AI tell · grade B
The "not only... but also" AI writing tell
The additive frame stacks two claims into one hinged sentence to sound comprehensive: it grants one point and then reaches for a second, larger one. Our model logs show it in most of the models we sampled. Readers flag it because it inflates a simple statement into a balanced-sounding pair without adding substance. The fix is to make two plain sentences, or keep the one claim that matters and cut the other.
What it is
A hinge that promises the sentence is about to over-deliver. The first half concedes a modest point; the second reaches past it for a grander one. The effect is a sentence that sounds thorough while saying roughly one thing.
How often it shows up
Our logs (emp-010) put the additive negative-parallelism family in 8 of 12 sampled models, and all three legacy style guides we cross-checked list it as a high-value strike. The checklist grades it B.
Why readers flag it
The frame reads as padding dressed as balance. A reader senses the second clause was added for symmetry, not because the writer had a second thing to say, and that reflex is what makes the whole line feel assembled rather than written.
The fix
Split it. If both halves matter, give each its own sentence and its own specific. If only one matters, keep that one and delete the hinge entirely.
Examples
This approach not only saves time but also improves morale, and not only that, it drives long-term results.
This approach saves about a day a week. The team also likes it better, which matters more than I expected.
Questions
How is this different from the denial reframe?
They are the same family. This version adds two things; the other denies one thing to assert another. Both are mirrored frames that read as a rhythm a model fell into.
underslop marks this tell and the rest of the consensus set in any draft you paste, using the same open lint that scores it here. The free tier caps at 500 words a run, and your text is held for 0 seconds.