ABSTRACT
The research for this article was stimulated by investigations into the elite settlement form of the early Japanese state (Barnes 1983). There, in Nara, many place and personal names survive in the historical literature that use the word ki, transcribed with characters meaning castle or timber. I hypothesized that these were remnants of a fortified site pattern and wondered if the settlement format had been transferred from Korea. Thus in the summer of 1987, I obtained a Korean Cultural Society Fellowship to study the settlement pattern of the early Korean states to see if a similar pattern was evident there. What I discovered was that there existed a strong tradition of fortified sites using large earthen ramparts or stone walls that are very apparent in the surviving settlement remains of Korea. In contrast, the Japanese sites show no overt evidence of such monumental construction. I thus concluded that moats and palisades would have characterized the early Japanese sites in contrast to the Korean walls and ramparts. This paper represents the first investigation into the functionality of walled sites on the Korean peninsula and consideration of their roles in the protohistoric settlement patterns. It highlights the difficulty of assessing site function without extensive excavation of site interiors; the walls themselves tell us relatively little, but these are the data usually used by historians in discussing the roles of these sites in the political life of the early kingdoms.
