PATTERNS OF AMPHORA STAMP DISTRIBUTION 209
make the attribution safe: Aigion, Leucas, Maroneia and Cephallenia, or more
precisely Kranioi (Panagou 2010 : 76–80, 285–9, 467, 468; Sotiriou 2012 : 372).
Leaving these seven cases aside, this chapter will examine the distribution of
amphora stamps from the remaining fifty-one cities. These stamped amphoras
have been dated from the fifth century BCE at the earliest to the first century
BCE at the latest, with the majority dated from the fourth century BCE to the
early first century BCE.
A quick glance at the map in Figure 9.1 reveals a large number of produc-tion centers (see also Empereur and Hesnard 1987 and Garlan 2000 : 50 for
simple lists of cities that applied stamping), which would suggest a very large
total volume of production if one were to assume that all these communities
were operating on the same scale as the better-known centers of Rhodes,
Cnidos or Thasos. All of these centers did not produce at the same level, as is
evident when one compares the huge differences in the numbers of known
stamps for each center (Figure 9.1).
To obtain a more accurate view of the practice of stamping amphoras,one must bear in mind that only a small, but by no means negligible, per-
centage of the more than one thousand poleis existing ‘at any given time in
the Classical period’ (Hansen 2006 : 31. Cf. Hansen and Nielsen 2004 : 53–4)
stamped amphoras. If we add centers from the rest of the ancient world (Black
Sea, Cyprus, Syro-Palestine, Cyrenaica, Sicily, Italy, Massalia) to those shown
on the map, we reach a total of approximately eighty centers.
Methodological Issues in Analyzing the Evidence of
Amphora Stamps
One must take into account several methodological issues when analyzing
amphora stamps as evidence for trading relations between different communi-
ties (Empereur 1982 ; Davies 1984 : 274; Gabrielsen 1997 : 65; Lawall 1998 : 74).
Before examining patterns of amphora stamp distribution, one must set out
the limits of the evidence. First of all, it is inevitable that the archaeological
record will always reflect only a portion of the material remains of the ancient
world. Chance and several other factors determine what stamps are found,
studied and selected for publication (cf. Lund 1999 : 188 for inadequate pub-
lications). The picture of ancient trade we reconstruct on the basis of artifacts
found in excavations is inevitably incomplete. The most important factor to
bear in mind when analyzing trade patterns based on stamp counts is the
fact that stamped amphoras comprise only a small percentage of the over-
all amphora production. Many amphora series were never stamped, and this
means that our sample may give a distorted view of the overall picture (Lawall
1998 ). If we add the communities that produced unstamped amphoras, we
would have a much longer list of cities producing and exporting amphoras. On