—
—
—
The section
This is the side of the trench, drawn to scale. Deposits are numbered in round brackets and cuts in square ones, the way they are on a real site. The numbers were handed out left to right as it was drawn, so they won't tell you the order. Click a deposit to dig it. You can only take out something that nothing later still lies on or cuts into. Being at the top of the section isn't enough. Heights are drawn half as tall again as they really are, so the thin layers can be seen.
Move over the section to read a context. Click one to dig it.
Harris matrix
U puts the last deposit back, and the left and right arrow keys walk to other days. Every day is a different site.
Finds register
Everything that came out, bagged by the deposit it came from, newest dig at the top. The date beside a find is when that kind of object was being made, not when it was dropped. A coin can lie about for centuries before it goes in the ground, but it can't go in before it was struck.
| Context | Find | Made |
|---|
- Date
- —
- Site code
- —
- Natural
- —
- Contexts
- —
- Deepest feature
- —
- Periods
- —
What is real here and what is not
The method is real, and so are the finds. The site isn't. Nobody should go looking for it.
-
01
The order is real
This is single-context excavation, the way most British urban sites have been dug since the 1970s. Every deposit and every cut gets a number, and each one comes out whole before anything under it is touched. A layer that runs across the top of a pit is later than the pit, and it comes out first. A pit that cuts a layer is later than the layer, even though the layer is higher up on both sides of it. Most beginners get this one wrong.
-
02
The matrix is real
Edward Harris drew the first of these diagrams at Winchester in 1973. Each box is a context, and each line says the one above is later than the one below. Only direct relationships are drawn. If A is over B and B is over C, the line from A to C is left off. The matrix here is worked out from what touches what in the section, and it fills in as you dig.
-
03
The dating is real
A deposit can't be older than the newest thing in it. That date is its terminus post quem, the date after which it must have formed. The rule also works through the matrix. A deposit is later than everything under it, so the newest find anywhere below it counts too, even if the deposit itself holds nothing but a flint that is three thousand years older.
-
04
The finds are real
Samian from South Gaul before Central Gaul, black-burnished ware from the 120s, a radiate of Tetricus struck in the 270s, Ipswich ware, long-cross pennies of Henry III, clay pipe stems from the 1580s, creamware, ring-pulls. Every one carries the date range the reference collections give it. Old things turn up in later deposits too, as they do on real sites, where residual finds are everywhere.
-
05
The rest is invented
The town, the street, the site code and everything that happened on the site are made up by the day's number, including the true dates the page shows you at the end. Real sites also have intrusive finds, which are later things that worked their way down a root hole or a rabbit burrow and give a false date. None have been planted here. On a real site, you would have to watch for them.
Twenty-nine rooms, and one number doing all the work. Today it spends a few thousand years on one strip of ground, digging and filling and building and pulling down. Then it hands you a trowel and asks you to undo all of it in the right order.