R: usage of ontolex:phoneticRep for phonological transcriptions

Dear Gilles, Christian and Dorus,

thanks a lot for your feedback.
From my perspective, simply using ontolex:phoneticRep regardless of the phonetic/phonological nature of the transcription is the simplest thing to do, so I am glad to hear that this is the standard practice, although I agree with Gilles that it seems sub-optimal.
Perhaps it should be useful to make the definition (if not the name) of poneticRep more inclusive by adding a specification that “The phonetic representation property indicates one phonetic (or phonological) representation of the pronunciation of the form using a scheme such as the International Phonetic Alphabet (IPA).”
If someone wanted to distinguish a specifically phonologic transcription (as Dorus would need), they could then introduce a specific subproperty for that, leaving phoneticRep as the property encompassing both phonetic and phonological representation.

If we want to discuss this further in one of the next Morph meeting, I wil be there as usual :)

All the best,

Matteo

Da: Fransen, Theodorus<mailto:theodorus.fransen@universityofgalway.ie>
Inviato: giovedì 2 marzo 2023 14:45
A: Christian Chiarcos<mailto:christian.chiarcos@gmail.com>; Gilles Sérasset<mailto:Gilles.Serasset@univ-grenoble-alpes.fr>
Cc: Pellegrini Matteo (matteo.pellegrini)<mailto:matteo.pellegrini@unicatt.it>; public-ontolex<mailto:public-ontolex@w3.org>
Oggetto: Re: usage of ontolex:phoneticRep for phonological transcriptions

Non si ricevono spesso messaggi di posta elettronica da theodorus.fransen@universityofgalway.ie. Informazioni sul perché è importante<https://aka.ms/LearnAboutSenderIdentification>
Dear all,

I also think that it would be useful to distinguish phonetic vs phonological at the least, for ex. by /.../ and [...]. I will start a project in October that will probably use this ontology for Old Irish. In fact, I plan to use more than one system of phonological representation. For morphological generation using an FST to build an inflected lexicon (with involvement of the mentioned Sacha Beniamine of Surrey), I'm relying on a collaborator's abstract system of phonological representation, from which we also like to produce a more commonly used system that is closer to phonetic representation for Old Irish. I must add that I'm not too familiar yet with the possibilities and ins and outs of OntoLex at this stage.

Perhaps this could be discussed in one of the Morph meetings?

Best,
Dorus


---


Theodorus Fransen
(#startsWithAName (Theo)dorus (rhymes with "chorus") Fransen (sounds like "man-sen") /(te:ɔː)dɔːrəs frɑnsən/)

Taighdeoir iardhochtúireachta | Postdoctoral researcher

[https://docs.google.com/uc?export=download&id=1NBycTSuJLfYD6ZYwuIslWkEgfOsAz8dW&revid=0B1fh7RXzcjE8MkJpVXlFbDdJK25wZlgwb01SaGZ4dCtuWmN3PQ]<https://eur03.safelinks.protection.outlook.com/?url=http%3A%2F%2Fwww.cardamom-project.org%2F&data=05%7C01%7Cmatteo.pellegrini%40unicatt.it%7C4bd54ca7e0a44072de7708db1b246856%7Cb94f7d7481ff44a9b5886682acc85779%7C0%7C0%7C638133615410008515%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=21Ed8m05Mj9B8EAjBJYuW%2BsGoo7U72%2B0KqRXixoEnJg%3D&reserved=0>
Tionscadal Cardamom | Cardamom project

Ionad Taighde Insight SFI um Anailísíocht Sonraí | Insight SFI Research Centre for Data Analytics
Institiúid na hEolaíochta Sonraí | Data Science Institute
Ollscoil na Gaillimhe (Éire) | University of Galway (Ireland)


From: Christian Chiarcos <christian.chiarcos@gmail.com>
Sent: Thursday 2 March 2023 04:52
To: Gilles Sérasset <Gilles.Serasset@univ-grenoble-alpes.fr>
Cc: Pellegrini Matteo (matteo.pellegrini) <matteo.pellegrini@unicatt.it>; public-ontolex <public-ontolex@w3.org>
Subject: Re: usage of ontolex:phoneticRep for phonological transcriptions

EXTERNAL EMAIL: This email originated outside the University of Galway. Do not open attachments or click on links unless you believe the content is safe.
RÍOMHPHOST SEACHTRACH: Níor tháinig an ríomhphost seo ó Ollscoil na Gaillimhe. Ná hoscail ceangaltáin agus ná cliceáil ar naisc mura gcreideann tú go bhfuil an t-ábhar sábháilte.

Hi,

Gilles' approach would have been my suggestion as well. I guess "phonetic" here comes from the use of IPA/Sampa for both purposes, and it's the International PHONETIC Alphabet. The name comes from the late 19th c., and also in literature of the time, phonology and phonetics were not clearly distinguished, yet. (So you have language descriptions where allophons are treated like full phonemes. This is because minimal pair tests were introduced only at a later stage, I guess by Kenneth Pike in the 195os or so.)

Best,
Christian

Gilles Sérasset <Gilles.Serasset@univ-grenoble-alpes.fr<mailto:Gilles.Serasset@univ-grenoble-alpes.fr>> schrieb am Do., 2. März 2023, 00:05:
Hi Matteo,

For DBnary, I am using ontolex:phoneticRep for both, and values are surrounds by /…/ or […] depending on the data I have to model.

I think this notation is quite standard.

It’s not really elegant though…

Regards,

Gilles,


On 1 Mar 2023, at 12:26, Pellegrini Matteo (matteo.pellegrini) <matteo.pellegrini@unicatt.it<mailto:matteo.pellegrini@unicatt.it>> wrote:

Dear all,

I write this email to ask for your advice on the possibility of using ontolex:phoneticRep for phonological – rather than strictly phonetic – transcriptions of forms.
Indeed, the name and definition of ontolex:phoneticRep suggest that it should be used for phonetic trancriptions. But what if a resource provides a phonological, rather than phonetic, transcription of wordforms? Should we take ontolex:phoneticRep to be intended to be used for any kind of sound representation, and thus also for such phonological transcriptions?

I am asking this because I am working at an ontology to be used for the conversion to ontolex-compliant RDF lexicons of paradigmatic lexicons released in a standard format that is being developed right now (Paralex, a project led by Sacha Beniamine at the Surrey Morphology Group).
In that standard, the idea is to have generic "sound" transcriptions, leaving up to data creators the choice whether the transcription will be phonetic or phonological: so there will be a generic column "phon_form" that can be used for both.
When converting to an ontolex-compliant lexicon, intuitively I think it would be reasonable to consider these transcriptions as ontolex:phoneticRep of the wordforms of the resource(s). However, it is possible that the transcription is phonological, rather than phonetic, strictly speaking. Would it be a problem to treat such phonological transcriptions as ontolex:phoneticRep? If so, what would your advice be on what to do?

Thanks in advance,

Matteo Pellegrini



[http://static.unicatt.it/ext-portale/5xmille%20firma%20mail%202022.jpg]