breaks down if there is no grateful populace to admire and applaud the phi-
lanthropy.
Only once did the system require the active commitment of its users.
The edict of the emperor Decius in 249 CEcommanded that all inhabitants
of the empire, presumably with the exception of the Jews, not only sacri-
fice but also acquire statements (libelli) certifying that they had done so.
Usually, a negative construction is placed on this requirement: it was a
police measure designed to identify Christians and by the test itself to get
them to desist or face the consequences. Recently, however, James B. Rives
(1995, 258–61; 1999) has forcefully reopened the alternative case that it
was a genuine attempt to institute a common empire-wide cult of the gods,
not by posting a slate of official deities to be worshipped (a hopeless task)
but by making the traditional mode of accessing them, i.e., sacrifice, a uni-
versal participatory requirement, thus a “compensator” in the full Starkian
sense. It was not sufficient just to opt out of Christianity; one must opt into
paganism or, rather, into religion as a contemporary would have seen it.
Whatever its intent, the experiment did not work and soon lapsed. Never-
theless, as Rives argues, it was the shape of things to come—in the Chris-
tian empire.
My final demonstration of the ill fit of Stark’s model of religion with
public paganism concerns, again, the issue of class. Class distinction, as we
have seen, was paramount in the workings of public paganism: the (minute)
upper class monopolized the priesthoods and paid the shot; the (huge)
lower class reciprocated with honour to the gods—and to the elite. Class is
also crucial to Stark’s explication of the rise of Christianity; he devotes his
second chapter to this theme.
For Stark, early Christianity is a good proving ground for the proposi-
tion that cults, in the modern sociological sense of new religions, draw
their membership from higher social levels than do sects or reform move-
ments within existing religions. These correlations work admirably in mod-
ern society, and Stark (1997, 30–31) uses the recent tendency to locate early
Christianity in the higher, if not tip-top, reaches of society to demonstrate
that the correlation works also in antiquity, since Christianity was a new
religion (i.e., a cult) and not a reform movement (i.e., a sect).
Stark’s ultimate purpose in chapter 2, made explicit in its conclusion
(1997, 45–47), is to make a point about the generalizability of theory,
whether it be the laws of gravity (his analogy) or sociological principles con-
cerning religious phenomena. One does not really need to demonstrate
empirically the class composition of early Christianity; one can infer it from
the principle that “cult movements overrecruit persons of more privileged
The Religious Market of the Roman Empire 249