to extrapolate from short sections of road to create orthogonal systems is
very strong, since it fulfi ls our expectations of what we want to fi nd, but
additional evidence when it comes can often show how our desires get the
better of us, as we will see when we look at Canterbury (below p. 145). All
we know we have at Silchester is two roads arriving at the centre of the site
at approxim ate right angles. We also have the road leading away from the
centre to the east, which is not on the same alignment as the Roman street
grid, so this may also be early. If so, the sum of the evidence is of three
streets coming together in the centre of the site, which is where the forum
eventually gets built. This may not be as exciting as imagining an Iron Age
street grid, with all its implications of pre-Claudian Romanisation, but this
more mundane interpretation is how I see the evidence. I have reconstructed
this suggestion in Figure 7.4, adding to it some of the ‘mis-aligned’ build-
ings, which may have early foundations. In this version, the erection of a
major new building in Period 4 at the junction of these roads in the Tiberio-
Claudian era, sweeping away pre-existing buildings, makes sense for a new
set-piece construction at the physical centre of the town.
Together with the Late Iron Age inner earthwork, the early layout envis-
aged here has some parallel and planned elements too, but is by no means an
Iron Age grid. If this arrangement originated in the late fi rst century BC, its
contrast with the Flavian street grid is in many ways similar to the contrast
between some Augustan period forts and their Flavian successors. Augustan
forts in Germany tend to be in large polygonal enclosures, which often have
right angles in some but not all corners. They also have planned but not
always orthogonal layouts – for example, the fortress at Marktbreit on the
river Main (Pietsch et al. 1991; von Schnurbein 2000) – whereas their Flavian
successors are the much more familiar playing-card arrangements. If early
Calleva was established by one of the Later Iron Age dynasts who had some
experience in the Roman army, then some structural similarity with Augustan
forts would not be too surprising.
In the Claudio-Neronian phase, it looks as if there was a variety of large
houses spread out across the town (of which the diagonal house in Insula IX
under excavation by Fulford and Clarke was one), to which were being added
the new proto-forum in the centre, a bath-house and the amphi theatre.
None the less, this cityscape was radically altered in the Flavian period by the
imposition of a new orthogonal street pattern. Nothing was allowed to get
in the way of this. If the front portico of the recently built bath-house was
on the line of a newly ordained road, then it would have to be demolished
and rebuilt, leaving the path clear. This tells us something about the level of
authority which an individual, or council, could wield in the Flavian period.
The contrast between early Silchester and the later more orthogonal layout
reminds me of the contrast between the Augustan oppida of Gaul and their
‘Roman town’ successors. In the Augustan period at Mont Beuvray, the early
capital of the Aedui, grand residences emerged as well as a temple/forum
THE MEMORY OF KINGS
138