{"canonical_url":"https://engineer.company/glossary/screen-reader/","content_markdown":"# Screen reader\n\n## What it is\n\nA screen reader is a program that presents everything on a computer or phone, from the text of a\ndocument to the buttons around it and the state each one is in, as synthetic speech or as a line\non a refreshable braille display, and lets its user work that interface from the keyboard or with\ntouch gestures. The word covers a whole family of screen readers rather than one product: NVDA,\nwhich is free and open source, and JAWS on Windows; VoiceOver, built into macOS and Apple's mobile\nsystems; TalkBack on Android. Its users are mostly blind or have low vision, and some people who\ncan see the screen use one because listening is easier for them than reading.\n\nThe part most developers miss is where the words come from. A screen reader does not look at the\npixels. The browser turns the HTML into a second structure, the accessibility tree, in which every\nelement is reduced to a role (button, link, heading, list), a name (the words that identify it)\nand a state (checked, expanded, disabled). The operating system's accessibility interface hands\nthat tree to the screen reader, which speaks it. A button whose visible content is a drawing of a\npaper plane and whose code has no text is, in that tree, a button with no name, and the reader\nsays exactly that.\n\nNor do its users listen to a page from top to bottom. NVDA's browse mode turns single letters into\nnavigation keys: `H` jumps to the next heading, `D` to the next landmark, `K` to the next link, `F`\nto the next form field, and the digits 1 to 9 to a heading of that level. `NVDA+F7` opens a list of\nevery link, heading or landmark on the page. VoiceOver has the rotor, opened with `VO-U`, which\noffers the same lists. A sighted reader skims by glancing at bold type and whitespace; a screen\nreader user skims by jumping between structural elements, and a page without that structure has\nnothing to jump between.\n\n## Why it matters\n\nThe decision a team actually faces is whether the interface it ships works for someone who\nreceives it only as the markup describes it. A screen reader is the most literal consumer a page\nhas. Layout, colour, position, an icon everyone recognises, a field that turns red: every cue\nthat lives only in the visual design is lost, and only what the code says reaches the listener.\n\nThat makes it the one test of the markup that no amount of looking at the page can replace. A\ndesign review, a screenshot comparison and a careful click-through all happen in the visual\nchannel, and the visual channel is where the defects a screen reader exposes are invisible by\ndefinition. Automated rule engines help, and they catch real problems in bulk: an image with no\nalternative text, a form control with no label, an attribute with an impossible value. What they\ncheck is that a name exists, not that it is the right name. Hearing a checkout button announced\nas “button” or a menu announced as “hamburger” takes a person and a few minutes.\n\nThe cost of that person and those minutes is lower than most teams assume. NVDA costs nothing to\ninstall, and VoiceOver is already on every Mac and iPhone in the office, so the first test needs\nno purchase and no specialist. It needs someone to close their eyes, or turn the screen off, and\ntry to complete the page's main task with the keyboard and their ears. The fix for what they find\nis usually cheaper still when it happens early: a native `button` or `a` element brings its role\nand keyboard behaviour with it, while a clickable `div` needs all of that built and maintained by\nhand.\n\nThe trade-off is that screen readers differ. Each one interprets the accessibility tree its own\nway, each browser builds the tree its own way, and a page tested in one pairing can read\ndifferently in another. Testing in two common combinations covers more than testing one\nexhaustively.\n\n## What it looks like in practice\n\nOn one application the defect was in plain hearing. The interface drew its controls with system\nsymbols, and a symbol with no accessibility label is read out under its own internal name, so the\nsend button announced itself as “arrow up circle, button”.\n[Seventeen icon-only controls were renamed for what they do](/portfolio/named-all-17-icons-for-screen-readers/),\neach label naming the action rather than the shape, and following the control's state where its\nmeaning changes with it. A linter now requires an accessibility label within eight lines of any\nicon-only control, a deliberately crude distance chosen by measurement, so an eighteenth cannot be\nadded without one. The page is plain about the limit: the rule proves that a label exists nearby,\nand whether that label is correct still needs someone to listen to it.\n\nDocuments raise the same question in another format. A PDF that looks right on screen says\nnothing about whether a screen reader can read it, so at the recorded validation stage, nine generated documents, the CV variants,\nreference sheet, portfolio and cover letter across languages, were\n[run through an independent validator for the PDF/UA accessibility profile](/portfolio/reached-pdf-ua-1-conformance-at-106-of-106/)\nand passed all 106 configured accessibility-profile rules. Passing required tagged structure, a declared document language,\nalternative text on every graphic that is not decoration, a real title in the metadata and an\nexplicit reading order instead of the one the layout happens to produce. Both numbers come from\none validator, and the page says so.\n\nThis site treats its own automated audits the same way. Its\n[credits page](/credits/)\ncalls those audits the floor rather than the proof, and asks anyone who uses a screen reader on\nit, and hears something read wrongly, to say so, because that is the one report the gate has no\nway to generate.\n\n## Where it goes wrong\n\nMost screen reader failures are invisible on screen, which is why they survive review. They tend\nto take a handful of shapes.\n\n**The control announces its artwork.** An icon button with no text is named from whatever the\ncode provides: a file name, a symbol name, or nothing, which leaves a bare “button”. Give every\nicon-only control a name that describes the action, and hide the drawing itself from the tree:\n\n```html\n\u003cbutton type=\"button\" aria-label=\"Send message\"\u003e\n    \u003csvg aria-hidden=\"true\" focusable=\"false\"\u003e\u003c!-- icon --\u003e\u003c/svg\u003e\n\u003c/button\u003e\n```\n\n**The page reads as one long stream.** Headings that are styled `div` elements, and a layout\nwithout landmarks, look fine and leave `H` and `D` with nothing to land on, so the listener has to\narrow through every line to find the part they came for. Use real `h1` to `h6` elements in order,\nand put the page's regions in `header`, `nav`, `main` and `footer`.\n\n**Something changes and the listener hears silence.** An error message appears under a field, a\nbasket count goes up, a toast slides in and out. Sighted users see it; the screen reader says\nnothing, because its focus is elsewhere. Move focus to the message, or put dynamic updates in a\nregion marked with `aria-live` so the reader announces them.\n\n**Hidden content is read, or visible content is silenced.** Text pushed off-screen with CSS is\nstill in the accessibility tree and is read aloud; `aria-hidden=\"true\"` on a visible element\nremoves it for the listener while everyone else still sees it. Both are useful on purpose and\nconfusing by accident, so check what the tree contains, not only what the screen shows.\n\n**Every check passes and the page still reads badly.** A label can exist and be wrong, an alt\ntext can describe the file and not the point, and a reading order can be valid and absurd. Put a\nshort listening pass on the release checklist for any page that changed its structure.\n\n## Further reading\n\n- [Tools and Techniques](https://www.w3.org/WAI/people-use-web/tools-techniques/) — the W3C's\n  overview of how disabled people use the web, screen readers among the other assistive tools.\n- [NVDA User Guide](https://download.nvaccess.org/documentation/userGuide.html) — browse mode,\n  single-letter navigation and the elements list, from the makers of a free screen reader.\n- [Use the VoiceOver rotor on Mac](https://support.apple.com/guide/voiceover/use-the-voiceover-rotor-mchlp2719/mac)\n  — Apple's guide to jumping by heading, link and landmark.\n- [Core Accessibility API Mappings 1.2](https://www.w3.org/TR/core-aam-1.2/) — how roles, names\n  and states travel from the browser to the platform interface a screen reader reads.\n\n\u003chttps://engineer.company/glossary/screen-reader/\u003e\n","date_modified":"2026-11-12T05:34:23Z","description":"How a screen reader turns a page into speech or braille, how its users move through it by structure, and what a team has to check by ear before it ships.","formats":{"gemtext":"https://engineer.company/glossary/screen-reader/index.gmi","gopher":"https://engineer.company/glossary/screen-reader/gophermap","html":"https://engineer.company/glossary/screen-reader/","json":"https://engineer.company/glossary/screen-reader/index.json","markdown":"https://engineer.company/glossary/screen-reader/index.md","pdf":"https://engineer.company/glossary/screen-reader/index.pdf"},"kind":"page","language":"en","title":"What is a screen reader?","url":"https://engineer.company/glossary/screen-reader/","version":1}