The detailed lists behind the report: the open questions science writers raise, the people and books they discuss, the papers they link to, the subfields and most common topics within each field, and the animals and diseases they mention. Free English science posts on Substack, 1 October 2023 – 30 September 2026: 47,562 posts from 2,785 newsletters. Year 1 = Oct 2023–Sep 2024, Year 3 = Oct 2025–Sep 2026.
From every science post I extracted the open questions it raises: the unknowns and debates the writer points to. Near-identical wordings are merged into one question (the wording shown is a standardized version). Listed: questions asked by at least three different newsletters. The bars show posts asking in Year 1, 2 and 3.
| Question |
|---|
Questions grouped into 41 recurring topics with published keyword rules (samples checked by hand). "Newsletters" counts different newsletters asking about the topic over three years. The yearly figures are posts asking per 1,000 science posts: science writing doubled over the period, so a rising number means the topic grew faster than science writing as a whole.
Everyone discussed substantively (not just name-checked) in 25 or more science posts, by at least three different newsletters. For each mention I recorded whether the post endorsed the person's work, critiqued their work (methods, claims, findings) or criticized their conduct. Critique of someone's work is a normal part of science: a high share means the work is argued over, not that it is wrong. Roles are as classified from the posts.
Books named in science posts, matched by title and author and checked by hand, ranked by the number of different newsletters discussing them. (The psychiatric manual DSM, discussed by 41 newsletters, is left out as a reference work.)
Research papers linked from science posts, counted by the number of different newsletters linking to the paper itself. Papers cited without a link are missed, so this favors papers that are freely available online.
| Paper | Newsletters |
|---|
Every science post was given one main subfield for each of its fields, from a fixed list per field. Shares are of the field's posts. Subfields under 2% and "other" are omitted.
| Subfield | Share (3 years) | Year 1 | Year 2 | Year 3 |
|---|
Each post lists the specific topics it is about (e.g. "long COVID", "dark matter"). These are the most common in the selected field, with the share of the field's posts that list them. Shares are small because topics are specific.
Every science post was searched for a curated list of animals and of diseases and conditions, matched as whole words (ambiguous words such as "bear" or "fly" were left out or only counted in unambiguous forms). A post counts once per animal or disease, wherever in the post it is mentioned, so these count what writers bring up, not what posts are mainly about. The yearly figures are posts mentioning per 1,000 science posts.
The rises in wolves, mammoths and Neanderthals come mostly from one prolific newsletter on evolution and prehistory.