What a tell is
A tell is an artificial mound created by people living in the same place for a very long time. The word comes from Arabic, and related terms are used elsewhere: “tel” in Hebrew and “höyük” in Turkish. Tells are especially characteristic of the Near East, from Mesopotamia and Syria to Anatolia and the Levant, though similar mounds exist in other regions, including the settlement mounds of West Africa such as Jenne-Jeno city in Mali.
The mound forms because of how towns were built. Houses were commonly made of sun-dried mudbrick on stone foundations. Mudbrick walls erode, and rain, fire, earthquakes or abandonment brought them down. When people returned or rebuilt, they usually levelled the rubble and built on top of it, and over generations floors rose, along with the debris of daily life. A mound that is several metres high, and in some cases much more, can therefore hold thousands of years of settlement. Sites are often in good locations for water, farmland or trade, which is why people kept coming back.
The principle of superposition
The key idea for reading a tell comes from geology. The principle of superposition says that in an undisturbed sequence, lower layers were laid down before higher ones. The Danish scientist Nicolas Steno stated a form of this principle in the seventeenth century, and it was later applied systematically to archaeological deposits.
In practice the oldest evidence is at the base and the youngest is near the surface. An archaeologist who digs down through a tell is therefore moving backward in time, and the order of excavation is the reverse of the order of building. Each deposit, wall, pit or floor is called a context, and the relationships between contexts, such as “earlier than”, “later than” or “the same as”, build up into a sequence.
Early recognition of layers
Understanding that tells could be read this way was a major step in the history of archaeology. In 1890 Flinders Petrie excavated at Tell el-Hesi in southern Palestine, and showed that different layers contained different styles of pottery, so that pottery could be used to date levels and compare sites. Mortimer Wheeler and Kathleen Kenyon later developed careful methods using narrow trenches with vertical sides, which allowed the section, or cut face, of the earth to be drawn and studied.
Troy is a well-known example of layering. The mound at Hisarlik contains many successive settlements, usually numbered from Troy I at the bottom to Troy IX at the top, with many sub-phases. Early work there cut deep through the mound, and the Hisarlik excavations later became a case study in how much careful recording matters. The story of Schliemann at Troy is often told to show how trenching without close attention to layers destroyed evidence that later researchers wished they could still read.
Recording the sequence
A modern excavation records each context separately, with drawings, photographs and written descriptions. Excavators also record the finds that come from each. Two kinds of drawings are especially important:
- Plans, which show the horizontal arrangement of walls and features at a given stage.
- Sections, which show a vertical slice through the deposits and how they relate.
The relationships can be summarised in a stratigraphic diagram called the Harris matrix, devised by Edward Harris in the 1970s. It shows which contexts are earlier or later than others and which are not directly linked, and it helps teams group contexts into phases such as “Period 4, first building”.
These phases then become the basis for the site’s chronology. Pottery and other finds in each phase are compared with neighbouring sites, and samples of seeds or charcoal give radiocarbon dating. Names such as the Uruk period are in effect labels for stages that were first recognised in stratified deposits.
When layers are not simple
Real tells are more complicated than a stack of pancakes. Several things disturb the sequence:
- Pits and foundation trenches: people dug down from a higher level, so younger material ends up cutting into older layers. A pit may contain a mix of old and new objects.
- Levelling and terracing: builders sometimes scraped away earlier deposits to create a flat surface, removing part of the record, or moved large amounts of earth from elsewhere on the mound.
- Erosion and slope: the sides of a mound slump, and the layers on the edges often slope and thin out.
- Robbing: later builders took stones and bricks from earlier structures, leaving only the trenches from which they were removed.
- Burrowing and decay: animals, roots and collapsing mudbrick blur boundaries.
Mudbrick is a particular challenge, because collapsed walls and the clay around them may be very similar in colour and texture. Excavators watch for subtle changes, sometimes aided by slight differences in moisture after rain or at the start of the day.
The height of a tell is also not a simple measure of how long it was occupied. Some periods left thick deposits and others almost none, and some phases were removed altogether. A gap in the sequence could mean abandonment, or it could mean that evidence was erased.
Why tells matter
Because they hold continuous sequences, tells give archaeologists a framework for entire regions. A dated sequence from one carefully excavated tell can serve as a reference against which other sites are compared. Questions such as how long an urban phase lasted, when a new technology appeared or whether a settlement was destroyed violently or abandoned slowly depend on this order.
The method also has a cost, since a mound can only be dug once. Most projects therefore excavate only part of a tell, leaving the rest for future researchers with better techniques. The sequence they build is a record of the site as seen through a small window, which is another reason to treat any single reconstruction with some caution.