notation of schwa

[Brian MacWhinney]

On 11/23/01 1:08 AM, "Catherine Davis" <cath at roddy.demon.co.uk> wrote:

> I am currently coding some data from a transcription of a language
> disordered child and am unsure of the best way to notate the schwa
> vowel.
> I have found in other data it has been notated as 6 at u - using the
> SAMPA transcription, or as 'uh' when not using SAMPA.  The disordered
> language I am coding is written in IPA in the transcription and as it
> is not the majority of the data and as the analysis is morphological
> I don't think it is worth using either SAMPA or IPA, but I'm not
> sure.
> Any help would be greatly appreciated.
> Thank you.
>
> Catherine Davis
> Speech and Language Therapist
> 25 Elm Grove
> London N.8 9AH
>
>

Dear Catherine,
  There are three ways to approach this problem: the easy way, the current
way, and the future way.
1.  The easy way.  Use SAMPA in CLAN.  But make sure that you create a
separate %pho line for this.  Mixing SAMPA on the main line is too
confusing.  If you aren't doing phonemic transcription throughout, I
wouldn't even worry about the schwa issue at all.  Given the fact that you
only care about morphology, I would go this route for now.
2.  The current way.  Use Hank Rogers IPA Font system on the %pho line in
CLAN, as in the Cruttenden data for example.  If you use this system, we
will eventually have to convert its IPA codes to Unicode, but we will have
to do this for several data sets anyway.
3.  The future way (not recommended except for the very brave).   Use
Unicode. If you are using Windows 2000, you can use the Arial Unicode MS
font which you can get from the CHILDES home page.  The nice thing about
Unicode is that it is pretty much guaranteed to be the font standard of the
future. SIL (Summer Institute of Linguistics) will be providing very
powerful tools for diacritic stacking and such.  However, if you only care
about notating on the phonemic level, then just putting your WordPad file
into the Arial Unicode font will work.  Currently CLAN does not support full
Unicode, but our new XCode program does.

--Brian MacWhinney