Principles of stratigraphic analysis

Giorgio Buccellati – August 2009, May 2026

The matrix

When approaching a set of cultural remains that have been abandoned and have collapsed over time, we face an amorphous matrix, a congeries of loose material and discrete fragments. A Mesopotamian tell like Urkesh is a prime example of this: the collapse process of took place over many centuries, and it was a dynamic process which went well beyond simple superposition and resulted also from a variety of intrusions and re-uses.

clarify better in relationship to Dismantling

section e.g. A16v25

The process

Excavation is the process through which we disentangle the matrix.

One approach is to extract what are perceived to be important elements within the matrix, without regard for the context. This is non-archaeological excavation.

What is unique to archaeological method is an excavation that lets us understand and document the way in which every element lays in the ground and then infer from their emplacement how they had come to be where they are.

Elements: stationary and movable

Elements are the things that are found in the matrix. We will see in typology how these elements can be defined and described. Within the realm of stratigraphic analysis we ditinguish two main categories:

stationary elements – their typological identity is tied to a location, and
movable elements – their typological identity is independent of place<.

This distinction is simple enough that it can be applied without hesitation in practically every instance. It is especially useful during the first moment of excavation, when other distinguishing characteristics may not be as apparent, or when specialists may not be present for identification. This is discussed in detail in xxx

wall
platfom
snake jar
andiron
sherd
tablet
sling ball
soil phytoliths for samples

Contacts

The notion of contact presupposes the identification of discrete entities, the elements. The contact is not mere juxtaposition. From various factors, such as texture, alignment of components, organization of planes and volumes, etc., we may plausibly argue for different types of contact. In other words, the clustering is hierarchical, and supports the basic inferences about the depositional process that is to be assumed.

Clustering

Stratigraphic clustering refers to the way in which elements in contact, in the physical space, can be linked so as to denote temporal sequences.

The physical contact is what can in fact be observed and documented. As such it is the starting point of all archaeological analysis. The individual instances of contact are defined as part of emplacement, and are measured as part of volumetry. What lies beyond is the clustering of these points of contacts into groupings that subsume all the pertinent primary observations. From this we derive, in a properly arguable fashion, the depositional process and the overall sequencing.

Space/time: emplacement and deposition

That space can be translated into time is the basic postulate that underlies the inferential process. For instance, having distinguished (through emplacement) two vertical planes as a cut vs. a face (texture, alignment, hardness, etc.), we may argue that a face is contemporary with the element behind it, whereas a cut is subsequent to it.

Strata

The notion of clustering finds its prime realization in the concept of stratum. This is a cluster of elements arranged according to the type of contact, and sorted according to nesting criteria that result in discrete wholes. These wholes are defined by the congruence of the elements in contact (e.g., a series of pits cut into a single accumulation), and by broad elements that extend to an entire volumetric unit (e.g., a floor that covers the entire surface of a locus).

There are three types of contact:

direct – physical contact between elements
indirect – contact mediated by a a third element (e. g., two floors on either side of a wall)
inferential – contact implied by the consonance of various factors, including typological analysis, across larger excavation areas (e. g. elements across several squares within the same unit).

For a detailed example see for now this file.

In the standard approach to archaeological publishing, the stratigraphic record remains by and large ungrammaticalized. The data are understood as a language might be for which no grammar is provided: we have a sense of what the data say, but their integration into a unified conceptual structure is incomplete and inarticulate as far as the .