To understand archaeology as a discipline, the evidence that it produces and the interpretations it puts forward, we need to begin with a matter that divides archaeologists from historians – the use of evidence and the combination of evidence texts and archaeological data. These are subjects that can cause scholars to get a little hot under the collar. Some have even suggested that combining these forms of evidence is only desirable after both sets of evidence have been studied independently (see Allison 2001). Part of the problem is caused by a belief that texts surviving from antiquity can be regarded or treated as data. What I wish to do for much of this chapter is to examine how texts need to be read, treated and analysed in the context of their production and consumption – i.e. the time when they were written and read – and how the time of production of a text reveals a temporal point that can be seen to be a context, or a point from which we may integrate the content with the archaeological record of the same time. The problems raised in discussion of how we read and combine texts

and material evidence may be embedded within the institutional power struggles that established the modern discipline of archaeology – that will be a focus of Chapter 2 of this book, but needs to be introduced here. To create a modern discipline of archaeology in the mid-late 20th century, it was seen as essential for archaeology to assert its place as much more than the ‘handmaiden of history’. In so doing the discipline developed an emphasis on the independent study of material culture mostly without reference, if possible, to texts – the province of historians. The teaching of historical archaeology in UK universities at times appears to feature a deliberate attempt to avoid an engagement with textual evidence, and to focus solely on the analysis of archaeological data. This may be pragmatic; after all few single honours archaeology undergraduates will be versed in textual analysis, but any joint honours student taking ancient history will find the avoidance of textual evidence baffling, and will be even more confused when their use of this evidence in essays is not rewarded. As

Anders Andrén (1998) stresses, there are archaeologists who do not wish to engage with any textual sources and instead derive their conclusions entirely from the archaeological data (see examples discussed in Sauer 2004a). There are also some archaeologists who still wish to use the archaeological record to ‘prove’ that the textual evidence is simply wrong or a misrepresentation of the past (Allison 2001). These approaches are similar to those of the historian, who sees no value in archaeology and suggests that archaeology can only show you what you knew already from your texts – all such practitioners are missing the point. As Andrén (1998: 4) points out, the two types of evidence, texts and artefacts, are two different human discourses. Relating these together can be straightforward or down-right impossible. The key to success in this area is keeping a wary eye on the contexts, dates and type of materials that are being integrated from these two quite different discourses. Part of the problem for ancient historians in dealing with archaeology is

that we have read the texts that seem to describe the places that survived as archaeology. We cannot disaggregate or completely lose our knowledge of texts prior to evaluating a piece of archaeological evidence. Nor would such a process be desirable, since texts can aid the development of archaeological knowledge. For example, in approaching the archaeology of King Herod’s new city at Caesarea most scholars will have read Josephus’ description of the new city (Josephus Wars 410-13; Antiquities 15.331-38). Caesarea, like Augustus’ Rome, was rebuilt in marble and its buildings were seen by Josephus to have been worthy of the person from whom the city gained its name, Caesar (Augustus). Josephus’ harbour at Caesarea is the size of the Piraeus in Athens. What we see here is a rhetoric that praises the city and its builder/monarch (compare the later writer Menander Rhetor’s How to Praise a City). However, what can be identified in the archaeological record coincides with the rhetorical description of the harbour itself, but there are monuments within Josephus’ text that have yet to be identified (see Vann 1992). This is an example of a text aiding archaeology to formulate a strategy for investigating a site – Josephus’ two accounts aid the fieldwork on site. Importantly, in this example, the texts of Josephus and the archaeological site are broadly contemporary. In effect, both the text and the archaeological data are derived from a very similar (if not congruent) context: the 1st century AD – hence in this example the convergence of the two sets of source material is less problematic.