Re: Ideograms

From: Stephan Stiller <>
Date: Wed, 30 Jan 2013 16:40:42 -0800

 From an earlier email:
> hanzi are perfectly adequate for the writing of a large number of
> Sinitic languages
A lot of people will disagree on that. (And Classical Chinese is an
exception just because it is directly based on Chinese characters. And –
yes – the omnipresent vestiges of Classical Chinese in the Sinitic
language – like the usage of kanji for Japanese – have created weak

I very much agree with what you wrote in your recent email. But I think
that DeFrancis' point is an ever-so-slightly different one. Most written
languages are vernacular-derived. People use written language to
represent spoken language, more or less. ...

(As a disclaimer, this isn't completely true. The scope of written
language largely overlaps with that of spoken language but isn't
identical. Prosodic matters are only partially expressed in writing. And
written language expresses, by convention, contrasts that spoken
language doesn't. And the existence of sign languages partially calls
into question the traditional view of linguists that the spoken
vernacular is (always) to be regarded as fundamental.)

... So when people produce the written dialect of a language, they –
mostly – start with something that can be spoken, which is then mapped
to a written form. Actually, let's assume that we start with some sort
of mental representation of what we want to communicate. (How this works
exactly is an unsolved linguistic mystery.) So, the mapping is something
like this:
     mental representation of L → spoken representation of L →
written representation of L

Okay, this was simplistic. Above I wrote "mostly". By written
convention, French indicates the plural of nouns even if it is not
pronounced. Japanese may indicate subtlety through the choice of kanji.
Most written languages distinguish certain homophones. Etc. So, the
semantics does matter. The true mapping is thus not
     M → S → W
but more like
     1. M → S
     2. (M, S) → W .

But what DeFrancis is saying is /I think/ that the S→W mapping (or the
S→W part of the (M, S) → W mapping) lies at the core of written
languages. Classical Chinese and such (Blissymbolics and ... anything
else?) – if these are not to be regarded as part of a different class of
language communication in the first place – aside, this S→W mapping is
how users of spoken language produce writing. Chinese characters – in
his very reasonable and imo correect opinion – unnecessarily increase
the M-component of the (M, S) → W mapping. For any written language
that is meant to be close to the spoken variety, anyway.

And I'm totally with him in so much of what he is writing. As far as
simplicity and practicality are concerned, it will be very difficult to
disagree with him.

But just now that I'm thinking about it – maybe we do need to honestly
ask ourselves whether a particular class of written representation is
really natural. There is a reason why early writing systems have no
phoneticity. On the other hand, one can argue that early written
language was merely a reading aid, for helping people remember
ritualistic uses of language. And the absence of established written
languages that go directly from M to W that have proven themselves to be
practical is evidence for the centrality of the S→W part.

Received on Wed Jan 30 2013 - 18:45:39 CST

This archive was generated by hypermail 2.2.0 : Wed Jan 30 2013 - 18:45:40 CST