Caravan Cities that is, a disposition to take any one item of evidence, especially archaeo-
logical evidence demonstrating the movement of goods, and then reject it
as not proving enough. By rejecting each item separately, Finley did come
rather close to guaranteeing the impregnability of his main hypothesis. For
if no item of concrete evidence were allowed to mean much, noaccumulation
of unexpected or unfamiliar items of empirical evidence could ever occur.
Consequently, one could never be led by the accumulation of empirical evi-
dence into revising one’s initial hypothesis.
In the course of talking about the ‘‘ancient economy,’’ by which he meant
the economic characteristics of the Graeco-Roman world, Finley did not say
a great deal about the eastern trade of the Roman Empire. It is significant
that thePeriplus Maris Erythraeiis never referred to. When trade with Asia
is mentioned, it is in sweeping terms: ‘‘To be meaningful, ‘world market,’
‘a single economic unit’ must embrace something considerably more than
the exchange of some goods over long distances; otherwise China, Indo-
nesia, the Malay Peninsula and India were also part of the same unit and
world market. One must show the existence of interlocking behaviours and
responses over wide areas’’ (). Certainly,ifonly some notion of a ‘‘world
market’’ or ‘‘single economic unit’’ will allow us to attributeanyimportance
to long-distance trade, we shall have to give up at once. For, given the re-
stricted means of transport, and the non-existence of credit-transfer systems,
operating at a distance, without cash, it is inconceivable that long-distance
trade could have had, overall, anything like the importance which would
have produced world-wide rhythms of boom or recession of the sort famil-
iar today. But that is hardly a reasonable requirement. What we need to ask is
two more modest questions. Firstly, did people perceive long-distance trade
with the East as an important phenomenon? Even Finley could not properly
deny that this would be a significant question: for on the very same page he
states that his justification for speaking of ‘‘the ancient economy’’ lies in its
‘‘common cultural-psychological framework’’—or, in other words (to put it
crudely), what people say in those ancient texts which happen to have sur-
vived. Secondly, long-distance trade, whose actual existence cannot possibly
be denied,mighthave been sufficiently important to give a different charac-
ter to certain places: in short to have produced, exceptionally, some ‘‘caravan
cities.’’ Oddly enough, theAncientEconomymakes only one passing allusion to
Palmyra, and then brushes it aside as an exception: ‘‘exceptional cities, such
as...thecaravan-cityofPalmyra’’().Butinfact,aswillbeseen, Palmyra
conformed rather better to his own model than he himself recognised.
If one is going to reject an approach based on the accumulation and analy-
sis of concrete items of evidence, and look instead for a ‘‘common cultural-