Download file in PDF format: TRAC 1994: Treasure: Interpreting Roman Hoards (pp. 99–106)
Treasure: interpreting Roman hoards
by Martin Millett
For three reasons it seems timely to consider how Romano-British treasures are generally
interpreted. First there have been a series of spectacular finds, most recently the Hoxne treasure
(Bland & Johns 1993) which have stimulated interest in the subject. These have also led to
misunderstandings, not least of the view I have publicly expressed that hoards cannot simply be
understood in terms of a desire to hide wealth in troubled times because they are but one aspect
of a wider phenomenon of ritual deposition. Second the revival of academic interest in ritual in
later prehistory (e.g. Bradley Hill 1992) has yet to have the impact it should upon the
study of the Roman period. Writing about this problem in 1987 Merrifield, himself a pioneer in
the study of ritual in Roman Britain, wrote: ‘It would probably be fair to say that archaeological
investigation of ritual activity is at about the same level today as were environmental studies in
excavations of 30 years ago’ . In 1994 this remains tme and I believe that until we attempt to
understand why this is the case it will remain so. Finally the whole subject of treasure has
recently been in the news because of the introduction into parliament of Lord Perth’s bill which
seeks to revise the medieval law of Treasure Trove. This public policy aspect is important since
there is no doubt that our present legal framework has sustained an atmosphere in which
academic debate about the nature of hoarding has been systematically curtailed through a
process of self-censorship within the archaeological profession. Although understandable this
has resulted in serious biases in our understanding of religion and society during the Roman
This paper aims to explore this background and aspects of how we should approach the
interpretation of hoards. Although I will not present any definitive conclusions I hope that in
opening a debate about interpretation this paper will help to lay the ground for wider
Before we can even begin to discuss interpretation we need to define what we are considering.
‘Hoards’ are widely recognised in archaeology and few think they would find it difficult to
identify one were they to uncover it on an excavation. For most of us a collection of valuable
objects deposited together, perhaps hidden, constitutes a hoard. This is unproblematic until we
consider what defines ‘value’; we use the term hoard of coins, silver plate and even ironwork
(Manning 1972), but not for instance of pottery or animal bones. In doing this we are applying
value systems which we ought to question. Farmers’ seed-corn or herd animals might have
been their most valued possessions, but these are not items which an archaeologist would grace
with the term hoard if found on an excavation. We thus need to explore the value systems
which have led us to define hoards if we are to make progress in understanding them. 100 Martin Millett
An exploration of these issues should start from the realization that we are affected by at
least three different value systems which have often been conflated. Confusion is caused by our
failure to distinguish between these and has been accentuated by a widespread belief amongst
those studying the Roman period that ancient values were similar or identical to those of the
modern world so their interpretation is unproblematical. I cannot agree with this. We need to
challenge this assumption if we are to make better sense of Roman provincial society.
The value systems with which we have to contend when considering the problems of
hoarding are first those of the people who deposited the objects, second those of society today,
and finally those of the society which produced our present Treasure Trove laws. Let us
consider these in turn to establish why the interpretation of treasure is so confused.
Values in the Roman Period
It is too often assumed that those who deposited hoards did so simply because they wished to
hide material of worth for reasons of security. These assumptions about values and motives are
difficult to assess because of those imposed by our own society but there are a series of
questions to consider. Foremost is the question of whether we can arrive at a satisfactory
estimation of values attributed to the objects buried. I have already indicated that our
assumptions may not easily be reconciled with those of a predominantly agrarian society in the
past. Even when we can apparently make reasonable estimates of the monetary worth of
materials like silver plate or bullion coinage using available information about wages and
prices in the Roman world there remains uncertainty. This will be all too evident to any who
have followed the recent debate about the social value attributed to silver plate in the late
Roman world (Cameron 1992: Painter 1993). Even where monetary values can be estimated,
we may too readily assume that objects were hidden because of their monetary worth when
other factors were evidently important and hoarding may reflect these more nebulous values.
Aesthetic merits appear important in the selection and perhaps the hoarding of the objects
buried in the Mildenhall treasure (Painter I 977a), although arguably absent when similar
objects were interred as ‘hack silver’ as at Traprain (Curle 1923). Symbols of religious
significance are common on objects found in hoards, be they from Pagan cults, such as Faunus
at Thetford (Johns and Potter 1983) or from Christianity (as at Water Newton Painter
I 977b). These suggest that a variety of different and overlapping values were attributed to all
types of objects so it is wrong simply to assume that a decision to hide a hoard was solely
related to its (assumed) monetary worth. We cannot just assume that cash value was of central
importance to the owners of hoarded objects. Even when monetary importance seems selfevident, as in the case of coin hoards it remains questionable. Many hoards of late Roman
bronze coins for instance seem to represent only modest purchasing power.
Even where coins were of bullion and thus presumably of value we cannot assume that they
were buried only for security. A close analysis of the contexts of Iron Age gold coin finds has
led Haselgrove (1993, 50) to the conclusion that they were often votive deposits. If this is true
of Iron Age coin deposits ought we not to at least consider that some Roman hoarding also had
a religious character? (see below p. 104)
Whatever the answer the interlocking values attributed to objects and the reasons for their
deposition must be a subject of investigation, not a matter of assumption. Treasure: interpretinK Roman hoards 101
Today ‘s Values
There are at least four different perspectives to consider in relation to the ways in which hoards
are defined and viewed in the modern world. Archaeologists have often not only assumed that
the values of the past are similar to those of today, but have also tended towards the view that a
constant rationality makes it obvious that hoards were buried because of a desire to protect
wealth from threats. They have thus implicitly naturalized the idea that the hoarding of
valuables was a normal behaviour in past societies, supporting their views with ‘commonsense’ parallels, the experiences of Samuel Pepys for instance (Kent 1988). This has also been
accompanied by a very coy attitude to treasure. Even the word treasure is avoided by most
academic archaeologists who prefer terms like hoard. There is an almost universal view that
professional archaeologists should not take an interest in treasure because it is ‘rather
common’, ‘vulgar’ or ‘makes us look too much like treasure hunters’. How often have you
squirmed at having to explain to an acquaintance that the ‘real value’ of the latest treasure
reported in the newspapers lies in the information it provides about the past, not its cash value?
We take this to the extent that ordinary books about the Roman provinces hardly mention
hoards as a phenomenon. Imagine the results of a similar view being taken by those writing
about the Bronze Agel The disdain for vulgar treasure is reinforced through its treatment by a
professional coterie which results from both its odd legal status (see below) and the poorly
developed state of artefact studies in our university departments
The archaeologists’ perspective forms a marked contrast with those of the general public
where I can detect three important different views. The bulk of the public are genuinely
interested in what archaeologists find and feel our coy embarrassment is rather curious.
Evidently we have not been very successful as a profession in explaining how we interpret the
past and why we are interested in it. Is it therefore any surprise that the public still hold the
view that we are primarily interested in finding treasure thus reinforcing our embarrassment
and their mystification when they want to discuss what is almost our professional taboo?
Because we have not put across any coherent message about the value of the past as a
fascinating challenge for the human mind we should not be shocked when the average person
on the bus only wants to know how many thousands of pounds a find is worth. In the western
world all seem to aspire to fulfil Oscar Wilde’s definition of a cynic by knowing ‘the price of
everything and the value of nothing’. Our failure to explain and espouse higher values than
those of the market make us guilty of complicity in this.
Whilst we fail to confront these issues many with a genuine lust to discover about the past
are seeking to find more objects on their own often with the aid of their metal-detectors. Their
success, and their abiding interest in treasure, reinforces the archaeologists’ anxiety, but
regularly refuels the public’s interest as new and spectacular finds are constantly reported in the
press. Whilst detectorists are criticised by archaeologists whose interests lie in the context of
finds, they are evidently encouraged by the antiquities traders and collectors who provide a
constant demand for objects which they value irrespective of their context. The respectable face
of the ‘ art market’ hides a much darker face, not for them ‘ art for art’ s sake’ but the darker ‘ art
(or anything else) for money’s sake’.
The power of their market creates modern monetary values for ancient objects particularly
those considered to have aesthetic worth or bullion value. The social acceptance of these values
reinforces the naturalization of the modern monetary values affixed to hoards of certain types of
objects. The belief in the unchanging nature of these values is so firmly fixed that they remain
almost wholly unquestioned. Whilst it is not self-evident that the values of Bond Street should 102 Martin Milletf
reflect those of the ancient world they have undoubtedly biased modern scholarly views about
the importance of many objects.
The values of the Roman and modern worlds are curiously connected in contemporary practice
(in England) through the application of the medieval law of Treasure Trove. The law as used
today is derived from the twelfth century desire of kings to ensure that bullion was reserved to
them (Hammond Sparrow Cookson 1992). The value in which medieval kings
were interested was that of cash to the Exchequer, so treasure was defined through
contemporary practice as bullion. In accordance with the privileges of the time such goods were
surrendered to the Crown if they had been buried with intent to recover and the original owner
could not be traced. Treasure Trove law has been subject to changes in legal practice up until
today, but the essence of the medieval value-system remains fundamental in shaping
contemporary archaeology. Bullion defines value, so objects of gold and silver have to be
treated differently to those of other materials irrespective of the values imposed by their makers
The separate treatment of gold and silver reinforces the view that their value has remained
constant through history. Secondly, archaeologists have often been called to convince coroners’
juries that the objects were ‘buried with the intention of subsequent recovery’. Courts continue
to use this test despite its absurdity and the impossibility of understanding such past motives (as
recognised in Colin Renfrew’s speech in the House of Lord’s during the passage of Lord
Perth’s Treasure Trove bill on 9 March 1994). Finally, as items classified as Treasure Trove
have customarily been claimed by the Crown but passed to the British Museum, they have
generally been studied by only a very small group of specialists who are constantly aware of the
importance of saving the objects for the nation
The desire to ensure that the objects of bullion are fully recorded and if possible saved for
the nation has meant that the profession as a whole has been reluctant to question or discuss the
academic problems raised by deciding whether objects are Treasure Trove. They have been
especially wary of publicly debating how far aspects of ritual were responsible for the processes
of hoarding precious metals. Thus although the deposition of Iron Age weaponry is widely
viewed as votive, there has been very little discussion about whether contemporaneous coin or
torc hoards were too (cf Fitzpatrick 1992). There has been a tacit assumption that such
discussion is best avoided if we wish to ensure that our national treasures end up in Great
Russell Street rather than Bond Street.
The Conjlation of Different Value-systems
In combination these different and often unrecognised value-systems have created an
acceptance of ideas about the constancy of values and the unproblematic nature of hoarding.
This is based on a series of implicit assumptions and tacit agreements which have together
stifled debate and left hoards outside the normal realms of discussion in Romano-British
There is little questioning of the assumptions that bullion was universally valued in the
Roman world and that value somehow explains hoarding. As a result some still hold to the
\ view that coin hoards can be used to map periods and places of ‘unrest’ or ‘trouble’ despite the
difficulties in assuming that this was so (cf Kent \988). Another consequence is that there has Treasure. interpreting Roman hoard<i 103
been very little general discussion of the contextual or possible ritual significance of hoarding.
Finally as coroners tend to consult the same people about Treasure Trove, the British Museum
has accidentally acquired a virtual monopoly in the study of the bullion items from RomanoBritish hoards. This final point leads me to a consideration of how academic organization has
served to isolate consideration of treasure from other aspects of archaeological interpretation.
There are two ways in which the organizational structure of archaeology has led to a lack of
contextual discussion about hoarded items. The first constraint results from who has studied the
material and the second from how it has been examined.
The often brilliant specialist interpretations of hoarded items produced by the British
Museum have not been subject to the same widespread discussion and dissent as publications
concerned with other archaeological subjects, like for instance the interpretation of pottery.
Because the Museum has most expertise in the subject only a small group of specialists know
the material sufficiently well to feel competent to comment upon its interpretation, especially
where issues of identification, style and dating are concerned. However there has also been little
discussion of the overall interpretation: why particular objects were selected for deposition
together and how they found their way into the ground. This contrasts with, for instance, the
interesting discussion of the depositional context of Iron Age currency bar hoards (Hingley
1990). The pressures on those forced to apply an inappropriately framed medieval law have of
course been intense. Through their efforts the great collections of objects found in recent years
have been saved for the nation and are being published but one of the costs has been a tacit
agreement by the whole profession not to discuss the problems of academic interpretation
created by the Treasure Trove law. One simply does not discuss the extent to which hoarding
may have been religious lest it appears to undermine the case for items being declared Treasure
Trove. With luck the passing of a more rational law will soon liberate the subject from these
shackles so that an informed and open debate can take place.
The second of the academic structures which has prevented full discussion of the
interpretation of material has been of our own creation, for Romanists in particular have been
fenced in by the boundaries of specialized study and have too infrequently stood back to take an
overview. Romanists looking at hoarded material divide it into categories, so for instance coins
are treated separately from silver plate. Equally we separate the material by period, so Iron Age
hoards are dealt with by different people from those who study the Roman period. This is no
different from other aspects of Roman archaeology. The curse is not specialization itself but the
failure to reintegrate results once studies have been completed to create an overall
understanding. As Merrifield (1987) has pointed out, this problem is probably responsible for
our persistent failure to identify votive deposits comprising everyday objects on settlement sites.
It is arguably necessary to atomize the study of finds because of the depth of knowledge
required to deal \-‘lith objects of different types. If this were only a stage in the processing of the
finds there would be little to concern us as a holistic view could be created by integrating the
ideas of a variety of specialists to create a collaborative whole. Unfortunately this rarely
happens and when it does implicit boundaries are left to isolate the study of hoards from the
rest of the discipline. Thus objects of lesser value are sometimes ignored and discussion of
context is all but unknown. 104 Martin Adi/lett
Principles for Future Study
I hope my analysis of the present state of study has not been too critical of those who have
produced some fascinating work. Equally I do not want to appear too prescriptive in suggesting
ways forward as r have little desire to appear Stalinist. However I do believe that there may be
certain lines which should be given some priority in future research if we are to recognise the
potential of the material and integrate it with other aspects of the study of Roman Britain.
r can see little rationale in separating the study of bullion or coin hoards from related
patterns of deposition. It seems evident that one of the principal characteristics of later Iron
Age and Roman Britain (and adjacent provinces) was the burial or deposition in rivers of a
range of collections of different types of objects. Although some would like to view this as the
result of a series of accidents, I prefer to see them as deliberate acts of deposition with a votive
aspect. To take a continental example, the corpus of information collected by Roymans (1990,
84-90) has shown clearly how rivers around the lower Rhine were consistently used as places
for such deposition during the Iron Age. I thus find it extremely difficult to view the
astonishing collection of all kinds of material found at Neupotz on the Rhine as loot from an
Alamannic attack in AD 277-78 as the report on it suggests (Kunzl summary in Painter
1994). Such patterns surely show the continuation of a long tradition of votive deposition into
the Roman world.
I also think we need to move away from the view that the interpretation of such deposits
has to be read according to our own rationality There is no need to consider hoarding and
deposition in the ground or in wet places as either sacred or profane – all actions were most
likely invested with significance in both spheres. It is probably only in modern secular societies
that we seek to make a distinction. In our interpretations we might be best to start from the
assumption that all actions were vested with some ritual significance rather than following the
current tradition that invokes ritual only as a last resort to explain the inexplicable.
When we approach the study of the materials deposited in hoards there seems little sense is
distinguishing between different materials. The fourth century pewter hoards (Poulton & Scott
1993), contemporary silver hoards (like MildenhaU), or the glass hoard from Burgh Castle
(Johnson 1983. 30–3 L Harden 1983) and collections of ceramic vessels placed in wells at
Neatham (Millett & Graham 1986,30–33: 159) are only treated as different because of our own
academic conventions. If we were able to integrate their study to seek patterns of deposition,
context and content we might reach a more satisfactory understanding of their meaning. Indeed
the material and literary richness of the Roman world provides one of the few contexts in which
a tndy contextual archaeology might be possible. Such broader investigation might equally help
us understand the peculiarities of some of the hoards. I accept the obvious religious symbolism
of certain items in many of the hoards, for instance the Christian inscriptions at Water Newton,
but am I alone in wondering why so many spoons occur in fourth and fifth century hoards?
Contextual study and comparison with other types of deposit might help us identify other
recurrent patterns even if they remain difficult to interpret.
Equally there are evident patterns of location which deserve investigation. 1 have long
wondered why the supposedly poor areas of East Anglia, largely devoid of villas have the main
concentrations of late Romano-British silver plate hoards. Is it a coincidence that there was also
a tradition of burying gold torcs hereabouts in the Iron Age? Equally, the late Roman lead tanks
which so often ended up in wells are also clustered in their distributions around the East
Midlands and East Anglia. Until we treat all hoards contextually and try to relate them to other
phenomena we will not begin to be able to identify slich patterns let alone understand them. Treasure, interpretil1f.? Roman hoards 105
To quote L.P, Hartley ‘the past is a foreign country they do things differently there,’ We
will not begin to understand it until we map it. The danger in Roman archaeology is that until
we recognise our disciplinary and social constraints we will be too hide-bound even to
recognise that it needs mapping,
I am most grateful to Simon James, Jenny Price and Sarah Scott who commented on an earlier
draft of this paper. Whilst they do not necessarily agree with any of the points I have made they
have done much to improve my presentation, 106 Martin Millett
Bland, R. & Johns, C. 1993. The Hoxne Treasure: a illustrated introduction. London: British Museum
Bradley, R. 1990. The Passage of Arms: an archaeological analysis of prehistoric hoards and votive
deposits. Cambridge University Press.
Cameron, A 1992. Observations on the distribution and ownership oflate Roman silver plate, Journal of
Roman Archaeological 5: 178-85.
Cookson, N. 1992. Treasure Trove: dumb enchantment or new law, Antiquity 66: 399-403.
Curle, A.O. 1923. The Treasure of Traprain: a Scottish hoard of silver plate. Glasgow: Maclehose,
Jackson and Co.
Fitzpatrick, A. P.1992. The Snettisham, Norfolk hoards of Iron Age torques: sacred or profane? Antiquity
Hammond, G.N. 1982. Treasure trove: ancient law to preselve archaeological sites, Antiquity 56: 58-60
Harden, D.B. 1983. The glass hoard, in Johnson, S. Burgh Castle: excavations by Charles Green,
/958-6/, 81-88. (East Anglian Archaeology 20).
Haselgrove, C.C. 1993. The Development of British Iron Age Coinage, The Numismatic Chronicle, 1993:
Hill, l.D. 1992. Can we recognise a different European past? A contrastive archaeology of later
prehistoric settlements in southem England, Journal of European Archaeology 1: 57-75.
Hingley, R. 1990. Iron Age ‘Currency Bars’: the archaeological and social context, The Archaeological
Journal 147: 91-117.
Johns, C. & Potter, T. W. 1983. The Thetford Treasure. Roman jewell elY and silver. London: British
Johnson, S. 1983. Burgh Castle: excavations by Charles Green, 1958-61. (East Anglian Archaeology
Kent, l.P.C. 1988. Interpreting coin-fmds, in P.l. Casey and RM Reece (eds) Coins and the
Archaeologist, 201-17. London: Seaby.
KiinzL E. (ed) 1993. Die Alamannenbeute aus dem Rhein bei Neupotz: Pliinderungsgut aus dem
r6mischen Gallien. Mail1z: Riimisch Germanischen Zentralmuseums.
Manning, W.H. 1972. Iron work hoards in Iron Age and Roman Blitain, Britannia 3: 324-50.
Merrifield, R. 1987. The Archaeology of Ritual and MagiC. London: B.T. Batsford
Millett, M. & Graham, 0.1986. Excavations on the Romano-British Small Town at Neatham, Hampshire,
/969-/979. Winchester: Hampshire Field Club and Archaeological Society.
Painter, K.S. 1977a. The Mildenhall Treasure: Roman silverfrom East Anglia. London: British Museum
Painter, K.S. 1977b. The Water Newton Early Christian Silver. London: British Museum Press.
Painter, K.S. 1993. Late Roman silver plate: a reply to Alan Cameron, Journal of Roman Archaeological
Painter, K.S. 1994. Booty from a Roman villa, Minerva 511: 22-27.
Poulton, R & Scott, E. 1993. The hoarding, deposition and use of pewter in Roman Britain, Theoretical
Roman A rchaeol ogy: first conference proceedings, 115-32. Aldershot: Avebury.
Roymans, N. 1990. Tribal Societies in Northern Gaul: an anthropological perspective. Universiteit van
Amsterdam: Cingula 12.
Sparrow, C. 1982. Treasure trove: a lawyer’s view, Antiquity 56: 199-200. 5