3.2 Diverging branches of excavation methodology

Carver (2009) identifies three methodological approaches of which the geometric schnitt and box trenching share similar traits of combining horizontal plans and vertical sections for documentation. These are in contrast to the third, stratigraphic excavation including single context planning, which operate primarily through plan drawings. Single context planning was first implemented in the UK through its adoption by the Museum of London as an extension of the Harris Matrix, and spread in the early 1980s to the Scottish Urban Archaeological Trust and the York Archaeological Trust through the initiative of Clark and Pearson (Clark and Hutcheson 1993; Pearson and Williams 1993). The principles are well described by Roskams (2001) and the method has proven its strengths, especially for deeply stratified sites. As Carver (2009) states, however, a sequence of single contexts without horizons and sections does not make our job of joining the fragments into a combined drawing any easier. It is indeed virtually impossible without the aid of a computer.

Stratigraphical or single context planning has two main objectives when it comes to spatial recording; divide everything into its different constituents or contexts related to prehistoric actions or events (cuts, fills etc.) to account for stratigraphy and chronology, and ensure the ability to collect and display these contexts according to the stratigraphical matrix or phasing. It is imperative to single context planning to aim at discerning as many details as possible in the horizontal plan, as opposed to recognising stratigraphy from vertical sections. To achieve this, a system is needed that allows for layering and, in digital terms, vector-representations of drawings so that we may overlay contexts and retain transparency between layers that would otherwise be blocked by the countless sheets of paper. Both GIS and CAD would allow for this approach. CAD would, however, allow this without the need to maintain an elaborate database, if the individual layers are named according to context numbers. CAD is also more equivalent to the process of finalising or 'inking' the drawings, rather than focusing on the creation of a cartographic map.

The schnitt or geometric excavation, on the other hand, has some other requirements. When excavating large sites with sparse stratigraphical information, the use of sections is '… a ruthless - but efficient – method' (Carver 2009, 117), which has the advantage of producing a three-dimensional model of a feature, but unfortunately the complete deposit is never documented in this destructive process. One would think that the three-dimensional aspect of geometric excavation would lead to CAD being the prevalent solution, but we actually see the opposite. GIS is widely used, maybe in part due to the geographically large areas of such excavations, which calls for a geographic representation, but maybe more so owing to the excavation method not explicitly producing a stratigraphical sequence, which must be represented as individual layers. The stratigraphy or phasing of contexts or features does not result from the layering, but must be assigned to the attributes of the individual features. GIS allows for elaborate data to be assigned context or feature numbers and relations to other larger structures for easy query and display. This is not possible in standard CAD solutions, which will hold only limited information about a geometric object or group of objects (Eiteljorg and Limp 2008). The choice of GIS or CAD, of course, also very strongly correlates to whether the focus of research is at a landscape-level or site-level – if our spatial features are considered geographies or geometries.

Figure 2: Comparing box-cut and single context planning of posthole excavation in Denmark and UK. The Danish example is of the excavation of Viking Age longhouses in Jelling (Nielsen 2015; Holst et al. 2013), the UK example is the Lyminge Archaeological Project, University of Reading

As an example, we can look at how both Danish and UK archaeology focus on recording the depositional history of a posthole (Figure 2). The questions we would usually pose include whether the posthole was part of a larger structure, if the post had burned, was pulled up or if it intersects with other features that would provide information regarding stratigraphy and chronology. In the UK, when adhering to the ideals of single context planning, the posthole would be approached from the top, emptied and recorded in plan through its individual contexts (post-pipe, packing and cut), and the sum of the contexts would constitute the posthole. It is generally considered a very reflexive excavation process, as it requires continuous evaluation of the nature of each context, and contexts are observed and recorded in their entirety, often as individual CAD layers of each context.

In Denmark, the starting point is usually based on initial surface observations, and a hypothesis that the feature is a posthole. This interpretation is then tested by making a box-section, effectively removing half of the posthole, to reveal a profile or section containing layers, which may be interpreted. Of course, either method is subject to adaptation according to the archaeological object in question, and in practice they sometimes converge by, for example, using a combination of stratigraphical top-down excavation and leaving half of the feature as a vertical section. In Denmark, the posthole is generally considered a feature that is part of a larger structure, a building, which is to be identified from similarities in the layers of the vertical sections. Similarly, Carver expanded on the Harris Matrix by introducing groupings of features and structures to single context planning (Carver 1990).

In either case, working with vertical sections has methodologically always been problematic. Vertical representations are not easily integrated with the horizontal plan drawing, and neither GIS nor CAD natively allows for this type of functionality. If at all digitised, vertical sections are often managed separately and in an arbitrary two-dimensional coordinate system, which is also why single context planning show much greater integration with these technologies.


Internet Archaeology is an open access journal based in the Department of Archaeology, University of York. Except where otherwise noted, content from this work may be used under the terms of the Creative Commons Attribution 3.0 (CC BY) Unported licence, which permits unrestricted use, distribution, and reproduction in any medium, provided that attribution to the author(s), the title of the work, the Internet Archaeology journal and the relevant URL/DOI are given.

Terms and Conditions | Legal Statements | Privacy Policy | Cookies Policy | Citing Internet Archaeology

Internet Archaeology content is preserved for the long term with the Archaeology Data Service. Help sustain and support open access publication by donating to our Open Access Archaeology Fund.