Knowing when a draft is finished

Is this a draft or a manuscript, and how would I know?

The distinction is practical rather than official. A manuscript is a complete file that somebody else can read from beginning to end without being told what is missing. A draft is everything before that.

The EFA's guide for new authors draws the same line in terms of hiring. It advises authors to complete or substantially shape a manuscript before looking for professional editing, and suggests a writers' group for feedback beforehand.¹ An editor asked to quote will ask whether the manuscript is complete for the same reason.

The test that settles it is unglamorous. Three things settle it. The file is a draft if it holds a chapter that exists only as a heading, a scene marked to be written, or a section a stranger could not follow without the author present. Nothing about that is a judgment on the writing.

Is this finished or am I avoiding the hard chapter?

No outside party can answer this, and an editor who claims to is reading the author rather than the pages. What an editor can do is describe the manuscript's own unevenness.

Chapters written under avoidance tend to share visible properties. They are shorter than their neighbors, they summarize where the rest of the book dramatizes, and they resolve quickly. A structural report that lists chapter lengths, and what each chapter changes, puts that pattern in front of the author without anyone having to interpret motives.

The author is the only person who knows whether the thin chapter is thin by design. Some books are meant to move fast at that point. The finding is offered as a finding, and it stays that way.

How many drafts is normal before anyone else reads it?

No one has measured it. We found no survey, association standard, or publisher statement giving a number of drafts, and any figure quoted for it is somebody's practice rather than a norm.

What varies more than the count is what a draft means to different writers. An author who revises continuously while writing may have one draft that has been through twenty passes. An author who writes fast and fixes later may have five drafts and less total revision.

The question that can be answered is what condition the manuscript is in, not how many times it was rewritten to get there. Consistency, completeness, and whether the chapters were last read in order are all checkable, and none of them requires knowing the draft number.

Should I read it aloud?

Reading aloud finds a specific class of problem and misses others. It exposes sentences that cannot be spoken in one breath, repeated constructions, dialogue that no person would say, and rhythm that flattens over a page.

It does not find inconsistency, and it is poor at structure. An author reading aloud is moving too slowly to notice that a scene on page 40 does the same work as one on page 180, and hears the sentence rather than the shape.

The version of this that costs nothing is reading only the dialogue aloud, which is where the technique earns most of its reputation. Whether it is worth the hours for a whole book depends on how much of the book is voice.

Should I print it?

Printing changes the unit a reader sees, which is why it works when it works. A screen shows a paragraph and a printed page shows a page, and faults of proportion become visible at the larger scale.

The mechanical version of this argument is stronger than the mystical one. Layout errors, uneven chapter lengths, headings at inconsistent levels, and long stretches with no paragraph break are all easier to see on paper. Proofreading is defined by Editors Canada as examining material after layout or in its final format, and the reason that stage exists at all is that a different presentation shows different faults.²

An author who does not want to print 400 pages can get much of the effect more cheaply. Changing the font, narrowing the column, and reading on a different device all disturb the familiarity that hides faults.

How long should I leave it before rereading?

No published source sets a period, and the advice in circulation ranges from a week to a year without evidence behind any of it. What the interval is for is losing the memory of what was intended, so that the pages have to do the work on their own.

That suggests a test rather than a duration. An author who can still recite what the next paragraph says has not waited long enough, and an author who has forgotten a character's name has probably waited long enough.

Setting a manuscript aside has a cost that is rarely mentioned. Momentum is real, deadlines exist, and a book left for a year sometimes stays left. The interval is a judgment about the author's own working habits rather than a rule.

Should I stop revising and start submitting?

Two different situations look identical from inside. Revision that keeps finding real problems is working. Revision that changes a sentence one way, then back, is not revision, and no further passes will make it so.

An editor can distinguish them from outside by comparing versions, which is one of the few genuinely diagnostic things a fresh reader can do. Where successive drafts differ mostly in word choice, the manuscript has stopped changing.

Writer Beware is worth knowing here for what it rules out. It states that saying a manuscript is professionally edited will not improve an author's chances with agents and publishers, and may harm them.³ Paying for another pass in the hope that it changes a submission's reception is buying something nobody has promised.

Should I set it aside and write the next one?

A career question is hiding inside this one, and an editor has no standing to answer that part of it. What can be said is what the alternatives cost.

Continuing to revise a manuscript that has stopped improving costs time and produces no second book. Starting the next one costs the current book its chance to be finished properly, and authors who leave a book at ninety percent frequently never return to it.

Some authors do both, alternating between the two, and that works when the books are different enough that neither voice contaminates the other. It fails when they are similar, because the revision starts importing the newer book's solutions into the older one.

We have no evidence about which choice produces better outcomes. Anyone offering some is generalizing from their own career, which is a sample of one.

Am I stuck on the book, or stuck on the writing?

The two have different symptoms and the distinction is usually visible in the pages. A book problem produces a manuscript that will not move forward from a particular point, where every attempted next chapter contradicts something already written. A writing problem produces sentences the author dislikes on pages that are otherwise doing their job.

A structural read separates them quickly, because a book problem shows up as an inconsistency in the plan rather than as bad prose. A manuscript stalled because chapter nine made chapter fourteen impossible is a structural finding, and it can be named.

Where the pages are sound and the author cannot bear them, no editorial pass will help. Selling one on that basis would be dishonest, and an editor who has read the manuscript can usually tell which of the two situations they are looking at.

Should a first novel be submitted, or set aside in favor of the second?

Both answers have serious people behind them, and the evidence for either is anecdotal. The case for submitting is that a manuscript nobody reads teaches nothing, and rejections carry information when they are personal rather than form responses. The case for setting it aside is that a debut is only debuted once, and a second novel written with everything learned from the first is often the stronger book.

What can be said from a published source is narrow. Writer Beware notes that no reputable editor will tell an author a book has huge commercial potential or is likely to become a best seller, so an editorial opinion is not the tiebreaker here.³

An editor can tell an author which of the two manuscripts is in better shape, and that is a smaller question than the one being asked. Every document a client sends us is treated in confidence, including a manuscript an author is undecided about.