While the Harris Matrix itself does not provide absolute dates, it can be used in conjunction with other dating methods to establish a relative chronology for the site. Typically this would involve reviewing specialist data relating to the dating of material culture and ‘spot dates’ would be plugged into the matrix to establish date ranges for the stratigraphic phases, and ultimately periodise the site. It may be that on less complex sites the underlying stratigraphic phases align with, or respect the overall chronological periods present on the sites, but it should be noted that phases do not have to align specifically with historical (or prehistoric) periods, and this is a key distinction.
Spot dating in archaeology refers to the practice of assigning specific dates or narrow date ranges to archaeological contexts or artefacts. Derived from the concept of identifying a particular point or moment in time, spot dating involves the use of datable elements or objects associated with known historical periods or cultural contexts. By identifying diagnostic artefacts such as coins, pottery types, or architectural styles, archaeologists can estimate the approximate age or chronological placement of a specific context. While spot dating provides a preliminary dating framework, more advanced techniques like radiocarbon dating or Bayesian analysis may be employed for more precise chronological assessments.
If one is not planning to conduct detailed (Bayesian) chronological analysis (see below) then this process of applying the dating evidence to the stratigraphy would ordinarily be used to do three things:
- Adjusting the Harris Matrix: Spot dates can be used to refine the sequence represented by the Harris Matrix by highlighting specific correlations and potentially emphasising certain relationships.
- Validating and Modifying Stratigraphic Phases: The dating evidence can help validate and potentially modify established stratigraphic phases or groupings as part of the iterative process of phasing the site.
- Assigning Time Periods to Groups and Phases: Spot dates assist in assigning chronological groups and phases to specific time periods, aiding in the temporal organisation of the site.
Depending upon the complexity of the site it may be that dating needs to happen at a group level (rather than at the individual context level).
However, it is important to note the tension that arises between using spot dates in this manner and employing more detailed Bayesian chronological analysis. Bayesian chronologists would consider it bad practice to use spot dates, particularly any derived from scientific dating methods, to underpin the phasing and organisation of the matrix and associated chronological modelling. This is due to the tautological problem it presents, where the date being refined is used as part of the a priori logic that underpins the final output. In simpler terms, using a preliminary or initial date to inform the Bayesian process can bias the final, more refined date, by creating a circular logic.
Methods
Ultimately all dates should be assigned to a context number in the stratigraphic sequence, being sure to check for possible issues of residuality and contamination of material culture (e.g. the latest spot date in a cesspit which was backfilled with domestic rubbish many years after it was built, would be misleading). At this point the matrix diagram can be annotated or colour coded for ease of reference. This will establish relative ages of the corresponding contexts in the Harris Matrix.
To accomplish this level of analysis, a comprehensive list of dates from specialists is required. These dates encompass spot dates on specific objects such as coins or brooches, or date ranges of assemblages of material culture such as ceramics, associated with specific contexts. Additionally, scientific dates obtained from specific samples should be collected and organised. Throughout this process, it is essential to recognize and differentiate between relative and absolute dates. Relative dates provide information on the age of an object in relation to other items, while absolute dates offer more precise numerical estimations for chronological placement.
Typical examples of relative archaeological dating involve establishing the sequence of stratigraphic layers and determining the relative chronology of different artefacts within a site. In contrast, absolute archaeological dating methods, such as radiocarbon dating or dendrochronology, provide numerical assessments of the age of archaeological materials or sites. It is also important to acknowledge that scientific dates often come with probability ranges. Thus, it is crucial to address these caveats and provide guidance on the appropriate utilisation of such dates in archaeological dating.
Once dating is established, at this point in the analysis it is often beneficial to cross-date contexts by seeking overlaps or associations between different contexts or stratigraphic units. These could share not only the same specific type of artefact but also artefacts of a similar character, thus allowing more flexibility and a broader understanding of the contexts in question. This process considers the issues of residuality and intrusion, wherein older artefacts may be found to have been redeposited in younger contexts (residual finds) and later objects are recorded in earlier deposits (intrusive finds) due to various natural or anthropogenic activities. By acknowledging these possible discrepancies in the dates of the finds, we can provide a more accurate analysis of the most likely dates for the deposition of the associated contexts.
If the stratigraphic relationship between two contexts is ambiguous, cross-dating with similar artefacts can offer critical clues about their relative order. This not only helps refine the relative chronology of the sequence but also aids in establishing temporal relationships between different parts of the Harris Matrix. By doing so, we can better understand the sequence and distribution of human activities in the area under study, considering all possibilities of artefact deposition and disturbance.
Throughout this process special attention should be paid to the problem of residuality. Each date should be considered in relation to its context within the sequence, with particular attention being paid to the security of its provenance, and the stratigrapher should ask ‘could this date be residual?’ If the answer is yes, then the date should probably be disregarded in terms of ordering the matrix or informing the phasing.
In a similar way it should be possible at this stage to assign absolute dating (if any has been acquired). This should suffer less from issues of residuality, because most absolute dates (with the possible exception of numismatic dates) will be derived from scientific methods (such as radiocarbon dating, dendrochronology, thermoluminescence), sampled from secure contexts. By dating specific contexts or associated materials, it is possible to establish fixed points within the relative chronology and refine the dating of other contexts, groups and phases accordingly.