202 THE ROMAN EMPIRE
for ways of gaining access to specifi cally religious roles from which they were generally
excluded by the hold exercised by local elites over civic religion (Gordon 2013). The
point certainly holds for early Christianity too.
The fi ftieth anniversary of the foundation of Vermaseren’s Études préliminaires
was the occasion for an analogous effort at re- thinking the old category (Bricault
and Bonnet 2013), which presents itself at the same time as an attempt to synthesize
the marked fragmentation of current work on religious change in the empire (see
also Rives 2007). The one notable attempt to re- assert the continuing validity of a
simplifi ed version of Cumont’s category has not made much head- way against these
de- constructive moves, and certainly relies far too heavily on the undemonstrable
claim that the cults of Mater Magna, Isis and Mithras are properly described as
‘mystery cults’ (Alvar Ezquerra 2008). This category of ‘mystery’, for decades a
central factor in accounts of the religious changes in the empire, has itself been
largely dismantled (Burkert 1987, Bremmer 2014), in my view correctly – ‘mysteries’,
a vogue term in the second century AD , are after all merely a specialized type of small-
group religion – yet the idea continues to exercise a fascination out of all proportion
to its historical importance. A different kind of challenge to the notion is offered by
an attempt to specify cognitive content, by asking exactly how such cults – in this
case Mithras – communicated their meanings (Beck 2006).
Turning now to the religion of Rome itself, a prosopography of all known priests
and other religious specialists from 300 BC to the synod in St Peter’s organized by
Pope Symmachus in AD 499 has deliberately challenged the traditional assumption
that ‘Roman priesthood’ is synonymous with the membership of the offi cial priestly
colleges – here it includes all known religious specialists, down to Isiac drummers
and Christian deacons (Rüpke 2008a). The interplay between social status,
negotiation with imperial pressures, and exploitation of local patronage are the
themes of a thoughtful account of senatorial religion up to Alexander Severus
(Várhelyi 2010), while the public face of that negotiation has been illustrated in
detail by the defi nitive publication of the Arval Acta (Scheid 1998). The history of
the development of the commented Roman calendar and its exploitation by the need
of successive imperial houses for auto- representation includes an account of the
emergence of historical refl ection upon the specifi c nature of Roman cult (Rüpke
2011b). In keeping with trends noted above, John Scheid has insisted on the complex
interrelation of ‘Roman’ and ‘foreign’ cults (Scheid 2005b).
Perhaps no Roman imperial practice shows more clearly the inadequacy of our
received notion of religion than the institution of the ‘imperial cult’, which needs to
be understood as a Foucaldian dispositif , whose dynamic combination of ‘deep’
discourse and diversifi ed praxis constantly re- composed the subjectively experienced
world (the instrumentalization of the sun offers a fi ne example: Berrens 2004).
Notable here, loosely in the wake of Simon Price (1984), have been studies of the
visual mediation of imperial virtues (Bergmann 1998), of the fatalis princeps ,
destined by the stars (Schmid 2005), of the exploitation of the idea ‘gods are humans
who do not die’ (Clauss 1999), and the commemorative and funerary architecture of
successive emperors in Rome (Davies 2000). In keeping with modern visuality
studies, this latter book deliberately attempts to adopt the viewpoint of an imagined
ancient spectator.
Just as in the case of Classical Greek religion, there have been expressions of
dissatisfaction with the model of ‘civic’ or ‘polis’ religion. Both are imprecise terms
that assume the centrality of the city as the organizational focus of religious practice,