Table of Contents
Six people are kneeling in a four-metre trench on a hillside, scraping at brown soil with the worn edges of pointing trowels. Nothing interesting has turned up for three days. Then a blade clicks on something hard: a rim sherd of fired clay, the curve of a pot mouth, still holding a faint thumbprint.
Within a minute the trowels stop. Someone fetches a camera and a scale bar. Someone else writes a number in a notebook and draws a small plan at 1:20. The pot is photographed where it sits, measured with a level, and only then lifted. That pause — the five minutes of recording before the object moves — is the whole difference between archaeology and treasure hunting.
Excavation Is Destruction, So the Paperwork Comes First
Every dig destroys what it studies. Once a deposit is lifted, the relationship between that soil, the object inside it and the layer above is gone forever. Archaeologists repeat a grim little mantra about this: the only way to preserve a site is not to dig it. Which is why months of preparation precede the first trowelful.
A project starts with a research question specific enough to be answerable. Was this hillfort occupied continuously or rebuilt twice? Did the harbour silt up before or after the town wall went up? Those questions shape where the trenches go, how big they are, and which samples get taken. Alongside the design comes a permit from the relevant heritage authority, a storage agreement with a museum that will curate the finds, and a conservator booked for anything fragile.
Before ground is broken, the team surveys. Fieldwalking picks up pottery and flint across ploughed fields. Magnetometry and ground-penetrating radar map buried walls and ditches without touching them. LiDAR, aerial photography and old estate maps fill in the rest. On the best projects, a dig is the last 5% of a very long investigation.
Setting Out the Grid
On the first morning, someone hammers a steel peg into the ground and declares it the site datum. Everything else is measured from it. Trenches are laid out in squares — one metre for tight, detailed work, five or ten metres for open-area excavation of a settlement — with string lines stretched between pegs and corners checked for square.
The grid isn’t fussiness. It gives every find a three-dimensional coordinate, lets the team sample rather than strip a whole site, and makes it possible for someone decades later to reopen the trench in exactly the right place. Modern teams record points with a total station or a GPS rover to the nearest centimetre, which beats the days of tape measures and graph paper, though both still appear when the batteries die.
Spits, Layers and the Difference Between Them
In deep, homogenous deposits, excavators remove soil in fixed slices called spits — often 5 or 10 centimetres at a time. That’s an arbitrary division imposed by the digger. Where layers are visible, the team follows natural stratigraphy instead, removing each deposit as a unit. Most sites need a mix of both.
Context Is the Real Find
A coin in a plough-soil scatter tells you almost nothing. The same coin sealed beneath a collapsed floor, next to a hearth, tells you who dropped it, when, and what the room was used for. Archaeologists call this the context, and they record it with a thoroughness that surprises visitors.
Each deposit — a layer, a pit fill, a wall, a cut — gets its own number. That number appears on a context sheet describing colour, texture, inclusions and boundaries; on photographs with a north arrow and a scale; on a scaled plan drawing; and on a section drawing of the trench edge. Soil colours are matched against a Munsell chart so that ‘brown’ becomes ’10YR 4/3′. At the end, all those numbered contexts are stacked into a Harris matrix, a diagram of what cut what and what sealed what.
The Tools and What They Actually Do
- Pointing trowel — the four-inch blade does most of the delicate work, often with the tip ground sharp.
- Mattock and shovel — for overburden and clearly sterile soil, well ahead of any finds.
- Sieves — 10 mm mesh for general recovery, 3 mm for features likely to hold beads, fish bone or flint chips.
- Total station — a laser instrument that fixes the position and height of a feature in seconds.
- Dental picks, brushes and wooden skewers — for cleaning around fragile bone or metal before it’s lifted in a soil block.
- Bags, labels and permatrace — unglamorous, and the reason the archive still makes sense in fifty years.
Reading the Layers
Stratigraphy follows a simple rule: in an undisturbed sequence, lower layers are older than the ones above. Trouble is, almost nothing is undisturbed. Pits, ditches and graves cut down through earlier layers, and their fills are younger than the soil they slice through. A wall foundation may be later than the floor it sits on but earlier than the rubble that buries it.
Digging a feature means recognising the edges of the cut before removing the fill. Experienced excavators spot them by texture and colour — a ditch fill is often slightly darker and softer than the subsoil around it. Get that edge wrong and you’ve mixed two deposits, and the finds inside lose their meaning.
Finding the Tiny Things
Plenty of evidence never shows up on a trowel. Soil from hearths and pit fills goes into sample buckets, usually 10 to 20 litres per context, and gets processed by flotation: water washes the sediment away while charred seeds and wood float off into a fine mesh. The heavy residue is dried and sorted for small bones, slag and flint debris.
This is where a site stops being about pottery and starts being about diet, fuel, farming and seasonality. A single floated sample from a Bronze Age hearth can produce hundreds of cereal grains, weed seeds and fragments of hazelnut shell — a whole afternoon of ancient cooking, recovered from a bucket of mud.
After the Trench Closes
The dig ends; the work doesn’t. Finds go to a processing lab for washing, marking and cataloguing. Conservators stabilise waterlogged wood and corroding iron. Ceramic specialists date the pottery assemblage, zooarchaeologists count cut marks on animal bone, and radiocarbon labs run samples through accelerator mass spectrometry before the dates are calibrated against tree-ring curves.
Isotope analysis of tooth enamel can reveal where a person grew up. Residue analysis on a cooking pot can identify dairy fats. Ancient DNA from a single tooth sometimes redraws a migration map. Results eventually land in a published report and a museum store, and increasingly in open datasets that anyone can download.
What the Method Has Turned Up
Good excavation technique is what turns a curious mound into evidence. At Göbekli Tepe in southeastern Turkey, careful stratigraphic work pushed monumental stone architecture back to roughly 9500 BC, thousands of years before farming villages — one of the archaeological discoveries in the Middle East that rewrote the human story. In East Africa, excavations at sites such as Lomekwi and Olduvai have recovered knapped stone tools dated to 3.3 million years ago, as catalogued in this survey of the world’s oldest archaeological sites.
None of those results came from a lucky shovel. They came from measured sections, sieved sediment and contexts recorded one number at a time — the same routine used on a Roman ditch in Yorkshire or a Taíno village in the Caribbean, where Caribbean excavations have rewritten island history by dating settlement and trade far earlier than colonial accounts suggested.
How to See One for Yourself
Most countries run archaeological field schools that take paying or volunteering participants for two to six weeks, and local societies often join test-pit digs on weekends. If you’d rather watch than trowel, many active excavations hold open days, and heritage bodies publish annual lists of digs accepting visitors. Bring boots, a hat and patience — the interesting find usually appears in the last hour of the last day.
Go, and you’ll learn the rhythm of a real excavation: scrape, watch, stop, measure, photograph, bag, label, repeat. The glamour is thin on the ground. The information recovered is anything but.
