Re: Language Shifts (Accessibility, metadata)

Hi all
Thanks for these useful responses. I've looked into them, but I'm unsure if they would work with what we are trying to do.

Maybe some more context could help: our books are primarily English or French with many of them having several instances of foreign languages, often with repeated phrases, or names with diacritics. We declare the primary languages but know that certain readers like Thorium don't have many available TTS languages, so even if they recognize language shifts tagged with <span xml: lang="">, these can be mispronounced if the reader does not have the language downloaded on their OS. We've put a lot of time and labour into language tagging our accessible titles and want to make sure a reader has all the information available to properly read these books.

I am searching for a way to let readers know which languages were used even if they are not the primary language. I know DAISY recommends that dc:language tags only be used for primary languages---does anyone know about an accepted standard for secondary languages other than the above <span>? But maybe this is not possible and using dc:language in a way other than it was intended would be the safest bet or we need to find another work around.

Thanks again!
Julie






________________________________
De : Ryan Levering <rrlevering@google.com>
Envoyé : Friday, July 11, 2025 12:44:23 PM
À : matt.garrish@gmail.com <matt.garrish@gmail.com>
Cc : Julie Jacques [Affiliate] <julie.jacques@affiliate.mcgill.ca>; public-schemaorg@w3.org <public-schemaorg@w3.org>
Objet : Re: Language Shifts (Accessibility, metadata)

Thanks Matt! That sounds much more informed. :)  I just went and read some of those docs for some context since I didn't understand the crossover between the different metadata syntaxes.

On Fri, Jul 11, 2025 at 9:20 AM <matt.garrish@gmail.com<mailto:matt.garrish@gmail.com>> wrote:

You can declare the language(s) of the content in the package document using dc:language tags.[1] That’s what EPUB reading systems will be looking for if they have the ability to preload TTS languages for a user. Reading systems (and vendors) will also typically expose the language information in the publication info if the user wants to manually preload languages.



[1] https://www.w3.org/TR/epub-a11y-tech-111/#publication-lang




Matt



From: Julie Jacques [Affiliate] <julie.jacques@affiliate.mcgill.ca<mailto:julie.jacques@affiliate.mcgill.ca>>
Sent: July 10, 2025 1:13 PM
To: public-schemaorg@w3.org<mailto:public-schemaorg@w3.org>
Subject: Language Shifts (Accessibility, metadata)



Hello,



I'm curious to know if anyone has thoughts about declaring which languages are used in an EPUB so that readers using TTS know which languages they would need to download on their OS or enable on their mobile devices to ensure proper pronunciation by a screen reader.



Our copyright pages are already quite long and we wonder if readers just skip over them anyways and would miss the information.



We have a 1,000-character limit in the accessibilitySummary field which is already getting quite long and would be restrictive for those titles which have over ten languages being used.



Would love to hear from anyone about thoughts on this.



Thanks



Julie Jacques (she/her)

Digital Publishing Assistant

McGill-Queen’s University Press

1010 Sherbrooke Street West, Suite 1720, Montreal, QC H3A 2R7

julie.jacques@affiliate.mcgill.ca<mailto:kathleen.c.fraser@mcgill.ca>

mqup.ca<http://mqup.ca/> | @McGillQueensUP

McGill-Queen’s University Press in Montreal is on land which long served as a site of meeting and exchange amongst Indigenous Peoples, including the Haudenosaunee and Anishinabeg nations. In Kingston it is situated on the territory of the Haudenosaunee and Anishinaabek. We acknowledge and thank the diverse Indigenous Peoples whose footsteps have marked these territories on which peoples of the world now gather.