Travaux de la Maison de l'Orient

The Roman Terracotta Lamp Industry. Another view about exports

Donald Bailey

Citer ce document / Cite this document :

Bailey Donald. The Roman Terracotta Lamp Industry. Another view about exports. In: Les Lampes de terre cuite en
Méditerranée. Des origines à Justinien. Table ronde du CNRS, tenue à Lyon du 7 au 11 décembre 1981. Lyon : Maison de
l'Orient et de la Méditerranée Jean Pouilloux, 1987. pp. 59-63. (Travaux de la Maison de l'Orient, 13);

https://www.persee.fr/doc/mom_0766-0510_1987_act_13_1_2143

Fichier pdf généré le 02/05/2018

 THE ROMAN TERRACOTTA LAMP INDUSTRY
ANOTHER VIEW ABOUT EXPORTS

Donald BAILEY

A recent paper published in The Journal of Roman Studies by Professor W.V.

Harris discusses the organisation of the Roman terracotta lamp industry from a
historian's and economist's point of view, taking into consideration the results of
many archaeological publications1. While there are many aspects of this study
which are of great interest and importance, one of the main arguments (and, indeed,
conclusions), that during Roman times the large-scale exportation of lamps was
economically unlikely, must be challenged. Professor Harris asserts that lamps were
too cheap a commodity for them to have been worth moving more than a few
miles from their place of manufacture ; that it is far harder than most archaeologists
have recognized to determine whether an object is locally-made or made elsewhere
and that scientific examination of clays would seem to be one way out of this
difficulty ; and that the majority of lamps which bear makers' names, but which
are found in areas away from the workshops in question, were not exports but
were locally made by branch workshops. He discusses mainly Firmalampen and
western signed Bildlampen and does not mention the well-attested widespread
exportation of, for example, lst-2nd century Cnidian lamps, of 2nd-3rd century
Corinthian lamps, and of Late Roman lamps from Asia Minor and Tunisia2.
As far as the cheapness or otherwise of lamps is concerned, the plain fact is
that we do not know how much they cost, except, as far as the present writer is

1 - JRS LXX, 1980, pp. 126-45/.

2 - For Cnidian lamps, see G. Heres in Staatliche Museen zu Berlin, Forschungen und Berichte X,
1968, pp. 185-211 ; for Corinthian lamps, see O. Broneer, Isthmia III, Terracotta Lamps, Princeton,
1977, pp. 65-6; for a random selection of Late Roman Lamps from Asia Minor, see NS 1909, p.
360, fig. 18, second row, first lamp, last row, first three lamps, and p. 361, fig. 19, all from Syracuse,
those in E. Joly, Lucerne del Museo di Sabratha, Rome, 1974, pi. LIII, those in C. Iconomu, Opaite
Greco-Romane, Bucarest, 1967, p. 26, figs. 49-50, from Romania, those in J. Perlzweig, The Athenian
Agora VII, Lamps of the Roman Period, Princeton, 1961, Nos. 349-63 ; for an equally random selection
of African Red Slip Ware lamps made in Tunisia, see L.A. Shier, Terracotta Lamps from Karanis,
Egypt, Ann Arbor, 1978, No. 389, Perlzweig, op.cit., Nos 322-32, from Athens, P. Bruneau, Exploration
Archéologique de Délos XXVI, Les Lampes, Paris, 1965, No. 4687, Joly, op.cit., Nos. 1026-1245, from
Sabratha, and the many found in the catacombs of Rome, now in the Museo Sacro of the Vatican
Library, largely unpublished, but listed by A. Provoost in Bull, de l'Inst. Belge de Rome XLI, 1970,
pp. 27-30. I would add that lamps from all the above groups, plus a huge number of Italian lamps,
were imported by the city of Berenice in the Cyrenaica, and will be published in the forthcoming
Excavations at Sidi Khrebish, Benghazi (Berenice) III, by P.M. Kenrick and D.M. Bailey.

Les lampes de terre cuite

TMO 13, Lyon, 1986
60 D. BAILEY

aware, for those mentioned in Diocletian's Maximum Price Edict3, where their
low price (either one lamp or ten lamps - the text is defective - should sell for
no more than 4 denarii) reflects the fact that lamps at the beginning of the fourth
century, throughout the Empire, were of exceptionally poor quality and essentially
local products, used locally. One might use Professor Harris's arguments here and
say that these lamps were not exported because they were so cheap ; but on the
other hand they might not have moved because they were of low quality and
unlikely to appeal to buyers elsewhere. Also the state of the markets after many
decades of military and political upheaval might well have discouraged exportation.
But although in AD 301 a maximum price was laid down, there is unlikely to
have been such a thing as a standard price for a lamp in the early Empire, as
Professor Harris suggests (Harris, p. 134). He says there is evidence to show that
lamps of that period cost only one as. But lamps of the early Empire had a huge
variety of size shape, quality and decoration, all of which must have commanded
different prices. Harris, p. 134, note 60, mentions inltynium, costing one as at
Pompeii, as possibly meaning a lamp, but it seems just as likely to have been a
comestible, as are many of the other items listed with it in the same painted
inscription (CIL IV, 5380). His only other evidence is inscriptions on lamps
suggesting that they cost one as each. As Professor Harris himself points out in
the same footnote, there is certainly some obscurity concerning the inscriptions
on these lamps, with their variants of EMI TE LVCERNAS COLATAS AB ASSE.
H. Leclercq4 came to the tentative conclusion that this inscription could be expanded
to read EMITE LVCERNAS COLATAS [DE 0FFICINA] AfVLIJ BASfllJ
SEfNECAEJ. Be that as it may, these lamps, made in the extreme west of North
Africa, cannot be used, as Harris uses them, as evidence that the general price of
a lamp was one as in the early Empire, as they are products of the fifth century
AD, when the as was not a unit of currency.
Lamp prices must always have been what the market could stand, and no doubt
the cost of transport would be taken into consideration when, arriving at the selling
price. But it is a hopeless exercise to argue, as Professor Harris does, using the
unproven price of a lamp and the cost of land-transport (Harris, pp. 134-5), known
only for a few places (mainly Egypt) for a few limited periods of time, that a
transaction would be uneconomic. One cannot say, with regard to antiquity, that
because a thing seems unlikely by modern standards, that it would not have been
done. But the transport of cheap items in bulk is eminently possible when one
considers (perhaps invidiously, in the light of the previous sentence) the vast
amount of worthless trash moved from one side of the world to the other by
modern entrepreneurs. Good quality lamps must always have been saleable (Athenian
lamps were purchased all over the Mediterranean world in the fifth and fourth
centuries BC because the oil did not seep through them), and such lamps had
novelty value, even where there were plenty of local products. Further, the
'consignment
in a single of
cart1000
or by
lamps
several
quoted
packbyanimals
Professor
' can
Harris,
only p.be 135,
a guess,
as being
and carried
in the
present writer's opinion is a grave error which undermines the whole of Professor

3 - See JRS LXIII, 1973, p. 103 and p. 108, where Joyce Reynolds describes in the Aphrodisias copy
of the Price Edict, found by E.T. Erim, the entry for clay lamps. I am grateful to Andrew Burnett
for this reference ; he tells me that a denarius at this period was worth very little. I would also like
to thank at this point Catherine Johns and Susan Walker, for reading the manuscript and making very
helpful suggestions.
4 - F. Cabrol and H. Leclercq, Dictionnaire d'Archéologie Chrétienne et de Liturgie VIII, Paris, 1928,
cols. 1196-7. For further examples of these lamps, some of which were made at Tipasa, see J. Baradez
in Libyca IX, 1961, pp. 147-9; P. Leveau in Antiquités Africaines XI, 1977, p. 234, fig. 25, from
Cherchell ; J. Deneauve, Lampes de Carthage, Paris, 1969, No. 1 137 ; M. Ponsich, Les lampes romaines
en terre cuite de la Maurétanie Tingitane, Rabat, 1961, Nos. 350-2.
THE ROMAN LAMP INDUSTRY 61

Harris's case for the economic non-viability of exporting lamps. In an experiment,

a box one cubic foot in size was packed with lamps, with sufficient packing material
to protect them. It held 160 lamps5. Thus, a small cart with 4 ft χ 6 ft of floor
space (24 sq ft), stacked to the not excessive height of 3 ft, has a capacity of 72
cubic ft: 72 χ 160 lamps = 11520. A cart could therefore carry more than eleven
times as many lamps as Professor Harris suggests. Pack animals are harder to
assess in this respect, but a modest load of 3 cubic ft on each side of a donkey
or a mule could contain 960 lamps, almost as many as the ' several pack animals \
As Professor Harris himself points out, ships and river-boats were cheaper than
land transport, and water transport was doubtless the main means used in the
exportation of lamps, particularly when one considers the closeness of much of
the Roman Empire to the Mediterranean coast, and the navigable rivers stretching
into much of the rest. Even with the example of sea-borne lamps mentioned by
Professor Harris (p. 135) there is a possibility that not just a hundred but thousands
of lamps were involved6. These are Caius Clodius lamps from a wreck in the
Balearic Islands. But land transport was used over very long distances to convey
objects, including lamps, imported from elsewhere. For example, tombs in the
Fezzan have yielded lamps imported from the Tunisian coast, of dates ranging
from the first to the third centuries AD, presumably carried by pack-animals for
distances of over 400 miles, and more than twice that in the Hoggar7.
Another factor to be considered is the Army. If the Army wanted lamps (and
it did) it got lamps, despite the cost, again mainly by sea and river, but if necessary
by land. Leibundgut has shown that the vast majority of lamps from Switzerland
are found on military sites, and not in civilian areas8.
Of course, the Army set up conditions for the local production of lamps9, but
many imports are recognizable. It is here that the present writer again parts
company with Professor Harris. Archaeologists who work with pottery are capable
of distinguishing between fabrics and deciding whether an object is likely to be an
import or a local product (and despite Professor Harris, p. 131, the concepts
w import ' and ' export ' do have unmistakably clear meanings within the archaeology
of the Roman Empire, and every archaeologist knows what is meant by the terms).
The laboratory analysis of clays (Harris, p. 137) is a most desirable and useful
tool and many interesting results have been obtained and will continue to be
obtained from scientific techniques of study. It is surprising, however, how many
of these results confirm opinions already held by archaeologists, whose main tools
are experience and the human eye, certainly the most useful tools we have.
Naturally, mistakes can be made, and the laboratory can refute accepted ideas,
but shape, decoration, details, colour, texture, hardness, macroscopic and microscopic
inclusions, etc... can all be used by the experienced archaeologist to determine
manufacturing sources and (one hardly dare admit it) a mere glance is often enough

5 - I used the largest group of equal-sized lamps to hand, Cnidian lamps of Loeschcke Type VIII.
These are about the same average length as Italian Firmalampen, but are rather wider.
6 - D.M. Bailey, A Catalogue of the Lamps in the British Museum II, Roman Lamps made in Italy,
London, 1980, pp. 92-3 ; A.K.B. Evans, in A.C. and A.S. Anderson, Roman Pottery Research in Britain
and North-West Europe, Oxford, 1981, p. 527.
7 - C. Daniels in Libyan Studies VIII, 1976-7, pp. 5-6 ; G. Caputo in Mon. Ant. XLI, 1951, cols. 329-
30, fig. 121, col. 343, fig. 133; R.E.M. Wheeler, Rome Bevond the Imperial Frontier, London. 1954.
pi. XVIII B.
8 - A. Leibundgut, Die römischen Lampen in der Schweiz, Bern, 1977, pp 100-129.
9 - For example, S. von Schnurbein in RCRF Ada XVII-XVI1I. 1977, pp. 38-50, at Haltern; M.
Vegas in Bonner Jahrbuch CLXVI, 1964, pp. 308-20 and in Novaesium II, Berlin, 1966, pp. 63-127,
at Neuss; D. Haupt in Das rheinische Landesmuseum Bonn, 1977, pp. 199-207, and 1979, pp. 151-
5, at Xanten.
62 D. BAILEY

to decide this. The present writer is chided in Professor Harris's Footnote 52 : 'To
suppose that red-brown and brick-red are the colours of lamps made in Northern
Italy... would simply be uncorrect '. It can be agreed that if north Italian lamps
in general were being discussed this would be so. However, the lamps in question
are north Italian Firmalampen, and the majority of north Italian Firmalampen
are, without doubt, in a ' brick-red, or red-brown, unslipped or self-slipped, fabric '10.
One of the difficulties here turns on the nature of verbal descriptions of such
features as texture, density and above all, colour: even the most elaborate colour-
chart (Munsell) is inadequate to convey the latter in archaeological work. These
nuances cannot be described in words, and yet, some generalizations must be made
if a work of synthesis is to be useful to the reader. Statements on such matters
as colour and texture must therefore be over-simplified ones, which do not begin
to convey the complexity of judgements made on the basis of years of experience
in handling and observing the artefacts themselves. Most practical archaeologists
are fully aware of the limitations inherent in putting into words these infinitely
subtle mental, visual and tactile judgements, and accept general statements as broad
guidelines rather than literal and strict rules. It is often forgotten that the human
mind, hand · and eye are every bit as sensitive as any scientific test : it is merely
that their conclusions cannot be presented in a print-out or histogram. To echo
the words of Professor Ashmole, writing about marble identification and scientific
examination : ' I fear that we shall have to rely on a method.. ..which was
employed.. ..for centuries...., namely, that of the naked eye and common sense*".
Professor Harris has come to the conclusion that ' most signed lamp found in
areas away from the original places of manufacture were in fact made locally *
(Harris, p. 144). (The question of branch workshops versus unauthorized copies
will not be discussed here : the facts are unknown, the ramifications so complex
and the possibilities so endless, that it will never be established what really happened
Empire-wide). While there are many examples of lamps bearing, for example, the
names of Italian or African makers, found in provincial areas and which were
undoubtedly made in those provincial areas12, there are also found, in the writer's
opinion, many more examples of lamps made in Italy or Africa but found elsewhere
than Professor Harris is prepared to admit. Deneauve is criticized for regarding
many of the signed lamps of Flavian to Antonine times from Carthage as being
imports: Professor Harris regards this as 'beyond belief (p. 132). Would he also
regard it as beyond belief that, except for a barely significant quantity made at
Colchester, all the terra sigillata found in such huge numbers in Britain was
imported ? As mentioned above, it can be comparatively easy for arr archaeologist's
trained eye to recognize fabrics, and when such a fabric, bearing all signs of being,
say, Italian, is found in a provincial context, and is, in addition, signed by a known
Italian maker, the conclusions are inescapable : it is an import from Italy. The

10 - Bailey, op. cit., p. 277. M. Cicikova in Actes du 9e Congrès international d'études sur les frontières
romaines, Bucarest, 1974, pp. 158-9, distinguishes a group of slipped Firmalampen exported from north
Italy (Groupe B), and much different from the dark red, red and grey unslipped lamps which she
regards as the normal imports from north Italy (her Groupe I A, pp. 156-8). But I am not convinced
that her footnote
Harris' Groupe I78,B lamps
wherearehe notstates
of provincial
that handles
manufacture.
on Firmalampen
Incidentally,
are I rare
do not
except
understand
in the Professor
German
provinces and in western central Italy and goes on to say that I go too far in saying that 'central Italian
Firmalampen normally have handles'. This appears to be what he himself has just said. I would repeat
that central Italian Firmalampen do normally have handles, from the Flavian products of the maker
Myron to the late Antonine-Severan lamps of Passerius Augurinus and Saeculus, when the type dies out.
11 - BSA LXV, 1970, p. 1.
12 - Compare, for example, adjacent kilns at Montans in southern France,' producing lamps with the
signatures of the Italian L. Munatius Threptus and the African M. Novius Justus : T. Martin in Figlina
II, 1977, pp. 51-78.
THE ROMAN LAMP INDUSTRY 63

opinions of the many archaeologists mentioned by Professor Harris who believe

this kind of attribution to be possible, and who know the material from an
archaeological rather than an economic point of view should not be lightly brushed
aside with demands for scientific proof. With regard to economists, the views of
Rostovtzeff, dismissed rather summarily by Professor Harris, seem still to be valid :
the lamp industry does fit ' into his overall theory about the decline of the Italian
economy in the second century, in consequence of the supposed loss of external
markets fc (Harris, p. 133). Despite his views about the unlikelihood of the exportation
of lamps, Professor Harris himself seems to agree that on occasion this occurred,
and says ' And more important, in many regions of the empire local production
of lamps must have been preceded by a period in which there was a market for
imported terracotta lamps ' (Harris, p. 1 36) : there is nothing to argue with in that
statement, but one can go further. During the first century AD, very few places
within the Empire were immune from the massive importation of Italian lamps.
The Army soon began to ensure its own local supply but this did not halt
importation, and although some civilian areas, particularly in the Greek East
(Cyprus, Egypt, Petra, western Asia Minor, and Central Gaul in the West) began
the large-scale manufacture of good quality lamps before the middle of the century,
it was not until the time of Trajan that many provincial workshops really became
active. Even after this, in the first half of the second century, Italian Firmalampen
went north and signed Italian Bildlampen went south13. From the middle of the
first century for many centuries, provincial centres, in their turn, exported lamps
over very wide areas. Historians, economists and archaeologists can all work towards
a greater understanding of this matter, but of only one thing can we be certain,
and that is that the terracotta lamp industry and trade of the Roman Empire was
far more complex than we shall ever know.

BRITISH MUSEUM

13 - And probably even later than this : a recent paper by C. Pavolini indicates that many signed
Italian Bildlampen were made much later in the second century than I had thought (Bull, della Comm.
Arch. Com. di Roma LXXXV, 1976-7, Tabelle I-II).