Of pots and layers: what stratigraphy and pottery can tell us

At the Black Friary, our pottery story tells us that the site was in use from the medieval period through to the early post- medieval period, and into the eighteenth century.Here, Black Friary Director and PI, Finola O’Carroll tells us how:

What is stratigraphy, and why is it important?

One of the most fundamental things that I try to teach students is to figure out how a site formed, and how can we make sense of the mounds of stone and earth that we’re confronted with and then tell its story? It’s a truism that archaeological sites are made up of deposits, laid down over time, which can then be excavated from the newest to the oldest. Theoretically, these can then be used to reconstruct the site and interpret it – how it looked and what functions it served. While these ideas were borrowed from geological laws relating to the formation of sedimentary deposits, this was not enough to explain and interpret the complexities of the average archaeological site. In 1973 Edward C. Harris developed a system now called the Harris Matrix (Harris 1975), which is a really useful tool used to depict the succession through time of archaeological contexts. Using this the sequence of depositions and surfaces on an archaeological site, otherwise called a ‘stratigraphic sequence’, can be drawn and makes understanding, analysing and interpreting the site easier and quicker. He developed this as part of his work on the really complex urban excavations at Winchester, England.

Harris believed that ‘the stratigraphic sequences of archaeological sites are made by the analysis of the interfaces between strata, not from a study of the soil composition of the strata. Without a stratigraphic sequence, the cultural remains of the strata cannot be contrasted except in a general ‘typological context’ (Harris 1979, 112).

Traditionally, excavators who worked in the first part of the 20th century such as Sir Mortimer Wheeler or Dame Kathleen Kenyon used a grid system to excavate which involved leaving baulks at regular intervals. A site would be divided up like a chess board, typically into 4m x 4m and 5m x 5m squares with baulks 0.5. wide between. The vertical faces on each side would be the reference point to analyse the stratigraphy. However, archaeological sites frequently contain many small deposits and features such as pits, which will not be represented in the vertical section faces at each baulk. This was the problem Harris was working on, and what his system was designed to cope with. Using his system each layer or feature is recognized, individually excavated and recorded descriptively and spatially before moving on to the next feature. In this way a matrix, or stratigraphic sequence can be built up, even though the deposits may not necessarily overlie each other directly. This practice was codified into the Single Context System as developed by the Museum of London Archaeological Service (MOLAS), which was designed to rapidly capture information onsite while digging, allowing for detailed analysis and interpretation post-excavations (see Spence 1993).

How do we use this method at the Black Friary?

At the Black Friary we use a slightly modified version of this system; understanding how to recognise features, and how to record a simple matrix which details the relationship of the feature to its immediate neighbours (on top of, beneath, beside or equal to) are fundamental lessons that students learn while digging. Integrating this information into the analysis of the artefacts and environmental material from each feature allows us to understand the events which trace an arc in time from the building, use-life, changes through enlargement or contraction of the friary, change of use post-reformation and dissolution, up to and beyond the time it was reduced to rubble by quarrying.

For those of you who’ve been following our progress, or have dug with us, you’ll understand the importance we attach to figuring out the stratigraphy and then analysing the artefacts found in each layer. We know, for example, that the rubble layers, almost always the upper layers over much of the footprint of the friary buildings, formed because of the quarrying of the site in the middle of the eighteenth century, (O’Carroll 2014). So, it follows that anything beneath those layers is earlier. We also know that the friary was built in the 13th century, so the earliest features should belong to that period. But figuring out what happens in between can sometimes be a little trickier.

How do we apply it?

View of east baulk of Cutting 7: cut F763 is visible to right, stone dump F765 is on the left, midden layers in the centre with F703 on top

Last summer (2016 season), excavations in Cutting 7 were extended to explore the north side of the north range; the deposits found suggested that a midden or rubbish pile had formed behind the north range in a series of very clear layers.  As a lot of animal bone was included it suggested that waste from kitchens formed the bulk of the pile. At the north end of the midden, where the wall should be, is a robber trench formed by the quarrymen digging down through the layers to extract the stone from the wall. That then filled up with loose soil in the area, in the place where the wall once was.

We recorded all these layers as they were excavated, by giving them unique numbers and by photographing, drawing and taking levels and co-ordinates under the supervision of Ian Kinch. All our feature, find and sample numbers incorporate the number of the cutting in which they were excavated. An interface is clearly shown here by the robber cut dug through the later deposits when the wall was taken down. This cut is visible as a clear disjunction between two sets of deposits to the right of the section face . In this stratigraphic sequence the deposits contained within the cut will be later than those in the midden, even where the midden layers are higher.

The features in the middle of the cutting can also be integrated with those in the section. The remains of two walls running N/S  have been cut by the robber trench, so we know they pre-date that event. But what is their relationship to the midden layers? The lowest exposed layer occurs on both sides of the narrower (eastern) wall. Other layers overlie this, some on to one side or the other, and some overlie the wall itself.

View looking north of walls F752 (left), F756 (right)

By using the Harris matrix, we can understand the sequence in which layers were deposited, cuts were made, and walls were built, even when they don’t directly overlie each other but occupy different areas within the cutting. When we have completed the excavations, we should know the exact relationship of the walls to these deposits.

What happens when we match the artefacts to the stratigraphic sequence?

We know that the sequence in the section face through the midden isn’t the whole story, but by looking at the pottery from layers in the section, we can see if the stratigraphic sequence is matched by a pottery sequence. Just by separating the pottery into medieval and post-medieval types from features F758, F755, F750 and F703, taken in order from the lowest to the highest will give us an idea of the pottery relates to the stratigraphy.

  • Feature 758 (F758 is the lowest midden layer. F758 is directly below F755, which in turn is directly below F750. The robber cut F763 and the fill of the trench formed by the cut, which is F753, cut these three features. All are capped by F703, which we can date to the destruction of the site in the 1750s.
  • F758, the lowest layer, contained a lot of pottery, 101 sherds in total. It was predominantly medieval in date, with 96 sherds of medieval pottery to only 5 sherds of post-medieval pottery.
  • The next layer, F755 contained a large percentage of medieval pottery, 28 sherds versus 3 sherds of post-medieval pottery were recovered.
  • F750 is being interpreted the upper layer of the midden deposit. Unlike the first two features, F750 has slightly more post-medieval pottery than medieval pottery, 16 sherds of medieval pottery were found as against 21 sherds of post-medieval pot. This is interesting as it shows the progression from medieval use to post- medieval use.
  • F703 is a destruction layer and made up of rubble collapse of the north range of the friary from the time of the destruction of the site in the 1750s. Here, medieval pottery again is dominant, with 37 sherds to 20 sherds of post-medieval pottery.

If we were to date the layers just on the cultural material contained within them, we would have to argue that all the layers are post-medieval or later in date, as all contain post-medieval pottery:

Table 1

By understanding the stratigraphic relationships, we know that F703 is much later than F758. We can explain the high proportion of medieval pottery in the latest layer by the fact that we know that earlier layers were dug through to access the wall to dismantle it. This would have meant that earlier deposits were mixed with the later material. The presence of the few post-medieval sherds in the lowest excavated layer may be a result of later contamination, possibly from the feature F765, the stone dump north of the midden. So, we see a change from mainly medieval pottery to late medieval/early post-medieval pottery (below, top left – bottom right), and a mix of medieval and post-medieval pottery dating in the layer associated with quarrying .

Most of the post-medieval pottery comes from the top two features. This pottery tells us that the site was in use from the medieval period through to the early post- medieval period and the pottery sequence, including clay pipes, seems to go right up to the eighteenth century and the time when it was quarried.


References and Further Reading:

  • Harris, E. C. 1975. The stratigraphic sequence: a question of time. World Archaeology 7 (1): 109-21.
  • Harris, E. C., Marley R. Brown and Gregory J. Brown, (eds) 1993. Practices of Archaeological Stratigraphy. Academic Press
  • O’Carroll, F. 20 11. ‘Interim Report, Archaeological Research Excavations at the Black Friary, Trim, (Unpublished report submitted to the National Monuments Service).
  • O’Carroll, F. 2014. ‘Interim Report, Archaeological Research Excavations at the Black Friary, Trim, (Unpublished report submitted to the National Monuments Service).
  • O’ Carroll, F., Shine, D., Mandal, M., Scott, R. and Mullee, B. 2016. Interim Report, Black Friary Archaeological and Community Report . (Unpublished report submitted to the National Monuments Service)
  • O’ Carroll, F., Shine, D., Mandal, M., Scott, R. and Mullee, B. 2017. Interim Report, Black Friary Archaeological and Community Report . (Unpublished report submitted to the National Monuments Service)
  • Spence, C. 1993. ‘Recording the archaeology of London: the development and implementation of the DUA recording system’ in Harris, E.C. et al (eds), Practices of Archaeological Stratigraphy. Academic Press.
This entry was posted in News. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *