Content uploaded by Julia Rabitsch
Author content
All content in this area was uploaded by Julia Rabitsch on Mar 27, 2020
Content may be subject to copyright.
495
REI CRETARIÆ ROMANÆ FAVTORVM ACTA 45, 2018
The present paper gives an insight into preliminary results of the analysis of ribbed bowls made of pottery based on the
findings from Brigantium/ Bregenz. Therefore, different aspects of research are taken into account: the paper will provide an
overview of the known finding spots and the possible origins of the ribbed bowls will be discussed. The indisputable similarity
with ribbed glass bowls Isings 3a/b suggests a close relation between them. Since the ribbed bowls from Bregenz are the only
ones in the Roman province Raetia pXRF-analyses were carried out to determine their origin. The results show that these
bowls were produced locally which is indicated by the same chemical composition as the pre-established reference group
from Bregenz. The very rare occurrence of this kind of pottery could stand for the personal preference of single costumers.
Julia Rabitsch
CERAMICS IMITATING GLASS?
RIBBED BOWLS MADE OF POTTERY FROM Brigantium/BREGENZ
Introduction
Fragments of at least six ribbed bowls made of pottery
were unearthed during excavations in 1974 in the Roman
settlement of Brigantium/Bregenz on the shores of Lake
Constance. The findings and features of this excavation are
presently analysed as part of the trinational project – “Limites
inter Provincias – Rome’s inner frontiers” which is located
at the Universities Zurich (CH), Freiburg (D) and Innsbruck
(A)1. The Austrian part of the project is financed by the FWF,
the Austrian research found. The excavations took place in the
area referred to as ‘Gmeinerwiese’, at the north eastern part
of the settlement. During the excavations remains of at least
five Roman buildings as well as a part of the ancient road
were documented. The finding spot of these ribbed bowls can
be located in the area where trench 1 meets trench 2, close
to building number 76, a so-called ‘Streifenhaus’ (a stripe
house), with seemingly only residential purpose, instead
of the more common multifunctional type with integrated
workshops. Two sherds were found at a more distant point,
in trench 4, an area where a waste pit was located. Sherds
from the same vessels (coarse ware and samian ware) were
found as well amongst the debris layers of building 76 as
inside the waste pit in trench 4. Therefore, we have a con-
nection between them. Although large areas of the Roman
settlement have been excavated during the last 200 years,
the ribbed ceramic bowls from the ‘Gmeinerwiese’ seem to
be the only specimen of that kind from Brigantium, maybe
indicating a single costumer.
1 Lead agency: University of Zurich (Ph. Della Casa, A. Kolb), partners:
University of Freiburg (A. Heising) University of Innsbruck (G.
Grabherr). FWF-project-number: I 1161-G19.
Description of the ribbed bowls made of pottery
The vessels (fig. 1) are best described as a hemispheric bowl
with a straight rim, decorated with separately formed, down-
ward pointing rips attached to the outer surface approximately
1 to 2 cm beneath the rim. The rips itself are triangular in
cross-section with rounded edges. Furthermore they are not
attached in straight lines to the body; instead they form slightly
arched lines. The bowls measure 19 to 25 cm
in diameter and
have an estimated height of 10 to 12 cm. Since no whole
vessels or base fragments were retrieved in Bregenz, it is
not possible to describe the full shape. However, fragments
of the lower body show common features: the rips form a
more or less pointed end which is followed by one or two
lines beneath them. Those lines can be flat or pronounced and
form the junction between the body and the base. Therefore
I would like to suggest a foot rim or perhaps a flat base. The
fabric corresponds with the so-called Terra Nigra – a grey
ware, slightly tempered, wheel-thrown and with a dark grey
or black slip. A very common ware, especially in the prov-
inces Raetia and Germania Superior used for indigenous
forms as well as for pottery imitating shapes of samian ware.
Pottery based on the model of Isings 3a/b?
The general shape of the bowls as well as their cross-sections
can be compared to Isings 3a/b. This common glass vessel
occurs in the entire Roman Empire from Augustan times
until the 2nd century AD and further on. Since the ceramic
bowls are an almost perfect imitation of the glass bowls, a
direct relationship between these two seems quite obvious.
This becomes particularly evident by looking outside the
Roman Empire: Many findings of glass vessels Isings 3a/b in
settlements across the Limes indicate a flourishing trade with
496
Julia Rabitsch
Germanic peoples. Most imports in the Germania magna
date back to the periods B1c until B2a, which corresponds
roughly with the time between the middle of the 1st century
until the 2nd half of the 2nd century AD2. This fact is relevant
because a large number of ribbed bowls made of pottery
was found in Germania. Both, the imports and the pottery
imitations, have more or less the same dating range. This ev-
idence suggests that, almost immediately after the first glass
bowls were imported, the first imitations emerged. What is
even more striking is that both distribution areas, those of the
glass vessels as well as the ribbed ceramic bowls are almost
identical, indicating a direct correlation between these two
variants of vessels3. The local pottery imitations of the bowls
were probably more affordable than the imported glass ves-
sels, which could explain their popularity. A big difference
between the ceramic ribbed bowls found in Germania magna
and those from Brigantium and from other Roman finding
spots as well, is the design of the base. All of the Germanic
bowls show star-shaped or cross-shaped ornaments on the
base, formed by the rips which run around the whole body
and join each other at the bottom of the vessel4. Judging by
2 H. J. EggErs, Der römische Import im freien Germanien. Atlas der
Urgeschichte 1 (Hamburg 1951) Karte 49; T. stawiarska, Naczynia
szklane okresu rzymskiego z terenu Polski (Warszawa 1999) 100.
3 HEgEwiscH 2005a, 206; HEgEwiscH 2005b, 209.
4 HEgEwiscH 2005a Abb. 3; HEgEwiscH 2005b, 80.
the fragments found so far, this shape can be ruled out for
the bowls in Bregenz since their rips form a pointed end on
the lower body.
Distribution of ribbed bowls in the Roman Provinces
Inside the Roman provinces the practice of imitating glass
vessels in ceramics, especially the Isings 3, seems quite
common as well (fig. 2). The main distribution area seems to
follow the river Rhine. However, there are a few outliers, for
example in Vindinum/Le Mans, where fragments of at least
six bowls with pronounced rips and a flat base were found.
On contrary to the vessels from Brigantium all of the samples
of Vindinum are moulded. There they occur from the 2nd half
of the first century until the beginning of the 2nd century5.
A similar mould technique based production is assumed
for the ribbed bowls made of pottery found in Noviomagus/
Nijmegen6. A significant amount of fragmented bowls
amongst the findings in a pottery kiln suggests a local pro-
duction. In contrast to the vessels from Brigantium they all
5 G. guilliEr, Un aspect méconnu de la céramique du Haut-Empire: La
vaisselle modelée, l’example du mans (Sarthe). SFECAG Actes Congrès
Rouen 1995 (Marseille 1995) 218–219.
6 R. P. J. kloostErman/M. Polak/m. J. Zandstra, Opgravingen op het
terrain van het voormalige Canisiuscollege in Nijmegen, 1987–1998.
Auxiliaria 14 (Nijmegen 2014) 123.
Fig. 1. Ribbed bowls made of pottery from the so-called ‘Gmeinerwiese’ in Bregenz (drawings: J. Rabitsch).
497
CeramiCs imitating glass?
Fig. 2. Distribution of ribbed bowls made of pottery. 1–29 Roman ribbed bowls; 30–102 Germanic ribbed bowls mapped
by HEgEwiscH 2005 (map data ©2016 Google, graphic: J. Rabitsch).
498
Julia Rabitsch
are made of an orange fabric with a thin reddish slip and
very pronounced rips with a triangular cross-section and
sharp edges7. The rips are evenly spaced and form straight,
downward pointing lines. The base is formed by a foot rim8.
Fragments from another finding spot close by, Naaldwijk-‘t
Zand Heultje, show a flat base and a slightly outward bended
rim9. The bowls in the area around The Hague (including
aforesaid finding spot) show a variation of production tech-
niques: some are wheel-thrown, some are handmade, and the
rips are either separately formed and attached afterwards or
impressed in the inner surface of the vessels10. The findings
from the modern Netherlands all date between the late 1st
century and the beginning of the 3rd century AD11.
Following the River Rhine the next finding spots can be
located in the Colonia Ulpia Traiana/Xanten12, at a burial
site in Pont by Geldern13 and in the Colonia Claudia Ara
Agrippinensium/Cologne14. The bowls seem to be locally
produced showing the typical light whitish fabric. No infor-
mation about the manufacturing technique has been published
yet; however, when looking at the pictures it is probably true
to say that a mould technique based production was used15.
An interesting aspect regarding ribbed bowls in the lower
Rhineland is that the finding spots correlate with those of
the so-called ‘lower Rhineland ware’, a fine ware of mostly
light coloured fabric and dark colour coating. The key shape
is a hemispherical cup with a variety of barbotine decors,
including the quite popular decoration with broad barbo-
tine-rips applied to the outer surface. These cups occurring
in Claudian times, can be encountered quite frequently until
Neronian times and still appear in later contexts16. Since their
distribution areas are almost congruent a connection between
the two types, the fine ware bowls and the local produced
bowls with pronounced rips is very likely. However, it cannot
7 HaalEbos 1995, 62 Abb. 38,8.
8 HEgEwiscH 2005a, 209.
9 gEErts 2014 fig.4.
10 gEErts 2014 fig.7.
11 gEErts 2014, 772–775; HaalEbos 1995, 60.
12
Unpublished material. Thanks to Bernd Liesen (APX) for the information.
13 cüPPErs 1962, 329–330.
14 FrEmErsdorF 1964, 22.
15 cüPPErs 1962 Abb. 19; FrEmErsdorF 1964 Taf. 66.
16 K. grEEnE, The Pre-Flavian fines wares. Report on the Excavations at
Usk 1965–1976 (Cardiff 1979) 56–60.
be excluded that the ‘lower Rhineland ware’ cups themselves
were an early adaption of the ribbed glass bowls in ceramics
and therefore have the same origins.
The ‘lower Rhineland ware’ is not the only fine ware
providing a vessel which resembles the Isings 3a/b in its
shape as well as in its decoration. A hemispherical bowl with
straight barbotine lines around the whole body belonging to
the ‘South-east England glazed ware’ which is distributed
during the 2nd century can be found quite frequently in
southern England. The fabric is micaceous, usually grey and
covered with a translucent dark green gloss. For the barbotine
decorations, a different whitish clay was used17.
Back to the river Rhine, in detail to the Upper Rhine
Valley, similar bowls were located in the area between Wies-
baden and Heidelberg18. The fabric of these vessels equals the
Terra Nigra, which is quite similar to the fabric occurring in
Bregenz. On the contrary to the earlier described ribbed bowls
the ones from this area do not have pronounced rips made
of separate clay strands. Instead, thin barbotine strings are
applied to the outer surface, comparable with the decoration
used on fine ware. Nonetheless the general appearance of
these bowls still resembles those from Brigantium/Bregenz.
The dating range lies between the 2nd half of the 1st century
and the beginning of the 2nd century AD19.
The southernmost distribution (so far) is in the area be-
tween Lake Constance and Lake Zurich. This area is particu-
larly interesting; firstly it includes the settlement of Brigantium
which is situated at the shores of lake Constance and secondly
both forms can be found here – the barbotine-ribbed bowls as
well as those with pronounced rips. They all seem to be locally
produced, wheel-thrown and mostly in grey, Terra-Nigra-like
fabrics20, only a few individual pieces have an orange colour21.
Both types seem to be equally common: the barbotine-ribbed
ones can be found e.g. in Seeb22, Turi cum/Zurich23 and Vitudu-
17 P. R. artHur, The lead glazed wares of Roman Britain. In: G. D.
Marsh/P. R. Arthur (eds.) Early Fine wares in Roman Britain. BAR
Brit. Ser. 57 (Oxford 1978) 193–356.
18 HEukEmEs 1964, 57–59.
19 HEukEmEs 1964, 56.
20 JaucH 1997, 126; mEyEr-FrEulEr 1990, 171; Vogt 1948, 197.
21 castElla/mEylan krausE 1994, 59.
22 mEyEr-FrEulEr 1990 Taf. 11,30.31.
23 Vogt 1948 Abb. 48,23.
Fig. 3. Barbotine-ribbed bowl from Brigantium/Bregenz (drawing: J. Rabitsch).
499
CeramiCs imitating glass?
rum/Oberwinterthur24. Those with the applied pronounced rips
with a triangular cross-section occur for example in Tasgetium/
Eschenz25, Aventicum/Avenches26 and of course in Brigantium/
Bregenz. There is evidence for the bowls with flat barbotine
rips in Bregenz as well (fig. 3):
two fragments quite similar
in style, measurements and fabric to those with pronounced
rips are decorated with thin barbotine strings.
Since ribbed bowls made of pottery are very rare amongst
the findings in each settlement in this area and often without
a clear stratigraphic context, it is difficult to date them. In
some cases a production time approximately between the
2nd half of the 1st century and the end of the 2nd century AD
is suggested27. The ribbed bowls from Brigantium cannot be
dated exactly as we do not know their stratigraphic context; it
seems they belonged to the inhabitants of building 76 though.
The evaluation of the findings from the ‘Gmeinerwiese’
shows that this part of the settlement dates from the early 1st
century until the middle (or maybe the end) of the 3rd cen-
tury AD at which point it was abandoned to be relocated to
a geographically safer site. The stone buildings date back to
Flavian times and show some modifications indicating a long
period of utilization. Therefore, the Flavian period serves as
a terminus post quem for the ribbed bowls. Apart from that,
the dating range can be narrowed down by the waste pit find-
ings which provide an approximate time period between the
late 1st century and the end of the 2nd century AD, which lies
within the suggested dating range of the other finding spots.
24 J. rycHEnEr/P. albErtin, Ein Haus im Vicus Vitudurum – die
Ausgrabungen an der Römerstrasse 186. Vitudurum 2 (Zürich 1986)
Taf. 47,918; 57,919.
25 JaucH 1997 Taf. 25,408–410.
26 castElla/mEylan krausE 1994, No. 183.
27 castElla/mEylan krausE 1994, 59; JaucH 1997, 56; mEyEr-FrEulEr
1990, 143.
Portable X-ray fluorescence analysis
Since the ribbed bowls from Bregenz count as a single find
in the province Raetia the question arose if they were locally
produced or imported. To determine their origin portable X-ray
fluorescence (pXRF) analyses were carried out by my project
partner Daniel Penz at the University of Freiburg. By analysing
100 different fragments of local coarse ware as well as a signif-
icant amount of Terra Nigra, a reference group for Bregenz was
established. This pre-established reference group is necessary
for the comparison of objects with uncertain origin such as the
ribbed bowls. Contrary to the expectations the analysis shows
that these bowls have the same chemical composition as the
local produced ware which indicates that the same clay deposits
were used (fig. 4). A single outlier (74.5130 in fig.1) can be
determined with higher portions of calcium (Ca), copper (Cu)
and lead (Pb). In addition to the varying chemical composition
it is the only fragment with a different fabric: slightly tempered,
micaceous, very dense and of orange colour. It has a smooth
surface and seemingly no colour coating. Since we do not have
suitable comparative probes the origin of this bowl cannot be
detected. Yet it is possible, that this one was the ‘prototype’ on
which the other bowls were based on.
Résumé
Looking at all the comparisons and results some conclusions
can be drawn. Firstly, there seems to be a direct correlation be-
tween the ribbed glass bowls Isings 3a/b and the ribbed bowls
made of pottery. Both appear around the same time and reach
their peak in the 2nd century AD. As shown by the occurrences
in Germania magna the imitations have the same distribution
area as the originals, respectively no ribbed bowl made of
Fig. 4. This diagram visualises the results of the pXRF-analyses. The grey dots mark the reference group, the black dots
the ribbed bowls made of pottery and the numbers indicate the sherds inventory numbers (graphic: D. Penz; modifications:
J. Rabitsch).
500
Julia rabitscH, cEramics imitating glass?
pottery has been found outside the glass vessels distribution
area. The practice of adapting this form in ceramics not only
appears in Germania, which is known for imitating Roman
vessels made of none-ferrous metals or glass in pottery, but
also in the Roman provinces; especially in the provinces Ger-
mania inferior and Germania superior. Even though we have
much fewer ceramic ribbed bowls in the Roman provinces it
is very likely that they have the same origins and therefore
can be seen as adaptions of the glass vessels.
Secondly the river Rhine appears to be of significant
importance as many finding spots are located in its valley.
In addition to the transfer of goods along this known trading
route some kind of cultural contact and transfer of ideas
seems to have happened since all of the known ribbed bowls
made of pottery are locally produced and were not distributed
over larger distances.
However, it is uncertain if the ribbed bowls made of pot-
tery are kind of a substitute for the glass original or a more or
less independent form. Beyond the Limes it probably was a
substitute, since the reproduction of Roman vessels was less
expensive than buying imported goods and that way a larger
number of customers were reached. The rare occurrence, not
only in the provinces itself but also amongst the findings of
each settlement, indicates a small number of costumers for
these vessels. This fact may refer to the personal preference
of individual people.
Acknowledgments
I would like to thank my colleague Daniel Penz (University
of Freiburg) for carrying out the pXRF analyses and assisting
me with reading and processing the data. Further thanks go
to all the Fautores who provided me with useful hints and
additional comparisons – especially Verena Jauch and Bernd
Liesen.
Julia.Rabitsch@uibk.ac.at
Bibliography:
castElla/mEylan krausE 1994 D. castElla/m.-F. mEylan krausE, La céramique gallo-romaine d’Avenches et de sa région: esquisse
d’une typologie. Bull. Assoc. Pro Aventico 36, 1994.
cüPPErs 1962 H. cüPPErs, Zwei kaiserzeitliche Brandgräberfelder im Kreise Geldern. Bonner Jahrb. 162, 1962,
299–390.
FrEmErsdorF 1964 F. FrEmErsdorF, Neuerwerbungen des Römisch-Germanischen Museums während der Jahre 1923–1927
(Köln 1964).
gEErts 2 014 R. C. A. gEErts, Creativity or statussymbol? Roman ceramic imitations in the Netherlands. Acta RCRF
43, 2014, 771–778.
HaalEbos 1995 J. K. HaalEbos, Castra und Canabae. Ausgrabungen auf dem Hunerberg in Nijmegen 1987–1997.
Libelli Noviomagenses 3 (Nijmegen 1995).
HEgEwiscH 2005a M. HEgEwiscH, Germanische Adaptionen römischer Importgefäße. Ber. RGK 86, 2005, 197−348.
HEgEwiscH 2005b M. HEgEwiscH, Germanische Nachahmungen römischer Metall-, Glas- und Keramikgefäße − eine
kurze Übersicht. Památky Arch. 96, 2005, 79−104.
HEukEmEs 1964 HEukEmEs, B. Römische Keramik aus Heidelberg. Mat. Röm.-Germ. Keramik 8 (Bonn 1964).
JaucH 1997 V. JaucH, Eschenz-Tasgetium. Römische Abwasserkanäle und Latrinen. Arch. Thurgau 5 (Frauenfeld
1997).
mEyEr-FrEulEr 1990 cH.mEyEr-FrEulEr, Gebrauchskeramik. In: W. Drack, Der römische Gutshof bei Seeb, Gem. Winkel.
Ausgrabungen 1958–1969. Ber. Zürcher Denkmalpfl., Arch. Monogr. 8 (Zürich 1990) 143–177.
Vogt 1948 E. Vogt, Der Lindenhof in Zürich. Zwölf Jahrhunderte Stadtgeschichte auf Grund der Ausgrabungen
1937/38 (Zürich 1948).
