Sequence and Date

Abstract

For the field researcher, a primary task is assigning a date and a sequence (order of occurrence) to the features and structures they record. Occasionally an archaeological site has already been recorded in history, for example, the celebrated urban excavation at Five Points, New York City, exposed a plan of buildings and streets that had appeared on a map in 1855. Even sites mentioned in documentary references seldom offer a date as precise as this, and dated events which might seem to refer to an excavated site have to be used with great caution.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Introduction

For the field researcher, a primary task is assigning a date and a sequence (order of occurrence) to the features and structures they record. Occasionally an archaeological site has already been recorded in history, for example, the celebrated urban excavation at Five Points (Chap. 31), New York City, exposed a plan of buildings and streets that had appeared on a map in 1855. Even sites mentioned in documentary references seldom offer a date as precise as this, and dated events which might seem to refer to an excavated site have to be used with great caution.

In general, very few objects, activities, or structures discovered by fieldwork can be given a precise calendar date, and archaeologists are obliged to build a chronological model, which balances all the available information (Fig. 10.1).

Fig. 10.1

Chronology in field work (Carver 2009, p. 267)

Dating Objects

As applied to objects, the methods at our disposal are typology, which offers a relative dating for artifacts (e.g., pottery) and structures (e.g., architecture). Typology uses the likely order of manufacture, based on form and style, combined with the dates given elsewhere. For example, pottery, one of the most useful of artifacts since it occurs widely and endures well in the ground, has an enormous range of types. Some periods of manufacture are known from kiln sites, but most are deduced from which types occur together; in this way archaeologists have built up extensive typologies that help to date every kind of site built by pottery users.

Artifacts may also be given an absolute dating, by scientifically measuring the age of materials they are made from. Well-known examples here include radiocarbon dating, which measures the age since the death of a living plant (i.e., wood) or creature (e.g., bone); dendrochronology, which measure the age of timber since it was felled, from the numbers and spacing of annual growth rings; archaeomagnetism, which measures the age of a hearth since it was heated; and optically stimulated luminescence (OSL) which measures the time that has elapsed since a layer of sand was last exposed to sunlight (Hedges 2001). These dates, which have an error range from 1 year (dendro) to 25 % or more, indicate when an organism died or a mineral was buried.

Dating Contexts

Absolute and relative dates for artifacts, or groups of artifacts, can be used to date the archaeological layers they are found in – but the relationship is not a simple one. A layer is always deposited later than the latest object found in it, for example, a floor with a coin of 400 CE beneath it must have been laid in 400 CE or later (since the coin must have existed before the floor was laid). This relationship is called a terminus post quem (TPQ). A wall which has a date written on it (say 1929) must have been constructed before 1929. This relationship is called a terminus ante quem (TAQ). However, these equations are by no means always valid or helpful. A coin may be not just earlier, but centuries earlier, than the floor that covers it. A coin within a floor may be intrusive and so later than the floor. A coin found on top of a floor may be later than the floor, if dropped on it soon after manufacture and never moved, or much earlier than the floor if carried around for decades in someone’s pocket. Similarly a coin or a potsherd found in a foundation trench is usually earlier than the wall in the same trench, but it may also be much earlier since it has been displaced – is residual – from an earlier phase or site. Since we rarely know the circumstances in which a coin was discarded, it is risky to use coins to date structures. They are at least very rarely contemporary. In spite of this, equating the date of a building with the date of the coins found in it remains a widespread practice.

Structures, features, and contexts may also be dated directly – by typology and by scientific dating. Typology may be applied to the shape of hearths or kilns, or the ground plans of houses, comparing them to others found elsewhere and so presuming that they can be assigned to the same culture and date. Absolute dating can sometimes be applied, for example, dendrochronology will date the timbers of a timber-framed house (Kuniholm 2001). However, it is frequently found that such a house, in the form it survives, is composed of structural timbers of different dates. Even the earliest of these may have been recycled from another usage – for example, in a ship. Radiocarbon dating is applied to carboniferous materials, such as charcoal or bone, but here the association with the deposit is of crucial importance (Taylor 2001). The charcoal in a hearth may represent the date of last burning but only if it derives from twigs or animals. Otherwise the wood may have already have been cut down long before it was burnt (the ‘old wood’ effect). Similarly the bones in a grave should date the digging of the grave very well, but animal bone may have been disturbed and redeposited and so give a date before, perhaps long before the deposition of the layer in which it was found.

An important method applied by excavators to contexts, features, and structures is stratigraphic ordering. This does not date them but provides a relative order for each deposit in the overall sequence. The traditional method of presenting the order of occurrence is the section, which shows the deposition of layers from the side, and is recorded by drawing (Fig. 10.2), and these may still be valuable even if they only report the sequence in a specific slice through the strata. A more comprehensive method of stratigraphic ordering that applies to the whole site is the stratification diagram – which models the sequence in two dimensions, the earliest contexts at the bottom and the latest at the top (Fig. 10.3). These have developed from pioneer examples in the 1970s (e.g., Harris 1989) to more comprehensive models which include features and structures and represent uncertainties in the sequence (Carver 2009, p. 296). These uncertainties form an important aspect of the modelling process and show where other interpretations are possible.

Fig. 10.2

A vertical section through consecutive layers at the early monastic site at Portmahomack, Scotland (M. Carver)

Fig. 10.3

Extract from a stratification diagram, showing part of the sequence encountered in the excavation of a church at Portmahomack, Scotland. Contexts are designated with the prefix ‘C’. Contexts belonging to features are contained within the feature box (prefix ‘F’). Some features are grouped with the structures they have been assigned to (‘Church 2’, ‘Church 4’). The features are shown as vertical arrows, locating their limits in time. The stratigraphy is modelled to show where chronological (vertical) variation is possible (Courtesy of FAS-Heritage Ltd)

The stratigraphic ordering of graves containing bone or hearths containing charcoal can be dated with much greater precision by using radiocarbon dating combined with Bayesian analysis. The radiocarbon dates give a range of error, but this range is reduced by knowing the order in which deposition occurred. The Bayesian analysis produces shorter ranges of probability for each date (posterior density estimates, Fig. 10.4). This is giving archaeological sequences of high precision dating back to 20,000 years ago (Buck 2001).

Fig. 10.4

A sequence of graves from the Anglo-Saxon cemetery at Wasperton, England, placed in their best order by Bayesian analysis. The outline shows the error range of the calibrated date; the solid black profile is the more precise “posterior density estimates” derived from Bayesian modelling (Carver et al. 2009: Fig. 4.1)

Relative Ordering by Space

The business of establishing a sequence makes also use of spatial analyses. For example, a settlement that is spatially coherent (like the grid of a planned town) suggests that all the streets and houses were laid out at the same time. Features that are aligned are also thought to reflect contemporary use. For example, inhumation graves in a cemetery may have different orientations, but those that are most closely aligned are said to be close to each other in date. Similarly, graves that mimic the orientation of a building are later in date than the building. On the scale of a landscape, alignments are important indicators of sequence. Roads and field boundaries seen from above may indicate a sequence where they “respect” one other. For example, the field boundaries may join up to a preexisting road, or the road may cut straight across the line of the fields, showing it was imposed on a preexisting agricultural landscape.

Modelling

At the scale of a landscape, archaeologists use such spatial mapping as an indication of sequence. It is often possible to apply typology, for example, to infer the likely culture and date of cropmark forms encountered in aerial investigation. Similarly, forms recorded in subsurface survey are sometimes recognized by virtue of their shape alone: a straight road or a circular ditch or a settlement grid. When using surface survey to find sites, archaeologists rely on typologies to provide a broad date for the pottery or stone implements or metal artifacts they are mapping. The distribution of this material provides the location of sites belonging to a particular culture and period.

On excavated sites, stratigraphic ordering provides the surest indication of relative sequence. The alignment of features and structures often give an inference of where these might be contemporary. The broad dates of objects and structures (and the very occasional documentary reference) allow a sequence to be anchored more closely in time. In well-stratified sites, such as towns, the stratigraphic sequence is treated as primary. Poorly stratified sites (the majority) rely more on relative and scientific dating of objects and samples. The method is to examine and record the detailed possible relationships of objects and strata in every case and then to use typology, stratification, spatial analysis, and absolute dating to build up a robust chronological model.

Entries in EGA relevant to this section include Dating Techniques in Archaeological Science; Dating Methods in Historical Archaeology; Radiocarbon Dating in Archaeology Stratigraphy in Archaeology: A Brief History.