How Professional Editors Actually Count Words (It's Not Always What You Think)

Traditional publishing (particularly fiction publishing) has long used a convention sometimes called "manuscript word count" or "publisher's word count," which doesn't simply count actual words the way a whitespace-split tool does. Instead, it's often estimated using a formula based on manuscript page count and a standard assumed words-per-page figure — historically, this traces back to the era of typewritten manuscripts formatted to a specific standard (12pt Courier or similar monospaced font, double-spaced, standard margins), where a "manuscript page" was assumed to contain roughly 250 words regardless of the actual word count on that specific page.

This convention persisted into the era of word processors and persists in some publishing contexts even now, because it originated as a practical printing and typesetting estimate — publishers needed a consistent way to estimate a book's final printed length and cost from a submitted manuscript before digital word-counting tools existed, and the page-based estimate became the industry standard rather than an exact character-by-character count.

Because the manuscript-count convention is based on an assumed words-per-page figure rather than the actual text on the page, it can diverge from a true whitespace-split count in either direction: a manuscript with unusually short paragraphs and lots of dialogue (which tends to have more, shorter lines with more white space) might have a lower actual word count than its manuscript-count estimate suggests, while dense, long-paragraph prose might have a higher actual count than the page-based estimate. Some publishers and genres also apply their own specific formula variations, so "manuscript word count" isn't a single universal figure the way a whitespace-split count is.

This is directly analogous to the page-count-to-word-count conversion problem covered on /blog/how-many-words-is-a-page/ on this blog — both involve converting between a page-based estimate and an actual word count, using an assumed words-per-page figure that's a reasonable average but not an exact match for any specific piece of text.

Beyond the manuscript-counting method itself, different fiction genres and formats have established, genuinely different expected length ranges as an industry norm — a novel manuscript commonly runs 70,000-100,000 words depending on genre (some genres like epic fantasy conventionally run longer, category romance often shorter), a novella typically runs roughly 20,000-50,000 words, and a short story is typically under 7,500 words, though these boundaries are conventions rather than strict rules and vary somewhat by publisher and genre community. These figures are separate from the manuscript-vs-actual-count distinction above — they're about expected total length by category, a genuinely different question from how that length gets measured.

Academic and business contexts (see /blog/word-count-guidelines-for-common-documents/ and /limits/essay-word-counts/ elsewhere on this site) almost universally specify word count as a genuine, literal count of words in the actual text — the same whitespace-split method /word-counter/ uses — rather than the page-based manuscript estimate publishing sometimes uses. If you're submitting to an academic journal, a business proposal, or a school assignment specifying a word count, a straightforward whitespace-split count is virtually always what's expected, unlike some traditional fiction-publishing submission guidelines, which may specifically ask for an estimated manuscript word count using the page-based convention instead.

If a submission guideline specifies a word count without clarifying which method it means, and you're in a fiction-publishing or literary-agent-submission context specifically, it's worth checking whether that publisher or agent has a stated preference for manuscript-count versus actual count — some do specify explicitly. For virtually every other context (academic, business, general web content, social media), a straightforward actual word count using a tool like /word-counter/ is the standard, expected method, and manuscript-count conventions are unlikely to apply.

A related, largely separate editorial convention worth knowing about: traditional print copyediting and proofreading use a standardized set of hand-marked symbols (a caret for an insertion, a delete mark for a cut, a specific mark for a paragraph break) that predate digital editing entirely and are still taught in some publishing and journalism programs today, even though most editing now happens directly in tracked-changes software rather than on a printed page. These marks aren't about word counting specifically, but they're part of the same broader professional-editing tradition that produced the manuscript-word-count convention — both grew out of a print-production era with its own specialized workflows distinct from how word counting works in a modern, purely digital writing and editing context.