- Chapter Ten -
4-5 ha; Levroux is large at about 20 ha (Biichsenschiitz et al. 1992),'The sites around
Clermont-Ferrand in the central Massif, however, present a different picture in both
the density of sites and chronology (Malacher and Collis 1992). In addition to the
three or four sites at Aulnat, surface finds hint at a number of other sites, one of
which, Gerzat, 5 km north of Aulnat, has been extensively excavated. Though these
sites lack the nucleated population of Manching (the individual clusters of houses are
separated from one another by 500 metres or so, within an area of about 10 square
kilometres), they represent the same concentration of industrial activity as Manching.
The site of Aulnat-La Grande Borne was extensively involved in working of iron,
gold, silver and bronze, including coin manufacture, as well as bone-and glass-
working, and probably textiles. Gerzat was more specialized in ironworking, but had
glassworking as well. This settlement pattern was already coming into existence in
the third century Be.
During La Tene Dl these open settlements were abandoned over much of Gaul.
Only in exceptional cases such as Roanne was there continuous development
into a Roman town. Instead there was a shift to defensive situations on hilltops,
peninsulae or islands, the oppida encountered by Caesar. Only rarely is there a case
of straight displacement of the settlement - Levroux and Berne are examples. More
normally the new defended sites are much larger than the earlier open settlements
and more densely occupied, implying that several settlements combined to defend
themselves, similar to the Greek process of synoecism. However, we know little of
the effect of this process on the overall settlement pattern. The Aulnat sites were
abandoned, whereas Gerzat seems to have survived longer, but there is little
comparable data from elsewhere in France.
The revolution in settlement pattern occurred while the Nauheim brooch was in
fashion for female dress - these brooches are generally the earliest types from the
major oppida such as Mont Beuvray, and have actually been found in the ramparts
of some oppida such as Berne. In absolute dates the change occurs around 120 Be.
We have dendrochronological dates from the earliest levels at Besan~on and Yverdon
(Orcel et al. 1992), and there is a similar date from one of the phases at Manching.
The Roman conquest of Provence in 125-123 Be and the subsequent explosion in
trade may not be unconnected.
Not all sites date back this early; indeed ramparts continued to be refurbished,
and even new sites founded, as at Gergovie, as late as the Augustan period. In some
areas such as Brittany and Normandy the enclosures were constructed, but no major
occupation took place. In other areas such nucleation may not have taken place, as
there are great gaps in the distribution, for instance in western France. In some cases
this may be because sites are masked by modern towns. The archaeology of some
sites mentioned by Caesar has yet to be found - Arras and Paris are examples - and
previously unknown sites are emerging as urban archaeology develops - Chartres,
for example. Nevertheless, urbanization was by no means universal even in areas
where social and political development seems to be relatively advanced, and
especially in northern areas such as Belgium and the Netherlands.
Another peculiarity is the ephemeral nature of some of the sites. Though some
sites, like Besan~on and Reims, and presumably Paris, once founded, have remained
in permanent occupation ever since, others, though densely occupied, were