This is a question that comes up a lot among writers who are working on academic books. Like so many aspects of scholarly publishing, there is not one correct answer! Or at least not a definitive, satisfying one.

The most accurate answer is that your book should be as long as it needs to be to fully develop and support its central argument. For some books, that’s 40,000 words; for some it’s 150,000. But what people are usually looking for is a ballpark average range for the kind of book they are writing. This is an eminently logical thing to want to know, because publishers, tenure committees, and readers all have pre-existing expectations for what a book should be like, and there can be real consequences if you deviate too far from convention.

So here’s a real answer: in my work with academic authors, the vast majority of manuscripts fall in the 80,000–90,000-word range. Books that fall on the more social sciencey or media studies side of things may be shorter. It’s not unusual for those manuscripts to run 65,000–75,000 words when I get a look at them. When I work on more historical or humanistic manuscripts, they can run longer, closer to 100,000 words. And I’ve certainly heard of manuscripts that run even longer than that.

The risk of a shorter-than-average book is that your peers and the people who are in a position to affect your academic career may perceive the book as not rigorous or substantial enough to meet professional standards. If you’re writing an explainer book or something else that you don’t really need to count toward tenure, this might not matter.

The risk of a long book is that it requires a greater investment from the publisher because it simply costs more to produce a book with more pages. When an acquiring editor makes profit and loss calculations before pitching your book to their colleagues at their press, a longer book may represent more “loss.” While scholarly publishers don’t always expect to make a ton of money from an academic title, they can’t afford to lose too much money either. A long book may therefore need a higher price point to cover its costs. And a long, expensive book may be intimidating to your target readers.

My advice is to write the book you want and are able to write. See how long that turns out to be. If you’re wildly over or under a reasonable word count—or the length specified in your contract, if you signed it before finishing the manuscript—you can cut things in revision or identify a new angle on the argument that could generate a new chapter and a deepening of the thesis. (A developmental editor can help with both these things, by the way.) Keeping word count at the front of your mind while you’re trying to draft the book just creates too much extra pressure on an already stressful process.

A corollary question that I see frequently is how long one’s chapters should be. Again, typical lengths can vary by field. I used to say that 8,000–10,000 words was a good length for an academic book chapter (because I’d read that in multiple advice books), but I have found that most academic authors write longer chapters than that. And sometimes it takes 12,000 or 14,000 or 18,000 words to develop a chapter-worthy argument. So I have backed off of prescriptivism with my clients when it comes to chapter length.

More helpful, I’ve found, are rules of thumb about what to include in a chapter. My first rule is that each chapter should have a cohesive, bounded argument of its own that supports the book’s central argument. And the only material that should make it into the chapter is material that supports that cohesive, bounded argument, without repeating anything. Imposing that rule usually trims a few thousand words out of any chapter draft that crosses my desk, so if you’ve got a long one on your hands, start there. That rule might even help you figure out how to split one very long chapter into two reasonable-length chapters, if you recognize that you are actually making two arguments in it with two separate sets of supporting evidence.

My second rule is that book chapters should be roughly equal in length. Some variance between chapters is fine, but a chapter that is way longer than its neighbors can indicate an argument that’s gotten away from its author, while an oddly short chapter can expose a lack of supporting evidence or analysis. (The introduction and conclusion are often shorter than body chapters, so don’t worry about these.)

The estimated length of your manuscript is just one of the “specs” that you’ll need to report when you submit your book proposal to a scholarly publisher. We’re demystifying specs and status and all the other little details that round out a book proposal this week in the Manuscript Works book proposal accelerator. If you’d like to join me for the accelerator’s fall session, there’s still time to enroll! And FYI, there are two “scholarships” available for authors who would like to join the accelerator but find the enrollment fee a hardship—check out my previous newsletter for instructions on how to apply for one of those spots by August 24th.