|A map of Amazonia with the Xingu river conveniently highlighted. The area we're looking at here is near the headwaters, so at the far south of the purple line. You'll find this map on a lot of sites dealing with Amazonian social complexity, a direct result of Heckenberger's work. h/t Wikipedia, user: kmusser|
|Terra preta, a kind of anthropogenic soil ('anthrosol') characteristic of Arawak settlements and likely part of the reason for Arawak success in populating nutrient-poor Amazonian land. h/t Gerhard Bechtold.|
The Upper Xingu region is an interesting ecological zone, combining many of the features of South American lowlands in a relatively small space. 80% of the land is covered by humid tropical forest, but the region contains plenty of gallery forest, buriti palm swampland, and natural savanna, according to Heckenberger's brief summary of the environment. It is currently equally linguistically diverse, with Arawak, Carib, Tupi-Guaraní, and Gê languages all found alongside one another. This makes it atypical within Amazonia (to some extent), but it makes for an interesting case study in using archaeological, linguistic, and oral historical data to make inferences about the past.
|Buriti palm, Mauritia flexuosa. Note the swampy terrain. h/t Wikipedia, user: Zimbres.|
Arawak and Carib migrations are relatively easy to identify because they are strongly associated with different village forms which stand out archaeologically. Associated with Arawak languages is the construction of circular villages centred on large plazas, with inter-connections between villages obvious from the wide, raised walkways linking them. Arawak languages also clearly correlate with the distribution of Barrancoid pottery, which began its life along the Orinoco, suggesting a northern South American origin of the Arawak family. Clay griddles used for cooking manioc bread are also part of the standard Arawak inventory. (See Comparative Arawakan Histories, Jonathan Hill and Fernando Santos-Granero (eds), for more on Arawak history and prehistory.)
Encircling ramparts and moats are also associated with Arawak villages, indicating that warfare played a role in these developments. Some scholars, including Eduardo Goes Neves, have suggested that the ramparts were built in response to the spread of more warlike Tupi-Guarani from the Atlantic coast into Arawak territory, believed to have begun around 1,500 years ago on the basis of a correlation with a ceramic culture (confusingly also named 'Tupi-Guarani'). Arawak communities have long been noted by ethnographers for their comparative lack of feuding and warfare, for their propensity to avoid inter-communal violence, and for their tendency to form alliances and confederations, in stark contrast to their Tupi-Guarani, Carib, and Gê neighbours. This also makes them, as you might guess, prime candidates for the development of complex chiefdoms.
Why did the Arawaks migrate to the Upper Xingu? Heckenberger points out that it is difficult to reach by river, in contrast to other Arawak-speaking zones, and may have involved overland movement. He hypothesises, in my opinion rightly, that the motive was division among the members of a clan or lineage, leading to some shunned party heading off to find territory elsewhere. This is a plausible motivation to explain prehistoric migrations among several different groups, and as with Austronesian and Niger-Congo movements within the Pacific and Africa respectively it fits well with what we know of Arawak sociology.
|The current distribution of the Arawak languages, where dark blue = southern Arawakan/Maipurean and light blue = northern Arawakan/Maipurean. You'll note that the Upper Xingu is home to some of the easternmost Arawak languages (the four dots inside the blob in south-central Brazil). h/t Wikipedia.|
Around the headwaters of the Xingu, the Carib and Arawak villages can be identified from archaeological remains. Linking archaeological sites with linguistic/ethnic groups is always tricky and attracts controversy, but here the identification of the different villages with Arawaks and Caribs is reasonably obvious. Arawaks are associated with villages of what Heckenberger calls the 'Western Complex' - villages with the attributes listed above, including circular form, a large central plaza, the presence of terra preta, and boulevards linking the villages in seemingly leaderless alliances. Caribs, by contrast, are associated with the villages of the Eastern Complex, an archaeological culture which appears near the Xingu headwaters in post-Columbian times as Caribs journeyed further into Amazonia.
The Eastern Complex villages were organized around one or a few circular malocas, which housed the entire community, reminiscent of the pattern of Guiana Carib villages, whereas the Western Complex villages were characterized by a multitude of dwellings situated across broad areas but gravitating towards large central plazas. The ceramic industry of the Eastern Complex is likewise distinctive from both Western Complex and contemporary Xinguano industry.So here we're looking at a pretty clean distinction between the two groups, at least in the early period of Carib-Arawak interaction (pre-1740-1770) in the Upper Xingu. By all archaeological metrics, Arawaks and Caribs can be identified as separate groups, with Carib villages characterized by multi-family residences and inside public gathering spaces (maloca) and Arawak villages characterized by plazas. Heckenberger gives wonderfully detailed information about the different village complexes of the Eastern and Western Complexes, making inferences from the remains about their social structures and attitudes, and this is easily the best part of the book. It is only lamentable that the formatting of the Kindle edition didn't allow me to make much of the village maps and diagrams.
|A maloca, this one apparently in Colombia. h/t Brodie Ferguson.|
|The modern Kuikuro village of Kuhikugu. The Kuikuro speak a Carib language but have apparently acculturated to Xinguano - i.e., Arawak - life. h/t this blog.|
He nevertheless comes to the same conclusions as everyone else: that where ethnographic, linguistic, oral historical, and archaeological investigations point to shared origins, we can use all of this evidence to interpret prehistoric remains and thereby make sense of human history in the absence of historical documentation. This is relatively uncontroversial and wholly unrelated to the other theoretical aspects of the book.
Much of the body of the text is a worthwhile and interesting exposition of Xinguano archaeology combined with insights from the work of South American linguistics. The only point worth criticizing - assuming you're more sceptical of linguistic-archaeological crossover than I am - is his taking of oral history at face value and assuming that contemporary Kuikuro oral history accurately records the original ethnicity of villages established hundreds of years ago. I think this is excellent evidence for prehistorians to use, but Heckenberger apparently didn't see fit to tell us how Xinguano oral history works or why we should trust it.
Certainly the claim - begun with the Villas-Boas brothers, who did fieldwork in the area decades ago - that oral history in the area is a unifying 'thread of Ariadne', linking the ethnically diverse groups despite their differing origins, is not self-justifying. We need to know more about oral history in the Upper Xingu, beyond its content. Is it transmitted father-to-son? Grandmother-to-granddaughter? Is it framed in any particular narrative structure? Does it rhyme or use similar memory-enhancing devices? Is it the privilege of a group of specialised oral historians or shamans?
His discussion of Carib origins amounts to little more than a single sentence, claiming that Karl von den Steinen had been correct when he wrote, back in 1894, that Caribs probably had a southern Amazonia or even coastal origin. I'm not sure how well this has stood up to inquiries post-2005, but either way, it pales next to Heckenberger's relatively detailed account of Arawak origins.
In summary, this isn't a bad book, but it's certainly not a great one, and it lacks focus. The archaeology is great, and as there are few books on any part of Amazonia delving into so much depth, I can't be against it. But the theoretical background is both too overdone and too underdone, with too little discussion of important chains of argument (i.e., the connection between oral history and archaeology) and too much musing on the back of generic continental archaeological theory. I suppose I have to conclude by saying that this is a curate's egg.