How it all began

By William Wetherall

A review of
Irving Rouse
Migrations in Prehistory:
Inferring Population Movement from Cultural Remains

New Haven: Yale University Press, 1986

A version of this article appeared in
Far Eastern Economic Review, 137(37), 10 September 1987, pages 50-51

A somewhat different version appeared as
"Migration or diffusion?" in
Asahi Evening News, 12 June 1987, page 9


Where was the ethnic womb of the Japanese people? Was it Central Asia, the spawning ground of the mounted tribes which began to migrate down the Korean peninsula about 2,000 years ago? Or did the rice-growers and hunters already in Japan acquire the martial trappings of Altaic culture, much like their modern descendants have been "Westernising" their customs, without a population invasion?

Yale University emeritus professor of anthropology Irving Rouse uses prehistoric evidence to surmise the ethnic origins of four peoples: Polynesians, Eskimos, Japanese and Tainos. A specialist in migration archaeology, Rouse believes that careful formation and testing of migration hypotheses is superior to "essentially fanciful postulations" such as the theories of Thor Heyerdahl and Namio Egami.

Heyerdahl has postulated that Easter Island's first settlers came from Peru, but Rouse argues that they were Polynesians. Egami has claimed that Japan was conquered by migrating horse riders from the continent, but Rouse finds no signs of large-scale migration to Japan.

Although Rouse's idiom is technical, his writing is lucid, and any armchair adventurist who likes panoramic prehistoric drama will find his six well-organised chapters quite readable.

Western archaeologists have been influenced by their own colonial migrations to "Africa, India, Australia, Oceania, the Americas, central and northern Asia, the parts of eastern and southeastern Asia . . . to use hypotheses of rapid, long-distance migration to explain the similarities in proto- and prehistoric remains they encountered in different parts of the world."

Asian archaeologists brought up in relatively stable populations have been "more inclined to postulate local development in the case of the Indians and Chinese, and acculturation in the case of the Southeast Asians and Japanese, who lived on the fringes of Indian and Chinese civilisation and were influenced by them."

So when Western archaeologists began to study South and East Asia, they postulated "migration and acculturation from the Western world in order to explain the rise of Indian and Chinese civilisations." This brought them into conflict with local archaeologists, though Western opinion has now "shifted to the side of the local archaeologists as new evidence has made it increasingly clear that Indian and Chinese civilisations are local developments, subject to transculturation from the West but to little, if any, migration or acculturation."

Both foreign and indigenous specialists in Southeast Asian and Japanese prehistory "are also beginning to place more emphasis upon hypotheses of local development and transculturation . . . Southeast Asian and Japanese civilisations are, therefore, syntheses of traits drawn from various sources."


Rouse believes that the Japanese probably originated from the continent because the main islands of Japan were once linked together, and to the continent, by land bridges. Migrating animals and their human hunters walked into Japan until the bridges were submerged by the seas.

Evidence shows that Japan has been occupied by people for at least 30,000 years. Some scholars have suggested 400,000 years, but Rouse thinks that such estimates exceed the available evidence. Earlier this year, a Japanese scholar claimed to possess a section of earth from which some human bones have been found to be between 50,000 and 80,000 years old.

Sometime around 10,000-8,000 BC, projectile points and more efficient axes were locally developed in central Japan and spread northward, but not to Kyushu, where about this time the world's first pottery appeared. But Rouse cautions: "We do not yet know enough about the situation on the mainland to be able to determine whether pottery-making was invented there or in Japan."

During the first millennium BC rice began to be cultivated in parts of northeastern Kyushu closest to Korea. By the 3rd century AD, it had spread down into Okinawa, and up to but not into Hokkaido. Some scholars have postulated a migration of rice cultivators across the Korea and Tsushima straits. But Rouse argues that the evidence points to diffusion between interacting peoples.

Rouse also refutes the horse rider theory of Egami, who once held that northern equestrian tribes came to Japan via northern Kyushu and advanced along the Inland Sea to the Kansai plain. With this theory Egami established the Yamato hegemony which "unified most of Japan."

If the horse rider theory is correct, Rouse argues, then archaeologists should have found remains of their culture, including their burial sites, along the presumed route of migration. But they have not. Instead, throughout the period when the horse riders are supposed to have moved in, "the indigenous complex persists in Japan . . . with accretions appearing so gradually that they are better considered a result of the process of borrowing from Korea that had begun" before the introduction of rice.


But Rouse's understanding of Egami's theory is based on English articles which appeared in the early 1960s. He did not mention Egami as one of the scholars he met during a three-month stay in Japan in 1979, though he did meet Koichi Mori and other leading archaeologists.

In 1982, Egami and Mori published a book called Tairon: Kiba Minzoku Setsu (The Horse Riding Ethnic Group Theory: A Discussion). In this book, Egami presents a considerably modified version of his earlier theory.

Egami now maintains that the bearers of the Altaic horse rider traditions need not have invaded Japan in large numbers, much less as part of a coordinated military assault. His reply to critics, who have wondered how a mounted army c'6uld have crossed the Korean straits in 4th century vessels, is to argue that the number of men, horses, and craft which made the transit were probably very small.

If even 100 polygamous Altaic warriors had stunned local farmers into giving them each five wives, the children, raised by their Yamato mothers, would have spoken Japanese but learned the customs of their warrior fathers. Within three generations there would have been more than 20,000 adults, over half of them men, constituting a completely assimilated elite with more native than foreign blood, but full of Altaic ambition to spread their rule.

Egami's modified theory does not appear to contradict Rouse's hypothesis of transculturation. Instead, it attempts to explain how the horse rider culture could have entered Japan, taken firm root, and spread throughout the islands.