resulting in the extraction of groups which appeared to provide full
temporal coverage; in some cases this could only be ascertained by
comparison of the pottery with stratified material previously studied, as
site plans had not been archived with the material (e.g. Brenley Corner,
Kent, and Pevensey, Sussex).
The major museum collections within the study area were studied
in depth, and several of the local professional and part-time excavation
groups visited, in the course of data collection. Some of the more important
site collections, as judged by site function or the extent of excavation,
could not be examined owing to the constraints of space and time experienced
by the excavators, thus the villa at Eccles and the Saxon Shore forts at
Reculver and Dover have had to be overlooked, except for the analysis of
publications.
III. SITE CLASSIFICATION
There are two approaches to classification that may be entertained here,
by the application of titles documented in classical sources such as ‘urbs,
‘oppidum’, ‘villa’ or ‘vicus’, and by the description
of the archaeological evidence alone. A system employing the latter approach
is less evocative and more verbose than one employing the former, and most
classifications adopt a compromise whereby the more ordered sites, where
planning is apparent in the street-grid or the design of a house and its
associated structures, are given titles of a classical nature, whilst the
remaining sites are described in terms of their location, structures, or
induced function (cf. Ordnance Survey 1956; 1978). Thus the titles of ‘civitas
capital’, ‘villa’ and ‘town’ are entrenched in the archaeological
literature despite the extreme rarity of a classical warrant for their
application to specific sites. Discussion over just what kind of site merits
|
|
these titles is frequent, and rarely conclusive (e.g. Johnson 1975; Percival
1976; Todd 1970; and Wacher 1975), as the functional interpretation
of structural remains is rarely unequivocal. The classification adopted by
the Ordnance Survey for its most recent ‘Map of Roman Britain’ (1978) is
not explained in detail in the text that accompanies the maps; thus, a villa
is defined as’ a ‘rural building of substance’ which has ‘at least
wall foundations Of stone’ with additional unspecified attributes that
distinguish villas from other buildings of stone-based construction ‘where
the villa character is suggested but not finally proved’ (ibid., 8)
(cf. Loughlin 1977, ‘104). Nevertheless the provision of a gazetteer,
albeit without bibliographic references, enables the researcher to ascertain
which sites are grouped under each major class.
The Ordnance Survey’s most recent scheme is favoured by the
present author as being comprehensive, if occasionally contentious; several
of the classes of site defined by the Ordnance Survey are not present in the
study region, notably all but one of the military categories (forts are the
exception), and certain of the civilian ‘communal settlements’ —coloniae,
spas, and burgi. Three types of site grouped by the Ordnance Survey under
‘Other finds of Roman material’ have been extracted by the present
author for the purpose of providing a fuller description of the context of
ceramic finds. These are listed in the following scheme as Classes VII —
IX. The following system of classification is thus arrived at:
I. Major towns (i.e. communal civilian settlements: civitas
capitals and ‘towns’ — Ordnance
Survey 1978);
II. Forts and associated settlements;
|