agricultural and pastoral regimes. Palynological analyses show that forest
cover in the mountain zone was still very dense. Some of the upland sites are
easily identifiable by the dry stone circuits on their summits.^2 Enclosures
afford a level of protection from other people, pens forflocks and herds, and
act as wind-breaks. So enclosure does not automatically mean that people
perceived dangers and uncertainty, and tried to defend themselves as best
they could. If we look at the wholefirst millenniumbc, and how settlement
patterns changed over this period of time, then it becomes clear that
defensive needs were often factored into the design, whether we are thinking
of upland sites that look like small fortresses, or lowland farmhouses, whose
most distinctive visual and planning feature was a multi-storey tower. The
need for protection was not new, but communities in the north were
evidently better able at this time (or had moreflexibility) to protect
those things as well as people that mattered most to them—their immedi-
ate families, kin, and social peers; the annual cereal crop and other stored
produce; and manufactured goods that might be considered valuable.
The success of this process of social expansion into new ecological
zones, in both upland and lowland locations, is reflected in a population
increase, expressed in the overall number of sites registered as dating to
the second half of thefirst millenniumbc. As we have seen in Chapter 2,
the intensification offield research in recent decades has multiplied the
number of known sites, particularly the range of smaller locations. We can
detect a new relationship in the landscape between traditional built struc-
tures—family farms and community-based facilities—and new forms of
social investment, including sanctuaries and funerary monuments. These
trends substantially contributed to the region’s economic framework in
the second half of thefirst millenniumbc.
The pattern becomes less easy to comprehend in the third and second
centuriesbc, when urban centres grew at the expense of rural ones, but
also became vulnerable to external pressures, in the form of aggression
by new enemies. Attacks by Gallic soldiers are most frequently men-
tioned as the cause, particularly around the time of the large-scale Gallic
attacks on Macedonia, Thrace, and the sanctuary of Apollo at Delphi.^3
Less often referred to, but probably of greater overall economic
(^2) Gerassimidis 2000 on palynological data; Baralis 2010, 247–56, for a survey of Early
Iron Age upland enclosures in Aegean Thrace; on the continued use of such upland
sites, see Misaïlidou-Despotidou 2008; E. Bozhinova and A. Andonova,AOR2008 [2009]
218 – 19;AOR2009 [2010] 136–7, for fourth- to third-centurybcactivities at Dragoyna,
Purvomay district.
(^3) Plb. 4.45.9–46.4, 52.1–2; Liv. 38.16; Just. 24.4.5–6, 6.1.5, 8.9.16, 25.1–3; Paus. 10.19–23;
see the contributions to Vagalinski 2010 for a recent reassessment of the Celtic presence in
Thrace.
130 Thelongue duréein the north Aegean