Where are live URIs for GOLD phonetic properties?

[W.P. McNeill]

I'm building a toy GOLD-enabled phonetic chart web app. I would like
to attach currently live GOLD URIs to some of the concepts I am using
in this app.

I am looking for GOLD URIs that correspond to phonetic features like
LABIAL, SYLLABIC, VOICED, etc. I am also looking for GOLD URIs that
correspond to individual phonemes like /d/, /a/, etc.


1. Right now I'm taking these from http://www.linguistics-ontology.org.
Is this the most up-to-date online GOLD website?

2. It looks like this site has URIs for some phonetic properties but
not all. Is that correct?

3. It looks like this site doesn't have any URIs for phonemes. Is
that correct?

4. Is the correct URI for the Labial phonetic property
http://www.linguistics-ontology.org/gold/2008/LabialProperty ?


I know for (1) Scott is doing active development work on a more
extensive private GOLD database, but for my purposes right now I want
to use one that's already on the web.

[bambooforest]

Scott has integrated properties from Linguistic Phonetic Descriptions
(Ladefoged 1997 in Handbook of Phonetic sciences, Hardcastle and
Laver).

I am suppose to integrate more of this work, but from other sources,
particularly phonetic feature theories (as I've mentioned before).
Remember, your phonetic properties (other than the articulatory
properties, arguably) are going to be defined by a particular theory.
A phoneme is a notion that will is defined by the theory.

If this is an app to input phonological inventories and create RDF
that's tied to GOLD then you're reinventing the wheel.

[bambooforest]

Bill,

I shared a couple of spreadsheets of feature theories with you
(Chomsky and Halle and Bates et al 2007; the latter being more driven
towards speech reco). There's also a document of phonetic descriptions
that would go into an OWL description, etc. They need to be curated
fonts might be off in the transfer), but they are the ones (minus a
few on another machine, include Ladefoged Handbook of Phonetics, and
Ladefoged and Maddieson) I promised you.

If anyone else needs access, drop me an email.

-Steve

[W.P. McNeill]

This app is just a toy I'm writing to figure out how to use Django.  I'm sure its functionality is all wheel reinvention, though insights I gain will inform useful new work.  All I want right now are some reasonable-looking URIs that I can use to test my user interface.

That aside, you bring up an interesting question about how theory-granular the concept of phoneme should be.  In our ontology we will have at least two kinds of entities:

1. IPA characters.  These will be the same across different phonological theories.

2. The phonemes hypothesized for a particular language according to a particular phonological theory.  This is because we might want to say that /t/ in English is a different phoneme than /t/ in Telugu, or two different linguists might put forward conflicting accounts of the Telegu phoneme /t/.

The question is do we have a third kind of element in the ontology?

3. A cross-theoretic notion of phoneme.  If we want to use GOLD vocabulary to make the claim "/t/ is different in English and Telugu" what does the URI for /t/ point to?

For my current purposes it doesn't matter because I'm just playing with UI and not trying to get the ontology correct, but it is an interesting issue.

On Fri, Sep 11, 2009 at 11:03 AM, bambooforest <bamboo...@gmail.com> wrote:

--
W.P. McNeill
http://staff.washington.edu/billmcn/index.shtml

[W.P. McNeill]

I may have asked the wrong question earlier. I'm looking for URIs
that point to RDF documents that I can process programmatically rather
than human-readable web pages. Do those currently exist online in
GOLD?

On Sep 11, 11:26 am, "W.P. McNeill" <bill...@gmail.com> wrote:
> This app is just a toy I'm writing to figure out how to use Django.  I'm
> sure its functionality is all wheel reinvention, though insights I gain will
> inform useful new work.  All I want right now are some reasonable-looking
> URIs that I can use to test my user interface.
> That aside, you bring up an interesting question about how theory-granular
> the concept of phoneme should be.  In our ontology we will have at least two
> kinds of entities:
>
> 1. IPA characters.  These will be the same across different phonological
> theories.
>
> 2. The phonemes hypothesized for a particular language according to a
> particular phonological theory.  This is because we might want to say that
> /t/ in English is a different phoneme than /t/ in Telugu, or two different
> linguists might put forward conflicting accounts of the Telegu phoneme /t/.
>
> The question is do we have a third kind of element in the ontology?
>
> 3. A cross-theoretic notion of phoneme.  If we want to use GOLD vocabulary
> to make the claim "/t/ is different in English and Telugu" what does the URI
> for /t/ point to?
>
> For my current purposes it doesn't matter because I'm just playing with UI
> and not trying to get the ontology correct, but it is an interesting issue.
>

> On Fri, Sep 11, 2009 at 11:03 AM, bambooforest <bamboofor...@gmail.com>wrote:
>
>
>
>
>
>
>
> > Scott has integrated properties from Linguistic Phonetic Descriptions
> > (Ladefoged 1997 in Handbook of Phonetic sciences, Hardcastle and
> > Laver).
>
> > I am suppose to integrate more of this work, but from other sources,
> > particularly phonetic feature theories (as I've mentioned before).
> > Remember, your phonetic properties (other than the articulatory
> > properties, arguably) are going to be defined by a particular theory.
> > A phoneme is a notion that will is defined by the theory.
>
> > If this is an app to input phonological inventories and create RDF
> > that's tied to GOLD then you're reinventing the wheel.
>

> > On Fri, Sep 11, 2009 at 10:49 AM, W.P. McNeill <bill...@gmail.com> wrote:
>
> > > I'm building a toy GOLD-enabled phonetic chart web app.  I would like
> > > to attach currently live GOLD URIs to some of the concepts I am using
> > > in this app.
>
> > > I am looking for GOLD URIs that correspond to phonetic features like
> > > LABIAL, SYLLABIC, VOICED, etc.  I am also looking for GOLD URIs that
> > > correspond to individual phonemes like /d/, /a/, etc.
>

> > > 1. Right now I'm taking these fromhttp://www.linguistics-ontology.org.

[bambooforest]

http://www.linguistics-ontology.org/gold-2008.owl

[scott farrar]

That's right. Bill, if you can't find the URIs your need, let me know. You can fake them by doing this:

GOLD namespace:

http://purl.org/linguistics/gold/

sample URI:

http://purl.org/linguistics/gold/MyConcept


Steve and Bill, let's meet as early  as possible next week for a PHOIBLE / 'Bill's django code' meeting.

Scott

[Dwight VanTuyl]

We also need to keep track of what's missing so that these concepts can be added to the next version of GOLD. If you have the time, please fill out this form:
http://linguistics-ontology.org/gold/2008/issue/edit
with any missing concepts you would like to see in the next version of GOLD.

Dwight