Money and society in pre-Roman Britain Barry Cunliffe
Nearly 40 years ago, in April 1940, Derek Allen presented most of the sample unusable because of lack of locational
his now-classic paper  The Belgic dynasties of Britain and detail, it is also regionally biased by the many factors
their Coins before a meeting of the Society of Antiquaries affecting discovery, and worse still, it is distorted to an
in London (Allen 1944). In it he set out to construct a unknown extent by the unscrupulous who wish to please
history of the late pre-Roman Iron Age of Britain, based on collectors (including museums) by providing false find spots
a geographical, stylistic, and chronological study of the for material which is either without a sound location or was
surviving coinage. This work, with periodic updatings, has acquired and dispersed under dubious circumstances
remained a standard part of subsequent considerations of (Rodwell, below, pp 43-52). While distribution maps can
the British Iron Age. The CBA Conference on the Problems quite reasonably be used in generalizing arguments, to
of the Iron Age in Southern Britain, held in London in attempt to use them too precisely to generate sophisticated
1958, provided Allen with the opportunity to reappraise the models can give rise only to a spurious and misleading
origins of coinage in Britain (Allen 1961). His considered impression of accuracy.
views, stated with great clarity, were readily accepted by Where individual site finds are concerned we are in even
subsequent writers (Frere 1967; Hawkes 1968; Cunliffe more difficulty. It is only in recent excavations like those at
1974), all of whom found that his historical approach the temples of Harlow and Hayling or the urban sites of
provided a satisfactory model against which to consider the Braughing, Colchester, and Canterbury that reliable data
rest of the archaeological evidence. One should, however, are at last being provided. The distorting effects which
call to mind Allen s perceptive warning:  It is essential in these collections have on our maps is a firm reminder of the
interpreting coin evidence to recall constantly that it is only inadequacy of much of the rest of the record. This is not
and not always the most important part, of the intended to be a counsel of despair but a warning that we
Part,
historical record (Allen 1961, 98). should not ask of the data questions which, in full
The historical approach to coinage has continued to knowledge of their limitations, they cannot be expected
develop. Rodwell s detailed restudy of the coinage of south- reliably to answer. In the following pages we will therefore
eastern Britain extends and refines the arguments, pre- use the evidence of the coinage at a general, rather than too
senting a meticulously argued  history for the period specific, a level.
based substantially on changes in coin type and distribution
(Rodwell 1976), while the work of Simone Scheers in
Britain and the Continent, c 120-40 BC
France and the Low Countries uses historical events as a
Most scholars will agree that coinage was introduced into
framework for understanding the coinage (Scheers 1972;
Britain during the period 120-50 BC, but the economic and
1977). The historical model is further examined in the
social situation in the south-east of the country, and in
recent work of John Kent (Kent 1978a and below, pp
particular the differences in the different regional systems,
40-2).
are seldom taken into account. Some aspects of these
Whilst the historical approach thus continues to thrive,
problems have been dealt with recently elsewhere (Cunliffe
the vogue for discovering and analysing economic systems
1976; 1978a) but several points deserve mention here.
in archaeology, which developed in the 1960s, led some
Foremost is the fact that in the early part of the period the
writers to focus attention on the potential of the coin
south-east of Britain can be divided into two distinct
evidence in studies of this kind (Collis 1971a; 1971 b; 1974;
regions: a hillfort-dominated zone stretching from Kent and
Haselgrove 1976). The writer, by virtue of his early
Sussex westwards to Wessex and the Cotswolds, and an
archaeological training, must confess to being more in
area of open settlement occupying the Thames Valley, East
sympathy with this school of thought. The present paper,
Anglia, and the Midlands. This same division is emphasized
however, is an attempt to consider the quality of the data
by a consideration of the ceramics of the area. Clearly, two
against the broad social questions which might reasonably
separate socio-economic systems are implied. In both zones
be asked of it, rather than to engage in the polemic which
coinage was adopted and a full-scale market economy
surrounds model building whether historical or economic.
eventually developed.
To suggest however that the idea of coinage in its various
The nature of the evidence
manifestations was completely novel might prove to be
Before we can begin we must briefly consider the nature of misleading. Widespread use of currency bars appears on
the available data. In all some 12 624 Iron Age coins are present evidence to have preceded the introduction of
recorded from Britain. Probably less than 50% survive coinage (if overlapping with it), while the discovery at
today. Of this impressive total c 3 100 come from the single several sites (including Winklebury and Danebury) of well
 hoard found at Hengistbury and a further 5 200+ from made stone weights implies that careful measurement was
other hoards. A mere 1 l00+ have been found in excavations being practised. It may well be that salt packed in ceramic
(the majority come from three sites: Camulodunum, containers formed another unit of value (Cunliffe 1977;
Braughing, and Harlow), and of these a substantial 214), while the possibility that storage pits for grain may
proportion are unstratified. Furthermore, it is estimated have been dug to a series of size standards is a further
that of the 5 000 or so coins recorded on the Index housed in reminder that accurate measurement, in the interests of
the Institute of Archaeology, Oxford, about two-thirds are exchange, may have been widespread. To this we might
without precise provenance. perhaps add that some at least of the large numbers of
Thus it must be realized at the outset that the data have Greek coins found in south-eastern Britain are likely to be
widely varying levels of reliability. To a numismatist dealing genuine Iron Age imports, thus familiarizing the natives
with metrology, typology, die linking, etc, the data are of with the idea of the coin as a unit of value. In other words,
reasonable quality but to an archaeologist working with at the time when large-scale long-distance trade was
distribution patterns they are far from adequate. Not only is re-established in the first half of the 1st century BC, it is
29
Link to previous section
30
Cunliffe: Money and society in pre-Roman Britain
Fig 11 Britain and the Continent showing the most convenient points of contact: ports of trade, Armorican coins, Gallo-Belgic coins
the traveller to spend a minimum of time our of easy reach
reasonable to assume that the communities of the south-east
already practised an ordered economy in which measure- of land (Fig 11), One axis linked the western seaways, via
ment by weight and possibly by volume formed an essential Armorica, to central southern Britain, and the other lay
between northern France and the Low Countries (Belgica)
part. In such circumstances the ready adoption of coinage
and the Thames estuary.
need occasion no surprise nor would it be exceptional if a
Evidence for the western axis, between Armorica and
money economy were to develop soon after.
Hengistbury Head, has recently been discussed by the
writer in some detail (Cunliffe 1978b) and need not detain
Pre-Caesarian contact
us here. Suffice it to say that there is ample archaeological
evidence for widespread trade involving the importation of
Two principal axes of contact between Britain and the
pottery from Armorica and wine from Italy in exchange for
Continent seem to have developed in the decades before
which metals are the most evident of the possible British
Caesar s invasions in 55 and 54 BC. Not surprisingly, the
exports. Apart from the establishment of what can
routes chosen spanned the shortest sea crossings, requiring
Cunliffe: Money and society in pre-Roman Britain 31
Fig 12 Distribution of Gallo-Belgic B and Potin 1 coins mapped after Allen 1961 and Haselgrove I978
reasonably be regarded as a port-of-trade at Hengistbury
Armorican types, but instead develop from Gallo-Belgic
and a general improvement in pottery technology in the models introduced from eastern Britain, strongly suggests
south-west of Britain (probably involving the introduction that Armorican coins in Britain represented little more than
of the potter s wheel), the trading axis had little lasting
valued items of precious metal: they do not seem to reflect
effect on the socio-economic structure of the Iron Age the introduction of a new trading system based on money
communities of southern Britain.
economy. The fact that the socio-economic system, as
From the point of view of the present discussion it is the exemplified by the continued development of hillforts,
numismatic aspect of the contact that is of interest. It is appears to remain unchanged is a further indication that
represented by 60 or so imported Armorican coins scattered trade with Armorica had little lasting effect on southern
over central southern England of which 25 come from Britain. The reasons for this are obscure. The contact could
Hengistbury, a distribution sufficient in itself to imply some have been (and indeed probably was) short-lived, but of
form of contact even if no other evidence had been
equal importance may have been the fact that the economic
available. That the subsequent local coinages owe little, and social systems in the area were not, at this time,
apart from adherence to a silver standard, to imported sufficiently structured to allow the easy adoption of the new
32 Cunliffe: Money and society in pre-Roman Britain
Fig 13 Distribution of Gallo-Belgic A-F coins (after Allen 1961 and Huselgrove 1978)
exchange system. The Armorican contact seems, then, to bution pattern very similar to that of the Potin I coinage for
provide an interesting example of one of the many kinds of which Allen has argued a pre-conquest date. Mapped
relationship, involving the transference of coins, which may together (Fig 12) the gross distribution of Gallo-Belgic B
have existed between communities. and Potin I probably reflects the territory within which
The relationship between the Belgic territories and coinage first came into regular use in Britain. The Potin
eastern Britain was quite different but in view of the coinage is of particular interest for not only was it minted in
current discussions concerning the chronology of the Gallo- Britain but its very existence must surely imply a system of
Belgic coin series (Kent 1978a) it is unwise to argue the currency involving two denominations. Collis has found it
sequence of events too closely. Most writers are agreed, difficult to accept that potin represents small change in a
however, that Gallo-Belgic B coins were probably in use in money economy (Collis 1974), but Rodwell has countered
Britain in the decades before Caesar s conquest, and some his arguments (1976, 207-8). While the matter is still open
of the Gallo-Belgic A examples may well have been in to debate, particularly in view of the uncertainty of the
circulation in this period. The distribution of Gallo-Belgic B dating evidence, we can tentatively suggest that the earliest
centres upon the Thames estuary favouring Kent, a distri- development of British coinage took place in Kent and that
Link to next section