The archaeology of the Roman military, unlike in some other areas, integrates the written record, whether in literary texts, epigraphy or on papyri and writing tablets, with the archaeological evidence from surveys and excavations. The scale of the enterprise should not be underestimated, since there is an abundance of documentary material that reveals the everyday experience in the military (see Campbell 1996; Fink 1971; Bowman and Thomas 1983, 1994; Bowman 1994; Salway 1965 is still fundamental for the study of epigraphy). This evidence provides a more informed textual context to approach archaeological material, yet there is no certainty regarding, for example, the dating of the writing tablets from Vindolanda and relating these to the archaeological evidence from the site itself. This should not diminish the value of this evidence from Vindolanda for the elucidation of the social system within the fort that included evidence for married officers, their families and slaves. However, the editors of the tablets took care to place women and children not connected to officers outside of the fortress in the vicus or nearby civilian settlement (Bowman and Thomas 1994). More importantly, the existence of these letters leads to the realization of the social and cultural changes (and also of that to the landscape) that were felt in the north of the province of Britannia at the end of the 1st and beginning of the 2nd century AD (Haynes 2002). The engagement with these changes was total for the soldier, as can be seen in the writing tablets from Vindolanda, but for the natives or civilians of the region, their conception of these changes may have been quite different resulting at best in a partial engagement with the cultural landscape that was dominated by the soldiers on the frontiers and their needs. Roman social historians tend not to focus on the history of the military, and leave such matters alone or to persons who might be described as military historians interested in the logistics, strategy and tactics of fighting for the empire. However, we need to accept that there is a social history to be written and the archaeological evidence

from Roman forts can provide a means to opening up the Roman fort as a milieu for sociological analysis. The first step to doing so is to alter our preconceptions of these forts and to accept that these form archaeological locales for the exploration of all the issues that might be evaluated in the study of civilian sites from the Mediterranean – for example, at Pompeii. Historians are very familiar with the concept of a division in Roman

society between those in the army and those who may be classified as civilian. The military was a different situation or different form of habitus from that experienced by the civilians in the Roman empire, symbolized and exemplified perhaps by the soldier wearing a sword and sword belt (Haynes 1999a). Soldiers, particularly in the empire, have been conceptualized as existing or being in a different habitus, that of being a soldier. Their difference is also encountered in Roman law, and a provision from the time of Augustus that soldiers could not enter into marriage in the same way that civilians did (see Scheidel 2007 for discussion). This distinction on marriage has in the past been interpreted within Roman archaeology as a simple ban on not just marriage but any form of cohabitation, and has caused interpretations of military archaeology to become focused on the male gender of the soldiers, their equipment and the organizational structure of the army as a whole. Underpinning the conceptual difference is not just this, but also a tradition of study that need not fully engage with the premises of social history and social archaeology. The separation of the archaeology of the Roman military from the history and archaeology of Roman civilians is unhelpful and is challenged by the evidence produced at military sites and through analysis that links explanations of the archaeology of the military to the civilian sphere. Some of these are discussed in this chapter in connection with the interpretation of the early 2nd century Roman fort at Vindolanda (Figure 7.1). Other material is easy to locate, but what I wish to set out here, quite briefly, is the scale of re-interpretation that is required to enable the re-engagement of the archaeological evidence for the military with current preoccupations in Roman social history and social archaeology.