represented, and the research strategy was thus subjected
to frequent revision as fresh material came to light which, on the grounds
of the spatial and temporal co-ordinates and (where hypotheses could be
formed) functions of the sites concerned could not be overlooked. The site
list, Appendix 1, and the site distribution maps (Figs. 2—9) record the
quality and location of the material examined, with a list of published
reports which were utilised without examining the pottery at first hand.
The number of major sites examined for which useful published information
on the pottery was available prior to the commencement of museum research
comprised slightly under half of the total of major sites that were
included in that research. In the cases of the most extensive sites, such
as Canterbury, Rochester, Springhead and Richborough, only a portion of
the material taken account of in the overall research programme was
available in published form.
The sites that have been the subjects of extensive
excavations presented a problem of assemblage selection. The response to
this problem varied according to various factors of publication,
stratification and the accessibility of the material. The numerous pits
excavated at Richborough provide most of the stratified, or at least
context relateable, pottery from that site, and a selection was made of
those with comparatively large assemblages in storage. The excavations of
the Canterbury Archaeological Trust, which were under way during the
research period, provided a data base for the city, and this base was
supplemented by material from earlier excavation programmes that was
available in both publications and the Canterbury Royal Museum. Much of
this material was securely stratified, but it was not always possible to
ascertain the retention strategies that had dictated the nature of the
surviving assemblage; full quantification of all periods of the Roman city
was thus not achieved. The imperfections of the resultant coverage should
be overcome during future research to be conducted on the
|
|
Roman pottery recovered by the Canterbury Archaeological Trust. The site at
Springhead presented a different pattern again, as most of the pottery from
the early seasons was discarded by the excavator, the late Mr W.S. Penn,
after summary publication (S. Harker, pers. comm.). The pottery reports of
this period of excavation employ an over-generalised type series (Penn 1957;
1958; 1959; 1960; 1962; 1964; 1967; 1968) deemed unsuitable by Mr Penn’s
successor, the late Mr S. Harker and this author as a basis for analysis of
this lost material. The Springhead research has thus been confined to Mr
Harker’s recent series of excavations (summaries in Frere 1977; Goodburn
1978; Wilson 1970; 1971; 1972; 1973), which have produced sufficient
material accessible at the time of research to provide full temporal
coverage, including quantified stratified assemblages for most of the Roman
period.
The material from two sites in Rochester was loaned to the
present author by the excavator, Mr A.C. Harrison, for the writing of
specialist reports. This material covered the late first/early second to
late fourth/early fifth century overall, and was quantified, rendering
personal examination of the material published from two other sites
(Harrison and Flight 1968; Harrison 1972) superfluous within the terms of
the research design. The published material has been taken account of,
nevertheless. Examination of the pottery from the Lullingstone villa and the
Classis Britannica fort at Dover was confined to the material selected for
publication, as this was considered by the excavators (Lt.-Col. G.W. Meates
and Mr B.J. Philp, respectively) to provide a full range of vessel types
over the whole of the periods of occupation. Subsequently, the pottery from
the former site has been re-analysed by the present author (Pollard 1987).
In other instances where a larger volume of material exists than it proved
possible to examine, assemblage selection was based upon a precursory study
of most or all of the collection,
|