What to do when they refuse to sign a release

The email came back, and the answer is no. Maybe it was one polite line. Maybe it arrived through a lawyer with adjectives attached. Either way, the chapter is load-bearing, the delivery date has not moved, and your publisher is waiting to hear your plan.

Before you touch the manuscript, get the size of the event right. A release was a shortcut: one signature swapping a stack of legal questions for a piece of paper. The signature is not coming. The questions are still answerable. What shrank is your margin for error, and this page is about spending the margin you still have.

A release, in other words, was insurance on top of a legal position, and the position exists whether or not anyone signs. The full map of that position is on the main guide, What You Can and Can’t Say. What follows is the release-shaped corner of it.

Do you legally need a release to write about someone?

No. American law does not condition writing about a real person on that person’s consent. The risk lives in what you publish: a false factual claim about an identifiable living person, a private fact publicized without legitimate public concern in states that recognize that tort, or material gathered in ways the law treats as intrusion.

So why did anyone ask for a signature? Because publishers buy certainty wherever it is for sale, and I did the buying for years as a publishing company’s general counsel. A signed release converts a legal analysis, strong but still expensive to defend, into a filed document that rarely needs defending. The request that read as bureaucracy was the house pricing one specific person’s likelihood of complaining.

Who gets asked, and why one manuscript can need three releases and zero name changes, follows a factor list I unpacked in Why you can write about some people and not others.

What did the refusal actually change?

Two things changed, and the legal standard on your passage is neither of them. The shortcut closed, and the subject started paying attention.

Every rule that governed the passage before you asked still governs it: the fault standards, the privacy analysis, the identification question. What the refusal removed is slack. A released passage can afford a loose edge; its claims were signed away in advance. An unreleased passage is carried by evidence, and evidence has no patience for a sentence that says even a little more than the file supports.

What are your options when the answer is no?

Five options, and they are not priced alike. Ranked by how much of the passage survives: narrow and attribute, de-identify, strengthen the file and proceed, cut, or escalate for a priced decision your publisher accepts in writing.

Each cell states a tendency rather than a promise, and the rows combine: most passages that survived a refusal in my review years used the first two rows together. Which mix fits turns on who the person is, what the passage claims, and the law of the states in play.

Renaming, the option authors reach for first and trust most, is the weakest form of de-identification: the subject’s own circle, where claims actually come from, reads straight past a new name. What identifiability turns on, and when it cannot be reduced at all, is the subject of Does changing the name actually protect you?

How do you make the passage stand on evidence?

Match every factual claim to its proof, and change every sentence that claims more than the proof shows.

The reason evidence carries unreleased passages sits in the fault rules. Under New York Times Co. v. Sullivan, 376 U.S. 254 (1964), a public official suing over a defamatory falsehood about official conduct must prove actual malice, knowledge of falsity or reckless disregard of the truth, a standard later extended to public figures. For private individuals, Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974), lets each state define its own fault standard so long as it does not impose liability without fault, and negligence is the common choice. Read both as a working writer and they converge on one sentence: how carefully you worked is legally relevant, and the file is what careful looks like from the outside.

So build the match. List every factual claim the passage makes about the person, one line each, and beside it the proof: source, date, form, where it lives. Each claim lands in one of three bins. Proven as written. Provable if narrowed. Not provable by you. The first stands, the second gets rewritten down to its record, and the third gets attributed to a source who can carry it, or it goes.

Use attribution as a precision tool rather than a shield. The Texas Supreme Court has described it as well settled that one is liable for republishing the defamatory statement of another, softened by the constitutional fault requirements and a family of privileges and defenses. Neely v. Wilson, 418 S.W.3d 52 (Tex. 2013). A sourced accusation still needs a source worth standing on. What attribution buys is exactness: the sentence now asserts that the report said the thing, rather than the thing.

"A release is a shortcut, and shortcuts close. The passages that survived refusals in my review years were the ones that stopped arguing and started citing: what the documents show, said plainly, at the length the evidence could carry."

Guy Muller, former general counsel of a US publishing company with global reach

Does asking and being refused make things worse?

No. A release request is not an admission that the passage needed one, and in my own review years no flag ever turned on the fact that an author had asked and been told no. The passage is judged on what it says and what supports it, as if the email had never been sent.

Where authors hurt themselves is after the no. Three habits, starting today. Keep your account of the exchange factual, written, and dated. Stop re-approaching; a second plea reads as pressure and adds nothing your file can use. And write nothing to the subject you would mind hearing read back later: no apologies, no promises about revisions, no previews. Further contact goes through counsel.

What should you do in the week after the refusal?

Five steps, in order. The first four are yours alone, and the fifth protects the passage.

1. Log the exchange. The same day. What you asked, what they answered, dates, channel, exact words, stored with the manuscript. No commentary, because the log may one day be read by people you have not met.
2. Re-scope the passage to documented claims. The claim-and-proof match above, applied line by line, every unsupported sentence narrowed, attributed, or cut.
3. Run the circle test on identifiers. List what the subject’s five closest readers would still recognize, then change or generalize until the list empties, or accept on paper that it will not, and plan on evidence instead.
4. Assemble the evidence file as a handable object. One folder, one index: each claim, its source, its date, where it lives. Built to hand to an editor or a lawyer without a walkthrough.
5. Tell your publisher, with the plan attached. The revised passage, the evidence index, and a one-paragraph factual account of the request and the refusal. The report converts the refusal from a landmine into a line item.

In a memoir project, this file joins the review letter and the insurance conversation in the package that makes the book acquirable; that machinery is in Why memoir is the riskiest category in publishing, and how to de-risk yours.

Can they actually stop the book?

A threat can raise the price of your passage. What it cannot easily buy is a court order stopping the presses, and the distance between those two is where most pre-publication panic lives.

American courts start from a heavy tilt against suppressing publication in advance. Striking down a state statute that let officials shut a newspaper as a public nuisance, the Supreme Court in Near v. Minnesota, 283 U.S. 697 (1931), described preventing previous restraints as the chief purpose of the constitutional press guaranty and confined the exceptions to exceptional cases, wartime military information among them. The same opinion left the second door open: publish, and you can be made to answer afterward. In my experience, the subject who threatens before publication is bargaining about that second door, loudly.

A pre-publication demand letter mostly moves work to the publisher’s side: counsel answers it, the passage gets re-read against its file, and the schedule holds or slips on the probability-and-cost math in “I don’t care if I get sued.” Your publisher does. Your job the day it arrives is speed. Your publisher hears it from you, immediately, with the evidence file attached.

---