Natural Language

What this test checks

This test checks whether natural language can be determined for most content inside the PDF document.

Why this maybe a barrier

Assistive technologies rely on explicit natural language information in the document metadata and structure for correct pronunciation, braille output, and language switching. Missing or ambiguous language information can severely reduce accessibility.

How this differs from the Document Language check?

The document language check finds the single language specified for the document from its catalog and checks its validity.

The natural language check is an extended check where it checks various elements like text content, alt-text and actual text, outlines, annotation content and form field tooltips etc for any separate natural language specification, in accordance with PDF/UA and the Matterhorn protocol.

This is because natural language may vary across the same document depending on context and content.

Note that this doesn't do a detailed check on every text element in the document for natural language. In other words, the PDF content stream is not processed.

WCAG 2.2 Success Criterion

3.1.1 Language of Page (Level A) — "The default human language of each Web page can be programmatically determined."

For more information, visit WCAG 2.2 Success Criterion 3.1.1 under the guideline Readable.

PDF/UA Principle

Section 7.2: Text — "Natural language shall be declared as discussed in ISO 32000-1:2008, 14.9.2 and/or as described in ISO 32000-1:2008, 7.9.2".

For implementation, it uses the Matterhorn protocol checkpoints namely:

1. 11-001 — Natural language for text in page content cannot be determined.
2. 11-002 — Natural language for text in Alt, ActualText and E entries cannot be determined.
3. 11-003 — Natural language for Outlines cannot be determined.
4. 11-004 — Natural language for Annotation Content cannot be determined.
5. 11-005 — Natural language for form field tooltips cannot be determined.