Ah, the mythical document outline! If you’re at least a little bit into HTML semantics, you might have heard of the phrase once or twice. Broadly speaking, the document outline is the content structure defined by the headings in a page. This structure is important as heading navigation is one of the main ways for screen-reader users to browse a website.
In this article for instance, the post title is a <h1> and then we have a bunch of <h2>. If any of these end up needing sub-sectioning, there would be <h3> and so on and so forth. The outline looks like this (as of writing):
1. A11y Advent Calendar
1.1. Day 1: What is Accessibility?
1.2. Day 2: Evaluating Accessibility
1.3. Day 3: Hiding Content
1.4. Day 4: Self-Explanatory Links
1.5. Day 5. Document Outline
To check the structure of a document, we can use the same accessibility bookmarklet we’ve mentioned yesterday. When activating it, one of the options is “Headings”, which lists all headings in order and level. From there, we can make sure the structure makes sense, headings are in the right order, and no level is skipped.
For years now, there have been discussions (and even proposals) about taking into consideration sectioning elements like section into the document outline to create sort of sub-structures where every root would go back to h1. This has never been implemented by any browser or supported by any assistive technology so this is basically moot at this point. Stick to appropriate heading levels.
For more information about the history behind the document outline and the proposed resolution algorithm, I encourage you to read the Document Outline Dilemna by Amelia Bellamy-Royds which is a fantastic overview of the topic.