LINGUIST List 3.2

Thu 02 Jan, 1992

Qs: Orthography, German Spell-checkers, Basque

Editor for this issue: <>

Directory

1. Mark Seidenberg, orthographies
2. Pamela W. Jordan, Info. on German spelling checkers
3. , films for elicitation

I have a question for readers of the list about orthographies.

There has been a large amount of psycholinguistic research on how
differences among orthographies affect reading. Most research has
focused on the extent to which orthographies encode phonological information.
Thus, in "shallow" orthographies, the correspondences between written
and spoken codes for words are supposed to be simple and direct. In "deep"
orthographies, the correspondences are more inconsistent, irregular,
obscure. Typically people have assessed "depth" in terms of characteristics
of the mapping between graphemes and phonemes, though of course there are
other ways of thinking about what orthographies actually encode.

The Roman and Cyrillic writing systems for Serbo-Croatian are said to be
"shallow," because within each of these alphabets the correspondences
between graphemes and phonemes are entirely regular (I am repeating here
what has been asserted by Lukatela, Katz, Feldman, and
others). English is said to be "deep" because of the existence of
minimal pairs such as HAVE/GAVE, SAID/PAID, and WERE/HERE. Hebrew is even
"deeper" because vowels are typically omitted.

I have often wondered about this characterization of orthographic depth.
For example, Serbo-Croatian is said to be shallow because of regularities
at the level of graphemes and phonemes; however, generating the pronunciation
of a word from print also involves assigning stress, and there I gather
that Serbo-Croatian is quite complex. I also wonder how well this
notion of orthographic depth applies to other writing systems.

So, my inquiry to the list is whether people might be willing to share their
knowledge of other writing systems. Are there writing systems in which
the pronunciations of words are entirely predictable from their written
forms? One test for this would be if every reader of the language agreed
upon the pronunciation of novel forms (e.g., MAVE or GABINAL in English).

I would also like to test a specific hypothesis. My prediction is that in
writing systems that admit exceptional spelling-sound correspondences,
the exceptions tend to cluster among the higher frequency
words in the language. Thus, in English, the irregular words are
overrepresented among the HF items (HAVE, GIVE, SAID, WERE, WAS, DOES,
NONE, GONE, etc.). Is this true in other languages?

Thanks for your help.

Mark Seidenberg

Could you give me some recommendations on German spelling checkers?
I'm interested in ones for either PCs, Macs or Suns. I know about one
that Alki Software has for $69.95 for Word 4.0 on a Mac.

Also, do you know of any German-English terminology banks for translators?

Thanks,
Pam Jordan

I am planning to do some controlled elicitation of Basque and i
was thinking about using the well known Pear Stories film (cf.
Chafe 1980) for that purpose (it's easily available for $27.50).
One advantage of using that film is that there should be a lot of
material in different languages to compare with (I am looking at
word order).

My question is: Is there anything comparable to the Pear film
out there which might be a good complement to it, or perhaps even
a better subtitute? Any information on this topic will be
greatly appreciated.

Jon Aske

jaskebat.bates.edu
jonaskegarnet.berkeley.edu