Screen reader

Also called assistive technology, AT, NVDA, VoiceOver, JAWS

This entry has not been translated yet and is shown in English.

Software that renders an interface as speech or braille from the accessibility tree — the platform's parsed model of roles, names, and states, not the visual layout. One member of a wider assistive-technology set that includes magnifiers, switch access, and voice control, each failing differently.

A two-column comparison. Left, the accessibility tree: roles, names, states, and structure parsed from the markup, in DOM order — this is what a screen reader renders as speech or braille. Right, the visual layout: position, size, colour, and the reading order the eye follows, produced by CSS and invisible to a screen reader. Reordering with CSS changes only the right column.

In practice

Testing one reader in one browser proves little: NVDA with Firefox and VoiceOver with Safari disagree constantly, and neither tells you whether anyone could finish the task.

Not to be confused with

Semantic markupUsing elements for what they mean rather than how they look — a button as `button`, a heading as `h2` — so role, state, and keyboard behaviour arrive from the platform rather than being rebuilt. It is what assistive technology reads. Class names mean nothing to anything except a stylesheet.

How this term connects

Related Confused with

In practice

Not to be confused with

Related

How this term connects

Referenced by