data – indeed, Hopkins called his study ‘an experiment’ and his
methods ‘frankly speculative and exploratory’ (1998: 155). Both
hypothesized that on the basis of a Christian population of 1000
inAD40, and with a growth rate of between 3.35% per year
(Hopkins) and 40% per decade (Stark), then the total number of
Christians could have reached 10,000 by the early second century,
200,000 by the early third, 6 million by the early fourth, and over
30 million by the mid-fourth.
Such calculations – like the guesses about Christian
numbers (either in total or as a proportion of the empire’s
population) made by earlier scholars such as Edward Gibbon,
Adolf von Harnack, and Ramsay MacMullen (cf. Stark 1996: 6;
K. Hopkins 1998: 191–2) – are necessarily speculative. Even for
periods for which we have better data, quantitative surveys can
yield only ambiguous results: two recent studies of conversion to
Christianity among the Roman empire’s senatorial aristocracy in
the fourth and fifth centuries, for example, have argued for very
different rates at which religious change took place (Barnes 1995;
Salzman 2002; cf. Humphries forthcoming a). Nevertheless, the
figures suggested by Stark and Hopkins have about them a certain
plausibility. Both suggest that the third century saw a massive
leap in Christian numbers. Other evidence suggests that this
was indeed a period of considerable growth: there was a greatly
increased output of Christian literature; a hierarchy of bishops
emerges much more clearly now than at any earlier time (see
chapter 5); and persecutions by the Roman state became more
frequent and more aggressive (see chapter 6). Certainly, many
more Christian communities can be identified in the sources for
the third century than for the period before it.
The hypothetical increase in Christian numbers suggested
by Stark and Hopkins is, in its own way, as crude an attempt
to explain the expansion of Christianity as the portrait of self-
explanatory success found in ancient Christian writings. There
must have been many variables in the process. Perhaps there were
decreases in Christian numbers at times of persecution. There
were very likely regional differences too. I noted above theCONTEXTS FOR THE ORIGIN OF CHRISTIANITY
112