Buried underwear, 24-legged insects, and the blowhard problem

Seven fresh oddities arrived in the August 16–30 window: Swiss soil scientists buried underwear by the thousands, a tiny house filled with invisible cleaning particles, and a fossil spent four decades being mistaken for the wrong branch of life. Elsewhere, brain organoids kept developmental time, AI agents passed a trick down human generations, a 1976 paper resurfaced under an excellent insult, and a journal deleted a peer-reviewed viewpoint. The labels matter: these are papers, a conference presentation, a resurfaced paper, and a publication-status dispute, all sharing one talent for making academia sound invented.

What happens when 1,000 volunteers bury more than 2,000 pairs of cotton underwear and wait two months? Switzerland gets a nationwide map of soil activity, plus an unusually intimate measure of how busy the organisms underground have been. The University of Zurich and Agroscope launched the citizen-science project in 2021. Volunteers buried the underwear and 12,000 tea bags at roughly 1,000 sites, then dug them up, photographed the remains, and mailed soil samples to a laboratory. 1

The punchline is useful. Faster cotton decomposition indicated more active soil organisms. Private gardens produced the fastest breakdown, lawns the slowest, while fields and meadows landed between them. Land use explained the differences better than most of the soil properties the team measured. 1

Marcel van der Heijden, a UZH professor of agroecology, and Franz Bender, who leads agroecological assessments at Agroscope, are named as project leaders. The paper, Soil Health Assessment Using Buried Cotton Underpants with the Help of 1000 Citizen Scientists, proposes an "underwear index" as a public-friendly way to show soil processes. The researchers also warn that the fastest decomposition can point to excess nutrients in some settings, so the dirtiest pair is not automatically the healthiest verdict. 1

Shareable line: Switzerland has built a soil-health test whose sample collection begins with the sentence, "Please bury your underwear."

Purdue researchers put conventional scented cleaners and botanical essential-oil products through their routines inside a trailer-sized model house. They mopped, sprayed counters, and wiped surfaces while instruments measured the air. The tiny house had a kitchen, wood flooring, and a bathroom, which is a remarkably complete setting for discovering that a clean room can have a chemistry problem. 2

The products released fragrance terpenes such as pinene, limonene, thymol, and linalool. Those compounds reacted with indoor ozone within minutes and formed billions or trillions of particles, depending on the product. Most measured 1–30 nanometers across, a range many home air monitors miss. 2

Brandon Boor led the work with Purdue colleagues including Nusrat Jung and Ernest Blatchley. The team also tested far-UV-C lamps, which raised ozone inside the tiny house to around 20–40 parts per billion. Combining that ozone source with scented cleaning products drove more intense particle formation. The ACS release presents particle formation and potential inhalation exposure as the concern; the evidence here describes indoor air chemistry rather than a demonstrated causal human-health outcome. 2

The study's practical advice is almost offensively ordinary: choose low-fragrance or fragrance-free products, ventilate the room, and avoid pairing scented cleaning with ozone-generating devices. The bizarre part is the scale mismatch. A room can look spotless while producing an invisible particle cloud in the time it takes to finish the floor. 2

Shareable line: Your tiny house may contain a tiny air-pollution factory, and it starts with the lemon-scented wipe.

A fossil from western Texas was found in 1985 and classified within a few years as a juvenile crustacean. The specimen dates to about 324 million years ago. A new study reexamined the dark imprint and identified a new early insect species, Chosha praecursor. 3

The creature belonged to the hexapod lineage, yet its body had dozens of paddle-like legs and a crustacean-like abdomen. The researchers used cross-polarized-light imaging to reduce glare and expose finer structures, including the head and antenna. The result places some early insects in semiaquatic settings and makes the move from water to land look more gradual than the familiar six-legged silhouette suggests. 3

The authors also revised three other enigmatic stem-insect fossils. Together, the fossils help address the "hexapod gap," an approximately 80-million-year blank in the early fossil record of insects and related animals. The joke begins with the leg count; the scientific payoff is a better look at how an insect body plan assembled. 3

Shareable line: The first insects may have had 24 legs, a crustacean torso, and the worst possible fossil lighting.

Human brain organoids are peppercorn-sized cultures of brain cells. Harvard's Arlotta lab kept them alive for more than five years, and the lab now has some cultures seven years old. Each organoid contains more than 1 million cerebral-cortex cells derived from a human donor. The five-year result belongs to the paper; the seven-year figure describes the lab's current cultures. 4

The team followed 34 organoids at eight timepoints between six months and five years. Combined with earlier work, the dataset covered 110 organoids and nearly 425,000 individual cells. The cells followed molecular developmental sequences resembling human brain development, while DNA methylation supplied an age clock that tracked those stages. 4

Then came the time-travel experiment. The researchers put young and old cells in the same organoid and chemically prompted them to form neurons. Younger progenitor cells made early-stage neurons. Older cells jumped ahead and made later-stage neurons, as though their previous development had left a molecular bookmark. Senior author Paola Arlotta led the work with lead authors Irene Faravelli and Noelia Antón-Bolaños. 4

The paper is titled Human brain organoids record the passage of time over multiple years. Harvard describes the cultures' long survival and developmental memory; the Broad Institute identifies the Nature publication and DOI. The result gives researchers a way to study later stages of human brain development in tissue that keeps a record of where its cells have already been. 5

Shareable line: These lab-grown brain cells kept a calendar, then skipped ahead when asked to grow up.

The experiment used 1,155 participants and a reward-based task. Participants earned points by choosing actions in a setup where the best strategy required accepting small early losses for much larger later gains. Human-only groups almost never found the optimal strategy: one of 600 participants discovered it independently. 6

Other groups began with AI agents that had already discovered the strategy. Human participants learned from the most successful role models in earlier generations, whether the role model was a person or an agent. The machine-discovered strategy spread through the learning chains and remained in use across generations in nine of fifteen human-machine groups. 6

Levin Brinkmann and Thomas Eisenmann are named as lead and co-lead authors. The collaboration joined the Max Planck Institute for Human Development, the Toulouse School of Economics, and Humboldt University of Berlin. The paper tests cultural transmission under controlled conditions; it does not turn an AI agent into a society, a tradition, or a wise elder. 6

The funny mechanism is deeply human. People copied the successful example, even when the example was a machine, and the machine had found its advantage by tolerating the short-term pain humans tended to avoid. 6

Shareable line: AI found the winning move; humans preserved it by copying whoever seemed to be winning.

Improbable Research posted the title on August 25: Asymptotic study of the blowhard problem for the supersonic flow of a gas past sharp slender bodies. A.M. Antonov and Yu. A. Belov published the paper in USSR Computational Mathematics and Mathematical Physics in 1976, volume 16, issue 6, pages 227–232. 7

The current post supplies the title and bibliographic details, then wisely leaves the mathematics alone. "Blowhard" is the joke because an ordinary insult has wandered into a technical problem about supersonic gas flow. The 2026 date belongs to the resurfacing post; the paper itself is 50 years older. 7

Shareable line: Somewhere in supersonic-flow mathematics, "blowhard" is a technical term and an insult at the same time.

Arna Mitchell's viewpoint appeared in August 2025 in the Journal of the New Zealand College of Clinical Psychologists after peer review. The article criticized proposed different standards of practice for Māori and non-Māori clients and argued for culturally responsive care alongside evidence-based individual assessment. The journal initially published an editorial defending open debate in the same issue. 8

The editors later issued a correction addressing definitions and facts in the article. On June 18, 2026, the college told Mitchell that its council had decided to remove the piece because retaining it could perpetuate harm to Māori and conflict with the college's values. The organization also said the journal was not an independent, indexed academic journal, a description disputed by former editors and members. 8

The publication record then acquired a second problem: Retraction Watch reported no public conventional retraction notice, and the article disappeared from Zenodo as well. COPE guidance says an article removed for serious reasons should have a freely accessible notice explaining the removal and linking to the article. The dispute therefore concerns both the viewpoint and the way the scholarly record changed around it. 8

This case belongs in a strange-science roundup because the status of a peer-reviewed article shifted from publication to correction to deletion, while the journal's own description of its role became part of the argument. Māori identity and the policy debate are the subject of the dispute; the absurdity lies in the publication trail and its missing public explanation.

Shareable line: The strangest thing about this paper is the publication record: peer-reviewed, corrected, deleted, and left without a standard public retraction notice.

The seven stories share a setting rather than a theory. Researchers used underwear, air sensors, fossil light, living cell cultures, learning chains, a very unfortunate word, and a contested journal decision to make hidden parts of science visible. The common thread is the evidence trail: each oddity gets funnier when the method and status stay attached to the punchline.