    and alveolo-palatals for a Mandarin mode, I'm still not sure how (or
                                    if) this distribution should best be reflected in an orthographic mode
                                    :).]

                                    Hísilómë
                                  • hisilome
                                    Well, mellon, I bow my head to you :). This looks pretty good to me...haven t carefully studied all the details yet, but I think as far as consistency and
                                    Message 17 of 23 , Nov 17, 2004
                                    • 0 Attachment
                                      Well, mellon, I bow my head to you :). This looks pretty good to
                                      me...haven't carefully studied all the details yet, but I think as far
                                      as consistency and accurate reflection of basic Tengwar principles go,
                                      I'd say this is vastly superior to my attempt. The work of a true
                                      loremaster. Thanks!

                                      Just two questions:

                                      1. Am I right to assume that you would also use nwalme for "ng" (velar
                                      nasal), which I think better, since it is _one_ sound, or would you
                                      use nuumen + ungwe?

                                      2. How would you spell the following syllables? If I understand you
                                      correctly, they would have to look like this:

                                      zhua = extended anga with left-curl on top + A-tehta on carrier (There
                                      is a preceding consonant sign, but since that already carries the
                                      tehta for <u/w> after consonant, I guess you would have to handle it
                                      like this, with a carrier for <a>? Analogous for the example below.)


                                      tuo = tinco with left-curl on top + O-tehta on short carrier

                                      lie = lambe with two under-dots plus E-tehta on short carrier (Or
                                      would you spell lambe with two under-dots and the E-tehta on top? Here
                                      this would be an option, since the sign for <i/y> is below the
                                      tengwa. Same question for syllables in -ia, such as jia.)


                                      xue (xüe) = hwesta with two under-dots and a left-curl on top +
                                      E-tehta on a short carrier


                                      quan (qüan) = extended quesse with two under-dots and a left-curl on
                                      top + A-tehta on short carrier + nuumen

                                      (As I said in an earlier mail, and as you are obviously aware of, in
                                      syllables such as juan and quan, we are really dealing with jüan and
                                      qüan, the same for xuan, jue, xue, lue, nue, which is why in all such
                                      cases I would use your suggested spellings for <ü, üe, üa>. This
                                      opposed to the spelling of <u> as in huang, zhuang, lu, shu, tu, du, etc.

                                      Pinyin is strangely inconsistent when it come to <u> vs. <ü>: it
                                      usually only uses <ü> in two cases, the syllables <lü> and <lüe>. In
                                      <lü> it is necessary as a distinction to <lu>, which is pronounced
                                      with the <u> as in huang, zhuang etc., but there is no corresponding
                                      syllable <lue> for <lüe>, so it makes you wonder why they bothered to
                                      indicate the actual pronunciation as <ü> in this case, while they
                                      didn't do so in the other syllables quoted [juan, quan, xuan, jue,
                                      xue, nue]. Of course, none of these syllables can be pronounced with
                                      actual <u>, either.)


                                      dao (daw) = ando + A-tehta on top of uure (Could one, in analogy to
                                      your spelling of tuo, also write ando with A-tehta on top + O-tehta on
                                      short carrier? I guess you'd rather write dipthongs as one unit
                                      whereever possible, though--see below.)


                                      xiong = hwesta with two underdots + O-tehta on short carrier + nwalme
                                      (or nuumen + ungwe, but his is a minor thing)


                                      Now I do _not_ have any problems with any of these spellings, but do
                                      they not go against your principle of not ripping diphthongs apart
                                      (i.e. representing them as one unit of vowel-tehta plus
                                      "vowel-tengwa")? Of course it means that one does not have to bother
                                      about finding tengwar to represent <o, e, a> as the final part of
                                      complex vowels, since one can just use the normal tehta on a carrier.
                                      For final <i(y), u(w)>, one has yanta and uure where required.

                                      (With triphthongs there is no other choice, unless one would want to
                                      work with two tehta, one on top and one under, one tengwa, which is as
                                      far as I know not attested anywhere.)

                                      Hísilómë


                                      --- In elfscript@yahoogroups.com, "j_mach_wust" <machhezan@g...> wrote:
                                      >
                                      > CONSONANTS
                                      >
                                      > t: tinco
                                      > d: ando
                                      > s: thúle
                                      > c: extended tinco
                                      > z: extended ando
                                      > n: númen
                                      >
                                      > p: parma
                                      > b: umbar
                                      > f: formen
                                      > m: malta
                                      > w: vala (1)
                                      >
                                      > sh: aha
                                      > ch: extended calma
                                      > zh: extended anga
                                      > r: anna
                                      >
                                      > k: quesse
                                      > g: ungwe
                                      > x: hwesta
                                      > q: extended quesse
                                      > j: extended ungwe
                                      > y: vilya (2)
                                      >
                                      > l: lambe
                                      > h: hyarmen
                                      >
                                      >
                                      > NOTES TO THE CONSONANTS
                                      >
                                      > (1): vala is only used when the <w> is the first sound of a syllable.
                                      > If it follows a consonant, then the <w> is represented by a left-curl
                                      > above that consonant. If it follows a vowel, then it's written with an
                                      > úre that bears that preceding vowel.
                                      >
                                      > (2): vilya is only used when the <y> is the first sound of a syllable.
                                      > If it follows a consonant, then the <y> is represented by two dots
                                      > below that consonant. If it follows a vowel, then it's written with a
                                      > yanta that bears that preceding vowel.
                                      >
                                      >
                                      > SIMPLE VOWELS
                                      >
                                      > The vowels are placed on the preceding consonant sign. If there's no
                                      > preceding consonant sign, then they're placed on a short carrier. For
                                      > exceptions see the notes to the consonants.
                                      >
                                      > a: three dots
                                      > e: acute
                                      > i: single dot
                                      > o: right curl
                                      > u: left curl
                                      >
                                      >
                                      > COMPLEX VOWELS AND Ü
                                      >
                                      > All complex vowels and ü are written by combinations of the following
                                      > kind:
                                      >
                                      > y/w/yw + vowel + y/w
                                      >
                                      > Like this:
                                      >
                                      > ua = w + a
                                      > uo = w + o
                                      > uai = w + a + y
                                      > ao = a + w
                                      > iu = i + w (y + o + w?)
                                      > ou = o + w
                                      > iao = y + a + w
                                      >
                                      > ia = y + a
                                      > ie = y + e
                                      > io = y + o
                                      > iao = y + a + w
                                      > ai = a + y
                                      > ei = e + y
                                      > ui = u + y (w + e + y?)
                                      > uai = w + a + y
                                      >
                                      > ü = y + u
                                      > üe = yw + e
                                      > üa = yw + a
                                      >
                                      >
                                      > TONES
                                      >
                                      > I suggest to use the same tone signs as in the pinyin romanizations
                                      > (like Dave 'Hísilómë' does). However, I suggest to place them below
                                      > the same sign where the vowel tehta is.
                                      >
                                      > ---------------------------
                                      > j. 'mach' wust
                                      > http://machhezan.tripod.com
                                      > ---------------------------
                                    • j_mach_wust
                                      ... Sure, nwalme, I forgot this one, thanks! ... No, wherever a tengwa carrying a modified left-curl should carry a vowel tehta as well, it does, e.g. in DTS
                                      Message 18 of 23 , Nov 18, 2004
                                      • 0 Attachment
                                        Dave 'Hísilómë' wrote:
                                        >
                                        > 1. Am I right to assume that you would also use nwalme for "ng"
                                        > (velar nasal), which I think better, since it is _one_ sound, or
                                        > would you use nuumen + ungwe?

                                        Sure, nwalme, I forgot this one, thanks!

                                        > 2. How would you spell the following syllables? If I understand you
                                        > correctly, they would have to look like this:
                                        >
                                        > zhua = extended anga with left-curl on top + A-tehta on carrier
                                        > (There is a preceding consonant sign, but since that already carries
                                        > the tehta for <u/w> after consonant, I guess you would have to
                                        > handle it like this, with a carrier for <a>? Analogous for the
                                        > example below.)

                                        No, wherever a tengwa carrying a modified left-curl should carry a
                                        vowel tehta as well, it does, e.g. in DTS 46 or in DTS 49. If I
                                        remember correctly, there are no contrary samples.

                                        I'd say this is because the modified left-curl is a consonant tehta,
                                        not a vowel tehta, and therefore rather analogous to the bar above
                                        (for prenasality) which also cooccurs with vowel tehtar. In DTS 46,
                                        there's actually a sample of a quesse with three tehtar above: a bar,
                                        a modified left-curl and a three dots tehta.

                                        The available computer fonts aren't very suited for a combination of
                                        left-curl and vowel tehta above, but that doesn't matter.

                                        I'm leaving away the other words you weren't sure about, since I
                                        assume it's now clear how I'd intend to spell them.

                                        > (As I said in an earlier mail, and as you are obviously aware of, in
                                        > syllables such as juan and quan, we are really dealing with jüan
                                        > and qüan, the same for xuan, jue, xue, lue, nue, which is why in all
                                        > such cases I would use your suggested spellings for <ü, üe, üa>.
                                        > This opposed to the spelling of <u> as in huang, zhuang, lu, shu,
                                        > tu, du, etc.

                                        You might do this, but I think it's not necessary, since there's no
                                        ambiguity, and since it's basically an orthographic mode.

                                        > Pinyin is strangely inconsistent when it come to <u> vs. <ü>: it
                                        > usually only uses <ü> in two cases, the syllables <lü> and <lüe>. In
                                        > <lü> it is necessary as a distinction to <lu>, which is pronounced
                                        > with the <u> as in huang, zhuang etc., but there is no corresponding
                                        > syllable <lue> for <lüe>, so it makes you wonder why they bothered
                                        > to indicate the actual pronunciation as <ü> in this case, while they
                                        > didn't do so in the other syllables quoted [juan, quan, xuan, jue,
                                        > xue, nue]. Of course, none of these syllables can be pronounced with
                                        > actual <u>, either.)

                                        On http://www.wfu.edu/~moran/Cathay_Cafe/IPA_NPA_4.htm there's also
                                        <nü>, <lüan>, <nüe>, and <lün>. Actually, the pinyin spellings aren't
                                        on that page; I've made them up with the help of
                                        http://en.wikipedia.org/wiki/Pinyin .

                                        I don't care much about inconsistencies of pinyin, but I could imagine
                                        that the reason why there's <lüe>, <nüe> and not *<lue>, *<nue> is
                                        that the replacement of <ü> by <u> is only made after the
                                        alveolopalatal consonants where the pronunciation of [y] is natural
                                        ([y] can be considered a palatal [u]), whereas in all other cases,
                                        this pronunciation is not natural and therefore indicated, no matter
                                        whether there's an opposition or not.

                                        > dao (daw) = ando + A-tehta on top of uure (Could one, in analogy to
                                        > your spelling of tuo, also write ando with A-tehta on top + O-tehta
                                        > on short carrier? I guess you'd rather write dipthongs as one unit
                                        > whereever possible, though--see below.)

                                        That's not my spelling of <tuo>, see above.

                                        > Now I do _not_ have any problems with any of these spellings, but do
                                        > they not go against your principle of not ripping diphthongs apart
                                        > (i.e. representing them as one unit of vowel-tehta plus
                                        > "vowel-tengwa")?

                                        No, in all cases, there's only one vowel tehta. All diphthongs are
                                        represented in one and the same sign. The short carrier is only used
                                        where a syllable begins with a vowel (unless the vowel tehta is placed
                                        on yanta or úre, of course).

                                        > (With triphthongs there is no other choice, unless one would want to
                                        > work with two tehta, one on top and one under, one tengwa, which is
                                        > as far as I know not attested anywhere.)

                                        Also the triphthongs require only one vowel tehta, though the
                                        consonant parts of the triphthong are written in two tengwar. I'd
                                        prefer to represent triphthongs in the following way:

                                        (consonant +) y/w + diphthong

                                        That means, I'd suggest to place the vowel tehta on yanta/úre. Like
                                        this, the tehtar are distributed more evenly.

                                        ---------------------------
                                        j. 'mach' wust
                                        http://machhezan.tripod.com
                                        ---------------------------
                                      • j_mach_wust
                                        This is a corrected version of the Chinese mode I proposed yesterday. Thanks to Dave Hísilómë for the corrections! CONSONANTS t: tinco d: ando s: thúle
                                        Message 19 of 23 , Nov 18, 2004
                                        • 0 Attachment
                                          This is a corrected version of the Chinese mode I proposed yesterday.
                                          Thanks to Dave 'Hísilómë' for the corrections!

                                          CONSONANTS

                                          t: tinco
                                          d: ando
                                          s: thúle
                                          c: extended tinco
                                          z: extended ando
                                          n: númen

                                          p: parma
                                          b: umbar
                                          f: formen
                                          m: malta
                                          w: vala (1)

                                          sh: aha
                                          ch: extended calma
                                          zh: extended anga
                                          r: anna

                                          k: quesse
                                          g: ungwe
                                          x: hwesta
                                          q: extended quesse
                                          j: extended ungwe
                                          ng: nwalme
                                          y: vilya (2)

                                          l: lambe
                                          h: hyarmen


                                          NOTES TO THE CONSONANTS

                                          (1): vala is only used when the <w> is the first sound of a syllable.
                                          If it follows a consonant, then the <w> is represented by a modified
                                          left-curl above that consonant. If it follows a vowel, then it's
                                          written with an úre that bears that preceding vowel.

                                          (2): vilya is only used when the <y> is the first sound of a syllable.
                                          If it follows a consonant, then the <y> is represented by two dots
                                          below that consonant. If it follows a vowel, then it's written with a
                                          yanta that bears that preceding vowel.


                                          SIMPLE VOWELS

                                          The vowels are placed on the preceding consonant sign. If there's no
                                          preceding consonant sign, then they're placed on a short carrier. For
                                          exceptions see the notes to the consonants.

                                          a: three dots
                                          e: acute
                                          i: single dot
                                          o: right curl
                                          u: left curl


                                          COMPLEX VOWELS AND Ü

                                          All complex vowels and ü are written by combinations of the following
                                          kind:

                                          y/w/yw + vowel + y/w

                                          Like this:

                                          ua = w + a
                                          uo = w + o
                                          uai = w + a + y (2)
                                          ao = a + w
                                          iu = i + w
                                          ou = o + w
                                          iao = y + a + w (2)

                                          ia = y + a
                                          ie = y + e
                                          io = y + o
                                          iao = y + a + w (2)
                                          ai = a + y
                                          ei = e + y
                                          ui = u + y
                                          uai = w + a + y

                                          ü = y + u
                                          üe = yw + e


                                          NOTE TO THE TRIPHTHONGS

                                          (2) In triphthongs, the vowel tehta (second element) is placed above
                                          the sign that represents the third element (see also notes to the
                                          consonants).

                                          Examples:
                                          WAI: vala, yanta with three dots above
                                          KUAI: quesse with modified left curl above, yanta with three dots above


                                          TONES

                                          I suggest to use the same tone signs as in the pinyin romanizations
                                          (like Dave 'Hísilómë' does). However, I suggest to place them below
                                          the same sign where the vowel tehta is.

                                          I'd also suggest not to place them immediately below the lúvar, in
                                          order to leave some space for a possible double dot below.

                                          ---------------------------
                                          j. 'mach' wust
                                          http://machhezan.tripod.com
                                          ---------------------------
                                        • j_mach_wust
                                          Here s one more clarification of my poor English (thanks again to Dave Hísilómë ) in the section on tones. CONSONANTS t: tinco d: ando s: thúle c:
                                          Message 20 of 23 , Nov 19, 2004
                                          • 0 Attachment
                                            Here's one more clarification of my poor English (thanks again to Dave
                                            'Hísilómë') in the section on tones.

                                            CONSONANTS

                                            t: tinco
                                            d: ando
                                            s: thúle
                                            c: extended tinco
                                            z: extended ando
                                            n: númen

                                            p: parma
                                            b: umbar
                                            f: formen
                                            m: malta
                                            w: vala (1)

                                            sh: aha
                                            ch: extended calma
                                            zh: extended anga
                                            r: anna

                                            k: quesse
                                            g: ungwe
                                            x: hwesta
                                            q: extended quesse
                                            j: extended ungwe
                                            ng: nwalme
                                            y: vilya (2)

                                            l: lambe
                                            h: hyarmen


                                            NOTES TO THE CONSONANTS

                                            (1): vala is only used when the <w> is the first sound of a syllable.
                                            If it follows a consonant, then the <w> is represented by a modified
                                            left-curl above that consonant. If it follows a vowel, then it's
                                            written with an úre that bears that preceding vowel.

                                            (2): vilya is only used when the <y> is the first sound of a syllable.
                                            If it follows a consonant, then the <y> is represented by two dots
                                            below that consonant. If it follows a vowel, then it's written with a
                                            yanta that bears that preceding vowel.


                                            SIMPLE VOWELS

                                            The vowels are placed on the preceding consonant sign. If there's no
                                            preceding consonant sign, then they're placed on a short carrier. For
                                            exceptions see the notes to the consonants.

                                            a: three dots
                                            e: acute
                                            i: single dot
                                            o: right curl
                                            u: left curl


                                            COMPLEX VOWELS AND Ü

                                            All complex vowels and ü are written by combinations of the following
                                            kind:

                                            y/w/yw + vowel + y/w

                                            Like this:

                                            ua = w + a
                                            uo = w + o
                                            uai = w + a + y (3)
                                            ao = a + w
                                            iu = i + w
                                            ou = o + w
                                            iao = y + a + w (3)

                                            ia = y + a
                                            ie = y + e
                                            io = y + o
                                            iao = y + a + w (3)
                                            ai = a + y
                                            ei = e + y
                                            ui = u + y
                                            uai = w + a + y

                                            ü = y + u
                                            üe = yw + e


                                            NOTE TO THE TRIPHTHONGS

                                            (3) In triphthongs, the vowel tehta (second element) is placed above
                                            the sign that represents the third element (see also notes to the
                                            consonants).

                                            Examples:
                                            WAI: vala, yanta with three dots above
                                            KUAI: quesse with modified left curl above, yanta with three dots above


                                            TONES

                                            I suggest to use the same tone signs as in the pinyin romanizations
                                            (like Dave 'Hísilómë' does). However, I suggest to place them below
                                            the same sign where the vowel tehta is.

                                            I'd also suggest not to place them immediately below the lúvar, but
                                            lower, so that the double dots below (see note to the consonants) can
                                            be placed immediately below the lúvar.

                                            ---------------------------
                                            j. 'mach' wust
                                            http://machhezan.tripod.com
                                            ---------------------------
                                          • hisilome
                                            ... [Well, that was really just a misunderstanding then :). I also saw you removed from the complex vowels list. I agree now that when writing Mandarin
                                            Message 21 of 23 , Nov 20, 2004
                                            • 0 Attachment
                                              --- In elfscript@yahoogroups.com, "j_mach_wust" <machhezan@g...> wrote:
                                              >
                                              > Here's one more clarification in the section on tones.
                                              >
                                              [Well, that was really just a misunderstanding then :).

                                              I also saw you removed <üa> from the complex vowels list. I agree now
                                              that when writing Mandarin with this Tengwar orthographic (!) mode,
                                              one should only spell <ü> (i.e. y + u/w) when Pinyin also uses <ü>, in
                                              all other cases (incl. <ju, juan, jue, jun, qu, quan, que, qun, xu,
                                              xuan, xue, xun) where Pinyin has <u> (though pronounced <ü>), one
                                              should spell <u> in Tengwar writing as well.

                                              This would also include <yu, yuan, yue, yun> (where <u> is also
                                              pronounced <ü>), and which would thus be spelt with "vala with U-tehta
                                              on top/vala with modified left-curl and A-tehta on top + nuumen/vala
                                              with modified left-curl and E-tehta on top/vala with U-tehta on top +
                                              nuumen", respectively.
                                              I assume it is right to use the U-tehta in <yu> and <yun> instead of
                                              the modified left-curl, since the modified left-curl [i.e. spelling
                                              <u> as <w>] is only needed where <u> is the first part in complex
                                              vowel combinations.

                                              The same goes for all other syllables: <u> "alone" = U-tehta; <u> as
                                              first sound in complex vowels = modified u-curl; and <u> as final
                                              sound in complex vowels = uure).

                                              And to round it off, in accordance with the above principles, the
                                              following would be spelt:

                                              lu = lambe with U-tehta on top
                                              lü = lambe with two under-dots and U-tehta on top
                                              lüe = lambe with two under-dots and modified left-curl and E-tehta on top

                                              nu = nuumen with U-tehta on top
                                              nü = nuumen with two under-dots and U-tehta on top
                                              nüe = nuumen with two under-dots and modified left-curl and E-tehta on
                                              top.

                                              The only case where <o> as the final part of complex vowels would
                                              _not_ be spelled <w> (modified left-curl) is in the combination <io>,
                                              so that <xiong> would be spelled "hwesta with two underdots and
                                              O-tehta on top + nwalme". Analogous for <qiong> and <jiong>.

                                              I gather these are the spellings your mode suggests for these sounds,
                                              and the ones which I will use. Please let me know if I've made any
                                              mistakes in my above spellings (just checking whether I've understood
                                              the principles correctly).

                                              I think we've covered pretty much everything now. (Well, should I come
                                              across any problems later, I won't hesitate to bring it up :). Not
                                              very likely though. The Chinese sound system is not quite as complex
                                              as the English one...)]

                                              Hísilómë
                                            • j_mach_wust
                                              ... All of your samples match perfectly with the suggestion I ve outlined! ... I m sure you wouldn t, and this makes me feel safer! ... j. mach wust
                                              Message 22 of 23 , Nov 21, 2004
                                              • 0 Attachment
                                                Dave 'Hísilómë' wrote:

                                                > Please let me know if I've made any
                                                > mistakes in my above spellings (just checking whether I've understood
                                                > the principles correctly).

                                                All of your samples match perfectly with the suggestion I've outlined!

                                                > I think we've covered pretty much everything now. (Well, should I come
                                                > across any problems later, I won't hesitate to bring it up :).

                                                I'm sure you wouldn't, and this makes me feel safer!

                                                ---------------------------
                                                j. 'mach' wust
                                                http://machhezan.tripod.com
                                                ---------------------------
                                              • hisilome
                                                ... ... syllable. ... syllable. ... a ... ... following ... [I have long wondered about one little problem . When using Tengwar to write Chinese,
                                                Message 23 of 23 , Feb 17, 2006
                                                • 0 Attachment
                                                  --- In elfscript@yahoogroups.com, "j_mach_wust" <machhezan@...> wrote:
                                                  >
                                                  >
                                                  >This is a corrected version of the Chinese mode I proposed
                                                  >yesterday.


                                                  <...>


                                                  >
                                                  > NOTES TO THE CONSONANTS
                                                  >
                                                  > (1): vala is only used when the <w> is the first sound of a
                                                  syllable.
                                                  > If it follows a consonant, then the <w> is represented by a modified
                                                  > left-curl above that consonant. If it follows a vowel, then it's
                                                  > written with an úre that bears that preceding vowel.
                                                  >
                                                  > (2): vilya is only used when the <y> is the first sound of a
                                                  syllable.
                                                  > If it follows a consonant, then the <y> is represented by two dots
                                                  > below that consonant. If it follows a vowel, then it's written with
                                                  a
                                                  > yanta that bears that preceding vowel.


                                                  <...>


                                                  >
                                                  > COMPLEX VOWELS AND Ü
                                                  >
                                                  > All complex vowels and ü are written by combinations of the
                                                  following
                                                  > kind:
                                                  >
                                                  > y/w/yw + vowel + y/w
                                                  >
                                                  > Like this:
                                                  >
                                                  > ua = w + a
                                                  > uo = w + o
                                                  > uai = w + a + y (2)
                                                  > ao = a + w
                                                  > iu = i + w
                                                  > ou = o + w
                                                  > iao = y + a + w (2)
                                                  >
                                                  > ia = y + a
                                                  > ie = y + e
                                                  > io = y + o
                                                  > iao = y + a + w (2)
                                                  > ai = a + y
                                                  > ei = e + y
                                                  > ui = u + y
                                                  > uai = w + a + y
                                                  >
                                                  > ü = y + u
                                                  > üe = yw + e


                                                  [I have long wondered about one little "problem". When using Tengwar
                                                  to write Chinese, I always spell complex vowels in accordance with
                                                  your suggestions above, presuming that when you use "u" and "i" you
                                                  imply that one should use the actual U/I tehta, but when you use "w"
                                                  and "y", the modified left-curl or uure/two underdots or yanta are
                                                  employed.

                                                  This works perfectly fine in most cases, but I believe there are two
                                                  ambiguous scenarios that involve the combinations "iu" and "ui",
                                                  featuring _both_ "i/y" and "u/w".

                                                  If I read your suggestions correctly, for "iu" you suggest "i + u",
                                                  i.e. "I-tehta on top of uure". Thus, a word like "diu" would be
                                                  spelt "ando + I-tehta on top of uure". Question: Wouldn't it also
                                                  comply with your NOTES TO THE CONSONANTS ("If <y> follows a
                                                  consonant...") to spell "ando with two under-dots and U-tehta on top"?

                                                  Similarly, I gather you'd spell a word like "tui" as "tinco + U-tehta
                                                  on top of yanta". In accordance with your rules ("If <w> follows a
                                                  consonant..."), couldn't one also spell "tinco with modified left-
                                                  curl and I-tehta on top" (however one would arrange them, from left
                                                  to right or on top of each other)?

                                                  In a nutshell, how did you decide in these two special cases that the
                                                  first part of "iu" and "ui" should be represented with a tehta and
                                                  the second element with a tengwa, instead of using combinations
                                                  employing two diacritic signs? I mean, here two choices seem possible
                                                  (unlike in cases like "ai", "ia", "ei", "ua", "ao [=au =aw]" etc,
                                                  since for "a" or "e" one has no choice but to use the "normal" A-/E-
                                                  tehta, and thus naturally resorts to the more "special" spellings
                                                  for "i/y" or "u/o/w", no matter if they make up the first or second
                                                  element).

                                                  Hm, hope this question isn't too obtuse...or confusing...:)]

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