Wikidata:Property proposal/Akkadian sign values

phonetic value edit

Originally proposed at Wikidata:Property proposal/Lexemes

Done: phonetic value (P12436) (Talk and documentation)

Description	phonetic value of signs/characters
Represents	cuneiform sign (Q23017336)
Data type	String
Domain	item, cuneiform sign (Q23017336)
Example 1	𒀭 (Q87555087)phonetic valuean
Example 2	𒆠 (Q87555819)phonetic valueki
Example 3	𒆠 (Q87555819)phonetic valueqi₂
Example 4	𒆠 (Q87555819)phonetic valueke
Example 5	𒆠 (Q87555819)phonetic valueqe₂
Planned use	Assign phonetic values to signs/characters
Expected completeness	eventually complete (Q21873974)

Motivation edit

Revised proposal:

Similar to different ways of reading the same Chinese sign in different languages, such as Japanese, Chinese, Cantonese a.s.o. there are different readings/phonetic values for languages written in the cuneiform script. While Lexemes can express that a word in a language written with a particular sign has a phonetic value equivalent to a transliterated version of the forms, not every cuneiform sign is expressed as its own word and phonetic values for a single sign might change when it is encountered in a different context on a word level.

The proposal is therefore to attach phonetic values to the entity/item of the cuneiform sign (or any other character expressed in Unicode) for that purpose. This allows to easily query every phonetic value which is attested for a particular character.

For cuneiform languages we know of about 11.000 phonetic values (https://github.com/tosaja/Nuolenna/blob/master/sign_list.txt) which may be qualified by usage in particular languages, time periods or locations and which may be linked to lexemes which use the particular reading.

Original proposal:

Adding Akkadian phonetic values to cuneiform signs. Sartma (talk) 18:51, 16 March 2023 (UTC)[reply]

Discussion edit

Support --عُثمان (talk) 18:10, 24 April 2023 (UTC)[reply]
Comment is "item" really the correct domain? and can't other existing property be used like IPA transcription (P898) ? if not, is it really specific to Akkadian? Plus, "Akkadian phonetic value" only gives 8 results in Google, is it really the right name and an appropriate data? @Situxx: who may tell us more. PS: is it the same thing as the sux-Latn representation and the value for transliteration or transcription (P2440) in L:L1000845#F1? Cheers, VIGNERON (talk) 11:39, 1 May 2023 (UTC)[reply]
Hi there!

@Item: Yes I think item is the correct domain, as signs themselves are currently represented using QIDs (the IDs of the Unicode code points) and from my point of view that is fine.

There are of course numerous paleographic sign variants one could consider (which could get their own QIDs in the future), but the phonetic value of the sign usually does not change as far as I know.

We have started to add more triples to e.g. https://www.wikidata.org/wiki/Q87555676 (cuneiform sign KA) for example (what it depicts, dictionary references a.s.o.) and would continue to do that for signs.

@PropertyProposal:

My question would be: What do we do with Sumerian, Hittite, Elamite and so on? Should we not rather define a property "phonetic value" and add the languages in which they occur with a qualifier?

"The" reference list about that I know is Nuolenna: https://github.com/tosaja/Nuolenna/blob/master/sign_list.txt which contains about 11000 phonetic sign values, but to my knowledge irrespective of language.

So I would think we would want a solution that fits all languages.

"is it the same thing as the sux-Latn representation and the value for transliteration (P2440) in L:L1000845#F1?"

---> The way it is written in the proposal it seems to follow the ORACC transliteration style. However, the phonetic value is just the value of one of the syllables in this example "zi" would be one phonetic compound.

@IPATranscription:

I would not say that this is appropriate, because as far as I know the IPA is not the basis of the transliterations we use for Cuneiform.

@Transliteration:

There are many competing notations of transliterations of cuneiform texts, two of which I use for Sumerian (CDLI and ORACC formats).

For phonetic values, these differ in the main following points:

- Subscript numbers vs. diacritics: Some transliterations use diacritics for subscript 2 and 3 like in French, some do not

- Usage of sz vs. š and some other similar cases of characters

In Sumerian I currently use the CDLI Notation for the sux-latn representation and add transliterations to the forms.

As I am in contact with a group of Digital Assyriologists, we are thinking about whether this is the right transliteration, but could, if we decide differently, apply some rules to convert from CDLI to other transliteration styles for sux-latn.

@Opinion about the property proposal:

We have talked about that it would be benefitial to have the list of phonetic values when querying a cuneiform sign in Wikidata directly, so in my opinion that would be a good addition.

An alternative way would be to query for Lexemes which contain exactly the cuneiform sign that we are interested in and get is phonetic values from the transliterations. But that seems way more cumbersome than adding a property like this for all cuneiform languages. Situxx (talk) 15:03, 6 May 2023 (UTC)[reply]
@VIGNERON:, would you like to give your opinion? Regards, ZI Jony ^(Talk) 06:52, 24 January 2024 (UTC)[reply]
This is something very specific... I don't feel I know enough to really give an educated opinion. I still guess that Lexemes would be a better place for these data than Items but as I said, I don't know enough so I won't oppose. @Situxx: the proposer is gone, would you take charge of this property if it is created. Cheers, VIGNERON (talk) 07:12, 24 January 2024 (UTC)[reply]
Sure, I can do that. Should I modify the proposal accordingly then? 2001:16B8:B14E:5C00:EC1F:F050:69E2:978D 08:05, 24 January 2024 (UTC)[reply]
@Situxx:, please do needful, also login and sign. Regards, ZI Jony ^(Talk) 08:21, 24 January 2024 (UTC)[reply]
Proposal has been revised! Situxx (talk) 12:53, 24 January 2024 (UTC)[reply]
Comment I agree with Situxx that it would be best to include a more general 'phonetic value' and qualify with the language at hand. This would then be useful for not only Akkadian, but for many other languages which use glyphs and ideograms in their writing system, such as Sumerian, Hittite, Ugaritic, Eblaite, Elamite, Old Persian, etc. I would also recommend qualifying such statements with the time period(s) when known, since these phonetic values do change depending on the time period, sometimes even collapsing different signs into the same reading in the first millennium BCE. Admndrsn (talk) 15:32, 6 May 2023 (UTC)[reply]
Support, an important property for lexemes.--Arbnos (talk) 20:52, 17 January 2024 (UTC)[reply]
@Sartma, Arbnos, عُثمان, Admndrsn, VIGNERON: Done phonetic value (P12436) Regards, ZI Jony ^(Talk) 07:36, 10 February 2024 (UTC)[reply]