Glossary — Tōdai

68 terms

A

Achromatopsia: Visual deficiency that reduces or eliminates color perception. In its complete form, vision is entirely in grayscale. In its incomplete form, the person retains some residual color vision, but it is reduced and varies from one individual to another depending on which cones still function. The WCAG contrast formula, calculated for typical three-cone vision, does not describe the reality of these atypical perceptions. Source: Orphanet — Achromatopsia.
ADA: United States federal civil rights law that prohibits discrimination based on disability. Title II (public sector) explicitly requires WCAG 2.1 AA compliance for government websites since the 2024 DOJ rule. Title III (private sector) has no formal web accessibility rule yet, but courts increasingly apply the same WCAG 2.1 AA standard. Source: ADA.gov.
Alt text: Textual description associated with an image via the HTML alt attribute. Enables screen readers to convey the image content, search engines (SEO) and LLMs (GEO) to index it, and users to understand the image when it doesn't load. An alt="" (empty) indicates a decorative image. A missing alt is an error — the screen reader announces the filename. Source: MDN — alt.
Anterograde amnesia: A memory disorder that prevents forming new memories after the triggering event. The person forgets what they just did, read or navigated. On the web, every page visited without a navigation landmark (breadcrumb, headings, landmarks) becomes a dead-end corridor: the user can't retain the path they just took. Source: Inserm — Memory.
AODA: Ontario's accessibility law, in effect since 2005. It requires Ontario organizations to comply with web accessibility standards, including WCAG 2.0 AA. It was one of the world's first laws to impose digital accessibility standards on the private sector. Source: Ontario — AODA.
AOM: An API under standardization that allows JavaScript to interact directly with the browser's accessibility tree, bypassing ARIA attributes in the DOM. AOM aims to address ARIA's limitations for complex dynamic components (canvas, web components). Still experimental in most browsers. Source: WICG — AOM Explainer.
ARCOM: French independent regulatory authority formed from the merger of CSA and Hadopi in 2022. In digital accessibility, ARCOM has authority to sanction public sector failures to comply with RGAA obligations. Source: ARCOM.
ARIA: Set of W3C specifications that define additional HTML attributes to improve accessibility of dynamic web components. ARIA attributes (role, aria-label, aria-hidden, aria-expanded, etc.) convey information to assistive technologies that semantic HTML alone cannot express. Golden rule: don't use ARIA when a native HTML element does the job. Source: W3C — WAI-ARIA.
aria-hidden: ARIA attribute that hides an element and its descendants from assistive technologies (screen readers, braille displays) while keeping it visible on screen. aria-hidden="true" removes the element from the accessibility tree. Not to be confused with display: none (hides visually AND from the tree) or alt="" (specific to decorative images). Source: MDN — aria-hidden.
Assistive technology: Any software or hardware that helps a person with a disability use a computer or the web. Screen readers (VoiceOver, NVDA, JAWS) are the most well-known, but assistive technologies also include braille displays, switch controls, voice recognition software (Dragon), screen magnifiers, and alternative keyboards. Accessibility standards (WCAG) aim to guarantee website compatibility with these technologies.
Autocomplete: Browser feature that pre-fills form fields from data the user has previously entered (name, email, address, credit card). Enabled by the HTML autocomplete attribute with standardized values (name, email, tel, street-address, etc.). Reduces input time and errors — particularly critical for users with motor difficulties. Source: MDN — autocomplete.
axe DevTools: Browser extension developed by Deque Systems for auditing page accessibility directly in developer tools. It uses the axe-core engine and offers more rules than Lighthouse. The free version covers automatable checks; the paid version adds guided tests. Source: Deque — axe DevTools.
axe-core: Open source accessibility testing engine developed by Deque Systems. axe-core analyzes the DOM of a page and detects automatable WCAG violations — insufficient contrast, images without alt text, forms without labels, etc. It powers Lighthouse, the axe DevTools extension, and many CI tools. It covers roughly 30–40% of WCAG criteria — the rest requires manual auditing. Source: GitHub — axe-core.

C

Cloaking: Abusive SEO technique that consists of presenting different content to search engines and users. Google considers cloaking a violation of its guidelines and may penalize sites that use it. Filling an alt with keywords on a decorative image comes dangerously close. Source: Google — Spam Policies.
CLS: Web performance metric that measures visual stability by quantifying unexpected layout shifts during page load. One of Google's three Core Web Vitals. An image without width and height attributes causes CLS: the browser can't reserve space before loading, and content jumps when the image appears. Source: web.dev — CLS.
CSS: Style sheet language that describes the visual presentation of an HTML document — colors, spacing, typography, layout. CSS does not change document semantics: a <div> styled like a button is still a <div> for screen readers. Source: MDN — CSS.

D

Design: The conception of a product, interface, or experience. In web context, design covers visual aspects (UI), user experience (UX), information architecture, and interaction. A design problem isn't an aesthetic problem — it's a communication problem between the interface and its user.
Deuteranopia: Form of color blindness that reduces or eliminates green perception. Affects about 5% of men and 0.4% of women. Along with protanopia, it's the most common form of color blindness. Source: Colour Blind Awareness.
DOM: Tree structure that represents an HTML document in memory. The browser builds the DOM from the HTML source, and it's this tree — not the CSS display — that determines keyboard navigation order. Source: MDN — DOM.

E

EAA: European directive on the accessibility of digital products and services, in effect since June 2025. It mandates accessibility compliance for digital products marketed in the European Union. In the US, the ADA serves a comparable role. Source: directive (EU) 2019/882.

F

Focus: In web navigation, focus designates the element currently selected to receive keyboard interactions. When you press Tab, focus moves from one interactive element to the next. A visual indicator (usually an outline) lets the user know which element has focus. See also: outline, skip link.
Focus trap: Technique that confines keyboard focus inside a component — typically a modal. Unlike a keyboard trap, a focus trap is intentional: the user can exit (via Escape) and the rest of the page is disabled. The HTML <dialog> element combined with the inert attribute implements this behavior natively.
Forced-colors: CSS mode automatically activated when the user enables a high contrast theme. Detectable via the media query @media (forced-colors: active). In this mode, the browser replaces the site's colors with system colors chosen by the user (ButtonText, Canvas, Highlight, etc.). Source: MDN — forced-colors.
FTC: United States federal agency responsible for consumer protection and competition regulation. In January 2025, the FTC fined accessiBe $1 million for deceptive advertising regarding its claims of automatic accessibility. Source: FTC.gov.

G

GDP: Economic indicator that measures the total value of goods and services produced in a country over a given period. Used in articles to contextualize the economic impact of disability: according to the WHO, people with disabilities represent 16% of the global population, a market whose purchasing power exceeds the GDP of several major economies.
GEO: Content optimization for AI-powered answer engines (LLMs such as ChatGPT, Claude, Gemini, Perplexity). Unlike SEO which targets traditional search engines, GEO aims for content to be cited, summarized, and correctly attributed by generative AI. Structured content (semantic headings, descriptive alt text, structured data) benefits both SEO and GEO.
GPS: Satellite positioning system that determines geographic location. On the web, the term is used by analogy: a breadcrumb, a landmark, or a heading structure functions like a GPS — it tells the user where they are in the site hierarchy, not the path they took to get there.

H

High contrast: Display mode offered by operating systems (Windows, macOS, Linux) that replaces website color palettes with a high-contrast theme chosen by the user. Designed for people with severe visual impairment, photosensitivity, or chronic eye fatigue. See also: forced-colors.
HTML: Markup language that structures web page content. Each HTML element (<p>, <nav>, <button>, <table>...) carries semantics that browsers, screen readers, and search engines rely on. Using the right HTML element is the first lever for accessibility. Source: MDN — HTML.

I

Inert: Global HTML attribute that makes an element and all its descendants inert: not focusable, not clickable, and invisible to screen readers. Used to disable content behind an open modal, replacing multiple aria-hidden="true" and tabindex="-1" declarations. Supported in all modern browsers. Source: MDN — inert.

J

JavaScript: Programming language executed by the browser, used to make web pages interactive. If JavaScript fails or is slow to load, components that depend on it exclusively become unusable. Progressive enhancement recommends building a functional foundation in semantic HTML first, then enriching with JavaScript. Source: MDN — JavaScript.
JAWS: Commercial screen reader for Windows, developed by Freedom Scientific. JAWS is the most widely used screen reader in North America according to the WebAIM 2024 survey (40.5%). It offers advanced navigation features (virtual buffer, scripting) and supports braille displays. Source: Freedom Scientific — JAWS.
JSON-LD: JSON-based structured data format recommended by Google for embedding schema.org metadata in web pages. A JSON-LD block in a <script type="application/ld+json"> enables search engines and LLMs to understand content structure (articles, breadcrumbs, FAQ, products, etc.) without altering the visible HTML. Source: Google — Structured Data.

K

Keyboard trap: Accessibility bug where keyboard focus enters a component and can't get out through normal means (Tab, Escape). Not to be confused with focus trap, which is intentional focus confinement. Keyboard traps are prohibited by WCAG criterion 2.1.2. Source: WCAG 2.2 — 2.1.2 No Keyboard Trap.

L

Label: In HTML, the <label> element that associates descriptive text with a form field via the for attribute. The label tells the user — and assistive technology — what the field expects. It also enlarges the click area (clicking the text activates the field). A field without a label is unusable for a screen reader. Not to be confused with placeholder, which disappears on input.
Landmark: A page region defined by ARIA roles or semantic HTML elements (<nav>, <main>, <header>, <footer>, <aside>, <search>, <form>, <section> with a heading). Landmarks allow screen reader users to jump directly from one region to another via a keyboard shortcut (D key in NVDA), without traversing all content sequentially. Source: W3C — Landmark Regions.
LCP: Web performance metric that measures the time needed to display the largest visible element in the viewport (image, video, text block). One of Google's three Core Web Vitals. A large decorative image served via CSS (background-image) is discovered later than an <img> tag, which can degrade LCP. Source: web.dev — LCP.
Lighthouse: Automated auditing tool built into Chrome DevTools, developed by Google. Lighthouse evaluates performance, accessibility, SEO, and best practices of a web page. Its accessibility score relies on a subset of axe-core rules and covers roughly 30–40% of WCAG criteria. A 100/100 score does not guarantee accessibility — it means the automatable checks passed. Source: Google — Lighthouse.
Live region: A page area whose updates are automatically announced by screen readers without the user needing to navigate to it. Defined by the aria-live attribute with values polite (waits for the current announcement to finish) or assertive (interrupts immediately). Essential for form error messages, notifications, and real-time search results. Source: MDN — ARIA live regions.
LLM: Large language model trained on massive text corpora — ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), among others. LLMs power GEO by citing and summarizing web content. They don't see images — they read alt text and structured data. Well-structured, accessible content is better rendered by LLMs.

M

MathML: XML-based markup language for describing mathematical notation in web pages. MathML allows screen readers to render mathematical formulas intelligibly, where an image of an equation conveys nothing. Supported natively by Firefox and Safari; Chrome has supported it since version 109. Source: MDN — MathML.
Modal: Window overlaid on the main content that blocks interaction with the rest of the page. An accessible modal confines focus inside it (focus trap), closes with Escape, and disables the content behind it (via the inert attribute). The HTML <dialog> element implements these behaviors natively.

N

NVDA: Free, open source screen reader for Windows, developed by NV Access. NVDA is the most widely used screen reader worldwide according to the WebAIM 2024 survey (37.7%), particularly in Europe and Asia. It uses a virtual buffer that allows navigating by headings (H), landmarks (D), tables (T), and form fields (F). Source: NV Access — NVDA.

O

Outline: CSS property that draws a contour around an element, most commonly used to indicate keyboard focus. Unlike border, outline doesn't take up space in the layout and doesn't shift neighboring elements. Removing the outline without replacing it (outline: none) makes focus invisible and prevents keyboard navigation. Source: MDN — outline.
Overlay (accessibility): Third-party JavaScript widget that claims to make a website accessible automatically by adding an interface layer (font adjustments, contrast changes, navigation aids). In practice, overlays don't fix the underlying code and are regularly denounced by disability organizations. accessiBe was fined $1 million by the FTC for deceptive advertising. UserWay is the subject of a class action lawsuit. The National Federation of the Blind publicly stated that accessiBe harms screen reader users' ability to navigate. Source: Overlay Fact Sheet.

P

Placeholder: Hint text displayed inside a form field when it's empty. The placeholder disappears as soon as the user starts typing — it cannot replace a label. Its support by assistive technologies is inconsistent across browser/screen reader combinations. Recommended use: a format example (e.g. name@domain.com), never as the only field indication. Source: MDN — placeholder.
prefers-contrast: CSS media query that detects the user's contrast preference. The value more indicates the user wants increased contrast. Unlike forced-colors which replaces the site's colors, prefers-contrast: more expresses an intent without imposing a palette: it's up to the developer to respond in their own design. On Safari and iOS, this is the only available lever since there is no forced-colors mode on Apple platforms. Source: MDN — prefers-contrast.
Progressive enhancement: A web design approach that builds a functional foundation in semantic HTML first, then progressively enriches the experience with CSS and JavaScript. If JavaScript fails or is slow to load, the content remains accessible, readable, and usable. It's the opposite of depending entirely on JavaScript to display any content. Progressive enhancement benefits accessibility (the screen reader reads HTML immediately), performance (content displays before JavaScript), and SEO (the crawler indexes real content) simultaneously. Source: MDN — Progressive enhancement.
Protanopia: Form of color blindness that reduces or eliminates red perception. Affects about 1% of men. Combined with deuteranopia, red-green deficiencies affect about 1 in 12 men. Source: Colour Blind Awareness.

R

Ranking: A web page's position in search engine results for a given query. SEO techniques aim to improve ranking. Manipulating rankings through deceptive practices (including cloaking) is penalized by Google. Source: Google — How Search works.
RNIB: UK charity dedicated to blind and partially sighted people. RNIB publishes web accessibility recommendations and conducts site audits. Their collaboration with Tesco in 2000 is an iconic case study: the accessible site redesign increased online sales by £350 million per year. Source: RNIB.
ROI: Financial indicator that measures the ratio between net profit and investment cost. In web accessibility context, ROI goes beyond revenue: it includes reduced legal risk, audience expansion (16% of the world's population lives with a disability according to the WHO), SEO improvement, and customer loyalty.
RTL: Right-to-left reading direction, used by languages such as Arabic, Hebrew, and Persian. CSS logical properties (inset-inline-start, margin-inline-end, etc.) automatically adapt to reading direction, unlike physical properties (left, right). Source: MDN — CSS Logical Properties.

S

Screen reader: Assistive software that converts on-screen content into speech synthesis or braille output. Primarily used by blind or visually impaired people. Most common: VoiceOver (macOS/iOS), NVDA and JAWS (Windows), TalkBack (Android).
SEO: Set of techniques aimed at improving a website's visibility in search engine results (Google, Bing, etc.). Good accessibility practices (descriptive alt text, semantic HTML, heading structure) are also good SEO practices. See also: GEO. Source: Google — Search Central.
SERP: The results page displayed by a search engine. It's what users see after entering a query in Google, Bing, etc. Structured data (JSON-LD, schema.org) can enrich the SERP display with snippets, breadcrumbs, FAQs, or star ratings. Source: Google — Search results features.
Skip link: Link placed at the very beginning of a page, invisible at rest, that appears when it receives keyboard focus. It allows the user to jump directly to the main content without traversing the entire navigation. Required by WCAG criterion 2.4.1. Source: WCAG 2.2 — 2.4.1 Bypass Blocks.
SPA: Web application that loads a single HTML page and dynamically updates content without full page reloads. Frameworks like React, Vue, and Angular produce SPAs. The main accessibility challenge: route changes don't trigger screen reader announcements and don't move focus automatically.
SSR: Rendering technique where the server generates the complete HTML of the page before sending it to the browser, as opposed to client-side rendering where HTML is built by JavaScript in the browser. SSR benefits accessibility (content is immediately readable by screen readers), performance (LCP is faster), and SEO (crawlers index real content). Source: web.dev — Rendering on the Web.
SVG: XML-based vector image format, suited for icons, logos, charts, and illustrations. Unlike bitmap images (PNG, JPEG), SVG scales without quality loss. For accessibility, a decorative SVG should have aria-hidden="true"; an informative SVG needs a <title> and role="img". Source: MDN — SVG.

T

Tritanopia: Rare form of color blindness that reduces or eliminates blue perception. Affects about 1 in 10,000 to 12,500 people, regardless of sex — unlike red-green deficiencies which primarily affect men. Source: Colour Blind Awareness.

U

URL: Address that identifies a resource on the web (page, image, file). A readable, descriptive URL (/shoes/trainers/) benefits SEO and user experience — it appears in the address bar, search results (SERP), and shares. Source: MDN — URL.

V

VoiceOver: Screen reader built into Apple platforms — macOS, iOS, iPadOS, watchOS, tvOS. On Mac, activate with Cmd + F5. On iPhone, via Settings > Accessibility > VoiceOver. It's the most widely used screen reader on mobile (71% according to the WebAIM 2024 survey) thanks to the iPhone ecosystem. On desktop, it ranks second behind NVDA. Source: Apple — VoiceOver.

W

W3C: International web standards organization, founded in 1994 by Tim Berners-Lee. The W3C produces the HTML, CSS, ARIA, WCAG specifications and hundreds of other standards. Its accessibility guidelines (WCAG) are the worldwide reference. Source: W3C.
WAVE: Web accessibility evaluation tool developed by WebAIM. WAVE displays accessibility problems directly on the page — insufficient contrast, images without alt text, incorrect heading structure, etc. Available as a free browser extension and API. It's the tool WebAIM uses to produce the annual WebAIM Million report. Source: WAVE.
WCAG: International W3C recommendations for web content accessibility. Define three conformance levels: A (minimum), AA (standard), AAA (enhanced). The ADA Title II rule explicitly requires WCAG 2.1 AA. Level AAA (7:1 contrast ratio for standard text, among others) is not legally required but provides a more comfortable reading experience, particularly in degraded conditions (screens in direct sunlight, eye fatigue, aging monitors). Source: W3C — WCAG 2.2.
WebAIM: Non-profit organization based at Utah State University, specializing in web accessibility. WebAIM publishes the annual WebAIM Million report (automated audit of one million homepages), the Screen Reader User Survey, and provides free tools (WAVE, Contrast Checker). Source: WebAIM.
Widget: Interactive UI component — date picker, dropdown menu, carousel, accordion, tabs, etc. Custom widgets (built with <div> elements) must replicate the keyboard behavior and ARIA attributes of the native components they replace. When a native HTML component exists (<select>, <details>, <dialog>), it's almost always preferable to use it.