Margaret Hollis
The Contracts Editor
22 years negotiating magazine contracts
"I've reviewed contracts for everyone from first-time contributors to staff writers at the top five science publications. The clauses that hurt writers most are the ones that sound routine — until they're not."
Specialises in:
Model Contracts & Rights Transfers
Three questions every science writer should be able to answer before signing anything. These are ungated — read them all.
An "all rights" clause transfers every form of use — print, digital, broadcast, translation, derivative works, and any medium not yet invented — to the publication, permanently, worldwide. Your name stays on the piece, but you lose the ability to reprint it in a collection, sell a foreign edition, or license it for a documentary without the publisher's permission (and usually a fee back to them).
Should you sign? Sometimes, if the fee is large enough to compensate for the foreclosed future income. The calculus changes if the piece has anthology potential (longform narrative science often does), if you're building a book, or if the topic is evergreen. The rule of thumb: all rights for a day-rate piece at a trade publication is reasonable; all rights for a 7,000-word New Yorker-style feature is a negotiation, not a given.
Work-for-hire is actually worse than all rights for a freelancer. Under U.S. copyright law (17 U.S.C. § 101), a work-for-hire arrangement means the publication, not you, is legally the author. There's no termination right, no reversion clause, and no moral rights argument to fall back on. The copyright clock starts running for the publisher's benefit, not yours.
The exception: if you're a staff employee writing within the scope of your employment, work-for-hire is expected and appropriate. As a freelancer, it's almost never in your interest to agree to it — even at a higher rate — unless the piece is genuinely disposable (a product description, a short news brief) and you have no plans to reuse it.
Flag for negotiation: Ask to replace "work made for hire" with "first North American rights with a 12-month exclusivity window." If they refuse, that tells you something about how they view the relationship.
A kill fee is the compensation owed when a publication assigns a piece and then cancels it — through no fault of yours — before publication. It's the floor, not the ceiling: you keep the kill fee and the right to sell the piece elsewhere, unless your contract says otherwise (check that clause).
Standard rates vary by publication tier. Legacy national magazines typically offer 25–33% of the contracted fee. Digital-first outlets often try 10–15%, which is negotiable upward. Anything below 20% for a reported piece is below-market and worth pushing back on at contract stage, not after the kill.
When they refuse to pay: send a formal demand letter citing the contract clause, give a 14-day cure period, then escalate to the NWU Grievance Committee or a media-focused attorney. The letter template below handles the first step.
The full Model Contract template — with every clause pre-annotated — is in the Writer's Shield Kit.
Get the KitThe One-Page Kill Fee
Demand Letter
A battle-tested template used by working journalists to recover kill fees that publications "forgot" to pay. Fill in four fields, attach your contract, send by certified mail. It works.
[Your Name] | [Date]
Re: Unpaid Kill Fee — Assignment Dated [DATE]
Dear [Editor's Name],
On [DATE], I accepted your assignment for "[ARTICLE TITLE]" at an agreed fee of $[AMOUNT]. Per Section [X] of our signed agreement, a kill fee of [X]% ($[AMOUNT]) is due within [X] days of cancellation…
← Full template continues for 3 more paragraphs with legal citation
Get the full template
Enter your email and we'll send the complete, editable version — no strings, no upsell.
Used by journalists at Science, Nature News, STAT, and dozens of independent science publications to recover fees ranging from $200 to $12,000.
Copyright & Open-Access Licensing
For staff writers navigating institutional mandates, and freelancers selling into multi-territory markets.
It depends on who wrote what. If the piece you're filing for a freelance outlet was written independently of your university role — on your own time, without institutional resources — then the university's open-access mandate doesn't apply, and you hold copyright as the author.
The complication arises when you're writing a piece for a science magazine that draws heavily on a report you wrote in your staff capacity. In that case, you may have a split-ownership situation: the institutional report is covered by CC BY 4.0, but your freelance article — even if it covers the same ground — is a new original work and is yours to license separately.
Multi-territory note: In the UK and EU, moral rights (including attribution and integrity rights) cannot be waived by contract for employed authors. This provides a floor of protection even under broad licensing agreements.
You're giving up the ability to sell translation rights — one of the most reliable secondary income streams for longform science writers, particularly those covering health, climate, and physics, where there are strong German, French, Spanish, and Japanese markets.
A standard alternative: grant first North American serial rights, plus non-exclusive worldwide rights in English only, with a 90-day exclusivity window. This preserves your translation rights entirely and is accepted by most mid-tier and above publications.
If they insist on worldwide rights, request a per-language license fee schedule in the contract — this converts their broad grant into a revenue-sharing arrangement rather than an outright transfer.
Only if your contract grants you reprint rights, or if you retained copyright and granted only first serial rights. If you signed away all rights or signed a work-for-hire agreement, you technically need the publisher's permission — and potentially a license fee — to include the piece in a collection.
The practical reality: many publications will grant reprint permission for book anthologies at no fee, particularly for non-competing markets. The key is to ask before you sign the book deal, not after, because your book publisher will require you to clear all permissions before the manuscript is accepted.
The Copyright Checklist — a one-page audit for every submission — is included in the Shield Kit.
Get the Kit
Dr. James Okonkwo
The IP Attorney
Media & publishing law, 14 years
"Open-access mandates have created a generation of writers who think their institutional employer owns everything they write. Usually, they don't. But you need to know exactly where the line is."
Specialises in:
Ray Delacroix
The Veteran Freelancer
31 years, science & environment beat
"I've had my byline stripped, my conclusions rewritten, my archive scraped, and my kill fee withheld. I've recovered from all of it. The difference every time was knowing exactly what the contract said."
Specialises in:
Byline Disputes & Scraping Defense
For when the piece is published and something has already gone wrong. These are the conversations no one told you to prepare for.
This is a byline integrity dispute, and it's more common in science journalism than anywhere else — editors without domain expertise sometimes "simplify" conclusions in ways that misrepresent the underlying research. You have several escalating options.
Step 1: Request a correction or rewrite before publication. Frame it as a factual accuracy issue, not a personal grievance — editors respond better to "this will create a correction risk" than "you changed my meaning."
Step 2: If published without your consent: request a formal correction, ask to have your byline removed (this is your right under most contracts), or publish a public note on your own channels clarifying your original conclusions.
Step 3: For systematic issues, file a complaint with the Society of Professional Journalists or the relevant science journalism organization. A paper trail matters.
If you hold copyright (you retained rights, or the original publication's rights have reverted to you), you have a clear legal path: file a DMCA takedown notice with the hosting provider. This is free, requires no attorney, and has a mandatory response window of 10–14 business days.
The process: identify the hosting provider (use a WHOIS lookup), find their DMCA agent (required by law to be publicly listed), send a compliant notice citing the original URL and the infringing URL, and request removal. Most scrapers fold immediately — they're not in the business of defending copyright claims.
If the scraper is in a jurisdiction with no DMCA equivalent (common for Eastern European and Southeast Asian content farms), you can still contact Google to deindex the infringing pages, which effectively kills their traffic.
If you don't hold copyright: You still have moral rights in many jurisdictions, and you can report the scraper to the original publication — they have both the rights and the incentive to act.
Almost certainly not explicitly — but whether they need your consent depends on the rights you granted. If your contract included "all rights" or "all electronic and digital rights," courts in the U.S. have generally allowed publishers to sublicense to AI companies under those broad grants, because AI training was not explicitly excluded.
The counterargument — currently being litigated — is that AI training is a transformative use that requires a new license, and that broad rights grants from 2005 could not have contemplated 2024 generative AI. This argument has traction in the EU under GDPR and the AI Act, less so currently in the U.S.
Practical steps: review your original contract, send a written objection to the new owner citing your specific rights reservation, and register with the Authors Registry or similar opt-out mechanisms. The NWU is currently running a class action on this issue — consider joining as a named or unnamed plaintiff.