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

2111Re: Dating Chinese writing Re: [ANE-2] Cherubim Origins

Expand Messages
  • Robert M Whiting
    Aug 7, 2006
    • 0 Attachment
      On Sun, 6 Aug 2006, Peter T. Daniels wrote:

      > ----- Original Message ----
      > From: Robert M Whiting <whiting@...>
      > To: ANE-2@yahoogroups.com
      > Sent: Sunday, August 6, 2006 6:06:49 AM
      > Subject: Re: Dating Chinese writing Re: [ANE-2] Cherubim Origins
      >
      >
      > On Sat, 5 Aug 2006, E Bruce Brooks wrote:
      >
      > > BRUCE: The typical Chinese character is made up of two elements, one
      > > conventionally called the "radical" in the belief that it represents
      > > the root meaning, and the other conventionally called the "phonetic."
      > > Best current thinking is that the "radical" is what at least some Near
      > > Easternists call the "determinative" or semantic-area designator, and
      > > the "phonetic" carries not only the sound, but [as in all languages]
      > > the meaning associated with the sound. It would be nice to know
      > > exactly what NE script is closest to this system, given that the
      > > "phonetic" part is pictorial rather than alphabetic. Peter implies
      > > Sumerian. Does that find general agreement on this list?
      >
      > BOB: A certain amount of simplification here, since if I remember
      > DeFrancis correctly (not having it here in front of me), this covers
      > only one category (although perhaps the largest and hence the "typical"
      > group) of Chinese signs.
      >
      > As I remember it, all the sign types identified by DeFrancis are found in
      > the Sumerian script, if not necessarily in the same proportions.
      > ********
      > PTD: Thank you for this exposition. (How many times do I have to ask you
      > to write down everything you know about cuneiform in a single monograph?
      > You know more about certain things than anyone else alive. Or are we
      > supposed to gather your Scripta Minora after you're gone, to piece an
      > account together?)

      What I know is dwarfed by what I don't know. Besides, I see no point in
      writing down things that should be obvious to anyone who has studied the
      script in contrast to reading one or more books about the script.

      > You don't need to refer to DeFrancis for the six categories -- they're
      > very traditional and can be found in just about any account of Chinese
      > writing, e.g. WWS p. 197.

      Okay, using the categories on p. 197 of WWS, we have the following
      (Chinese script and transcription omitted):

      1) 'indicating the matter': These are indexical signs that suggest
      the "semantic" through some association (e.g., Sum. DUMU, A, etc.).

      2) 'representing the form': These are iconic representations (e.g.,
      SAG, UTU, etc.).

      3) 'forming the sound': These are compounds of the KA + /me/ = EME
      type.

      4) 'conjoining the sense': These are compounds of the KAxA = NAG2
      ('mouth' + 'water = drink) type (what I call "word pictures").
      According to WWS, in Chinese, "In origin actual characters are never
      formed this way; this is an artificial, retrospective category." In
      Sumerian, however, signs were often formed this way, at least until
      the signary was closed.

      5) 'redirected characters': A usage category, not a category of
      character formation typology; Boltz seems unsure what this category
      reflects, and I confess that I'm not entirely sure either. In any
      case, I can't think of a Sumerian parallel. I would think that the
      parallel should be the use of a sign for associated words in other
      semantic domains (e.g., the use of the sign for 'sun' to write the
      words for 'day', 'time', 'bright', 'shining', 'hot', 'dry', etc., but
      this is not clear from what Boltz says.

      6) 'borrowed characters': Another usage cagegoy. This is rebus
      writing, where a sign is used to write a word homophonous with the
      word the sign usually represents (e.g., [famously] Sum. TI 'arrow'
      --> TI 'life').

      So according to Boltz, Chinese did not create signs of the "semantic" +
      "semantic" type but only of the unitary "semantic" and "semantic" +
      "phonetic" types. Since he does not mention it, presumaby it did not
      create signs of the "phonetic" + "phonetic" or "blank" + "phonetic"
      either.

      Sumerian, on the other hand created signs of all these types (before the
      signary was closed, of course).

      > ********
      > <...>
      > BOB: Finally, there are Sumerian signs that are more or less completely
      > phonetic. Such signs usually involve a "bearer" sign (which you might
      > consider as corresponding to a "radical") which does not enter into the
      > ********
      > PTD: They're not really parallel to anything in Chinese, where both
      > components always contribute something to the reading of the character,

      Sumerian even created polysyllabic combinations by ligaturing two
      individual signs, interestingly, always, to my knowledge, in reverse
      order: ZU+AB --> ABZU; GAL+LU2 --> LUGAL; GAL+U$UM --> U$UMGAL. These,
      however, are quite rare.

      > and there don't seem to be any phonetic components that never at all
      > function as a separate character.

      The same is true of Sumerian. This would seem to be logical since the
      phonetic value of a sign is only determined by its use as a character.
      Sumerian does have some values of some signs that are only phonetic (i.e.,
      there is no logographically represented word with that value represented
      by that sign. Sumerian referred to such values as KA.KA-si-ga.

      > ********
      > BOB: equation but merely provides a platform for the "phonetic" which provides
      > the entire determination of the semantic content of the sign (although
      > sometimes the "bearer" contributes part of the phonetic content). Common
      > bearer signs typically have open space where the "phonetic" can be
      > inscribed. Signs like KA, GA2, EZEN, URU and LAGAB are frequently
      > "bearers". An example is GAZI (= GA2xPA) or GAZI2 (= GA2xSUM) (where PA =
      > si29 and SUM = si3). Sumerian gazi corresponds to Akkadian kasû and the
      > word is probably a loan in both Sumerian and Akkadian since it refers to a
      > kind of condiment or spice. On the other hand, BAD3 ('wall') is simply
      > EZENxBAD and the phonetic value is provided solely by the inscribed BAD
      > sign and the EZEN part contributes nothing except to carry the BAD sign.
      >
      > As far as I know (as I say, I don't have DeFrancis here), Chinese signs
      > corresponding to all these types are known, but the fact that the Sumerian
      > signs with explicit phonetic content are much rarer than the Chinese may
      > be explained by another factor, which brings me to the next point:
      >
      > On Sat, 5 Aug 2006, Peter T. Daniels wrote:
      >
      > ********
      > The _only_ typological difference between Chinese and Sumerian is that
      > the combination of radical and phonetic was fused into a single item in
      > Chinese -- the characters -- while in Sumerian, they remained separate
      > (thus the thousands of Chinese characters vs. the several hundred in
      > cuneiform).
      > ********
      >
      > BOB: This, of course, is simply false. As can be from the discussion
      > above, or from a knowledge of the Sumerian script, there are numerous
      > examples in
      > ********
      > PTD: You have an odd definition of "simply false"! It's one of your
      > favorite phrases ...

      Actually, it is not one of one of my favorite phrases. It is just, sadly,
      often applicable. I am much fonder of "poppycock and balderdash", but I
      only use that for arrant nonsense. What you have said is not arrant
      nonsense. With some modification it could be made into a valid statement.
      But as it is stated, it is simply false. There are a large number of
      fused signs in Sumerian. There are just not any from after the signary
      was closed. So it's not a typological difference, it is a procedural
      difference. What caused the limited number of signs in Sumerian cuneiform
      is not the typology of sign creation, but the fact that the signary was
      closed forcing new combinations to be written as separate signs (generally
      known as diri-type logograms).

      > but it rules out productive generalization.

      I don't consider a generalization that is false over much of its range to
      be productive.

      > Define "numerous."

      Possibly as many as several hundred. To get an idea, look at a realy good
      sign list (I recomment Deimel, $L part I; Labat leaves too much out and
      Borger lacks the paleography). Look at the common "bearer" signs (KA,
      URU, EZEN, LAGAB, GA2 [PISAN], NINDA2 etc.) and see how many combinations
      there are. In a quick check today I counted over 40 combinations with KA
      and over 60 with GA2 as "bearer". All of these signs are not of equal
      evidentiary value, of course; some are only attested in archaic texts and
      their meanings and readings are unknown; some are attested from late sign
      lists and may reflect "scholarly reinterpretation over the millennnia.
      But there are sufficient transparent examples of "semantic" + "semantic",
      "semantic" + "phonetic" and "blank/bearer" + "phonetic" combinations to
      show that the technique was productive before the signary was closed.

      > There is also a small number of Chinese characters that are
      > semantic-only, and an even smaller number that really are
      > semantic+semantic (despite folk etymologies that have obscured their
      > semantic+phonetic background, see e.g. David Prager Branner in the new
      > ed. of _Encyclopedia of Lang & Ling_ [ed. Brown, Elsevier 2006]).

      That would be interesting since Boltz seems to be of the opinion (as
      described above under 4) that semantic+semantic was not used for character
      creation but is a "learned" category of reflective analysis.

      > That doesn't nullify the fact that the vast majority of Chinese
      > characters are semantic+phonetic.

      I don't doubt it. And if Nisaba or whoever was in charge of the Sumerian
      script hadn't put a stop to it, I don't doubt that a large part of the
      Sumerian signary would be too. I have always maintained that it is
      extremely fortunate that knowledge of the Chinese script was never lost,
      because the only way to decipher a script is to be able to break into the
      phonetic code and the phonetic part of the Chinese characters is so well
      hidden that it would have been a real bear to decipher.

      > ********
      > BOB: the Sumerian script where "radical" and "phonetic" were fused into a
      > single sign. The factor that accounts for the difference is the fact that
      > sometime fairly early on the Sumerian signary was closed. This meant that
      > after this point no new signs could be added to the inventory and blocked
      > the fusion of existing signs into a new sign. In logography, new signs are
      > needed for new words. These may be new coinings, but most commonly,
      > foreign names and loan words that must be expressed in writing. Since no
      > new signs can be added to the signary, such words must be written
      > syllabically (phonetically) or with "word pictures". The former really
      > needs no example since it is found passim. The latter is illustrated with
      > the writing of the Sumerian word for 'horse'. The horse came into
      > Mesopotamia long after the cuneiform signary was closed. The Sumerian
      > word for 'horse' is sisi (or zizi, since Sumerian did not have a
      > voiced/voiceless distinction it's hard to tell). Since no new sign could
      > be created for it, it was written either syllabically (AN$E.zi.zi or
      > simply zi.zi) or with a "word picture" (AN$E.KUR or AN$E.KUR.RA; "equid of
      > the mountains" or "foreign equid). As the examples of NITAxKUR, SAL+KUR,
      > and GUDxKUR given above show, had the word for 'horse' come into Sumerian
      > before the signary was closed, the most likely development would have been
      > *AN$ExKUR --> SISI 'horse'.
      >
      > So the thing that sent the development of Sumerian script on a divergent
      > path from the development of Chinese script was the closing of the
      > Sumerian signary which prevented the fusion of existing signs into new
      > ones. Sumerians could coin new words, but not new signs.
      >
      > Peter describes the situation in Sumerian after the closing of the
      > signary, but not before. If Chinese writing developed by "stimulus" from
      > Sumerian, then the closing of the Sumerian signary, whenever that may have
      > been precisely, would seem to be a likely terminus ante quem for the
      > stimulus.
      > ********
      > PTD: If you can't suggest when this closing happened, the suggestion
      > that it occurred isn't of much help for any such "stimulus." But surely
      > this closing happened before the mid 3rd mill.?

      I would say so; probably somewhat earlier than that but not more than a
      century or two.

      > And please don't try to claim that Gelb advocated for this "stimulus" --
      > his argument for outside influence on Shang China comes down to nothing
      > more than the view held in the 1930s (when most of Gelb 1952 was
      > written) that Shang China was "characterized by so many foreign
      > innovations that many scholars regard it as a ready made imported
      > civilization" (219); and what he very tentatively suggests -- not in the
      > text, but only in the chart (1952 endpapers, 1963 front matter) -- is
      > Proto-Sumerian to Proto-Elamite to "Proto-Indic" (= Indus Valley) to
      > Chinese (involving two unknowns, so any typological similarity between
      > PSum and PChin isn't relevant).

      Gelb just started with Sumerian and started walking East. The only real
      argument for the sequence is the correlation between the time and distance
      displacement of the appearance of the script as one moves eastward. The
      real problem is lack of evidence because Proto-Elamite and "Proto-Indic"
      are undeciphered, the sterotypical "Proto-Indic" inscriptions are problaby
      just the tip of an iceberg that was written on perishable materials; and
      the earliest Chinese script shows a stage that already reflects what is
      likely to be several centuries of development of which we have no record.

      Bob Whiting
      whiting@...
    • Show all 27 messages in this topic