Convergence Principle

The Convergence Principle of the Zero Information Casualty standard holds that agents and screen readers are functionally equivalent at the structural level. Both navigate by landmarks, headings and links; both need link text that describes its destination, heading hierarchy that reflects content, and reading order that matches the page; and neither can interpret meaning carried only by colour or position.

The convergence is not incidental. Information held only in the visual layer is hidden from a large and growing share of all information consumers, and as agents become primary readers alongside people, "the case for visual-only encoding collapses entirely." It is the web-page form of the convergence the Tyrer Framework describes under generative accessibility.

Where it stops

"The equivalence is real, and it is local." ZIC's requirements are the union of two overlapping sets, neither inside the other. A sent agent needs delivery, admission, a certificate that validates and a robots policy that doesn't fail closed, all invisible to a screen reader user, who arrives inside the building. A human reader needs visible focus, keyboard operation, skip links, text scaling and contrast, which no agent has ever needed. The overlap is big enough that a publisher building for machines removes most barriers facing blind people, "and it needs no apology." But "neither reader sees all of the terrain."

Behind the scenes

A real idea, from Zero Information Casualty v1.7, the design standard by Charli-Jo Tyrer and Claude, §1.3 and §1.13. This entry summarises it; the standard is the authority.