LINGUIST List 3.11

Sun 05 Jan 1992

Disc: Orthography, German Spelling Checkers

Editor for this issue: <>

Directory

1. Michael Covington, Re: 3.2 Queries: Orthography, German Spell-Checkers, Basque
2. Rick Wojcik, orthographies
3. , orthographies
4. Henry Kucera, Re: 3.2 Queries: Orthography, German Spell-Checkers, Basque

Message 1: Re: 3.2 Queries: Orthography, German Spell-Checkers, Basque

> Are there writing systems in which the pronunciations of words are
> entirely predictable from their written forms?

How about Spanish? Stress is _always_ predictable by rule, or marked
with an accent mark when not predictable.

Modern Greek, too; stress is marked with accent marks.

Regarding Mark Seidenberg's query on deep and shallow orthographies, I would
recommend one of Baudouin de Courtenay's obscure essays: "The Influence of
Language on World-View and Mood" (1929--translation in Stankiewicz's
A Baudouin de Courtenay Anthology). Despite the title, this excerpt was about
two types of orthography--"phonemography" and "morphemography". The two
categories somewhat resemble Mark's distinction between "shallow" and "deep"
orthographies. Phonemographic writing, however, seems more closely tied to
shallow phonemic representation in Baudouin's conception--e.g. Sanskrit
devanagari script, which represented the output of sandhi rules. Morph-
emographic script represented words more as isolated pronunciations--e.g. the
representation of final devoicing with voiced symbols in various European
languages. So morphemographic script represented phonemic (physiophonetic)
neutralizations for Baudouin, whereas phonemographic script disallowed
phonemic neutralizations. English orthography was singled out for its rather
extreme representation of historical pronunciations--which Baudouin mistakenly
attributed to the influence of Old English. Anyway, despite the speculative
nature of Baudouin's musings, his thoughts were among the earliest in the
linguistic literature to classify relationships between phonemes and graphemes.

This is a reply to the query on orthographies. Spanish is said to have a perfec
t unidirectional correspondence, in that given a written word there is only one
 way to pronounce it (even stress is marked). This is by and large true, leavin
g aside a few Mexican place names, including the word M'exico. However, contras
tive syllabification is not marked and here one finds some disagreement among s
peakers (although most people are completely unaware of the fact that there is
more than one possible pronunciation).There are two such cases: a/ Syllabificat
ion of stop+liquid clusters across certain boundaries (e.g. sub.ra.yar or su.br
a.yar. b/ Prevocalic high vocoids (e.g. via.je or vi.a.je).

To Clarify: In sequences of two vocoids where the first one is high, most
Spanish speakers have a contrast that is not reflected in the orthography.
Thus, for many people a word such as -dueto- is trisyllabic [du.'e.to] and
contrasts with, for instance, -duelo-, which is bisyllabic ['dwe.lo].
But in many cases there is disagreement among speakers on whether a
particular word belongs to the V.V or the Glide+V group. This is not too
surprising, since in fast speech the underlying contrast can be lost.
The second case where the orthography is not completely explicit is in
the syllabification of stop+liquid groups. Such groups are always
tautosyllabic morpheme-internally. Across prefix and compound
boundaries, on the other hand, they are usually heterosyllabic. Thus
-sublime- [su.'bli.me] and -sublunar- [sub.lu.'nar] differ in the way
the group /bl/ is syllabified. But speakers not always agree when the
morphological structure of the word is only semi-transparent. Jose I.
Hualde.

Message 4: Re: 3.2 Queries: Orthography, German Spell-Checkers, Basque

 THIS IS WITH REFRENCE TO THE INQUIRY ABOUT GERMAN SPELLING CHERECKER :
 The latest press release I have from Microsoft says that Microsoft Word for
Windows (TM) Version 2.0 and Microsoft Word for the Mac, version 5.0, have a
fully integreated International CorrectSpell(TM) which (I guess optionally)
 includes 14 languages, among them German. I don't know the details since Word
5.0 for the Mac is not out yet and I am a Mac user (but the Windows version is
 available). I can't guarentee anything except that I have seen the prototype
 of the German version some time ago and it was rather good.It even did some
 useful things with compounds. Hope that helps (with the proviso that I am
not pushing anything!). Henry Kucera