Most research is written as if the first job of a reader were to prove that they already belong. The Paper proposes the opposite: every serious scientific publication should first explain itself fully to the public, then to an informed reader, and only then ask anyone to enter the technical record.
Scientists usually publish one specialist version of their work. The Paper proposes three connected versions of the same research: a public story, a deeper explanation of how we know, and the exact technical record. You can also choose simpler or more advanced wording without changing the science.
Use the explanation buttons above. This paragraph changes; its meaning does not.
A scientific result can be explained with ordinary words first, then connected to the methods, numbers and evidence that support it.
Very simple · ELI5 Scientists can tell you what they found in easy words first. You can still look at all the hard details if you want.
Clear · ELI12 A science result can start with a clear explanation, while keeping links to the methods, numbers and evidence underneath.
Young adult · ELI18 Scientific findings can be presented in accessible language without removing the evidence or uncertainty behind them.
General A scientific result can be explained with ordinary words first, then connected to the methods, numbers and evidence that support it.
The web changed. The paper mostly did not.
Imagine that newspapers had invented websites, searchable archives, permanent links and live data — and then decided that every story should still be written like an internal legal memo.
That is roughly where scientific publishing landed. The network changed. Search changed. Citation linking changed. Data became digital. Experiments became computational. Figures could move. Readers could click, run, compare and inspect. But the main object offered to a human being often remained a document whose shape had been optimized for professional exchange among specialists.
This is not because scientific papers have always looked this way. They have not. The earliest Philosophical Transactions in 1665 mixed letters, book reports, observations, summaries and experiments. The familiar Introduction–Methods–Results–Discussion structure only became dominant in major medical journals in the second half of the twentieth century. The format changed when institutions needed it to change.
So why should a reader in 2026 have to enter science through a form that assumes the reader is already a scientist?
The usual answer is precision. Technical subjects require technical language. That part is true. The hidden leap is the idea that precision therefore has to be the front door.
A wiring diagram can be precise. So can a repair manual. So can the cockpit checklist of an airliner. None of them needs to begin by making a non-specialist feel that the subject has been deliberately placed out of reach.
Much of science is paid for, protected or used by the public. The public explanation should not arrive last.
Much science is publicly financed, publicly protected, publicly taught, publicly regulated, or eventually applied to public life. Yet the public-facing explanation often arrives last: a press release, a university news post, a journalist’s reconstruction, or nothing at all.
That arrangement quietly says something about who the “real” reader is. The technical paper counts. The explanation is outreach.
The Paper reverses that priority without deleting technical work. A public article must be a complete account of the research at the level appropriate to a general reader. It tells the reader what question mattered, what was done, what changed, what did not change, what remains uncertain, and what the result does and does not justify.
The public article is not a press release. It may contain bad news, null results, caveats, disagreement, failed attempts and boring but consequential details. Its job is not to promote the work. Its job is to make the work available to the people outside the profession.
If an educated non-specialist cannot work out what a publication is about, why it matters, what happened and what remains uncertain without first learning the authors’ professional dialect, the publication is unfinished.
Some complexity belongs to the subject. That complexity stays. Complexity added by exposition is different. It is a design defect.
One piece of research. Three connected entrances.
Start with a person, a problem, an object, a decision or a surprise. Define terms as they become necessary. Explain the result in ordinary language without hiding uncertainty.
For the reader who now wants to know how we know. Bring in method choices, alternative explanations, prior work, disagreements, uncertainty and enough detail to challenge the interpretation.
The technical record keeps the exact detail science needs: methods, equations, data, code, identifiers, references, limitations, versions, reviews and correction history.
Yes. Their successes and failures tell us what not to repeat.
Distill showed what interactive scientific explanation could look like. Its editors later put the journal on indefinite hiatus, citing volunteer burnout, journal overhead, author effort and the rare mix of scientific and design skill required.
What we learn →Lens let readers keep figures and references beside the text instead of jumping around a PDF. It survived as a useful viewing idea — but the scientific article underneath remained recognizably the same object.
What we learn →The EU-funded LiquidPublication project proposed scientific objects that could evolve, combine and receive distributed credit. The project ended in 2011; the paper-centered ecosystem did not.
See the full field →The lesson is not that innovation “failed.” Some ideas were absorbed. Some remain active. Some solved infrastructure instead of exposition. Some discovered that incentives can defeat an excellent interface.
The Paper therefore does not begin by inventing another clever document widget. It begins with a publishing obligation: the same research must be genuinely readable at more than one depth. Presentation technology comes second.
It asks what actually travels from one paper into another.
A citation is usually drawn as a line: Paper B points to Paper A. But that line hides almost everything interesting. Did B inherit a result? A measuring tool? A definition? A dataset? A method? A famous mistake? A sentence copied from a review?
The Lineage Project treats science as a history of transmission. It follows chains of papers and asks what survived each step. Retractions make this especially visible because the status of an ancestor changes on a known date. We can then watch which parts of the ancestor continue downstream.
That project produced the first working Transductive publication at Retractions, Unpacked: a public story, an interactive evidence layer and a reproducible technical record. The Paper generalizes the publishing lesson.
A scientific publication should not merely store what the authors know. It should make the path into that knowledge traversable.
No. Precision stays; being the only serious entrance does not.
The classical paper still does several jobs well. It freezes a dated account of what authors claim. It gives reviewers a common structure. It lets specialists scan familiar sections. It can be archived and cited. None of those functions require it to be the only serious representation of the work.
In the Transductive format, the technical paper becomes the deepest layer of a larger publication. Its equations and exact terminology remain available. But every important conclusion has a path upward to an explanation, and every simplified statement has a path downward to evidence.
The result should feel less like choosing between “popular science” and “real science,” and more like choosing how far into the same building you want to walk.
An author begins by writing the human question before the abstract. A reviewer checks whether the public layer accurately reflects the technical result. A figure is designed once, then given explanatory and technical views. A correction updates all three layers. A reader can move downward without losing their place.
A funder can ask for public intelligibility as a deliverable rather than funding communication as a later add-on. A journal can publish the three layers under one research identity. A citizen can read the work without first borrowing the language of the profession that produced it.
Serious science should have no single front door reserved for insiders.
A new format can create new problems. Those belong in the proposal too.
No. The proposal is one research record with several reader-facing versions. AI and editorial tools can lower the translation cost, but researchers remain responsible for meaning.
Yes. That is why claims, numbers, uncertainty and causal strength must stay synchronized. If one version changes, the others need review.
Keep it. Technical papers are good at precision, review and archiving. The proposal changes their role from the only serious doorway to the deepest doorway.