Content analyzer
What is an anaphoric opening (and why it breaks AI citations)?
8 trigger words — it, this, that, these, those, they, such, and “as mentioned” — fire the ANAPHORA_OPEN flag when they open a sentence. The passage verdict becomes “orphaned,” meaning AI engines cannot lift it as a standalone citation because it depends on prior context.
Seeklyn flags this in every content analysis — run your page free at /dashboard/content.
What is an anaphoric opening?
An anaphoric opening is a sentence that begins with a pronoun or reference word that points back to a prior sentence — creating a dependency on context the reader has not yet seen. Seeklyn's ANAPHORA_OPEN flag fires when a sentence starts with any of 8 trigger words: it, this, that, these, those, they, such, or the phrase “as mentioned.” The 8 words above are the exact triggers defined in Seeklyn's lint.go anaphoraOpeners list.
When a passage contains an anaphoric opener — especially as the first sentence — the passage cannot be read in isolation. A reader (or AI engine) encountering “This reduces crawl efficiency by 40%” without the preceding sentence has no referent for “This.” The passage verdict is set to “orphaned,” the most severe extractability outcome in Seeklyn's rubric.
Why do anaphoric openings break AI citations?
AI answer engines retrieve passages individually, not pages. Google AI Overviews, Perplexity, and ChatGPT each select a passage from your content and quote or paraphrase it as a standalone answer. When that passage begins with “This,” “It,” or “They,” the engine is quoting a dangling reference — the citation is misleading or incomplete without the paragraph that preceded it.
Seeklyn models this as a binary structural failure: a passage with an anaphoric opener has verdict “orphaned” and self-contained = false. The engine cannot extract the passage cleanly, so it is passed over in favor of passages that stand alone. Princeton's GEO research (KDD 2024) identifies self-containment as 1 of the 6 core AEO levers; ANAPHORA_OPEN is the primary flag that breaks it.
How does Seeklyn detect and fix an anaphoric opening?
ANAPHORA_OPEN fires when any sentence in a passage begins with one of the 8 trigger words. The check is performed per-sentence in Seeklyn's sentenceFlags() function; if the offending sentence is at position 0 (the passage opener), it also triggers NO_ANSWER_FIRST. The verdict for that sentence becomes “orphaned,” and the whole passage inherits the self-contained = false status.
The fix is to restate the subject explicitly so the sentence stands alone. Before: “This reduces crawl efficiency by 40%.” After: “Lazy-loading assets reduces crawl efficiency by 40% (Google, 2024).” Restating the noun eliminates the backward reference and gives the engine a self-contained, quotable fact. Clearing ANAPHORA_OPEN lifts the orphaned verdict — if no other structural flag fires, the passage becomes self-contained.
Frequently asked questions
Can a passage still be cited if one sentence opens with a pronoun?
No — if that sentence is the opener, ANAPHORA_OPEN fires and the passage verdict becomes orphaned. Mid-passage pronoun openers reduce the citability score but do not orphan the whole chunk; only a sentence-0 anaphoric opener flips self-contained to false.
Is “as mentioned” also an anaphoric opener?
Yes — Seeklyn includes “as mentioned” in its 8 trigger phrases because it explicitly references prior text, making the passage non-standalone. Any sentence opening with “as mentioned” fires ANAPHORA_OPEN regardless of the sentence's content.
Does fixing anaphora guarantee a citation?
No — clearing ANAPHORA_OPEN removes the extraction blocker, but other flags like HEDGE or LOW_FACT_DENSITY may still reduce the citability score. A passage with verdict “liftable” is eligible for citation; whether an AI engine actually cites it also depends on query relevance and BM25 retrieval score.
What is the severity of ANAPHORA_OPEN?
High — ANAPHORA_OPEN is 1 of 2 structural flags (along with NO_ANSWER_FIRST) that break self-containment entirely. Every other flag in Seeklyn's rubric reduces the 0–100 citability score without voiding the self-contained badge; only these 2 structural flags set self-contained = false.