Matterhorn 01-002 · PDF/UA-1 UA1:7.1-1
Real content marked as an artifact: text a reader skips
Content that carries meaning has been marked as an artifact, which removes it from the structure tree entirely. This is the most damaging single mistake in PDF remediation because the content becomes invisible rather than merely mis-labelled.
The condition, as the protocol states it. “Real content is marked as artifact.”
Matterhorn Protocol 1.1, checkpoint 01, index 01-002, section UA1:7.1-1. The protocol marks this as requiring human judgement.
What a screen reader actually does
The content is simply not announced. The reader has no way to know it exists.
What the condition is
An artifact is content a page paints and the structure tree leaves out. It is marked in the content stream rather than in the tree:
/Artifact <</Type /Pagination /Subtype /Footer>> BDC
BT /F1 8 Tf 1 0 0 1 72 30 Tm (Page 4 of 12) Tj ET
EMCNo structure element points at it, so assistive technology never reaches it. For a page number, that is correct. Condition 01-002 is the same marking around something that means something:
/Artifact BMC
BT /F1 11 Tf 1 0 0 1 72 654 Tm (Clause 7: the tenant shall keep the premises in good repair.) Tj ET
EMCIt is the mirror of 01-001. Decoration tagged as content is noise; content marked as decoration is a hole.
What the user hears
Nothing, and no gap where it was. A reader moving through the document goes from the sentence before the clause to the sentence after it. There is no “graphic”, no blank line, no hint that the page holds more than was read. Other failures misdescribe content; this one removes it.
Where it comes from
Whole regions artifacted at once. A producer that treats everything in a header or footer area as pagination also artifacts the one line there that is not — the version date, the reference number, the disclaimer.
Remediation shortcuts. The Reading Order tool’s Background/Artifact button and the Tags panel’s Change Tag To Artifact both remove content from the tree in one click. Applied to a region drawn a little too large, they take a caption or a footnote with the ornament.
Stamps and overlays. A tool that stamps text onto finished pages can mark it as an artifact so as not to disturb the existing tree. For DRAFT that is defensible. For Superseded — see version 3 it is not.
How to check it free
Listen. Adobe’s Content panel documentation notes that artifacts are not read by a screen reader or by Read Out Loud, so any text you can see and never hear is either an artifact or untagged. A screen-reader pass, or Read Out Loud, is the most direct test there is.
Acrobat Pro’s Tags panel. Options → Find searches for artifacts, among other things; Adobe describes it as searching for “artifacts, OCR suspects, and unmarked (untagged) content, comments, links, and annotations”, by page or across the document. Step through what it finds and ask of each one whether a reader would miss it.
Acrobat Pro’s Content panel (Show/Hide → Side Panels → Content) shows the page’s content objects in order, artifacts included, and Highlight Content outlines each on the page.
The Accessibility Checker will not help: an artifact satisfies its Tagged content rule by definition.
How to fix it free, in Acrobat Pro
Using the commands Adobe documents on its Content and Tags panel page and its accessibility checker page:
- Open the Content panel and find the object. Right-click it → Remove Artifact, which removes the artifact definition. The text is now untagged — neither content nor artifact.
- Tag it:
All tools → Prepare for accessibility → Fix reading order, select the text on the page, and click the button for what it is — Text/Paragraph, a heading level, Note. This is the route Adobe gives for fixing a failed Tagged content check. - Open the Tags panel with Highlight Content on, find the new element, and drag it to where it belongs in the reading order. A tag created this way lands where Acrobat puts it, which is not necessarily next to the sentence it continues.
- Run the Accessibility Checker again. Tagged content should pass, because the text is now in the tree.
If the producer artifacts a region systematically, fix the template rather than every page: in InDesign, the Tags panel (Window → Utilities → Tags) shows which frames carry the Artifact tag (Adobe), and a text frame that holds content should not be one of them.
What Taggart does
Taggart’s check is a heuristic, and it looks for volume rather than meaning. On each page it counts the text-showing operations inside artifact sequences, and the number of text blocks in the top and bottom 12% of the page, where running heads and footers live. A page is reported when it has more than twelve artifacted text operations and more than four for every block in those bands:
Page 1 marks 14 text-showing operations as artifacts, far more than its header/footer bands contain. Content hidden this way is invisible to assistive technology.
A page with a running head and a page number is not reported, and a page with twelve artifacted text operations never is. That makes the check good at the large failure — a body of text artifacted wholesale — and blind to the small one: a single disclaimer line in a footer is well below the threshold, and a threshold low enough to catch it would flag ordinary footers. Images marked as artifacts are not counted at all. It counts operations, not lines, so a producer that paints one word per operation reaches the threshold sooner.
The finding is a warning and goes into the review queue with its page, where you record whether the content is really decoration. There is no repair — deciding which artifacted text is content is the whole judgement — and recording the decision changes nothing in the file. Make the change in Acrobat, as above, and check again.
The screen-reader preview helps with the small case. When anything in the document is artifacted text, the preview adds one line to its reading saying how many runs a reader will skip, across the whole document:
14 run(s) of text are marked as artifacts, so a screen reader never announces them. That is correct for a page number or a rule, and wrong for anything a sighted reader would call content.
On a one-page letter that says 2, you can guess what they are. On a letter that says 14, go and look.
How Taggart handles it
The protocol marks this condition as needing human judgement, so no validator can decide it outright. Taggart applies a heuristic to find the likely cases and reports each one as warning for you to settle in the review queue. A heuristic finds what it was built to find and can miss the rest; what this one looks for is described above.
Whatever Taggart changes, the page still looks identical. Fixes edit the PDF object model, not the content streams, and every remediation is re-rendered and compared against the original at structural similarity ≥ 0.999 before you get the file back. A tool that quietly reflows your document is worse than no tool.
Frequently asked questions
How is this worse than leaving the text untagged?
Untagged text is condition 01-005, a machine-checkable failure that validators report. Text marked as an artifact has been declared meaningless, and that declaration is valid: the file passes every machine check while the reader is told to skip the content. Nothing flags it except a person looking.
Aren't headers and footers supposed to be artifacts?
Running heads and page numbers, yes — that is condition 18-001. But a footer can also hold the only copy of something: a document reference, a version date, a contact address, a legal notice. Repeated on every page, it is pagination. Printed once and nowhere else, it is content, and somebody who cannot see the page needs it once.
Will Acrobat's Accessibility Checker catch it?
No. Its Tagged content rule asks whether content is either in the tags tree or marked as an artifact, and artifacted text is the second. Nothing else in the rule list Adobe publishes asks whether an artifact should have been content.
Can Taggart put the text back automatically?
No, and it does not offer to. Deciding which artifacted text is content is the judgement this condition is about, and putting it back means choosing an element type and a place in the reading order. Taggart reports the page and leaves both decisions to you.
Where this sits in the standards
| Standard | Reference |
|---|---|
| Matterhorn Protocol 1.1 | Checkpoint 01 (Real content tagged), index 01-002 |
| PDF/UA-1 (ISO 14289-1) | Clause 7.1-1 |
| WCAG 2.1 | 1.3.1 Info and Relationships — Level A |
| EN 301 549 / Section 508 / ADA Title II | All three point at WCAG 2.1 Level AA for non-web documents, so a Level A or AA criterion here is in scope for each of them. |
Related conditions
- 01-001 Artifact tagged as real contentDecoration — a rule line, a page-border image, a watermark, a background flourish — has been given a real content tag, so assistive technology announces it as if it were part of the document.
- 01-005 Content is neither marked as Artifact nor tagged as real contentEvery text-showing, image-painting and path-painting operator on a page must sit either inside a marked-content sequence that maps to the structure tree, or inside an artifact sequence. Content that is in neither is orphaned.
- 18-001 Headers and footers are not marked as pagination artifactsRunning heads, footers and page numbers are tagged as real content, so they repeat in the reading order on every page.
- 09-001 Tags are not in logical reading orderThe order of elements in the structure tree does not match the order a human would read the page. This is the single most common substantive defect in tagged PDFs and the one Acrobat's own tooling handles worst.
Other conditions in checkpoint 01
Checkpoint 01 covers real content tagged. These are its other failure conditions; the ones without a link are in the rule set but do not have a written page yet.
- 01-001 Artifact tagged as real contentDecoration — a rule line, a page-border image, a watermark, a background flourish — has been given a real content tag, so assistive technology announces it as if it were part of the document.
- 01-003 Content marked as Artifact is present inside tagged contentA BMC/BDC artifact sequence has been opened while a marked-content sequence for real content is still open. Artifacts and real content may not be nested inside one another.
- 01-004 Tagged content is present inside content marked as ArtifactA marked-content sequence for real content has been opened inside an open artifact sequence. The artifact swallows the real content.
- 01-005 Content is neither marked as Artifact nor tagged as real contentEvery text-showing, image-painting and path-painting operator on a page must sit either inside a marked-content sequence that maps to the structure tree, or inside an artifact sequence. Content that is in neither is orphaned.
- 01-006 Structure type is not semantically appropriateThe tag does not match the nature of the content: a heading tagged as a paragraph, a data table tagged as a layout container, a caption tagged as body text. This is the judgment call that no validator can make for you.
- 01-007 Suspects entry has a value of trueThe MarkInfo dictionary carries /Suspects true, which is the producing application's own declaration that its tagging may be unreliable. PDF/UA-1 requires the value to be false.
Check your own file. Taggart’s validator is free and unlimited — 75 of the 87 Matterhorn conditions software can decide, with the other 12 listed for you to confirm; no watermark, no expiry, and no account needed up to 60 pages.
Checked with Taggart on a generated file that fails this condition on . The Acrobat steps follow Adobe’s documentation and were not walked through by hand. Last updated .