FRONTIER WITH THE HOUSE OF GOLD 27and North-South connections throughout the history of various civi-
lizations. It is precisely because of this rich history of contacts across
millennia that the term "Indian Ocean" obfuscates more than it re-
veals; it is imprecise in its usage and uncertain in its range of usage
across time.
The Indian Ocean arc, thus, needs further delineation. One can
single out some of the various seas that constitute the arc: the Red Sea,
the Arabian Sea, the Persian Gulf, the Bay of Bengal, and the South
China Sea. There are at least four distinct networks of maritime
contact and movement between coastal cities and communities:
between the Gulf and East Africa (Hadramawt to Somalia); the Gulf
and the western shores of India (Muscat to Daybul or Surat or Ca-
licut); the eastern coast of Bengal and the Andaman and Malay islands;
the southern China shore and Southeast Asian archipelago. These net-
works varied a great deal over time in terms of activity and exposure.
They were augmented by smaller maritime trade networks and supple-
mented by land routes such as the Red Sea routes up the Arabian penin-
sula or the Yemen, Sind, and Gujarat routes. Biscussions of the "Indian
Ocean network" in the ancient or the medieval era, the early modern
era or, even the modern eras are really discussions of particular nodes
on particular networks, all with their own contingent histories. For ex-
ample, the two dominant foci of examination in existing scholarship
are the Harrapan and Mesopotamian trade connection and the ancient
Rome-to-India trade route. Both of these would constitute one arterial
network among the many Indian Ocean nodes.^3
Nile Green has argued for carving up another monolithic category,
"Middle East," into a subcategory, "Indian Ocean Arena," by "placing
the Arabian Peninsula, the Gulf, and parts of Iraq into an Indian Ocean
arena that positions these regions into contact with the Horn of Af-
rica, East Africa, India, and even Southeast Asia."^4 This is a sensible
argument that also allows us to address temporal discontinuities in our
archival materials. There is no doubt that the complexity of source ma-
terial (from 'linguistic variation to major lacuna in scholarly coverage),
the diversity of nodes, and the resultant fissures in historiography make
it impossible to paint a comprehensive picture of the "Indian Ocean-
Middle East arena" that makes sense from the fifth millennium BCE
to the modern era. The archaeological (as well as numisfilatic and epi-
graphic) evidence for premedieval periods is sketchy due both to the