Writing tablets

Before turning to review these themes, it is worthwhile making a few points about the key new evidence provided by the writing tablets from the Bloomberg site. The writing survives on wooden tablets (see opening figure), once covered with wax, where the metal stylus used to write on them has penetrated through to the wood, creating scratch marks. Reading these texts requires incredible skills, not only the ability to read cursive Latin handwriting, but also the patience to decipher incomplete or superimposed texts. Such reading is inevitably a recursive process with the knowledge of the fragments read being tested against understandings that rely on both an exceptional grasp of Latin and an in-depth comprehension of the Roman imperial world. Deciphering, reading, re-reading and thinking about the fragments thus leads to a final understanding of the text. Roger Tomlin has made a truly outstanding contribution, providing authoritative editions and readings of 79 texts that date to between c. AD 50 and 90. These provide an entirely new perspective on Roman London, and one that will take a considerable time to assimilate fully.

Given Tomlin's skill and expertise in presenting his work, it is a great pity that the volume provides only a comparatively short discussion from him about the contribution of the tablets to the history of Londinium. There is an excellent introduction to the epigraphy and the form of the tablets, as well as a summary of their context, but the broader historical discussion is limited (Roman London's first voices pp. 51–58), even if a lot of further detail is provided in the commentaries on the individual texts. It would have been interesting to read a little more discursively about his wider-ranging thoughts on their significance, as provided briefly in an interview with Current Archaeology (Tomlin & Jackson Reference Tomlin and Jackson2016).

Latin specialists may wish to debate some of Tomlin's readings, but I am not qualified to do so. I think, however, there are one or two issues about the reading of proper names that should be noted. The recognition of a Latin word in these difficult and incomplete texts depends on the reader's knowledge of both vocabulary and grammar. Reading place names and tribal names, however, is different, and here we are restricted by both our more limited knowledge (we do not know the names of all the places in Roman Britain) and by variations in the spellings of those that are attested. This is important as certainty about the readings of tribal names and place names in these tablets may have great historical significance. Thus, tablet WT45, dated to 21 October AD 62, records the transport of provisions from Verulamium to Londinium very soon after the destruction of both settlements in the Boudiccan revolt, and so has implications for our understanding of both the date of the revolt and its impact (Roman London's first voices p. 158). The text here is clear, so the reading and interpretation are not in doubt. In other instances, readings are less certain. In tablet WT59, the context of which is dated to AD 62–65/70, Durob. . . is read as Durobriuae and identified with Rochester in Kent without discussion. This is problematic since the prefix Duro. . . is widely used in Celtic place-names (see Rivet & Smith Reference Rivet and Smith1979 who list nine known places in Britain), while the name Durobriuae was used for two settlements in Britain, the other being Water Newton near Peterborough, a site which, unlike Rochester, has sound evidence for a fort (at Longthorpe) in occupation at about this date (Frere & St Joseph Reference Frere and St Joseph1974: 38–39). Similarly, the tentative identification of Viroconium (Wroxeter) in tablet WT23 is of note as there was also a fort in occupation here at this date (White & Barker Reference White and Barker1998: 32–50). Although the reading of this place-name is uncertain, there is no obvious known name as an alternative. Possible links to military bases in these texts are potentially of wider interest (see below), so their correct identification is crucial. Historical significance also rests upon Tomlin's identification of the First Cohort of Vangiones (WT48), although it is not clear why he takes this reading to be certain (p. 164). In several of these instances, a little more discussion and justification of the conclusions drawn would have been helpful. I make these points tentatively, not to detract from the value and self-evident quality of Roger Tomlin's contribution, but because it is important that we evaluate this new evidence properly before it passes without question into the literature.

Finally, we should note that Damian Goodburn's discussion and analysis of the manufacturing of the writing tablets, which are largely of silver fir (Albies alba), are also impressive (Roman London's first voices pp. 8–13). His conclusion that they were manufactured from barrel or cask staves has broader relevance, suggesting that barrels were a key mode for the transportation of commodities from the earliest stages of the development of London.

The origins of Londinium

There has long been a debate about the origins of Londinium, with some arguing in favour of its establishment around a fort of the invasion period, and others suggesting that it was a planned urban foundation. Most recently, Perring (Reference Perring2011, Reference Perring, Fulford and Holbrook2015) has pressed the case for a military origin, while Wallace (Reference Wallace2014) has drawn together evidence that supports the planned-foundation model, and has carefully questioned the evidence Perring (Reference Perring2011) cites in favour of the military case (Wallace Reference Wallace2013). I still find the argument presented for even a short-lived invasion-period fort unconvincing, but irrespective of this, the now extensive excavated evidence from the AD 40s and 50s clearly shows that any initial military presence at London was marginal to the growth of the settlement prior to its destruction in the Boudiccan revolt of AD 60/61. This evidence, as synthesised by Wallace (Reference Wallace2014), demonstrates the processes by which the planned town had been established and thrived prior to the revolt. It is also evident that commonly held models for the relationship between the army and London's population need re-thinking (see below).

The publications under review here offer new evidence about the character of the early settlement, which is of direct relevance to the debate about urban origins. Both the Archaeological Studies volumes on excavations south-east of the later forum and in the north-west of the City of London add small pieces to the jigsaw of excavated evidence for this period, but the Bloomberg tablets provide an entirely new dimension. The tablets largely date to the period after the Boudiccan revolt (down to the AD 90s), but 12 predate it and these include one (WT44) that is very important, a legal record of a sale dated 8 January AD 57. This is the earliest-dated document, about a decade after archaeological evidence shows that the planned town was founded. These early texts have little direct bearing on debates about any possible short-lived fort, but they do emphatically support the evidence from Tacitus that Londinium in AD 60/61 was a place “well-known because of its large numbers of merchants and great quantities of merchandise” (Annals 14.33; Birley Reference Birley2005: 45). As a group, the tablets show not only that the population represented was heavily engaged in trade (and trade conducted within the structure of the Roman legal system), but also that it was predominantly civilian and included a significant number of Roman citizens in addition to freedmen and slaves. It thus seems certain that this was an immigrant community, supporting earlier suggestions that London was founded by a group of traders from Gaul. It should of course be emphasised that these texts do not represent the whole community: no women appear and the texts are administrative records that are biased towards economic and legal transactions. It may also be that they are representative of only one geographic part of the settlement, but as the evidence stands, they lend weight to the conclusion that whether or not there was a fort here in AD 43, the principal dynamic for the growth of the settlement came from a trading community by AD 60 (see also below). This provides an important new perspective on how the province as a whole developed, arguably shifting the balance away from the Roman state and towards the agency of independent incomers taking advantage of new opportunities.

It should also be noted that the tablets shed some further light on the related debate about the status of London (summarised by Wallace Reference Wallace2014: 18–20), with Tomlin (p. 56) concluding that two of the texts (WT51, dated to AD 76, and WT37, no later than AD 70/80) support the conclusion that London was neither a municipium nor a colonia at this date. Given the evidence noted above for the systematic planning of the early settlement, this emphasises the fact that London does not fit easily into the general pattern of urban development assumed elsewhere in the provinces.

The role of an immigrant Gallic community is also witnessed over a century later, at the sanctuary excavated at Tabard Square. An inscription (Temples and suburbs pp. 192–93) found during that project records a dedication to the god Mars Camulus (whose origins lie in Gaul), most probably in the period AD 161–169 or AD 177–180 (Tomlin et al. Reference Tomlin, Wright and Hassall2009: RIB III, 3014). The dedication was made by a Gaul, a citizen of the Bellovaci, who is recorded as a moritex, most probably a member of a guild of traders (see below). The epigraphic date range coincides with the archaeological evidence for the foundation of the temples (Temples and suburbs p. 35), and perhaps indicates that the whole complex was the creation of a trading community. Whether or not this was the case, the inscription provides clear evidence that trading relations with Gaul continued well after the period represented by the Bloomberg tablets. We should, incidentally, note how the Tabard Square temples were located adjacent to a navigable channel of the river and so would have created a visible landmark for those approaching by boat, reinforcing the link with the Gallic traders, who may have remained key to the society as well as the economy of Londinium.

The role of the military

Related to the question of the origins of Londinium is the broader issue of the relationship between the army and the city. Discussions of Roman Britain as a whole often draw a clear distinction between the civilian and military populations, and invoke the latter (or undefined Roman ‘officials’) as the prime agents of change, denying the indigenous population any significant role. This harks back to an era of our own imperial past when ‘the natives’ were seen as inferior and incapable of achieving ‘civilisation’ without ‘help’ or ‘direction’.

Surprisingly, vestiges of such attitudes can be found in some of the work under review. Thus, in the Tabard Square volume, military engineers are seen as responsible for land-drainage operations (Temples and suburbs p. 238), soldiers or ‘officials’ are suggested to have been the occupants of buildings supposedly located to control the local road network (p. 240), and interpretatio Romana is described as “the interpretation by the superior Roman culture of indigenous deities as Roman gods” (p. 254). Similarly, in the volume on glass-working, an army garrison based in the Cripplegate fort is assumed to have had a defensive role in relation to London's port, as well as a function in provisioning the province, hence the suggestion that glass production developed under military influence (Glass working on the margins of Roman London pp. 101–102). This seems unlikely given the evidence collated long ago by Mark Hassall (Reference Hassall and Strong1973), which shows that the fort housed soldiers and officials on secondment from the provincial army who were in the service of the governor or fulfilling other administrative roles.

In fact, these recent excavations provide important evidence to help us develop a new understanding of the complex relationship between the army and the rest of the province. We should begin by appreciating that Roman Britain was a province in which there was always a substantial military presence. We can see this in the evidence of the Vindolanda tablets, which record soldiers from that fort's garrison posted in London (Bowman & Thomas Reference Bowman and Thomas1994: Tab. Vind. II, 154; Bowman et al. Reference Bowman, Thomas and Tomlin2010/2011: Tab. Vind. IV, 858; also perhaps Bowman & Thomas Reference Bowman and Thomas1994: Tab. Vind. II, 310). Second, the Roman state was militarised, so that many roles were fulfilled by soldiers rather than civilians, and this will have been especially significant in a place such as London that had a major administrative function. Against this background, we need to consider the evidence from two of the sites reviewed here: Bloomberg and that to the south-east of the forum. The evidence from both relates principally to the period after the Boudiccan revolt when we know that a fort was constructed in the area to the south-east of the later forum at Plantation House (Dunwoodie et al. Reference Dunwoodie, Harward and Pitt2016). The area reported in the Archaeological Studies volume by Wroe-Brown discussed here covers the adjacent site to the west, within the interior of the fort that was constructed perhaps in AD 63 and continued until c. 85. This excavation dealt with very truncated and damaged deposits, and did not extend far enough either to the north or to the west to reach the line of its postulated defences (Dunwoodie et al. Reference Dunwoodie, Harward and Pitt2016: figs 58 & 62). A number of fragmentary buildings were examined (Roman occupation south-east of the forum pp. 21–25, fig. 19), but were insufficient to provide a detailed understanding of the fort plan. Two observations can, however, be made. First, the evidence from this site together with that from Plantation House does not support the idea that this was a regularly planned fort with standard building types. Second, the buildings that have been recognised appear to have lined the margins of the fort, perhaps leaving the area in the centre relatively clear. Although one needs to be cautious, both in expecting a standardised fort plan and in reading such fragmentary evidence, this does raise questions about what this fort was for and how it might have worked.

The Bloomberg texts provide some important information about the military at this period, with evidence from three tablets (WT33, WT55 and WT48), each of which may be associated with military units sent from Gaul to reinforce the army in Britain following the Boudiccan revolt (Roman London's first voices p. 56). They include reference to Classicus, Prefect of Cohors VI Nerviorum (WT33), as well as soldiers from the Vangiones (WT48) and Lingones (WT55), in addition to three others from an unspecified unit (WT62). Given that they do not all come from the same unit, they do not provide help in the identification of a specific garrison that may have manned the fort. Additionally, we have a singularis (WT56), a member of the governor's bodyguard, and a veteran soldier, perhaps serving in an administrative role (WT20), adding to the picture of a military presence drawn from a variety of units—a pattern consistent with that described by Hassall (Reference Hassall and Strong1973) as noted above. Set alongside the other evidence for the presence of the governor in Londinium (Roman London's first voices p. 56), and for residence of the procurator after the Boudiccan revolt (Collingwood & Wright Reference Collingwood and Wright1965: RIB I, 12), we might suggest that the Plantation House fort, rather than housing a garrison, was a predecessor to the Cripplegate fort, providing a secure compound for soldiers passing through the town and for those on secondment to the governor's staff. If so, the destruction of the fort may have coincided with the construction of a postulated timber predecessor to the Cripplegate fort, and the construction of the adjacent amphitheatre (Bateman et al. Reference Bateman, Cowan and Wroe-Brown2008: 19–38), placing the military base out of the way of the developing civic focus of the city. Although a recent review has provided no direct evidence for a timber predecessor to the Cripplegate fort (Shepherd Reference Shepherd2012: 153–54), one does seem probable.

Economy, death and belief

Another military aspect of the Bloomberg tablets relates to their economic implications. If we are right in thinking of London at this early period as being a trading hub, it must surely be that the merchants were doing much of their trade with the army. This fits with the idea that military supply was not directly organised by the army, with monolithic ‘supply contracts’, but rather that individuals and units entered into arrangements with a plethora of private contractors, creating something of a trading bonanza for merchants after the annexation of a new province. If this hypothesis is correct, then the attested link to a military castello in the territory of the Iceni (WT39), as well as the possible contacts with Wroxeter (WT23) and Durobriuae (WT59), both sites with forts at the time (see above), may have additional significance.

The broader economic evidence from excavations in London is also important in the context of current debates about the nature of the Roman urban economy. If the settlement began as a boom town reliant on the state in the wake of the Claudian invasion, how far did its economy become more self-sustaining and how did this relate to the broader development of the province? The conventional answer has been to see the city's premature decline by the third century AD as a product of the rebalancing of provincial economies while production grew and reliance on imports declined (Millett Reference Millett1990: 123). The evidence from the sites examined here does little to detract from this explanation, but the discussion of glass production in the Walbrook Valley provides clear insights into the complexities of specialist production in the period AD 140–160/70. Wardle and Freestone document with care the import of raw materials (in three batches) from the Eastern Mediterranean, alongside which there was also a large-scale and systematic process of gathering broken glass (cullet), perhaps collected from a wider region around London. They also convincingly demonstrate that the glass-blowers were peripatetic. On the one hand, this implies a sophistication of industrial practice and a developed network of supply and production across the Roman world. On the other hand, it offers something of a corrective to the common notion that the Roman economy developed large-scale markets with high volumes of goods: were this to have been the case in London, the largest centre in Roman Britain, why could it not sustain a permanent glass-blowing industry? These results tend to support the view that the Roman provincial market remained less developed than some current commentators assume.

Finally, let us turn to think about the culture of the people who inhabited Londinium, where the volumes under discussion also provide valuable new evidence about burial and religion. The careful work on the cemetery in the upper Walbrook Valley at Finsbury Circus—an area that fell outside the Roman city wall when it was constructed probably early in the third century AD—provides some fascinating new perspectives on burial practices.

First, the cemetery itself is somewhat unusual in that the graves examined occupy a linear zone, lying alongside a road and into a partly canalised stream running roughly parallel with the line eventually followed by the city wall. Most of the graves were carefully aligned, without intercutting. Unlike many other cemeteries, however, there is little evidence for the definition of burial plots or groupings. As the area was at least seasonally wet, located as it was in an area of braided streams forming the headwaters of the Walbrook, it is tempting to see this as a marginal location and to assume that those buried were probably of lower status. The detailed analysis of the human remains and the grave goods, however, offers little support for this hypothesis, as they apparently form a reasonably representative sample of Roman London's population (Upper Walbrook Valley cemetery pp. 134–35). We are nonetheless left with some sense that those buried may have been liminal, an impression reinforced by the presence of leg-shackles on two of those buried (pp. 95–96). The cemetery has other unusual features. The inhumation rite rather than cremation was predominant from the outset, and this may perhaps have more significance than the authors believe (p. 87). There is also consistent evidence for the disposal of horse bones in the area of the cemetery, a phenomenon noted in certain other cemeteries (pp. 83–86) and which must surely be of significance. For me, the cemetery provides a clear reminder of how misleading it is to think that there was any norm for burial in Roman Britain, and of how much more we still have to learn about the meaning and variety of burial practices.

Second, the careful study of the erosion of burials by the streams in this area leads the authors to conclude that the much-debated finds of isolated skulls in the Walbrook are a product of the disturbance of formal graves by water action rather than as acts of votive deposition (Upper Walbrook Valley cemetery pp. 127–32). This is soundly argued and, at least in the case of the Walbrook, removes any implication that ‘Celtic’ religious practice remained central to life in the heart of the Roman city.

This brings us back to the religious complex at Tabard Square, where we have, for the first time, the large-scale modern excavation and publication of a temple-complex from Roman London. Certain points about the site are absolutely clear, particularly the evidence that the sanctuary was laid out de novo after the levelling of previous structures around AD 160. The plans of two so-called Romano-Celtic temples, as well as the Mars Camulos inscription (see above), leave no doubt about its identification as a religious site. Heavy robbing, however, has left only traces of some very splendid statuary and remarkably little evidence for any votive activity associated with the site's monumental phase. There is some convincing evidence of ‘special deposits’ in ditch 8, which defines one side of the temenos, but this predates the temple buildings. A very careful analysis and discussion of the finds provides little else that convincingly indicates votive activity, apart from a small number of burnt animal burials. The one feature that is not discussed, but which might possibly be significant, is the group of four seal boxes (Temples and suburbs p. 160), as Derks (Reference Derks, Metzler, Millett, Roymans and Slofstra1995) has argued that these may have been related to the ritual of the vow at such temples. A dissonance between structural and finds evidence at religious sites is not unusual, but might have been given more prominence here.

The discussion of the site (Temples and suburbs pp. 231–67) takes in a very broad sweep, but not all aspects of the fascinating evidence are dealt with in depth. Highly unconvincing suggestions, such as an association with the suovetaurilia (the sacrifice of a sheep, a pig and a bull; p. 247) are included, while the implications of the unique and important inscription are hardly touched upon. Similarly, there is an odd discussion of the idea of pilgrimage (p. 257), but no debate either about the site's position beside the river channel, which may well relate to the Gallic trade (see above), or its possible boundary location on the margin of the Thames wetlands. Noting the observation that the term moritex, which appears in the inscription, relates to a guild (Dondin-Payre & Loriott Reference Dondin-Payre and Loriott2008), I would tentatively suggest that we might think of this complex not as a public temple as the authors assume (pp. 253–54), but rather as a private guild-complex with its associated shrines and statuary. This might account for its marginal location in Londinium, the nature of the hall-like structures (buildings 18 and 19), which were evidently to allow large gatherings of people, and the paucity of votive material.

It will be obvious from the discussion above just how far our understanding of Londinium has progressed in recent years as a result of developer-funded archaeology. The issues for debate raised here illustrate the quality of this work and the publications discussed. We look forward to the continued flow of similar volumes in the future. At the same time, it is also clear that the Roman city would greatly benefit from an up-to-date synthesis of the mass of evidence so that future research can be focused on improving our understanding even further.