Made the same way every time: read across language editions, give every entry a source, leave disagreements standing as disagreements, let the evidence set the time axis rather than the other way round, and publish what is missing beside the plate.
Timeline plate · learning
education — India and the world
40 entries · 8 open questions · 8 named gaps · 12 disagreements, 2686 BCE to 2020. The argument over the world's oldest university is carried as content rather than settled: English hedges, Hindi asserts Nalanda, Arabic raises al-Zaytuna. The largest gap we name is exclusion — this method records when things were founded, not who was kept out.
deep-timetimelineeducation
Timeline plate · learning
vital essence
29 entries, weighted toward the modern end — humoral and doshic systems, through vitalism, to modern biochemistry. The first edition was too hard to read, and that failure rewrote the method: densities worked out from the data, labels packed to fit, and living practice given its own strip.
deep-timetimelinemedicine
Timeline plate · learning
Indus script
Four thousand objects, four hundred signs, and the longest text is thirty-four characters. Edition 2 runs about 30 entries from Mehrgarh to transformer models, keeps the decipherment argument live in both directions, and lists nine open questions and nine named gaps.
deep-timetimelinescript
Timeline plate · skill
knock
The sound a fuel makes when asked to do more than it can. Two kinds of stored sunlight — one buried a hundred million years, one fermented last season — meeting inside one cylinder, and a hundred-year argument about it. WCAG 2.1 AA, and every entry is also published as plain running text.
deep-timetimelineenergy
Timeline plate · culture
the moving image — storytelling in light
Lascaux's flicker-animated deer, the camera obscura from Mozi to Ibn al-Haytham, wayang's 840 CE inscription, then the magic lantern through to Toy Story. Katha and living wayang appear as presence rather than dated points. Generative video sits on the horizon as a position we are taking, not a prediction.
deep-timetimelinefilm