The time layer · standing up · 16 rival years

The Aquarium

Geography on the floor, time rising through the tank. Nine states of the Deccan and Bengal stop being lines on a chart and become solids you can walk around — and a date the sources cannot agree on stops being a footnote and becomes a length you can measure.

1950time rises13001565

drag to turn · pinch or scroll to zoom
red rings are years other books give — click one

1565Bijapur · Ahmadnagar · Berar · Bidar · Golconda · Vijayanagara · Bengal Sultanate

Nine solids

Nine columns in one frame is already a thicket, and the archive is going to keep growing. So the tank is scoped by region rather than thinned: show the Deccan, show Bengal, or hold both side by side to see how out of phase they are. Within a scope, click any column to pull it out of the crowd — the rest fall back to ghosts, its cloud thickens, and its rival-date rings brighten. Click any mark in the tank to read the event it stands for.

How to read the tank

The floor is the map, to scale: the same coordinates as the flat view and the same 32 places. Height is the year, 1300 at your feet and 1950 overhead. A state is therefore a column standing on its own capital — and on the day a capital moved, the column leans.

The leans are why it is worth turning. Golconda’s column is nearly plumb — eight kilometres of drift in a hundred and sixty-nine years, a dynasty that stayed put. Bengal’s kicks two hundred and ninety kilometres west in 1453, from Sonargaon to Gaur. The Nizams’ throws four hundred and thirty-one in forty years, Aurangabad to Hyderabad, and never settles again.

And the reason to build it at all: where the sources disagree about a year, the disagreement becomes a distance. The Barid Shahis of Bidar never held a coronation to date from, so five different published years mark their beginning — and here those five are rings you can count, spaced along the axis in proportion to how far apart the authorities actually are. On a flat map that is a footnote. Standing up, it has size, and you can see at a glance that the argument spans half a century.

What you are looking at

Solid column — the span this archive uses. Its height is the state’s life.
Fading end — where a state’s beginning or ending is argued about, the column frays across every year that is claimed for it.
Ringa year someone else publishes. A different authority dates the same beginning or ending to that height instead. Bidar has five rings round its base: five books, five answers, spread over half a century. Click one.
Soft cloud — ground near a capital. Where two clouds overlap they brighten, and that brightening is the only thing in the tank drawn twice.
Point — a dated event no source the archive opened disputes.
Vertical smear — a dated event printed several ways. Its height is the disagreement: Malik Ambar’s four death dates, measurable.
The pale sheet — the year on the slider, cutting through everything standing in it.
The nine states, their spans, and how many rival dates each carries
StateAdopted spanYearsRival dates published
Bahmani Sultanate134715271803
Bijapur — the Adil Shahis149016861961
Ahmadnagar — the Nizam Shahis149016361462
Berar — the Imad Shahis14901574841
Bidar — the Barid Shahis149216191275
Golconda — the Qutb Shahis151816871692
Vijayanagara133615652291
Bengal Sultanate135215762240
Hyderabad State — the Nizams172419482241

Why a cube, and only one

Torsten Hägerstrand proposed this in the early 1970s and called it an aquarium. It has exactly one measured advantage and one measured cost, and both matter. Asked about the overall structure of a dataset, readers using a cube were twice as fast — 121 seconds down to 60. Asked where something was in a particular year, they made more errors than readers given a flat map.

So this is a companion and never the way in. Ask it about shape — who overlapped whom, which century is crowded, how long a thing lasted beside the thing next to it. Ask Contested Ground what was happening in 1687.

What standing up makes visible

Two things, neither of which survives being flattened. The first is duration as a shape: Vijayanagara and the Bahmani sultanate are near-equal columns rising together from the fourteenth century, and the eye reads that parity instantly in a way two bars on a chart do not deliver. Berar is a stub. The Nizams are a spike that outlasts everything around it by two centuries.

The second is uncertainty with dimensions. Every archive footnotes its disagreements; this one collects them, and standing up they take up room. Bidar’s five published start dates are five rings spaced along the axis in proportion to how far apart the authorities really are — you can see the argument covers half a century without reading a word of it. Malik Ambar’s four death dates are a smear you can hold a finger against.

Shape on one side, sequence on the other

A cube is good at shape and bad at order. Standing among nine columns you can see at once which centuries are crowded and which state outlasted which — and you cannot easily say what happened next. So the same events run down the side in order, and the two halves point at each other: run your eye down a line to light its mark in the tank, or click a mark to find its line.

The list carries the doubt too. A year printed one way sits alone; a year the authorities argue about is shown as the span of the argument, so 1487–1542 tells you before you click that Qasim Barid’s beginning has been placed anywhere across fifty-five years.

Still no borders

Nothing here is a territorial claim. A column stands where a capital stood, because that is a coordinate the archive can cite; a cloud fades outward because influence does, and because no source read supports a line where it would end. Where two clouds overlap they brighten, and that brightening is the only thing in the tank drawn twice — it is the nearest this archive comes to saying two states wanted the same ground.

Empty space is no claim, not empty land. And the same 1688–1723 hole that shows on the flat map shows here as a band with nothing standing in it.