Skip to main content

Writing guide · 6 min read

The Passive Voice Audit: How the Engine Finds It and Why It Matters

Passive voice is not a crime. It is a structural choice with specific costs and uses. Here is how Inkbreaker detects it and what the percentage means.

By Alyssa Glasco, Founder · Published

Passive voice is not a crime. It is a structural choice with specific costs and uses. The cost is diffusion. Passive constructions remove the actor from the sentence. “The door was opened” versus “She opened the door.” The first version has no agent. The reader cannot tell who acted. In fiction, that obscures accountability. In journalism, it sometimes enables it deliberately.

How the engine detects passive voice

Inkbreaker detects passive voice using a pattern match: auxiliary verb followed by a past participle within a fixed token window.

The auxiliary verbs are: is, are, was, were, be, been, being, plus the compound forms has been, have been, had been, will be, would be, can be, could be, should be, may be, might be.

The past participle is identified as a word ending in -ed or an irregular form from a curated exception list (written, taken, built, found, and approximately 200 others).

This is syntactic parsing, not semantic judgment. The engine does not assess whether the passive is appropriate in context. It counts how often the pattern appears and compares that count to a genre-specific threshold.

The Inkbreaker benchmarks

  • Fiction: tolerates up to 10 percent passive constructions.
  • Technical writing: tolerates up to 20 percent.
  • Screenwriting: tolerates up to 40 percent. Action lines are conventionally passive.

When the engine flags your passive voice percentage, it is not telling you the passive is wrong. It is telling you the ratio. What you do with that ratio is yours.

The predicate adjective exception

Predicate adjectives are not passive. “She was tired,” “the room was bright,” “he seemed distracted” are linking verb constructions, not passive voice. Inkbreaker maintains a stoplist of common adjectival past participles (tired, excited, bored, confused, surprised, worried, satisfied, and dozens more) to prevent false positives. That stoplist lives in the engine source and is checked on every match.

The stoplist grew through use. The first version of the engine flagged “she was tired” as passive. After it kept flagging that pattern on my own drafts, I started keeping a list. The list is not theoretical. Every entry is a false positive I hit while writing.

A practical test

Pick the most recent paragraph you wrote. Paste it into the Prose Grade tool and look at the “What the engine found” panel. Every passive construction is highlighted in teal. Decide, for each one, whether the missing actor is doing work. If it is, leave it. If it is not, rewrite the sentence active.


References

Quirk, R., Greenbaum, S., Leech, G., Svartvik, J. (1985). A Comprehensive Grammar of the English Language. Longman. / Strunk, W., White, E.B. (1959). The Elements of Style. Macmillan.

Inkbreaker uses cookies to keep you signed in and, with your permission, to understand how the app is used. Nothing is sold or shared. Privacy Policy.