Cutting a scene you love

Subject: Revision

The hardest cuts are never the weak passages. Weak passages go quietly; you know they are weak, you have known since you wrote them, and removing them feels like tidying. The hard cut is the one that works on its own and does nothing for the piece: a passage that is well made, that you would happily read in someone else's writing, and that the piece is carrying rather than using.

It is worth being precise about why these passages are dangerous. They are not merely neutral. A strong passage that does no structural work actively misleads the reader about what the piece is for. It sets an expectation, it raises a subject with apparent significance, and then the piece moves on and the promise is never redeemed. Readers rarely articulate this. They just report that the middle sagged.

A test that does not rely on taste

Ask what the reader could not do without it. Not what would be lost in flavour, but what would break. Would a later moment stop making sense? Would a claim later on lose its evidence? Would a change in the piece's direction become unmotivated? If the answer to all of those is no, the passage is decoration, however good the decoration is.

Run the test by actually removing the passage rather than imagining removing it. Cut it into a separate file, then read the surrounding pages straight through. Most of the time the join is invisible, and the fact that it is invisible is your answer. Occasionally you will find a small hole where the passage was carrying something you had not noticed, and then you know exactly what the passage was for, and you can put back that function in three sentences rather than three pages.

The separate file matters

Keeping an offcuts file is not sentimentality, it is a mechanism for lowering the cost of the decision. Deleting outright makes the cut feel final and therefore frightening, and a frightened reviser keeps things. Moving the passage to a file called offcuts makes the cut feel provisional, and provisional cuts get made. That the file is almost never reopened is beside the point; its job is done at the moment of the cut.

It has one genuine second use. When you are three pieces further on and need a passage on that subject, the offcuts file already contains a finished one, written when you cared about it. Some of the best paragraphs arrive in a piece they do not belong to.

When the piece is the wrong piece

There is a version of this problem where the beloved passage is not decoration at all. It is the actual subject, and the piece around it is the scaffolding. You can usually tell because cutting it leaves something coherent but pointless: the argument still stands, and you no longer care whether it stands.

That is not a cutting problem. That is a signal that you wrote the frame first and found the subject late, which is a common and entirely survivable outcome of drafting. The move is to invert: keep the passage, discard the frame, and write the piece the passage implies. It will cost more than a cut, and it is nearly always the right call.

Doing it on a schedule rather than by mood

Cuts made in the flush of finishing are too timid, and cuts made when you have soured on a piece are too broad. Neither state is a reading state. The workable compromise is to make structural cuts on a separate pass, on a separate day, with a printed or otherwise unfamiliar version of the text, and to write down every cut you make and why in a single line.

That log is worth keeping. Read back over a year, it shows you the passages you habitually write and habitually have to remove, and that pattern is more useful to your writing than any individual cut. Most writers have one: the long throat-clearing opening, the second example that repeats the first, the summary paragraph that says what the piece just said. Once you can name yours, you begin to stop writing it.