When the figure won't hold still: the quiet contest over who illustrates the science we read

On 17 July 2026 at 13:45 UTC, a short note from the editors at the Russian science publication N+1 went up on their Telegram channel and read, in effect, as an admission of guilt. The post opened with a confession familiar to anyone who has tried to popularise a paper: "It happens that a scientific article contains nothing but boring graphs and diagrams, but the news needs beautiful illustrations!" The line was self-deprecating, but it pointed at a working arrangement most science desks would prefer not to discuss in public, the regular borrowing of imagery from the visual arts to dress up findings that, on their own, look like a spreadsheet.

The N+1 post frames a quiet contest over who, exactly, gets to illustrate the science a general reader ends up seeing. It is a contest the publications themselves do not always win, because the work of translating a paper into a hero image is more often done by a freelance illustrator than by the lab that produced the underlying data. The result is a peculiar split-screen product: the technical figures inside the article remain the property of the authors, while the cover picture belongs, by default, to whoever could draw fastest.

What the journal owns, and what the website borrows

Inside a peer-reviewed paper, the figures are tightly bound. They are the authors' own; they are checked against the underlying data; and they reproduce badly outside the PDF they were designed for. A heat-map tuned for a journal's two-column layout will turn to mud on a phone screen. A forest plot dense enough to satisfy a reviewer is unreadable in a social card.

So when a science desk turns a paper into a news story, the figures get replaced. The N+1 note acknowledges the substitution without naming a culprit: "Sometimes in such cases we use works of art that are already in the public domain, or commissioned anew, or, and this happens most often, borrowed." The borrowing is rarely documented in the byline, which is why readers who click through from a striking illustration to the underlying article often find that the striking illustration is not, strictly, a picture of the science at all.

This is not unique to N+1. Science communication has long treated the hero image as a separate production task from the body of the work. The same paper can travel under a Renaissance painting in one outlet and a colourised microscopy plate in another, and both decisions will be defended as legitimate exercises in visual translation.

The case for borrowing, and the cost

The defence is straightforward. A graph of gene expression is correct; it is also, to a general reader, nearly invisible. A Brueghel sky above a story about planetary atmospheres is not strictly about planetary atmospheres, but it does something the graph cannot: it tells the reader that the writer takes the topic seriously enough to spend money on the page. News outlets that refuse the substitution tend to look threadbare against competitors who embrace it.

The cost is subtler, and it is what the N+1 post is really airing. Borrowed imagery imports the connotations of its original setting. A sky from a sixteenth-century Netherlandish painter carries the freight of religious painting; a photograph of a nineteenth-century laboratory carries the freight of a particular myth of discovery. Neither is neutral. Neither was made to illustrate the science. When the borrowings stack up across a feed, the cumulative impression is of science being narrated by pictures that the science did not earn.

There is also a quieter economics at work. Commissioned illustration costs money that freelance-friendly science desks do not have; public-domain art is free but finite, and the same Vermeer turns up on three competing explainers in the same week. The pressure to keep producing, with shrinking budgets and rising reader expectations, pushes editors toward the familiar compromise: a beautiful picture that is not, strictly, a picture of the thing.

A field without a name, and a credit it rarely gets

The people who do the actual work of bridging these two registers, research illustrators, science editors, freelance visual translators, are rarely credited under any of those titles. The journal article names the authors of the paper and, sometimes, the designer of the figure. The news story names the writer and the photographer, if there is one. The illustrator who turned the paper into something a reader would recognise sits somewhere in between, paid in flat fees and named, if at all, in a small line at the foot of a webpage.

This is what makes the N+1 note quietly radical. By framing the borrowing as a confession rather than a workflow, it treats the gap between journal figure and news illustration as a problem of credit and craft, not merely a problem of taste. If the picture is doing interpretive work, the post suggests, the picture should answer for it.

The implication for the wider science-publishing field is uncomfortable but clarifying. The reader who finishes an explainer and cannot tell what was drawn specifically for the story, what was borrowed from public-domain archives, and what was added by the original authors of the paper has been quietly steered. That steering is not necessarily dishonest. But it ought to be legible.

What changes if the credit gets written

The practical change would be small and overdue: a line at the end of every science story that names the illustrator, the source of the image, and the relationship between the image and the paper it accompanies. Commissioned, borrowed, adapted, derived, these are the four honest categories. Anything that cannot be placed in one of them is, by the standard N+1 is now implicitly proposing, a fifth.

It is not hard to imagine the resistance. Editors will say the reader does not care; publishers will say the column inches are not there; illustrators themselves will sometimes prefer to be paid rather than credited. But the reader does notice. The reader notices that the same picture appears under three different headlines in a week; the reader notices when the picture contradicts the article; the reader notices, without quite knowing what to call it, when the science feels illustrated from the outside.

N+1's small confession, posted to a Telegram channel in mid-July, does not solve the problem. It does, however, name it in a register most science desks still avoid. The figures inside the paper are accurate, the figures outside the paper are persuasive, and the distance between them is filled, increasingly, by people whose names do not appear anywhere on the page.

Desk note: where wire coverage of science tends to flatten the visual layer into a single credit line, Monexus treats the image as a separate editorial artefact with its own provenance. The N+1 post is the rare wire-level admission that the gap between journal figure and news illustration is a craft problem, not a taste problem.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/nplusone
• https://t.me/nplusone/12001