Every tool that reports unusual words has to decide what counts as usual, and that decision is a list. The list is invisible in the output by definition, since its whole purpose is to remove things, and it determines what you'll never be told. Knowing it exists is what lets you understand a report's silences.
The trade
| List size | Effect |
|---|---|
| Too small | Results swamped with ordinary English; obviously broken |
| Right | Invented and specialist vocabulary, cleanly |
| Too large | Real terms silently excluded; not obviously broken |
The asymmetry is the point. A list that is too small produces an obviously useless report; one that is too large produces a report that looks clean and is missing things, which is the dangerous failure.
Where it bites in fiction
- Names that are ordinary words. A character called Grace, a place called Hope, a ship called Endeavour.
- Titles that are ordinary words. The Keeper, the Warden, the Hand.
- Invented terms built from real words. A compound of two common words.
- Common surnames. Which appear in name lists as ordinary vocabulary.
- Words that are ordinary in one sense and specialist in another.
Row one is the practical limit an author should know about. A character named with an ordinary English word will not be surfaced by frequency-based extraction, and checking that name's consistency stays a search job.
Working with it
The response is not to distrust the report but to know its shape: it covers the unusual vocabulary thoroughly and does not cover names made of ordinary words. Those go on a manual list, and searching for them is quick because you know which they are.
The Novel Kit reports proper nouns and unusual terms, which means exactly this trade is being made on your behalf. A character named Grace is in the ordinary vocabulary of English and will not appear as an unusual term; the check for that name is a search, and knowing so stops you assuming otherwise.
A practical next step
List the characters and places in your book whose names are ordinary English words. That's your manual list, and it's usually short.
Sources checked
Frequently asked questions
What is a stop list?
A list of words a tool deliberately never reports, usually the common vocabulary of the language, so the results are not swamped.
What happens if it is too large?
Genuine terms get excluded. A name that is also an ordinary word will never appear, which is a silent miss.
What if it is too small?
The report fills with ordinary English and becomes unusable, which is the failure everyone notices immediately.