Tomas Romson asked me a good question. Why do archaeological layers form in such a way that older things end up buried below newer ones? The answer is, because people and natural processes deposit dirt on the ground. Using a word borrowed from geology, we call the study of such layers stratigraphy.

Imagine a Neolithic family who build a house on virgin land and live there for some decades. At the end of this phase there are a number of stinky middens and remains of the house on the site. These materials get compacted by microbial activity and mucked around by wildlife, and eventually form a culture layer full of artefact fragments and bones and charred plant remains across the site.

Then after a lightning strike a nearby wood burns down, exposing a sand deposit to strong wind from the sea for a few decades. The wind covers the Neolithic layer with drift sand.

Now a Bronze Age family comes along and builds a house on the drift sand, and the process repeats itself.

Then a nearby river changes its course and floods the site, covering it with a layer of river muck.

Now the site becomes grassed over and Iron Age farmers from a nearby village use it as pasture for centuries. Hardly any deposits accumulate on top of the old river muck, just a couple of hearths and thin waste scatters from transient shepherds.

Then a Medieval potter sets up shop nearby and uses the site as a dumping ground for broken and misfired pottery.

By now there’s a stratigraphical layer cake on site. But dating the stuff you may find there is not as simple as measuring the level above the natural subsoil. Because people dig pits and place the spoil on their era’s ground surface.

Imagine the Iron Age folks digging a pit on the site to bury an old man with a dodgy reputation who isn’t allowed to rest in the village cemetery. They dig his grave deep enough that the base of the grave cut reaches all the way into the natural subsoil. This means that when they backfill the grave, they leave a number of redeposited Neolithic and Bronze Age artefact fragments on top of the Iron Age land surface. As Amnon Ben-Tor taught us volunteers at Tel Hazor in 1990, every major layer on a tell contains finds from every phase of the tell’s use. In fact, during the month I spent at Hazor we managed to extend the tell’s known life back into the Early Bronze Age despite the fact that we were digging much later layers dating from the Persian Empire. There was some EB pottery in that layer too.

But of course we no longer dig stratigraphy in site-wide layers and do not attempt to keep the excavation surface level. Imagine removing the Medieval potter’s dump layer in a good-sized trench on site, exposing the top of the river muck. A number of discolorations are visible in the surface of the muck layer. Some are Iron Age hearths. One is an Iron Age grave. And next to it is a small diffuse deposit left from the grave diggers’ spoil heap. Before we remove any of the river muck, we carefully remove and sieve each of these later deposits, which in the case of the grave provides us with a ready-made section through the site’s stratigraphy. Though we haven’t actually dug into the Neolithic or Bronze Age deposits on site yet, we will now already be aware that they exist and are separated by geological layers in a manner you almost only ever see in theoretical thought experiments.