368" GUNS, GERMS, AND STEEL
families, each consisting of up to several hundred related languages. For example, the Indo-European language family, which includes English as well as French, Russian, Greek, and Hindi, comprises about 144 languages. Quite a few of those families occupy large contiguous areas—in the case of Indo-European, the area encompassing most of Europe east through much of western Asia to India. Linguistic, historical, and archaeological evidence combines to make clear that each of these large, contiguous distributions stems from a historical expansion of an ancestral language, followed by subsequent local linguistic differentiation to form a family of related languages (Table 18.2). Most such expansions appear to be attributable to the advantages that speakers of the ancestral language, belonging to food-producing societies, held over hunter-gatherers. We already discussed such historical expansions in Chapters 16 and 17 for the Sino-Tibetan, Austronesian, and other East Asian language families. Among major expansions of the last millennium are those that carried Indo-European languages from Europe to the Americas and Australia, the Russian language from eastern Europe across Siberia, and Turkish (a language of the Altaic family) from Central Asia westward to Turkey.
With the exception of the Eskimo-Aleut language family of the American Arctic and the Na-Dene language family of Alaska, northwestern Canada, and the U.S. Southwest, the Americas lack examples of large-scale language expansions widely accepted by linguists. Most linguists specializing in Native American languages do not discern large, clear-cut groupings other than Eskimo-Aleut and Na-Dene. At most, they consider the evidence sufficient only to group other Native American languages (variously estimated to number from 600 to 2,000) into a hundred or more language groups or isolated languages. A controversial minority view is that of the linguist Joseph Greenberg, who groups all Native American languages other than Eskimo-Aleut and Na-Dene languages into a single large family, termed Amerind, with about a dozen subfamilies.
Some of Greenberg's subfamilies, and some groupings recognized by more-traditional linguists, may turn out to be legacies of New World population expansions driven in part by food production. These legacies may include the Uto-Aztecan languages of Mesoamerica and the western United States, the Oto-Manguean languages of Mesoamerica, the Natchez-Muskogean languages of the U.S. Southeast, and the Arawak languages of the West Indies. But the difficulties that linguists have in agreeing
HEMISPHERES COLLIDING • 369
table 18.2 Language Expansions in the Old World
Language
Inferred Family Ultimate
Date or Language Expansion Driving Force
6000 or 4000 B.C. Indo-European Ukraine or Ana– food production
tolia –* Europe, or horse-based
C. Asia, India pastoralism
6000 B.C.-2000 b.c. Elamo-Dravid– Iran-*India food production
ian 4000 B.c.-present Sino-Tibetan Tibetan Plateau, food production
N. China-*
S. China,
tropical S.E.
Asia 3000 B.C.-1000 b.c. Austronesian S. China-»Indo– food production
nesia, Pacific
islands
3000 b.c.-a.d. 1000 Bantu Nigeria and food production
Cameroon
-*S. Africa
3000 b.c.-a.d. 1 Austroasiatic S. China-*tropi– food production
cal S.E. Asia,
India
1000 b.c.-a.d. 1500 Tai-Kadai, S. China-*tropi– food production
Miao-Yao cal S.E. Asia
a.d, 892 Hungarian UralMts.-* horse-based
Hungary pastoralism
a.d. 1000-A.D. 1300 Altaic (Mon– Asian steppes-» horse-based
gol, Turkish) Europe, Tur– pastoralism
key, China,
India
a.d. 1480-A.D. 1638 Russian European Russia food production
—» Asiatic Siberia
370 " GUNS, GERMS, AND STEEL
on groupings of Native American languages reflect the difficulties that complex Native American societies themselves faced in expanding within the New World. Had any food-producing Native American peoples succeeded in spreading far with their crops and livestock and rapidly replacing hunter-gatherers over a large area, they would have left legacies of easily recognized language families, as in Eurasia, and the relationships of Native American languages would not be so controversial.
Thus, we have identified three sets of ultimate factors that tipped the advantage to European invaders of the Americas: Eurasia's long head start on human settlement; its more effective food production, resulting from greater availability of domesticable wild plants and especially of animals; and its less formidable geographic and ecological barriers to intracontinen-tal diffusion. A fourth, more speculative ultimate factor is suggested by some puzzling non-inventions in the Americas: the non-inventions of writing and wheels in complex Andean societies, despite a time depth of those societies approximately equal to that of complex Mesoamerican societies that did make those inventions; and wheels' confinement to toys and their eventual disappearance in Mesoamerica, where they could presumably have been useful in human-powered wheelbarrows, as in China. These puzzles remind one of equally puzzling non-inventions, or else disappearances of inventions, in small isolated societies, including Aboriginal Tasmania, Aboriginal Australia, Japan, Polynesian islands, and the American Arctic. Of course, the Americas in aggregate are anything but smalclass="underline" their combined area is fully 76 percent that of Eurasia, and their human population as of a.d. 1492 was probably also a large fraction of Eurasia's. But the Americas, as we have seen, are broken up into "islands" of societies with tenuous connections to each other. Perhaps the histories of Native American wheels and writing exemplify the principles illustrated in a more extreme form by true island societies.
after at least 13,000 years of separate developments, advanced American and Eurasian societies finally collided within the last thousand years. Until then, the sole contacts between human societies of the Old and the New Worlds had involved the hunter-gatherers on opposite sides of the Bering Strait.
There were no Native American attempts to colonize Eurasia, except at the Bering Strait, where a small population of Inuit (Eskimos) derived from
HEMISPHERESCOLLIDING " 3 7 I
Alaska established itself across the strait on the opposite Siberian coast. The first documented Eurasian attempt to colonize the Americas was by the Norse at Arctic and sub-Arctic latitudes (Figure 18.1). Norse from Norway colonized Iceland in a.d. 874, then Norse from Iceland colonized Greenland in a.d. 986, and finally Norse from Greenland repeatedly visited the northeastern coast of North America between about a.d. 1000 and 1350. The sole Norse archaeological site discovered in the Americas is on Newfoundland, possibly the region described as Vinland in Norse sagas, but these also mention landings evidently farther north, on the coasts of Labrador and Baffin Island.
Iceland's climate permitted herding and extremely limited agriculture, and its area was sufficient to support a Norse-derived population that has persisted to this day. But most of Greenland is covered by an ice cap, and even the two most favorable coastal fjords were marginal for Norse food producrion. The Greenland Norse population never exceeded a few thousand. It remained dependent on imports of food and iron from Norway, and of timber from the Labrador coast. Unlike Easter Island and other