Is this a book, and who is it for?

Is there enough here for a book, or is it a long article, a short story, or a blog?

Length is the part of this question with published answers. The Nebula Award rules, effective January 1, 2025, set four boundaries. A short story is fewer than 7,500 words and a novelette is at least 7,500 and fewer than 17,500. A novella is at least 17,500 and fewer than 40,000, and a novel is 40,000 words or more.¹ Those four categories are what the field means by the words, and a manuscript can be measured against them in a minute.

What length cannot settle is whether the material sustains a book. That is a judgment about the material, and an editor reaches it by reading rather than by counting. We can tell an author where a manuscript falls against the published definitions, and we can say which sections carry the argument and which restate it. We do not tell an author that an idea is too small, because that is a judgment about their subject and not about their sentences. Every document a client sends us is treated in confidence.

Is it one book or two?

Publishers ask this in a form an author can borrow. The University of California Press asks a proposal to show what each chapter contributes to the logic of the argument.² Princeton University Press asks for a list of planned chapters with a paragraph describing the contents of each.³ A chapter list that needs two separate explanations, one for the first half and one for the second, has answered the question on the author's behalf.

Splitting is not free. Two books need two readerships, two sets of comparable titles, and two arguments that each stand without the other. A book divided because it grew long usually produces one strong half and one thin one.

An editor reading for structure reports where the subject changes and what each chapter is doing. Whether that division becomes two books is a decision about a career and a market, and it belongs to the author.

Should this be a novel, a memoir, or a nonfiction book about the same events?

The trade files these separately, and the filing has consequences. The BISAC Subject Headings are the codes publishers use to categorize a publication, and the Book Industry Study Group issues them in editions, of which 2025 is the current one. Fiction and Biography and Autobiography are separate categories within the list.⁴ A book is sold, shelved, and reviewed according to which one it carries.

The practical difference is what the book is allowed to do. A memoir is read as a claim about what happened, which brings naming, permissions, and legal review with it. A novel drawn from the same events is read as invention, and it can compress, reorder, and combine people.

Neither choice is more serious than the other. Authors who cannot say what happened, or who need a scene nobody witnessed, are describing a novel. We are editors and not attorneys, and where a manuscript names real people, the question of what may be published belongs with a lawyer.

Has someone already written it?

No one owns a subject. The United States Copyright Office states that copyright does not protect ideas, procedures, methods, systems, processes, concepts, principles, or discoveries. The same circular states that it does not protect titles, names, short phrases, and slogans.⁵ A book on the same subject, or with the same title, is not a legal obstacle.

Commercially it is a different matter, and publishers ask about it head on. Princeton asks authors to list comparable books with title, author, publisher, and year, and to state briefly how the proposed book differs. It also tells authors to conduct the appropriate research before claiming that no comparable titles exist.³

That is a catalog search rather than an impression. The list a publisher wants is of real books with real publication details, and an author who has not looked is easy to catch.

Who reads this kind of book?

Princeton puts the question in three parts: is the book for nonspecialist general readers, for scholars, and if so in which fields, or for students. It adds that few books, if any, fall into all three.³ The University of California Press asks the same three ways, and asks for the course level where the answer is students.²

The answer that a proposal cannot use is everyone. Two presses independently asking an author to choose is a fair sign that choosing is expected.

The answer changes the book rather than only the pitch. It sets how much is explained, how much vocabulary is assumed, and whether the book needs notes, a glossary, or neither. An editor works from the answer the author gives, and marks the passages that assume knowledge the stated reader was not given.

Am I writing for readers, for a committee, or for people I know?

A committee, a circle of acquaintances, and a stranger want different things from the same pages. A committee reads for completeness and wants every qualification present. People who know the author read with the missing context already supplied. A stranger has no reason to keep reading except what is on the page.

Manuscripts written under the first two conditions share a symptom, which is material that only makes sense to someone who was there. It appears as unexplained names, private references, and qualifications that protect the author against an objection the reader has not thought of.

That symptom is checkable. An editor can mark every place where the text assumes something the stated reader has not been told, without deciding whether the assumption should stay. Some books are written for a committee on purpose, and the marking is still useful.

What shelf does it go on, and does that shelf still exist?

Shelves have codes, and the codes are public. BISG offers the BISAC Subject Headings as a free service, and publishers select a BISAC code to categorize a publication; the list is issued in editions, of which the 2025 edition is current.⁴ The code travels with the book through the systems that decide where it is stocked and how it is found.

Because the list is versioned, a category an author remembers may have been renamed or absorbed. The edition a publisher is working from is the one that decides the question, not the one in print when the book was started.

A book that resists every code is not thereby a bad book. It is a book whose publisher has a harder job, and that difficulty is worth knowing before the proposal is written rather than after.

Which books are its closest comparisons, and were they published recently enough to matter?

The recency question has a published answer. The University of California Press asks authors to describe existing books in the field published in the past three to five years. It then asks them to discuss how the proposed work expands on the coverage those books give.² Princeton asks for title, author, publisher, and year, with a brief statement of how the book differs.³

Those two requests agree on more than they differ. Both want named books rather than a category, both want the difference stated by the author, and one of them puts a date range on it.

A comparison drawn from a book published fifteen years ago describes a readership the publisher is no longer selling to. Where an author genuinely cannot find recent comparisons, that is worth saying plainly in the proposal, because a publisher reading a list of old titles will otherwise conclude the research was not done.