PREVIOUS   NEXT   CONTENTS   HOME 

4. Negotiating Scales

The publication of Theodore Peña's Roman Pottery in the Archaeological Record (2007) was an attempt to bridge these issues of scale for one category of material culture. It sought to explore how the normative, generalising model of the pottery life-cycle might be exploited in specific contexts to elucidate both the specific elements of that life-cycle and what that says about human behaviour in the past (Peña 2007, 1-2; 2011, 136). A similar sort of thinking can be applied to rural landscapes but by taking labour, rather than ceramics, as the central variable.

Most site classification and categorisation take ceramic 'ware' as the predominant defining criteria, and though surveys may count total ceramics they usually only collect rims, bases, handles, and decorated sherds — primarily to facilitate ware identification (Alcock et al. 1994, 141; Fentress et al. 2004, 147-9; Renfrew 2003, 312-13).

Yet these are modern criteria. In Eleni Hasaki's ethnoarchaeological work in Tunisia (2011, 20), for example, the potters themselves took vessel height rather than vessel shape as a defining specialist criterion. Coupled with this is the uncertain use-life of particular ceramics. Mark Lawall's work has shown that, in the case of Rhodian amphorae, it takes 20 years for the jars to be sufficiently integrated into 'trash' for their discard distribution patterns to reflect import patterns. In other words, the more recently a deposit was sealed, the less likely it is that contemporaneous storage/transport vessels reflect widespread import/use patterns (Lawall 2011, 41-3). Both these studies show that there are factors other than 'ware' that are driving the character of the archaeological record.

One of the impacts of the Peña lifecycle has been to force us to think about the duration and longevity of the recovered material culture in the ancient world, and survey archaeology in particular is well placed to address these issues. Ancient agricultural activity may have been focused on the 'end point' of harvest, but agricultural labour was a year-round necessity that fluctuated in intensity and duration (Goodchild 2007, 100-3; Gallant 1991, 84; Kron 2000; Frayn 1975; Mason 1995). There would be particular times of year with increased demands for local, regional and supra-regional transport, for the use of filling stations for agricultural products, for the mobility of labour (potentially) to handle increased intensive workloads at key harvest and production times, and for static and mobile forms of storage.

Despite this, the nature of most of our evidence makes it very difficult to try to identify the phases in the lifecycle of a particular ceramic corpus or individual wares as recovered from the rural landscape. Or, in simpler terms, at what point in its use-life does material enter the archaeological record, and how can we retrieve that? This is very difficult for survey material, certainly for the legacy data of existing surveys (for example, Gaffney et al. 1985).

An examination of the ceramic data from the Laconia Survey in Greece highlights some of the problems. The Laconia Survey was carried out between 1983 and 1989, with the aim of setting excavations conducted within Sparta by the British School at Athens in a regional context. Further aims included increasing the broad topographic knowledge of Laconia and to understand the broad interplay between physical landscape, cultural circumstances and human society over time. It had an explicitly diachronic focus (Cavanagh et al. 2002a, 1). The survey is almost unique among Greek surveys in publishing its ceramic data in full (in Cavanagh et al. 1996; the other notable exception is the Pylos Regional Archaeological Project, which has published its raw data in an online gazetteer). This survey has generated a vast body of legacy data, in an accessible format, yet even so its ceramic data are conditioned by the collection methodology used by the survey (Allison 2008).

The methodology used by the Laconia survey, and indeed the methodology of many projects of the 'New Wave' of archaeological surveys in the 1980s and 1990s, was to divide identified 'sites' into transects and sample them (in this case with 1 square metre sherd counts at 2 metre intervals along four transects from a provisional site centre; Cavanagh et al. 2002a, 43-5). While all sherds were counted, only identifiable 'diagnostic sherds' were recovered consistently. As Table 2 shows, this means that there can be a broad disparity between the total numbers of sherds and those sherds used to identify predominant periods and activity types. For example, site C114 was interpreted as a Hellenistic-period farmstead. Sixteen of the 134 sherds counted were identified as Hellenistic. Of those sixteen, eight were broadly recognisable as either finewares (6) or cooking wares (2). On the Roman farmstead of N191, 256 sherds were counted. Of these, 23 were distinctly Roman, and only two were diagnostic (finewares). Sites were identified and classified on very small numbers of sherds, but the presentation of survey data masks these tiny numbers. This pattern is consistent throughout the data, and is an acknowledged issue with survey data (discussion of the bibliography can be found in Stewart 2013a, 10-14; Foxhall 2004, 251; Pettegrew 2001). Current methodology attempts to correct for this (for example, Tartaron et al. 2006; Lolos et al. 2007), but increased intensity of survey methodology has tended to exacerbate the disparity between the ceramics that are counted, the ceramics that are collected, and the ceramics that are identified. In other words, increased survey intensity can compound problems of accessing the use-life or life-cycle of ceramic assemblages.

A potential corrective may be to approach the problem from another direction and think about the 'taskscape', as Ingold deployed it, and the knock-on effect this has for our understanding of recovered assemblages. Ingold's 'taskscape' was defined as the socially constructed space of human activity: spatially bounded arrays of related tasks that, crucially, place human behaviour at the forefront of the analysis (Ingold 1993; 2000).

Figure 1: Taskscape of a ceramic storage vessel. After Lawall 2011, 46-50.

For rural landscapes, there are two ways to apply this type of thinking:

1. We can think about this in terms of the 'tasks' represented by a type of vessel, and how this relates to its final deposition (Figure 1). At what point in the vessel's 'life-cycle' did it enter the archaeological record?
2. We can approach the 'taskscape' as a way of conceptualising the behaviours required, over time, in order to run the economic elements within the landscape successfully. In other words, we can think about the social and material requirements of agricultural labour.

 PREVIOUS   NEXT   CONTENTS   HOME