Internet Archaeol. 13 Harland et al. Approaches to recording

Approaches to recording

A coherent strategy is needed when recording an assemblage, especially if it is to be subject to any form of later analysis. At a very basic level, there are two approaches to recording. The first can be termed the 'specimen approach', while the second is the 'context approach'; each has benefits and disadvantages.

The 'specimen approach' is widely used in computer-based recording. Examples include the recording protocols of the original Environmental Archaeology Unit database (Dobney et al. 1999) and of NABONE (Newton et al. 2001). Recording systems using this approach treat each bone fragment as a separate entity, recording all available details about each before moving on to the next specimen. This produces a comprehensive primary record of the assemblage and it is possible to link particular specimens with particular measurements. The specimen approach has the advantage that it doesn't require any pre-recording analysis, and can correspondingly have several zooarchaeologists recording material from the same source at any given time.

The 'context approach' is that illustrated by O'Connor (2000 fig. 5.9), Reitz and Wing (Reitz and Wing 1999, fig. 6.2) and others (e.g. Grigson 1978; Meadow 1978; Winder 1994). Instead of recording each bone in isolation, a group of specimens of the same species is recorded together, listing details of elements present, their degree of completeness, the quantity of each, etc. Pathological details are listed, as are metrical data, although it is not always possible to link each measurement back to the specimen from which it was taken. The reasoning behind this method of recording is that the recorded data are not the primary record – the bones themselves are the archive. Each bone does not need to be recorded in isolation, as anyone who wishes to re-evaluate the zoological assemblage could return to the original assemblage (O'Connor 2000, 52). Although this method can be quite quick, it also requires a degree of pre-recording sorting and analysis. This needs both time and space, as well as preferably only one zooarchaeologist performing the recording in order to maintain consistency throughout the recording process.

These two approaches to recording are not mutually exclusive. A flexible recording system should be able to incorporate both methods within its structure. It is a truism among zooarchaeologists that '[t]he perfect method does not exist and no single technique is appropriate or even necessary for every study' (Reitz and Wing 1999, 143).

The York System allows both recording methods to be used. Context-level recording allows general observations about fragmentation, butchery, gnawing, weight, preservation and recovery to be recorded for all bones from that context. Moreover, the 'quantity' field in the main data form allows several bones from the same species and element to be recorded at once. Within this single database record of several bones, multiple entries can then be made of pathologies, butchery, measurements, tooth wear and other variables (accepting that the specific association between a variable and a single specimen will be lost). If preferred, one specimen can be recorded at a time, and preservation, fragmentation, recovery, etc. can be recorded on an individual basis.