People Who Changed How Science Records and Shares Knowledge

Conrad Gessner, a Zurich physician, compiled the Bibliotheca universalis in 1545, an attempt to list every book known in Latin, Greek and Hebrew with summaries, often called the start of modern bibliography. His illustrated Historiae animalium, from 1551, gathered observations from correspondents across Europe. Both showed how knowledge could be collected from many hands and organised for others to search.

Marin Mersenne, a French friar, ran a letter network from his Paris convent in the 1620s to 1640s that linked Descartes, Fermat, Pascal, Galileo's supporters and many others, copying and forwarding results and problems between them. Before scientific journals existed, his correspondence acted as a clearing house for new work. The meetings around him helped lead to the founding of the French Academy of Sciences in 1666.

Henry Oldenburg, the first secretary of the Royal Society, launched Philosophical Transactions in 1665, the longest-running scientific journal. He printed letters and reports from correspondents across Europe, dated their arrival to settle questions of priority, and began sending submissions to knowledgeable fellows for comment, an early form of peer review. His network included Leeuwenhoek's microscope letters.

Carl Linnaeus introduced consistent two-part Latin names for species in Species Plantarum in 1753, and for animals in the tenth edition of Systema Naturae in 1758, replacing long descriptive phrases. His students, sent around the world as collectors, applied the system as they went. It gave scientists everywhere a shared reference for the same organism, and those books remain the formal starting points for naming plants and animals.

Denis Diderot edited the Encyclopédie from 1751, at first with the mathematician Jean le Rond d'Alembert, gathering articles from well over a hundred contributors and hundreds of detailed plates of trades and machines. Cross-references linked related entries into a web of knowledge. Despite bans by the French authorities, it was completed in 1772 and spread technical and scientific knowledge to a wide readership.

Vannevar Bush, who ran American wartime science, warned in his 1945 essay As We May Think that researchers were drowning in publications, and imagined the "memex", a desk that stored documents and linked them by associative trails. The idea inspired later pioneers of hypertext such as Douglas Engelbart and Ted Nelson. His report Science, the Endless Frontier, the same year, shaped how governments fund research.

Eugene Garfield proposed a citation index for science in 1955 and launched the Science Citation Index in 1964 through his Institute for Scientific Information, letting researchers trace which later papers cited an earlier one. He also devised the journal impact factor to help libraries choose journals, and later warned against its misuse to judge individual scientists. Citation databases such as Web of Science descend from his work.

Margaret Oakley Dayhoff published the Atlas of Protein Sequence and Structure from 1965, collecting every known protein sequence in one book and updating it regularly. She devised the one-letter amino acid codes still used today and scoring matrices for comparing sequences. Her Atlas was a forerunner of the protein databases and sequence-search tools that underpin modern biology, and she is often called a founder of bioinformatics.

Helen M. Berman was among the crystallographers who campaigned for the Protein Data Bank, set up at Brookhaven in 1971 with Walter Hamilton and others to archive protein structures, and from 1998 she led it at Rutgers. She pushed for journals and funders to require structure data to be deposited and for common standards. The open archive, now with over 200,000 structures, trained the AlphaFold system that predicts protein shapes.

Donald Knuth was so unhappy with the typesetting of the second edition of his book The Art of Computer Programming that in 1977 he began writing TeX, a typesetting system for mathematics, and released it free. Leslie Lamport's LaTeX, built on top of it in the 1980s, made it easier to use. TeX and LaTeX became the standard way mathematicians, physicists and computer scientists write and share papers.

Tim Berners-Lee, working at CERN, proposed in 1989 an information system so physicists could share documents across different computers, and built the first web browser and server with Robert Cailliau's help in 1990. CERN released the technology royalty-free in 1993, which let it spread far beyond physics. He founded the World Wide Web Consortium in 1994 to keep web standards open, and campaigns for people's control of their own data.

Paul Ginsparg, a physicist at Los Alamos, set up an automated email server in 1991 for colleagues to share preprints of high-energy physics papers before journal publication. It grew into arXiv, now run by Cornell University and covering physics, mathematics, computer science and more. Posting there first became normal in those fields, and its model inspired preprint servers in biology and medicine.

John Sulston, head of the Sanger Centre near Cambridge, helped draft the Bermuda Principles in 1996 with Robert Waterston and others, committing the Human Genome Project's labs to post new DNA sequence publicly within 24 hours. The rule kept the human genome freely available when a private company sought to patent and sell access. Rapid public release of data became a norm in genomics.

Brewster Kahle, a computer engineer, founded the Internet Archive in 1996 to save copies of the web, and its Wayback Machine now lets researchers cite and check pages that have since changed or vanished. The archive also scans books, journals and scientific literature, and its Fatcat project preserves open-access papers whose publishers have disappeared. Court cases over its book lending have tested how far such libraries can go.

Harold Varmus, as director of the US National Institutes of Health, proposed in 1999 a free online archive of biomedical papers, which opened in 2000 as PubMed Central. With Patrick Brown and Michael Eisen he then co-founded the Public Library of Science, whose open-access journals, from PLOS Biology in 2003, showed that research could be free to read and paid for by publication fees. NIH later required funded papers to be deposited openly.

Jimmy Wales launched Wikipedia in 2001 with Larry Sanger, after their expert-written encyclopedia Nupedia stalled, letting anyone edit and requiring claims to be backed by published sources. Run since 2003 by the non-profit Wikimedia Foundation and written by volunteers, it became one of the most consulted reference works in the world, and its sister projects Wikidata and Wikimedia Commons supply open data and images used across science and education.

Ben Goldacre, a doctor and epidemiologist, co-founded the AllTrials campaign in 2013 with Sense About Science and others, demanding that every clinical trial be registered and its results reported, because unpublished negative results distort what doctors know about treatments. His team's TrialsTracker and COMPare projects checked whether trials met those rules. He now leads the Bennett Institute for Applied Data Science at Oxford, which built OpenSAFELY for secure analysis of health records.

Brian Nosek, a psychologist at the University of Virginia, co-founded the Center for Open Science in 2013 with Jeffrey Spies and built the Open Science Framework, where researchers can preregister plans and share data and code. He coordinated the Reproducibility Project, whose 2015 report found that fewer than half of 100 psychology findings replicated clearly. Preregistration and registered reports have since spread across many fields.