You are on page 1of 16

Lib)'al/ SludieJ I K( I <JK 7 ,

,:-~

'Tripolitanian Sigillata': North African or Campanian?


By G. Soricelli I
Abstract
In this article Campania is proposed as the provenance for a class of terra sigiIJata, somewhat frequent at
Berenice and on other North-African sites, which was recently termed by Kenrick 'Tripolitanian SigiIJata'.
Particularly frequent on Campan ian sites, above all at Pompeii and Naples, it was present in this region
from the e nd of the first century BC, occurrin g with high perce ntage values in assemblages of fine potte ry from
excavations. This new evidence for its distribution would appea r to exclude a North African provenance . A
Campanian o rigin , in the area of the Bay of Naples, is, moreover, supported both by the results of thin section
analyses of the clay and by the discovery of possible kiln wasters at Naples.

Introduction
Recently, P. Kenrick has proposed the term 'Tripolitanian Sigillata' to indicate a class of
terra sigillata well represented at Berenice, Sabratha and other localities in Libya (Kenrick

1985,283-302; 1986, 179).


The production of this ware seems to be concentrated between the Augustan era and
about the middle of the first century AD. At Sabratha, fragments of Tripolitanian SigiUata'
appear in a level dated to around 10-1 BC (Kenrick 1986, 175-199, Pit 2), while at Bere
nice, although the earliest stratigraphic evidence comes from contexts dating to the
Tiberian era (Kenrick 1985, Deposit 46 ), the whole group recove red would seem to
belong to a rather short period between c. AD 1 and 40, not excluding the possibility that
some pieces could be of a later date (Kenrick 1985,285).
The chemical analyses of some fragm ents of Tripolitanian SigiUata' and similar frag
ments from Carthage attributed by Hayes to an 'Imitation of Italian SigiUata' (Hayes 1976,
75-77, Hayes 1978,64 ) while furnishing sufficient evidence to confirm that both of these
products could have originated in the same place (Kenrick 1985, 510), did not clearly
define the exact locality of their manufacture (some of the samples analysed, nevertheless,
formed a subdivision dose in composition to Black Glaze wares of Southern Italy, see
Kenrick 1985,5 10 ). The main factor in favour of the North Aftican origin of this ware, as
postulated by Kenrick , has been its known distribution in Libya and the fact that, outside
of Carthage, it appears to be less common in neighbouring Tunisia and Algeria (Kenrick
1985,284 ).
A distribution in regions of Southern Italy, in particular Sicily and Campania, was
known to Kenrick, though not in quantified terms, through the presence of some of the
potters' signatures which recur on this type of pottery; e.g. BAACTI/MOYN, identified not
only at Berenice (and perhaps Misura ta ), but also at Catania, Pozzuoli, and Cuma (Kenrick
1985,298,300, X 154, CVA,.,.1 033).
Kenrick deplores the lack of studies dedicated to material culture in Africa for the
Roman period; it must be stated that this criticism also holds good for Southern Italy and
Sicily. In these regions greater emphasis should be placed on the whole subject of pottery
production. The potential is demonstrated, f hope, in a study which I am conducting on
the production and circulation of terra sigiUata in Campania between late Republic and
early Imperial times: preliminary results of this study are offered in the present article.

Ulli l'enily ofNaples.

73

G. SOruCELU

The analyses of diverse groups of ware coming from excavations carried out in .the las t
few years at Pompeii (the Impianto Elettrico excavation: for preliminary results see Arthur
1986), at Naples and at o ther Campanian sites (Ischja, Sin uessa, Cales) have, in fact,
permitted a quantitative evaluation of the presence of 'Tripolitanian Sigillata' in Campania,
and as regards its origin, an al ternative hypothesis to that proposed by Kenrick can now be
advanced .

'Production N sigillata in Campania


Among the sigillata wares appearing in levels datable between the first century BC and the
s cond century AD, it has been possible to isolate a number of fragments of a ware which
I shall refe r to as 'Production /1\ whose homogeneous morphological characteristics co r re ~
spo nd closely with those of 'Tripolitanian Sigillata', whilst being easily distinguishable both
fro m the best Italian sigillata and from other imported sigillata wares (Soricellj 1984a ).
The identification of these fragments as 'Tripolitanian SigiLlata' is verified either by the
shapes of the vases or by the presence of a few potters' stamps in common. An example o f
this is the base of a cup found at Pompeii in a level of the Impianto Elettrico excavation,
the profile of which is comparable with the form Berenice B42 5,2. This bears the
rectangular stamp L.PVLLljCARPI (Fig. 1.1 ), the same potter who stamped the base of a
plate of 'Tripolitanian Sigillata' recovered at Berenice (Kenrick 1985, 30 1, X164,

L.PVLL /CA RPl).


In contrast wi th Bereruce and Sabratha, a t Pompeii -Production A/Tripolitaruan Sigil ~
lata' (hereafter refe rred to as 'Prod. /1\), appears in several contexts of the Impianto
Elettrico excavation with a preliminary dating of around the middle of the first century
BC, a period characterised by a greatly reduced prop0l1ion of sigillata (22 fragments in all,
of which more than half are attributab le to Eastern Si~illata A or to productions imitating
its forms), with respect to other fragments offine tableware, cillefly Black Glaze.
These products recur at Pompeu in much larger quantity in group dated abou t 30-20
BC, associated with the 'archaic' Arretine forms of Goudineau (1968, form 2, 5 and 6.)
and with the radial Arretjne stamp C. SER./OCE (eVArr 177 9). In till group, composed
f 68 fragments, 'Prod . /1\ I' presents approximately 61.7% of the sigillata recovered,
against 27.9% of Italian si!:,riLlata and 5.5% of Eastern Sigillata A. The form most fre~
quentl y encountered in this group (Fig. 1.2) has a hanging lip (wruch could be typologically
connect d with the rims of 'Service la' of Fellman), apparently with a later da te of arou nd
20 BC (Goudineau 1968,47). It has not been recognised at Berenice or Sabratha. For the
other two forms present, a plate and a bowl, they may be compared with the plate Berenice
B399.2 and the cup Bereruc B42 3. 1 (Fig. 1.3 ).
Well stratified level of later Augustan date have not been found save for a group of
sherds amongst willch was the stamp L. PVLLI/CARPI. T hese fragments, though not
belonging to any specific context, are chronologically homogenous (their provenance was
the back ~fill of a trench dug by A . Mai uri in the 1940's between the Aerarillm and the
Capifo/ium, almost certainly composed of material distu rbed by his excavation). In tills
group consisting f 200 fragmen ts, 'Prod. A' again represents approximately 38.5% of the
total tableware, against 51.5% of Italian sigillata, and 9(/'0 made up of J 5 fragments
grouped together under the name of 'Production C, being imilar to -Prod. ~ either in
fab ric or type. Many of the -Prod. /1\ types in tills gro up compare with those noted at Bere ~
nice. T he two pla tes (Fig. 1.4 and 1.5) can be compared respectively to the shapes Bere
nice B399 and B4{)O willie o ther examples (Fig. 1.6 and 1.7) can be likened to the plates
Beren ice B409 and B4 10. Nevertheless some new designs are present, such as the plate
(F ig. 1.8) with bipartite interna l wall and anothe r plate (Fig. 1.9), willch may be a deriva74

Figure f

tion of
only gl
In a
and, 0 1
letterir

'TRlPOUTANIAN SIGILLATA' FROM CAMPANIA


~ Last
rthur
fact,

.' G
'-I I

,ama,
Iwbe

~hicb
prreboth

'84a).
y the

~-

pLe of
ation,
s the
e of a
<1 64

,0-2 0
nd 6.)

. 7.) "

=;7

~~~~r
6

~-

1-----------17.
7

~~

~ical1 y

IUp of
:h not
:e was
ld the
n this
of the
ments
he r in
BereBereplates
: p!ate
enva-

f -,:::?

.,.

posed
iered,
;t fre

roun d
or the
renice

~,

Id the

igil)ian to
'ntu ry
in all,
tating

(j

"

C?'

J
J

/
=====7/
'9

Figure I. Nos 1-9, Impilllllo Elellrico, Pompeii. Scale I: 2.

tion o f the 'Service l' of Haltern (equivalent to Kenrick's group A). Furthermore this is the
only group of stamped fragments from the excavation .
In addi tion to that of L. Pullius Ca rpus, a radial stamp CHI/RE in a circular cartouche
and , on the base of a cup Berenice B 42S.2, a rectangular stam p with two lines of uncertain
lettering (perhaps EROS on the lower line ?) are also to be noted.
75

G. SORIC LLl

The percentage values obtained for 'Prod. A' are lower in two contexts of the excavation
dated respectively to the Tiberian and Claudian eras. [11 the Tiberian context ( 185 frag
ments ) 'Prod. P\ reaches approximately 25.5%, against 65.4% of Italian sigillata and 3% of
Eastern Sigillata A. The most complete type in this group corresponds to the bowl Bere
nice B427 (Fig. 2.1 0), though other rims, comparabl with the plate Berenice B409 and
carinated bowl Berenice B417 also occur. In the Claud ian context the p roportion of 'Prod.
P\ drops again, resulting in only 19.7% of the total 66 sherds, as docs the proportion of
Italian sigillata (approximately 57.6%); Eastern Sigillata A however, rises to approximately
7.5 %. Overall, the composition of this group is anomalo us with a high number of frag
ments of unidentified production (around 13':;'0 ) and a strong component of residual
material (this seems to be especially true for Eastern Sigillata A ).
Data from a group of sherds from excavations at S. Sofia, Naples, seem more reliable.
These are datable from the Tiberian/Claudian era and consist of 365 fragments. The
percentage of 'Prod. A' is approximately 14.5%, against 73.8% of Italian sigillata and 1.5%
of Eastern Sigillata A. To these must be added c. 50/" of fragments attributed to other local
production. The most common types are the large plate Berenice B407/B409 (F ig. 2.11 ),
the carinated bowl Berenice B417 and the conical cup Berenice B427. Some fragments
can be attributed to the hemispherical bowl Berenicc B419, while one base (Fig. 2.12) can
be likened to the plate Berenice B403.
A group of sherds from the excavation at the Girolamini (Naples), may be assigned to
the Claudio-Neronian period. This has yielded a fair quantity of fragments of 'Prod. P\
with divers examples capable of being almost completely recomposed . The most common
types are still the large plate Berenice B409 (Fig. 2.13 ) and the conical cup Berenice B427
(Fig. 2.14 ). The carinated bowl Berenice B41 7 is also frequently found, to which may be
added two examples of the hemispherical bowl Berenice B419 (Fig. 3. 15) and one
example similar to the small plate Berenice B410 (Fig. 3.16 ).
A group of 94 sherds from the excavation of Carminiello ai Mannesi (Naples) may be
dated to the middle of the first century AD . Here 'Prod ~ represents approximately
12.8% of the sigillata, against 73.4% of Italian sigillata. A further c. 10.5% is made up of a
group of fragments of 'Production D', identified for the moment only in Naples (appearing
for the first time at S. Sofia, where they represent 3.6% of the entire collection of pottery
found ) and whose forms resemble those of 'Prod. ~. 'Prod ~ is represented by the forms
which appear to be the most frequent from Tiberian times - the large plate Berenice
B409, the conical cup Berenice B427 and the carinated bowl Berenice B417.
'Prod. P\ is again present in another group, consisting of 147 fragments, dated to the end
of the first century A D, also found in the excavations of Carminiello ai Mannesi. It is
represented by approximately 16.8% of the total fine pottery, against 6.1.7% of Italian sigil
lata and 10% of imported products.
In the groups of later date, and particularly in those assigned to the first half of the
second century AD, the presence of 'Prod. P\ is always meagre, sometimes totally non
existent, as in the assemblage from Cratere Senga at Pozzuoli (SoriceUi 1984b). This leads
me to believe that circulation of this ware in Campania did not continue beyond the firs t
century AD. A reasonably precise chronological peg is offered by the Pompeian material
in use in AD 79. These products were still in circulation, as demonstrated by a plate Bere
nice B407 with a rectangular stamp BLAST (inv. 10139 ) and a few unstamped carinated
bowls attributable to the form Berenice B417 (inv. 15258 and 15271 ). To these one may
perhaps also add a fragmentary plate, attributable to the type Berenice B410, with a
rectangular stamp EPAPRA (inv. 37145). However, one is dealing with a reduced percen
tage of examples when compared with the rest of the sigillata discovered at Pompeii and
they may be considered residual specimens still in use at the time of the eruption.
76

ation

'TRIPOLlTANlAN SIGILLATA' FROM CAMPAN IA

frag
\% of

Bere
~ and
frod.

,....
,....

II II

on of

~
~

,....

nately
frag

<'-i

......
~
\:;

<..

~idual

J)

.:.-;
~

<:0..

liable.

, The

,-"

1.5%
. local
2.11 ),

t:

ments

2) can

,g

(.J

,....

C')

':1-'

......
I

r")

......
'"
a

led to
od. A'
nmon
B427

:<::
~
~

<:0..
-~

<:

~~

lay be
:i one

~
J)
~

r-i'
......

lay be

nately

---

N
,....

Ip of a

<:

eanng

.:......
'<;;

ottery

forms

c...

remce

goo,

le end

L:.l

i. It is
1 sigil
of the

{ non

; leads
Ie first
aterial
Bere
inated
e may

:::

.~
0
,....

111 _1 ~

I.

t::

--

a'

~
<"i

1::;::

::.r

;_J

"'

with a

ercen

:ii and

77

G. SO RlC LLI

~;'

C3

und

'f2
""
~

oN

Poz
thol
to 0
1

.~

""

..",.

<:

Vi

198

Ir7

(")

Xl:

"-I

.:...."
'53

radi

~
::::,..

the

::::,..

/0,

,,''-'

Xlm :

"

67 .L
F

-'"
:..:.J

at PI
'Pro
frag
sigiJ
freq
Obtll
T
sigil:

::::,..

,.....

"'~

"I

.:....,

g
"I

~;

the f

co

blag
corr,

-2"'

'"

"<.

:::
~

\\

Vi

large

:::

B40

cup:
the E
bow l

"

.....
"-I

'"
~

in th
to ta
supp
tho (

...

-'"::::,..
~

187~

=::

810)

. ,. .

Tl
early

(3
"'<i

"-I

afO lIJ

"""=::

fact,
appe
seem

" I

" I

-...c,"'

......

lA,

r:'~

W
whicl

Arez
POZZI
'Prod
Girol

-. -.
'"c -'"

. C/)

'"~" ~.
r

78

~~
L:: ~

'TRlPOLITANlAN SIGILLATA' FROM CAMPANIA


.-:;
:::

...

sr
gf
~

v-;
Ir,

G,

~
.<:;
~

,,~

.~~

-2
~

.S

E:

~.

c:)

:.:::
'~

t<

,~,

:d'
>.>
Cl)

'=
'<:
~

-G,
~

~
':

-'::

.::

~'

(3
~

'"

::::

""""
-c

I I'"'j

~ ~
~

.~

~ '3
oJ")
,..."
.

~{
Cl);;:;

i.i:: <:

In addition, two large groups of 'Prod. P\ found at Ischia, in the excavations carried out
under the church of S. Restituta, and at Monte S. A ngelo, in the immediate hinterland of
Pozzuoli, are worthy of consideration. Both of these lack precise stratigraphic contexts,
though a few sherds bear potters' stamps, whilst the class is well represented in proportion
to other contemporaneous sigillata wares.
The Ischian material provides two stamps: EROS, in a rectangular carto uc he (Monti
1980, Fig. 95), probably the same potter represented at Berenice (Kenrick 1985 , 30 1,
X 157), and SEX, on a palm branch, in an oval cartouche (Monti 1980, Fig. 95 ).
From Monte S. Angelo we have a rectangular stamp reading NlrEP (Fig. 3.17 ) and a
radial stamp on two lines, unfortunately illegible (in the upper line only the last two letteL,
10, can be read ). These fragments (a few of which are illustrated in F ig. 3.18- 21) come fro m
the area of a villa, the object of clandestine excavations, where 'Prod. A' represents appro
ximately 25.5% of the total of 184 fragments of sigillata gathered from the surface, against
67.4% made up of fragments of Italic sigillata, and 7.1 % of other types.
Figure 4 summarises graphically the percentages of the groups fo und in the excavations
at Pompeii (Impianto Elettrico) and Naples (S. Sofia and Canniniello ai Mannesi ), between
'Prod . A', Italian sigil lata, other sigillata wares attribu ted to local production (in cluding the
fragments grouped together under the names 'Production C' and 'Production D ' ) and other
sigillata wares either inlported or simply unidentified (miscellaneous). To this is added a
frequency table of the material from Monte S. Angelo which illustrates the q uantities
obtained from an analysis of the surface finds.
These resul ts appeal' to indicate a progressive fall-off of 'Prod. N. with respect to Italian
sigillata. After having obtained particularly high percentage values in contexts o f the end of
the first century BC, from the Ti berian era 'Prod. N. represents only 15- 25% in the assem
blages, against 65 - 75% of Italian sigillata. This strongly marked fall-off would seem to
correspond with a decline in the number of types .
While groups of sherds from the Augustan period would seem to be characterised by a
larger number of vessel types, in later groups these seem to be limited to plates Berenice
B409 and B4 10, (which represent two diverse sizes of the same type-for m), the con ical
cup Berenice B4 27 (which would make up the 'service;' with the two preceding forms ), and
the Berenice carinated bowl (841 7 - to which perhaps may also be added the hemispherical
bowl Berenice B419, relatively welJ represented in the pottery group of Girolaminj).
All this could be interpreted as a conseq uence of the opening of the Pozzuoli workshops
in the last years of the first century BC which, with their high-quaHty products, were able
to take over in a short space of time, a large slice of the Campanian market previously
supplied by potters who were producing 'Prod. A'. The Pozzuoli workshops, or at least
those which were noted during excavations carried out by Di Criscio in 18 73-74 (Bruzza
1875; Comfort 1964 ) must have begun their activity around 15- 10 Be. (Comfort 1970,
810), at least twenty five years later than the appearance of Prod. A'.
This chronological gap between the two wares can be clearly perceived , not only in the
early appearance of 'Prod. N. in those levels of the Impianto Elettrico excavation dating to
around the middle of the first century BC, but also in various typological differences. In
fact, while some forms of 'Prod. N. are correlated with those of 'Service r of Haitern, thus
appearing to indicate their contemporaneity, the typological repertory of Puteolan sigillata
seems to consist almost exclusively of later types or 'Service II' (von Schnurbein 1982, 84 ).
What is more, while the potters who produced 'Prod. N made use of the radial stamp
which, according to Goudineau (1968,3 52- 353), was substituted by the central stamp at
Arezzo around I5 - I() BC, radial stamps were never employed for the products of the
Pozzuoli workshops. A n even earlier date could be assigned to a series of fragments of
Prod . A' which present a 'lozenge' stamp. In addition to the base of the plate fo und at
Girolamini (Fig. 3.2 2 ), two different examples of this stamp may be noted: the first comes
79

G. SORICELLI
654

61.7

from
576

515

27.9

Pompe ii I. E.

pompeii I,E.

Pompeii I. E.

Pompeii I.E.

30-20 Be

augustan

liberian

claudia"

738

73 .4

61.7

3713
B-tn

(Mo
sigi1l2
Th
five p
ductil
Black
Cap
area,
those
indic<
fragm
third
Co
earlie
highe
into C
FUI
in bot
Camr
Gree~

S. Sofia

Carminiello ai Mannesi

Carminiello ai Mannesi

' i beria - claudia"

mid

late

I AD

I AD

67.4

c::J

Produzione A-TripOlitania" Sigi ll at a

It~MWWJ

Italian Sigillata

Other local Products

Imported and

Miscellaneous

Monte S . Ange 10

Figure 4. !'ercel1li1ge vallles 0/ dijTerell1 Iypes o/ sigillalo pUllely ("hod. A', !laliOI1 , ()Ih er loca l prodllcis alld
ill1por/ed or unidellliJied ware.l ) III Pompeii (/mpianlO Llellrico sile), Naples (S. Sofia alld Canniniello ai
MIIllnesi) (lild PozUlo li (IHollie S. Angelo).

80

noted
fourth
BLA!J
also b,
tion tc
be not
Naple
even l

(ClLl
3609)
The

ductio
exclus
were f!
tures j
prodw
(Pavoli
The
angula
Catani
'Prod.
the sta
fragme
E aster
Exa
togetht

Xl 56,
of thei

'TRlPO LITANIAN SIGILLATA' FROM CAMPANIA

s and
I/o ai

from Carminiello ai Mannesi, the second is preserved in the storerooms at Pompeii (inv.
37136). Very similar lozenge stamps characterise a production of Black Glaze ware of
B-type located in northern Campania, probably at Cales and well distributed in the region
(Morel, 1975,279-280; 1981,82). On the other hand, this type of stamp is rather rare on
sigiUata.
T his provides an opportunity to refer to a unique example from Rome of a lozenge with
five points, hearing T/ at the centre, which has been attributed hypothetically to local pro
d uction (Schindler Kaudelka 1984,24, No. 81 ) and which resembles a series of stamps on
Black Glaze ware (themselves bearing five-pointed lozenges with A T (from Capua, CVA
Capua-Museo Campano , fasc. HI, plate 29, No.9 ), or P, at the centre (from the Vesuvian
area, D'Avino-Parma 1981, Fig. 1 )). The presence on sigillata of lozenge stamps similar to
those which appear on Black Glaze ware justifies the belief in their contemporaneity and
indicates that the design was passed from Black Glaze ware to sigillata. In this case the
fragments of 'Prod. A' (and the Roman example ) could be dated towards the middle or the
third quarter of the first century BC
Considering, therefore, that these products appear in the region of the Bay of Naples
earlier than their appearance at Berenice and Sabratha, and the fact that they possess
higher percentage values in Italy, it is difficult to believe that these wares were imported
into Campania from North Africa.
Furthermore, the distribution of the few stamps attributed to this ware, their appearance
in both G reek and Latin, and the names of the potters themselves, all seem to indicate a
Campanian origin for this class of pottery, particularly in the Phlegraean Fields, an area of
Greek culture and language. The potter's stamp of BAACT//MOYN, as has already been
noted, occurs on four vessels, coming respectively from Pozzuoli, Cuma and Catania and a
fourth from Berenice. A further example from Tarragona gives the same name in Latin-
BLAST//NIVNA T/ (CVAr,. 1033d ). The plate from Pompeii with the stamp BLAST can
also be attributed to this potter (for the Grecian Blastus see Solin 1982,962-963 ). In addi
tion to the relative concentrations of these stamps in the Phlegraean-Vesuvian area, it must
be noted that the genlilicium Munatius is very common in the region. It has been noted in
Naples even ill mid-Imperial times (CIL 10.1492; NOI.Sc. 1887,291; Maiuri 1913. Perhaps
even CILIO.2573, attributed to Pozzuoli, is Neapolitan). The name occurs at Pozzuoli
(CIL10.1767, 2048,2285,2751 , 2752,2754, Not.Sc.1913, 25 ), Miseno (CILlO.3505,
3609 ), Herculaneum (CIL1 0.14(3 ), and Pompeii (Castren 1975, 193 ).
The Greek/Latin alternation on the stamps may reflect a bilingualism in the area of pro
duction. Potters who signed their ware in Greek, but whose products were distributed
exclusively in the West are known at Pozzuoli, where the stamps LJ/ONYEION and flYAA
were found by Di Criscio (Co mfort 1964, 24. No. 89 ), while examples of bilingual signa
tures are known also at a later date on the lamps with the stamps KELCEI/CELSI,
produced for approximately a century at a workshop located most probably in Campania
(Pavolini 198 1, 06 ).
The base of a plate from Nuceria referred to as form Goudineau ] 4 and with a rect
angular stamp AMlwN, (Budetta et al. 1984 ) may be equatable with the potter who at
Catania, Carthage and Berenice signed himself AMPHIO or ANIPHIOI on vessels of
'Prod. A' (C VArr 65b and c; Kenrick 1985, 300, X 152 and X 153 ). Bearing in mind that
the stamp AMlwN recurs also on Eastern Sigillata A (see Hayes 1985, 36, form 48 ), the
fragment from Nuceria (not examined by the writer) could equally, therefore, have an
Eastern origin.
Examples of signatures known, at the moment, only in Greek, are the stamp NlrEP
together with the stamps signed EPlfl from Berenice (letters which Kenrick 1985, 300,
X156 , prefers to EPID ): the use of Greek by these potte rs lends support to the hypothesis
of their localisation in an area of Greek culture. One may bear in mind that even in
81

G. SORlC LLI

Im perial times magistrates at Naples re tained their Greek names, whilst also keeping alive
the typically Greek institutions like the fratriae and using Greek for pu blic documents.
In Campania, other family names which appear on these prod ucts turn out to be rather
co mmo n, uch as Pompeius, Marius and P ullius. The first two do not appear to be
pa rtic ularly indica tive since, altho ugh they are widespread in Campania, they are also fre
quently attested in other areas. Sex. Pompeii are, howev r, attested at Pozzuoli
(CIL IO .1 594, 2862 ) and in the Vesuvius area (ClL10 .1 403 fro m Herculane um; 8157
fr m Pompeii), while a O. Pompeius appears amongst the Ii t of pott rs fro m Pozzuoli (0.
Pompeius Serenus, C VA ,.,. 1354- 1359). Similarly, M. Marius is no ted many times in the
Phlegraean area (ClL IO.27 12 from Pozzuoli; Denni on 1898,385, No. 29 from Pozzuoli;
3 2, No. 49 fro m Baia; IGXrv, 720 from Naples).
T he name Pullius, on the other hand, is frequ ntly met with in Campania, but seldom
elsewh re: it is found at Pozzuoli by the end of the second century B (ClLl 0.1 781,
datable to 105 BC). Pullii are noted at Ca pua in early Imperial times (CILl 0.4309), at
Sorrento (CILl O.742 ) and possibly in th v: suvian area (CIL1 0.2900 - 2902 , attri buted to
Pozzuoli, bu t fo und in the Vesuvian area ). The name i usually widespread in Campania
under the form of Pollius, already attested at Capua in 94 BC (CILI O.3772), and fre
quently at Pozzuoli (CIL1 0. 1574 , in the Neronian era; 10.2252, 2856 ). F inally, two
examples of sigillata stamp d POLLl (eVAr,. 1349) probably refer to the gentilicium of
Polliu , noted on possibly locally p rod uced vessels from D i Cri 'cio 's excavation at
Pozzuo li.
The h pothesis that the so-called 'Tripolitanian sigillata' originated in the Bay of Naples
receives additional support, however fro m p trological analys s and from the discovery of
possible kiln waste which aUow Naples to b pin-pointed a' a potential centre of
produc tion.
Petrology
Result from the thin section analyses ca rried out by Dr. D. F. Williams on five samples f
'Production A' (Sam ples 1, 2,3, 4, 6) and one of 'Production C' (Sample 5), are given here:
'Sample I : the most prominent non- plastic inclusions are grains of grcen colourless clinopyroxe ne scattered
throughollt the fabr ic. Also prese nt are some fragmen ts of volcanic rock , sa nidine felspar, ye llow-b rown
gam t, some grains of quartz and l1e ks of mica, and with a little limestonc. This sherd is d istinctive in thc
hand specim n as the fabric appears to contain much 'black sand ' - d ue to the presence of dark-colou red
augite, T his re ails the fab ric of Pompeian Rcd Ware and the ampho ra typcs Dress I 1 and D ressel 2-4,

though t to have been made in the region of Pompcii and Herculaneum (Peacock 1971 , fabr ic 2; 1977, fa bric

1). A similar origin seems likely for the sigillata she rd.

Samples 2 -6: samp les 2 (F ig, 1,7 ) and 3 appear fai rly simila r under the microscope with a scatte r of sub
angular quartz grains, nec ks of mica, some lime to ne , sanidine fel spar and a few grains of clinopyroxene, The
other three samples also contain some grains of clinl>pyroxe ne and sanidin e but lack the li mestone and are

slightly coarser, with samplcs 5 (F ig. 3.23 ) and (j (F ig, 1.6) in particular di splaying frequent grains of quartz
and tlecks of mica. Thc presence of clino pyroxene and sanidine in all fi e she rds sugge,. ts a source, o r more
like ly so urces, along thc It lian volcanic tract. The fab ric of th se [j e sherds is diffe rent from previously
sampled sigill ata and moulds fro m Cu ma (So ricelli 1n2 ), while the Po Valley seems geologically unlike ly (tw
sugncsted origi ns). The pctrology wou ld fit a local Campanian origin, and Puteoli is a possibility (W illiams
J 97 8), howe er othe r sources cannot be ruled ou t (including eapolis),'

Some possible kilo wasters


At least two p robab le wasters from kilns have been discovered at Naples. The first was
found in the pottery assemblag from the Girolam ini xcavations. This con i ts of a small
cup, similar to the form Berenice B427, wi th the body strongly deformed by a firing acci
dent (Fig. 3.24 ). The clay where it is not overfired , i' brown, with flecks of mica and grains
of limestone. The slip completely burnt on the rim and on part of the bowl, has tu rned
brown and shows flecks of mica.
82

T
squ(
to tt
able
not I
also

than
pres
worl
clay
pro(
cent
use I
It
Nap
ther
pite
sugg
grm
type
It

CoV(

duct
A

sigi/,

larly
las (
groLl
froIT
shar
tiom
coul
Distl
Lea\.
indi\
wide
(Fig.
Tyrn
I

ance
(nea
I
be pi
ti on
Fran
mati

'TRIPOLITANIAN SlGlLLATA' FROM CA MPA NIA

g alive

s.
rather
to be
so fre
Izzuoli
81 57
IOli (Q.
in the
zzuoli;

ieldom
1.1 781,
at
~ted to
hpania
rId fre
y, two
itun of
tion at

P9),

Naples
very of
1tre of

lples of
here:

:catterecl

.v-brown
>'e in the

:olourccl
isel 2-4 ,

T he second waster has been found in the course of recent excavations carried out in the
square in front of the church of S. Maria Maggiore, in a back-fill given a preliminary dating
to the fi rst years of the second century AD. This example is a fragment of a plate attribut
able to the fo rm Berenice B409 (Fig. 3.25). The characteristics of the clay and the slip are
not definable because of overfiring. In the same stratum a second fragment of 'Prod. /\ has
also been fo und, exhibiti ng reduction on the internal surfaces.
W hilst it must be admitted that vessels can be distorted by heat in circumstances other
than their initial manufacture and that in neither case was any kiln structure located, the
presence of possible wasters underlines the belief that Naples was the seat of one or more
workshops which made this pottery. It must be added that, at least microscopically, the
clay of these p roducts is different from that of the Black Glaze ware 'Campana /\, also
produced at Naples between the end of the third century BC and the middle of the first
century BC, either because of the employment of different techniques, or because of the
use of different clay sources.
It is probable, however, that, in addition to the workshops suggested by the wasters at
Naples and a hypothetical one in the Vesuvian area presumed on the basis of clay analyses,
there may have been other workshops located in the area of the Bay of Naples. In fact, des
pite a substantial morphological uniformity within 'Prod. /\, sufficient variations occur to
suggest the presence of various wo rkshops: variations which, as in the example of the
groups from Ischia and Monte S. Angelo, may be noted by the thickness of the wails, the
type of slip, the surface finish and the grade of clay elutriation.
It is clear that precise answers to this problem will be obtained only through the dis
covery of more rich kiln waste (and consequent identification of eventual centres of pro
duction ) and by a more thorough programme of petrological analyses.
Another problem which needs clarifying is that of the connection with other terra
sigillata production in northern Campania (where, however, 'Prod./\ is recorded), particu
larly with the 'Local imitations of terra sigillata' signalled amongst the sherds from the vil
las of Posto and San Rocco (Morris 1979, 123- 126; Bird 1985, 188-190 ). This last
group seems to possess a rather different typological repertory from that of the products
from the Bay of Naples, (though this may be due to rather scanty documentation), but also
shares certain traits, possibly due to analogous production techniques. If the 'Local imita

tions' from Posto and S. Rocco belong to 'Prod. /\ the typological differences encountered

could indicate the existence of further workshops sited in the north of Campania.

7, fab ric
" of sub

Distribution

~JJ e.

Leaving aside the mo re or less limited circulation so far ascertained for the products of
individual workshops, the field of distribution of 'Prod. A' on the whole appears to be
widespread, taki ng in Southern Italy, the coast of North Africa and, to some degree, Spain
(Fig. 5). T he distribution in South Italy appears limited mostly to the regions bordering the
Tyrrenean Sea.
In addition to being p articularly widespread in the area surrounding the gulf of Naples,
'Prod. /\ is also recorded, in apparently more modest quan ti ties, also in Northern
Campania, at Liternum, Sinuessa and Cales. For the moment, the most northerly ap pear
ance of this ware is represented by a small group of fragments found at the villa of M atrice
(near Campobasso), in Molise.

Information fro m southern Campania has not yet been gathered though the class may

be present at Paestum (Pedley - Torelli 1984,375, pointed out the presence of an 'imita

tio n sigillata of unknown provenance' without, however, describing its characteristics ).

From Calabria other examples of 'Prod. A.' have been recorded: from Vibo Valentia (infor
mation from Dott. B. Sangineto ); from Reggio Calabria (material exhibited in the

The

and are
)f quartz
or more

"eviously

(ely (two

Williams

rst was
a small
19 accl
i grains
turned

83

i)

111

G. SORICELLI

mus
Mar

t
0

:J

j.

A
283

T
men
perh
niqu

clas~

AC
~

tage
Itali,
gestt

F<

111
I

283

;[/
'"'"

equc
of N

"

strOJ

494;

may

1971
Tl

repo
over.
atelit
easte
coul(

.::!
~

:o

VJ

.g
~

~
<:>

.S

~
"C

-t::

.~

t;

Com
In co
man
the (
favOl

OJ

ever,
reglO
and (
14-1
Naplt
whid

;:;

e..l:

'c
t::

'S
~
's

'"
Q
lr)

"-'

:::

5fo

; ~

114

Ackn.

AlthoL
tanian
Meditc
I wi:
and fo
Matric
amp,
thank I

Prof

Casert.

'TRlPOLlTAN lAN SIGILLATA' FROM CAMPANLA

museum ); fro m Scolacium (excavation by Prof. E. Arslan ) and perhaps also from Ciro'
Marina (material exhibited in the museum of Reggio Calabria).
As regards Sicily, at present one must refer to finds reported by Kenrick (19R5,
281-28 4 ).
The p resence of 'Prod. N in Adriatic coastal regions may be confirmed by several frag
ments disco vered at O tranto (information from Dott.ssa G. Semeraro ). To these may
perhaps be added a small group of sherds from Ordona (Vanderhoeven 1976, 81, Tech
nique C) which, from a typological point of view, are fairly similar to fo rms up tiU now
classed as 'Prod. A: (e.g. Vanderhoeven 1976, Plates XLIV, 130; XLVlll , 186; LVIIJ, 404 ).
A Ca mpanian origin for these fragments could be backed up by the relatively high percen
tage of Puteolan sigiJlata ware found at Ordona (approximately 13% of the stamps on
Italian si,gillata published by Vanderhoeven), though a possible local origin has been sug
gested on somewhat dubious grounds (Vanderhoeven 1979,83 ).
Fo r the North African discoveries one turns once again to Kenrick's reports ( l985 ,
283- 284). The conspicuous presence of 'Prod. N at Benghazi exists side by side with an
equally no table quantity of other products from Campania and , in particular, from the Bay
of Na ples (fine pottery, coarseware and amphorae ). These finds point to the existence of
strong trade links between Campania and Benghazi in the fi rst century AD (Kenrick 1985 ,
494), whic h would explain the large market which 'Prod. N found in this area. The ware
may also have been more common at Carthage than was originaUy thought (Hayes 1976;
1978; M. Fulford p ers. comm ).
T he presence of this ware in Spain is indicated by the stamp-BLASTI/MVNATI,
reported from Tarragona (eVArr 1033d ). It could also be present in France, where, more
over, local wares exist with morpl1ological characteristics very close to 'Prod. N (e.g. the
ateliers of Bram ). Two plates recovered from the wreck of the Tradelier, attributed to an
eastern production, could be 'Prod. A' forms (Liou, 1975 , Fig. 41 ; in particular Fig. 41.1
could be identified with form Berenice B399, and Fig. 41.2 with form Berenice B401 ).

:::

t>J:)

i;:
;::

.;::

::;

,-.....
.~

-"<:

.2;:;

Conclusion
In conclusion, the prob ~em remains as to how to define, in future , 'Production A/Tripolita
nian Sigillata'. If one accepts the localisation of the production of this ware in the area of
the Gulf of Naples, it will be necessary to abandon the term 'Tripolitanjan SigiJlata' in
favour of another, more appropriate name.
One possibility would be to rename it 'Campanian Sigillata' - a definition which, how
ever, could generate confusion with other sigillata wares which were produced in the
region and which show noticeable morphological differences, such as those of Pozzuoli
and Cuma and, in particular, those of the 'Campanian' potters reported by Pucci (1977,
14-15 ). lt is perhaps preferable to continue to call the ware 'Production A' of the Bay of
Naples, whilst awaiting mo re precise evidence regarding its centres of production, of
which Naples is currently the strongest possibility.

-
~

~:-
~

.~
~

.E

Q
10

.~

----.J" L.:....

Acknowledgements
Although this articl e has d isagreed wit h Philip Kenrick's hy pothesis concern.ing the proven ance of 'Tripoli
tanian Sigillata', it must be stressed that many o ther aspects of hi s outstanding and pioneering resea rch on
Mediterranean fin e wa res wi ll remain of fund ament al importance for many years to come.
I wish to thank Dr Jo hn Lloyd for having invited me to contribute thi s paper to th e journal Lihyan Sill dies
and for having offered me the opportunity to pre pare for publication the sigill ata from his excavatio ns at
Matrice. This study forms pa rt of a research project o n the productio n and distribution of Roman potte ry in
Campania, coordinated by Dr Paul Arthur, whom I thank for he lpful criticism. Furthermore I also wish to
thank Prof. Raffaela Pierobo n, Dr M. Fulford and Dr P. Kenrick for having read and commented o n this study.
Prof. Enrica Pozzi and Prof. Baldassare Conticello, Supe rinte ndents of the provinces of Naples and
Caserta, and Pompeii res pectively, as well as Dott. G iuseppe Vecchio, Sergio Cascella and Don Pietro Monti ,

8' )-

G . SORlCE LU
have permitted me access to unpublished mate rial. Dr David W il liams kindly carried out the petrological
analyses.
Last, but not least, Do tt.ssa Alessandra Pollio prepared the illustrations and Mrs linny Hutchinson under
too k both the burde n of translating my Italian text into E nglish , and the task of typing it.
All rrors arc my own.

Bibliography

f ,"

Solil
Sori
Sori
Sori
Van,
Van,
von
Will

xcavations 1980-198l. Antiquaries Journal 06:


29-44
Bird , l . 1985. Local Terra Sigillata. In M. A. Cotton and G. P. R. Met rau x The S(/II Rocco Villa (/1 Francolise,
London: 188-190.
Bruzza, L. IS75. Scoperta di figuline in Pozzuoli. Bullelino dell ' InSlilUiO di Corrispondenza Archeologica.
242-24 6.
Budetta, T., Laforgia, E . a nd Miniero, P. 19 84. La sequenza stratigrafica ed i materiali d ello scavo d e lla torre
III di Nocera. l. Selllillario de! cicio: Sulle ! orliftcaziolli in Italia centro-meridionale Ilel 1I e lleI I secolo
o. C. Centre Jean Berard apoli.
Castren, P. 1975. Ordo Populusque Pompeianus. Po lily and SocielY in roman Pompeii. Romel.
e lL = Corpus In scripliol1ul11 LalillalTllI1.
Co mfort. H . 1964. Puteolan SigiUata at the Louvre. Rei Crelariae Romane F([U/orum A Cla. 5/6: 7-28.
Comfort, H . 1970. Ceramica Puteolana. sotto voce Terra Sigillata. In Enciclopedia dell'Arle Anlica, Supple
mento, Roma: 8 10-81 1.
C VA = Corpus Vasomm Allliqllomm.

eVA rr = A. Oxe and H . Comfort (eds ), Corpus Vasorum Arretil1orllll1. Bo nn.

D'Avino, R. and Parma, N. 1981. Una villa rust ica romana in localita' upa Olivella a S n t'Anastasia. In 1I

COllvegllo dei Gruppi Archeologici della Campania. (Maddaloni, 24-25 aprile 1981 ): 9-53.
Dennison, W. 1898. So me new inscriptions from Puteoli, Baiae, Misenum and C umae. Americall Journal or
Archaeology 2: 373-398.
Goudineau , Ch. 1968. La cerarnique aretine lisse (Fol/illes de I'Ecole Fral/{'aise de Rome ci Bolsena-Poggio
Moscini - 1962-/96 7, IV ), in Melanges de {'Ecole rran('oise de Rome, Suppl. IV.
Hayes, 1. W. 1976. Pottery, stratified groups and typology. In J. H . Humphrey (ed.), Excavations at Carthage
1975 conducted by the University of Mi chigan. Vol. 1. Ann Arbor: 47-123.
Hayes, 1. W. 1975. Pottery repo rt. In 1. H. Humphrey (ed. ), EXClivatiol1s al Carlhage 1976 condl/cted by Ihe
Ulliversity a/Michigan. Vol. rv. Ann Arbor: 23 - 98.
Hayes, 1. W. 1985. SigiJlate Orientali. In Atla/7/e delle forme ceramiche II. Ceramicajlne romona nel bacino del
Medilerraneo (ICIrdo ellen ism 0 e primo illlpero). t;n cicIopedia dell'Arte Antica. Supple mento, Roma:
1-95 .
IG = Imcriptiones Graecae.
Kenrick, P. 1985. Excavaliolls at Sidi Khrebish, Benghazi (Berenice), 1I1, I: The jine pOltel)'. Tripoli.
Ke nrick , P. 1986. Excavalions at Sabratha 1948- N51. London.
Liou , B. 1975. Reche rches Archeologiques sous-marines. Galha 33; 570-605.
Maiuri, A. 1913. La nuova iscri zione della fratria Napoletana degli Artemisii. Studi Romani 1: 21- 36.
Monti P. 1980. Ischia, archaeologia e sloria. Napoli .
Morel, 1. P. 1975. Aspects d e I'artisanat dans la Grande Grcce Romaine. In La J\llaglla Grecia in ela' romal/o.
Alii del XV Convegno di sludi sulla lVlagn(/ Grecia (Taranto ), 5-lO ottobre 1975: 263 - 324.
More l,1. P. 198 I. Giramique campal1iel1ll e. Les Formes. Roma.
Morris, 1. 197 c . Te rra Sigillata Wares and Imitations. In M. A. Cotton The Late Republical/ Villa al POSIO,
Fr([llcofise. London : 177-126.
Not. Sc. = NOlizie Salvi Antichila.
Pavolini, C. 198 1. Le lucerne nell'ltalia romana. In A. G iardina and A. Schiavone (eds), Merci, mercali e
scambi nel tvledilerran eo. Bari: 139-184.
Peacock, D . P. S. 1971. Roman amphorae in Pre-Roman Brit ain. In M. Jesson and D. H ill (eds ), The Iron Age
and its lIill-Forts. Southampton: 169-188.
Peacoc k, D. P. S. 19 77. Pompeian Reel Ware. In D. P. S. Peacock (ed. ), Pot/elY and early cOlllmerce. Southamp
ton : 147-162.
Pedley, 1. G. and Torelli, M. 19 84. Excavations at Paestum 1983. American Joul'I1al of Archaeology 88:
367-376.
Pucci, G. 1977. Le te rre sigillatae italic he, galliche e o rientali. In A. C arandini (cd .), L'lnstmmelllwn domesti
CUIIl di Ercolano e Pompei nel/a Prima eta' imperiale. Quademi di cullum materia Ie, I: 9-21.
Schindler Kaud elka, E. 1984. Terra Sigillata aus Rom. D ie Sammlung O lco tt. Rei Cretaria e ROlllallae
Faulomm Acta 26: 13-36 .
Arthu r, P. 198(). Problems of the Urbanisation of Pomp ii :

86

'TRJPOLITANI AN SIGLLLATA' FROM CA MPA NIA


gical
Ider-

Solin, H . 1982. Die G riechischen Persollenllomen in R om. E in Iiandbll ch. Berlin-New Yo rk.
So rieelli, G . 1982. U n'offieina d i N. Naevius Hilarus a Cuma. A rcheologia C1assica 34: ] 90-195.
Soricelli, G . 1984a. Ricerche Preliminuri sulla Produz ione e la circo[azione della Terra Sigillata ill Campania
Ira la [(/I'da repubblica e la prima eta ' imperiale. University of Naples, undergraduate thesis.
Soricelli, G. 19 84b. L e sigilJate. In F. Garcea, G. Miraglia an d G . Soricelli , U no scarico di materiale ceramico
di eta' ad rianeo-antonina d a C ratere Senga (Pozzuoli ). PuleoIi 7/8: 245-285.
Vanderhoeven. M . 1976 . La terre sigillee !isse, In J. Mertens (cd.), Ordona V. Bruxelles : 79-182.
Vanderh oeven, M. 1979 . La terre sigillee re lief. In J . Me rtens (ed. ), Ordona VI. Bruxe llcs: 83-104.
von Schnurbein, S. 1982 . Die unverzierle Terra Sigill(ll(1 aus H.allern. Bociellalleriimer Westfalens 19, Mtinste r.
W illiams, D . F. 19 78 . Petrological analysis of Arretine and early Samian: a pre liminary report. In P. Arthur
and G. M arsh, (eds.), Early fine wares in Roman Britain. British Archaeological Report s 57: 5-12.

" 66:
olise,
!gica.

!orre

ecolo

Jple-

In /I

'aloj'
Jggio

'hage

)' the

o del
.Jma:

mza.

OSlO,

a1ie
Age

Imp

8il:

leSli

([nae

87

G. SORICELLI
'Tripolitanian SigilIata': A Response

By P. M. Kenrick
I am grateful to Dr. Soricelli and to thc Editor for the opportunity to see this intercsting article in advance of
publication. The evidence prcscnted here of chronology, distribution and petrology seem to me to offer com
pelling reasons for supposing the ware discussed to have been produced in Campania, and this is further
supported by the names of the potters and their occasional use of the G reek alphabet. (This was a factor to
which I had not given adequate weight in suggesting an origin in P unic North A frica ). As to the precise origin
of the warc within the region, I am dubious of the significance of two wasters found on their own in separate
excavations: an accident in the later life of a vessel may on occasion be indistinguishable from an original fault
in firing, and a greater concentration of W'asters would be necessary to have the force of definite proof. Besides
which, the foot of the cup illustrated in Figure 3.24 is not typical for Production A/'Tripolitanian' Sigillata.
Soricelli is undoubtedly right in suggesting that a plurality of similar wares was produced in different parts of
Campania (and probably other parts of Southern Italy and Sicily) during the first ccnturies BC and AD. Their
individual characterisation will depend on the publication an d study of a great deal more material from the
region and the present article represents an important step along that road. And as we know that the products
of Campania could so readily find their way into distant markcts throughout the Mediterranean, so the
identification of more of these products is likely to be of relevance to xcavators working in many different
parts of the Roman World.

88

You might also like