← The Journal · Revision Department
The Last Version Before the Next Version
A manuscript version can be complete, coherent, and temporary at the same time.
By Kevin Snell
A manuscript version can be complete, coherent, and temporary at the same time.
"Final" is a wish disguised as a filename
Writers name files with optimism: final.docx, final_revised.docx, final_revised_2.docx, final_USE_THIS_ONE.docx. The names become increasingly emotional because the manuscript keeps generating new information after the writer has attempted to declare an end.
That does not mean the earlier version was fraudulent. The first complete draft solved the problem of existence. A structural revision responds to knowledge that became available only after the ending existed. An edited version responds to another reader's experience. A proof version catches errors created or exposed by formatting. Each file can be finished relative to the work then understood.
A new version needs new information
Completion requires a boundary because every sentence remains changeable in principle. Without one, the writer can mistake the endless availability of alternatives for evidence that the book is unfinished. A next version is justified when new information reveals a meaningful problem: an editor identifies a broken motive, a proofreader finds a contradiction, an export corrupts the formatting, or publication exposes a consequential error.
The mere existence of another possible sentence is not new information. The writer will always be able to produce a different cadence after enough time away. Revision research describes writing as recursive, but recursive does not mean infinite.[1] The loop has to remain governed by purpose.[2]
Version control protects the book
File management feels administrative because it is administrative. It is also part of authorship. A brilliant revision uploaded from the wrong folder is still the wrong book in public. The writer needs to know which file contains the accepted changes, which file went to the editor, and whether a correction reached every format.
Use dates or sequential version numbers, keep one current working location, archive superseded files, and record major changes. Distinguish editable sources from exported EPUB, PDF, and print files. The U.S. National Archives recommends filenames that are unique and consistently structured, with leading zeroes when numbers determine sorting.[3] A novelist does not need a federal records system. He needs a convention that makes uncertainty difficult.
Publication raises the threshold
Before publication, revision can be broad because the manuscript has one private history. After publication, a change touches retailers, files, editions, and readers who already possess the work. Correcting a factual error, broken format, damaging typo, or serious continuity failure differs from rewriting a paragraph because the author's style has continued to develop.
A published book should not be forced to impersonate every later version of its author. Old versions preserve useful ghosts: abandoned structures, different motives, scenes remembered more vividly than they ever deserved. They show that the final shape arrived through decisions rather than inevitability. The archive is evidence of learning, not an invitation to keep the book permanently open.
Current is a real status
The phrase last version before the next version carries useful humility. This file is current. It is the one to read, edit, format, or send. It does not have to claim that no future correction or discovery can exist.
That openness should not become indecision. Commit fully to the current version. Reopen it only for a reason proportionate to the disruption. Fatigue is not proof that the book is finished, but endless possibility is not proof that it is unfinished either.
The last version before the next version may become the final one. The writer will know only when there is no longer a reason to make another—and, for once, the filename will not have to do the knowing for him.
Sources
Reader comments
Loading comments…
