Geography Reference
In-Depth Information
demoted or merged, though three-quarters of them were subsequently
restored. By contrast, in the old course region, less than a fifth of counties
were abolished, but almost all of the abolitions persisted.
The course change did influence reform-era administrators consid-
ering spatial renovation. In one case, the Hebei county of Qianning
乾
寧
in the new course region, the historical record explicitly notes that
the jurisdiction was demoted in the early 1070s as a result of continu-
ing devastation in the wake of the Shanghu levee break and restored
in 1088 when refugees returned to their homes and the population of
the former county surpassed ten thousand households.
37
W hile we have
not yet identified other documents explicitly linking changing politi-
cal geography and Yellow R iver catastrophe, it is reasonable to believe
that this case is representative. Flooding disrupted setlements equally
around the old course and the new course; but long-term environmental
degradation affected the old course more severely and probably refugee
decisions as well. In addition, beginning in the 1070s, the government
initiated ambitious and labor-intensive water-control measures in the
new course region; these activities required a more dense state presence
there.
38
This, even more than environmental degradation, may explain
the differential rate of county restoration in the two regions.
This section of the essay is an early-stage experiment in integrating
political geography and demographic data from the DGSD with widely
available environmental information. The preliminary results are prom-
ising in demonstrating that environmental change, though not cata-
strophic crisis, affected the political landscape. However, much more
remains to be done, particularly to incorporate census data. Further
analysis of this phenomenon would also benefit from more mathemati-
cally sophisticated spatial analysis that would allow for the evaluation of
multiple variables simultaneously - including population, proximity to
the international border, and proximity to the old or new course of the
river - rather than considering them sequentially and making inferences
about their relationship. In order to support more complex spatial statis-
tical analysis, the data that we have utilized thus far also need to be inte-
grated with other spatial data, including soil, terrain, flood events, and
transportation routes, and with additional georeferenced demographic
and historical information, including tax rates, tax exemptions, popula-
