I have recently begun reading David Grann's The Lost City of Z, a popular and critically-acclaimed book about the English explorer Percy Harrison Fawcett, who went missing in the Mato Grosso*, Brazil, in 1925. Fawcett was on an expedition to locate 'Z', an alleged pre-Columbian city in the jungle. The expedition was famous, Fawcett became a celebrity, and his disappearance spurred a number of people to go looking for him, all of whom disappeared, died, or otherwise failed. The number dead as a result of Fawcett's disappearance may be as high as a hundred.
Grann suggests that the search for Z, sometimes conflated with 'El Dorado', is ultimately about disproving 'environmental determinism', the idea that the development of civilisation is determined by environmental conditions. The idea that the Amazon basin, which has some of the poorest soils in the world, could possibly support societies more complex than confederations of agricultural villages has always seemed a bit of an impossibility, and so finding towns or cities, or public architecture, would appear to invalidate the view that environment determines complex social structure. That's a straw man of a position - there are few academics who believe that the environment is the sole determining factor in developing complex society - but clearly environment is a factor, and an important one.**
It is often said that there are no laws in the human sciences (besides the laws of physics, that is). There is no necessary (ie, systematic, law-like) correlation between, say, feudalism and slavery. Humans don't work like that; they are susceptible to such a wide range of influences and variables that their actions do not correspond to formal laws identifiable cross-culturally.
This is true; there are no such laws. But there are still some absolutes that limit humans, and it would be absurd to believe otherwise. If you do not get enough water, you will die. If you do not have enough energy in the form of edible nutrients, you will die. If you do not have oxygen to breathe, you will die. And the same applies to all humans who have ever lived, as far as can be told. People need water, sufficient food, and oxygen or they can't do anything at all, let alone create cities and temples and who knows what. These things can hardly be disagreed with, unless you accept the ridiculous assertions of New Agers who claim to live on 'prana' or 'qi' (and if you've gone down that route, there's no point in arguing with you as you've already left the reasonable world behind).
This has obvious implications for the development of complex society. For instance, it isn't possible to have grand public buildings without a large population to build them (in absence of modern mechanical earth-movers, that is), and a large population requires food and water, or else they will die. It isn't possible to have a specialisation of labour without there being sufficient food produced by the specialist agriculturalists/pastoralists/whatever else to allow for other people to make goldsmithing or basketry their professions. In order to have an all-conquering army, the army must be fed and watered, requiring either plundering of the places conquered (assuming they have sufficient food) or a surplus at home.
It's not just that people need to eat to live, either. People prefer to eat; they like eating. They like being in situations where they don't have to worry about whether they will get to eat, or whether the water they drink will give them dysentery. Given a choice between a complex society that can provide a large population of people with a small amount of food (like a single manioc cake per diem) and a tribal society with a surplus of different foods fit for a smaller population, most people will choose the latter, all things being equal. (If all things are not equal - if the person has another reason for associating with the resource-poor complex society, like belief in the power of its cult or the charisma of its leaders, or a belief in the overall value of states, laws, and disinterested leadership - then that can change things, again attesting to the fact that such in-built preferences do not produce law-like behaviour in humans.)
In order for there to be any of the features associated with complex society, there has to be a relatively large population, and in order for there to be a large population, there has to be sufficient food and water for them. This is why agriculture seems to correlate with the development of a more complex society, as it allows for population growth. It's not inevitable; complex societies appear to have developed along the west coast of South America in the absence of agriculture using a surplus generated primarily by fishing, and also possibly by trade with foraging (and maybe agricultural) groups from the highlands.*** Even though agriculture allows for a larger population, it isn't the only thing that does, and so there is no law correlating agriculture and complex society - only a generalisable principle that usually holds true but isn't an inevitability.
Having sufficient food and water to support a large population likewise fails to guarantee its development. New Guinea's soils are very fertile, and combined with the horticultural practices of the inhabitants, sufficient food is produced to fuel the development of complex societies - which until recently did not exist on the island. It is also possible for people to artificially limit, through social mechanisms, the number of babies produced by the society. Modern-day China is an example of this, with limits placed by law on the fecundity of the population, but Japan and much of Europe show a similar trend despite the lack of a legal prohibition.
Examples from non-industrialised societies can be found. Polyandrous groups (groups in which a single woman can marry several men, often brothers) show a lower birth rate and population density than would be supported by their food gathering practices, for instance. They also typically demonstrate a higher quality of life than their non-polyandrous neighbours, albeit a generally less complex society in most respects (again, not necessarily; it is not a law). It also possible for humans to develop resource-gathering practices that produce a deliberately smaller surplus, as with the raising of pigs in pre-contact Hawai'i, something that did not produce more food for the population (the pigs were raised for ceremonial purposes).
So if we can establish that some parts of the Amazon basin could support a large population, we cannot necessarily infer that complex societies - what we might be tempted to call 'civilisations' - developed there. We can ask, however, whether complex Amazonian societies are a viable possibility. Are they?
It turns out that they are.
The soil of most of lowland South America is terrible for growing crops: it is nutrient-poor and disappears within a generation of the destruction of the forest it supports, leaving bare rock behind it. The rivers are full of edible fish, and there is plenty of ethnographic support for the preferential placement of Amazonian villages on their banks. But rivers and thin soil aren't sufficient to support a large and organised population. This appears to have been realised by the native peoples of the lowlands, who developed their own soil. This soil, named terra preta in Portuguese ('dark earth'), is incredibly rich. It was developed by successive generations of lowland farmers beginning around the start of the first millennium BCE (c. 900), who used charcoal, manure, and bones to create a dark, fertile mush that could be added to enrich or replace the impoverished forest soil.
It's kind of a Wonder Soil, transforming the landscape and capabilities of lowland South America. Wherever terra preta is found, there is the potential for a large population and significant complexity.
Now we can ask whether there is any empirical evidence for complex society in the area, besides that indicating the viability of it. There is some. Some of the old ideas behind Z and other Amazonian cities are known to be untrue. 'El Dorado' seems to have existed, but not in the
Amazon and not in the way formerly believed, and it can be discounted. It was likely the name given to
certain traditions of the Muisca people, who lived in what is now
Colombia from c. 800 CE. Some early Spanish accounts, on the other hand, including (if I remember correctly) the account written by Gaspar de Carvajal of Francisco de Orellana's 1542 adventure down the Amazon, mention towns and confederations on the riverbank. But besides these written items, is there any more evidence for Amazonian civilisation?
The Spanish accounts were earlier dismissed as unrepresentative of the basin, and the orthodox view among anthropologists was that there was a division between the fertile floodplains (the varzea) and the rest of the land, the unflooded interfluvial hinterland, assumed to be poor and not at all fecund. Anthropologists claimed that only the varzea could support the large villages seen by Orellanna et al. It has been accepted for a long time that there were some relatively complex societies in the Amazon, but they were believed to have been restricted to flood-dependant varzea (see Donald Lathrap's The Upper Amazon, 1970). That was the view until quite recently. This view has been called into question, however.
The stereotype of the Amazonian landscape is that of the 'Green Hell', a pristine environment in which humans without history ('Noble Savage' Bororo and Kalapalo and whomever else) live in some sort of harmony with nature. This is far from the truth. In reality, the history of these tribes is trackable through linguistics, genetics, and archaeology, however poorly these may now be known, and they are not generally egalitarian foragers with no sense of the past (although there are exceptions, including the now-famous Pirahã people). They are humans whose presence has been felt in many ways in the Amazonian environment, and they and their ancestors are not of a species different to those of us who drive cars and drink Coke.
Brazil is the most bio-diverse nation on earth (the second, in case you were wondering, is Indonesia), but this is in spite of considerable human intervention and stewardship, not because humans lived in perfect harmony with nature (a very spurious notion indeed). Studies by anthropologists and ecologists, like Michael Heckenberger of the University of Florida, Gainesville, have shown that much of the Amazonian environment is anthropogenic (human-created). It is not pristine, and it likely supported some large settlements with perhaps as many as 4-5,000 inhabitants, especially along the upper Xingu (a southeastern tributary of the Amazon). That isn't to say that ancient Amazonians created cities; Heckenberger's theory is that large villages of around 2,000 people or so would be linked to one another, and to other smaller settlements, using straight cleared roads rather than all of the people of the community living in a single conurbation. 'Z' would not have been much like Fawcett's vision.
Manioc agriculture, in particular that practiced by speakers
of Arawakan languages from pre-Columbian times to today, could have
supported a much larger population in the past, the kind of population needed to establish the large villages and inter-village roadways. Other evidence includes the presence of earthworks in the hinterland far from the varzea, indicating the presence of large sedentary populations across the basin from Bolivia to Ecuador and Brazil. Some of these earthworks were initially identified using, yep, Google Earth!
An overview of these developments in the evidence for complex Amazonian societies may be found here, in an article in the journal Antiquity on earthworks along the Purus river (a southwestern tributary of the Amazon). Similar overviews can be found in various parts of Heckenberger's oeuvre, and an earlier overview is also present in Karen Olsen Bruhns' Ancient South America (1994), which I have mentioned previously. Charles Mann's popular 1491: New Revelations of the Americas Before Columbus (2006) uses Heckenberger's data to produce an account of pre-Columbian South America emphasising the likelihood of developed civilisation in the Amazon basin, although new data, including the paper on the Purus river, have come in since the book's publication. A shorter but more recent look at South American archaeology can be found in Chris Scarre (ed.), The Human Past (Thames & Hudson, second edition 2009). The section on South America was a collaboration between Heckenberger and Michael Moseley, an archaeologist of the Andes. The Handbook of South American Archaeology (2008) also has plenty of information on Amazonia, but I don't own a copy and have only gone through small sections of it.
Interestingly, Claude Levi-Strauss had also theorised that lowland South America must have been the site of considerable social complexity, and it now appears as if this idea (made by Levi-Strauss on what I consider to be quite spurious grounds) may be true. As with much of American archaeology, studies of this region progress incredibly quickly, and it seems as if any overview written now will be invalidated within a few years. It's all rather wonderful, and while I have to stress that I am not an expert on this region, I'm following it as closely as I am able. My amazon wishlist is currently packed with books on the Amazon, in fact, and I'll get through some of them in the next few months.
The question of where these large Amazonian populations went is relatively easy to answer, even in the absence of an abundance of archaeological evidence. Eurasian people and diseases killed them off. This process continued in earnest until recently. Perhaps the best twentieth century documentation of it is Norman Lewis' essay Genocide in Brazil. It is shocking.
And, so that I don't end on such a terrible and disturbing note, I'd recommend David Grann's Lost City of Z, which will satisfy you if you're looking for a brilliantly-written account of a remarkable set of true adventures.
* Mato Grosso means 'thick scrubland/bush' in Portuguese, which should give some idea of the terrain.
** I believe in determinism, in the sense that I do not believe in free will; your will is not free to choose any option, and you are not free to choose to do something that you do not have a reason to do. But it is obvious that the number of reasons, and the variables influencing them, cannot be counted; people have the capacity to be inspired to action by their beliefs about a simply enormous number of different things. Environment is merely one.
***This is a little controversial, or it was when it was first discovered. But it seems to be true. The discussion of west coast South American towns c. 3500 BCE has been complicated by discoveries in Ecuador which suggest an early date for the introduction of ceramics, lacking at the Peruvian sites, but the Peruvian towns centered on river valleys flowing out into the Pacific, like the Supe, do appear to have been pre-ceramic and non-agricultural.