trines, and that which produces abstract metaphysical, epistemological, and
ethical conceptions.
The underlying issue is the way in which intellectual life rests on material
and social supports of the surrounding world. The abstract level of philosophy
is determined by the dynamics of inner networks among intellectuals. Here we
have a two-step causality: the external world is indirectly influential insofar as
it supports those organizational structures which allow an intellectual network
to face inward upon its own struggles. This structural autonomy for intellec-
tuals does not always exist. We have seen it build up during the pre-Han
Warring States; we will see it again during the heyday of Buddhist philosophical
networks in the T’ang, and yet again in the neo-Confucian networks of the
Sung. At other times network autonomy is greatly diminished, for example, to
a considerable extent during the Han, and during much of the following period
of political disintegration. During such times intellectual life does not neces-
sarily come to a halt. But its inner dynamics are greatly weakened. The content
of intellectual doctrines is now filled much more directly by the cultural
interests of the social classes and external institutions which support intellec-
tuals.^17
There are, then, two sociological questions about the relationship between
intellectuals and the external world. We can ask about the conditions under
which the internal autonomy of the intellectual network is high, low, or at
points in between. And we can ask about the patterns of intellectual production
when autonomy is low. The latter is the question of the effect of class cultures
on intellectuals, the traditional topic of the sociology of knowledge. I have
argued that class determination is not a very useful theory for dealing with the
highest levels of creativity, the sequences of abstractions produced within the
core of the intellectual community; but class determination is applicable in
periods when structural bases of autonomy are absent. These are typically
periods of intellectual stagnation for an abstract discipline such as philosophy.
What innovation occurs will be at a more concrete and particularistic level,
such as the developments of anthropomorphic religions and magical doctrines.
Let us be clear: periods of low intellectual autonomy can also have moments
of innovation; only they are not driven by the inner network structure of the
field. Left to themselves, intellectuals produce their own factions and alliances.
Their competition over intellectual attention space leaves behind a trail of
abstractions which constitute the inner history of ideas. When intellectual
autonomy is low, this self-propelling dynamic is absent. Instead, new ideas
occur at the moments when the class structure changes, when there are new
external bases for intellectual life—new political conditions fostering religious
movements, new economic and administrative conditions raising or lowering
the salience of court aristocrats, state bureaucrats, or propertied gentry, and164 •^ Intellectual Communities: Asian Paths