amd

An #accessibility question for folks who use a #screenreader, I occasionally see people using alternate Unicode letter forms to get an italic effect on servers that don’t support rich text.

How does that play with a screen reader? Are those alternate characters read properly?

In other words: Is “𝑟” read as “r” or as ‘MATHEMATICAL ITALIC SMALL R’?

Jun 20 at 1:19amWeb
Show threadNV AccessJun 20
@amd It's not ideal, and I would recommend against doing it. Although for a user, screen readers are starting to get better at working with SOME of these in some cases. NVDA has an option called "Unicode normalisation" to try to accomodate for this. Here is some more info in our user guide: https://download.nvaccess.org/releases/stable/documentation/userGuide.html#SpeechUnicodeNormalization
NVDA 2025.1.1 User Guide

Show threadamdJun 20
@NVAccess Thank you for the info!
Show threadDrew MochakJun 20
@amd @NVAccess It will depend on the screenreader/TTS combination used. Without unicode normalization, in NVDA with espeak, this reads as “letter 1 d 4 5 f,” in other words “WTF IDFK”
Show threadDrew MochakJun 20
@amd @NVAccess With the setting on it says “r” but not sure what others will do.
Show threadGeorgiana BrummellJun 20
@amd If I just read it, they both just say "r". If I go letter by letter, one says "normalised r", whatever that means. I am using NVDA and Windows.
Show threadAdrian RoselliJun 20
@amd Videos to show the effect: https://adrianroselli.com/2025/03/dont-use-fake-bold-or-italic-in-social-media.html
Show threadHamishJun 20
@amd Using voiceOver on my iPhone in the Mona app it is not reading at all :-)
Show threadAuthor-ized L.J.Jun 20
@amd I've heard those are generally bad for accessibility, but you can check by turning on screen reader settings on your device to have such a post read out. I know my brief experience using the one on my Android phone was revelatory.