The Dawn of Everything: A New History of Humanity

by David Graeber

ancient ‘arts of memory’, where those trying to memorize stories, speeches, lists or similar material would each have a familiar ‘memory palace’. This consisted of a mental pathway or room in which a series of striking images could be arranged, each a cue to a particular episode, incident or name. One can only imagine what might happen if someone were to draw or carve one such set of visual cues, and a later archaeologist or art historian were to discover it, with no idea of the context, let alone what the story being memorized was actually about.

In fact, nothing in Chavín’s monumental landscape really seems concerned with secular government at all. There are no obvious military fortifications or administrative quarters. Almost everything, on the other hand, seems to have something to do with ritual performance and the revelation or concealment of esoteric knowledge.54 Intriguingly, this is exactly what indigenous informants were still telling Spanish soldiers and chroniclers who arrived at the site in the seventeenth century.

Despite initial scepticism, archaeologists have gradually come round to accepting that they were right.55

Chavín – a remote precursor to the Inca – was an ‘empire’, it was one built on images linked to esoteric knowledge. Olmec was, on the other hand, an ‘empire’ built on spectacle, competition and the personal attributes of political leaders. Clearly, our use of the term ‘empire’ here is about as loose as it could possibly be.

What makes more sense, we suggest, is to look at these otherwise puzzling cases through the lens of our three elementary principles of domination – control of violence (or sovereignty), control of knowledge, and charismatic politics

neither case gives us reason to think anyone was asserting a strong principle of sovereignty.

The obvious next question, then, is whether examples of the third possible variant can also be found: i.e. cases of societies which develop a principle of sovereignty (that is, grant an individual or small group a monopoly on the right to use violence with impunity), and take it to extreme lengths, without either an apparatus for controlling knowledge or any sort of competitive political field.

By general consent, the Natchez (who called themselves Théoloël, or ‘People of the Sun’) represent the only undisputed case of divine kingship north of the Rio Grande. Their ruler enjoyed an absolute power of command that would have satisfied a Sapa Inca or Egyptian pharaoh; but they had a minimal bureaucracy, and nothing like a competitive political field. As far as we know it has never occurred to anyone to refer to this arrangement as a ‘state’.

The Great Sun was a sovereign in the classical sense of the term, which is to say he embodied a principle that was seen as higher than law. Therefore no law applied to him. This is a very common bit of cosmological reasoning that we find, in some form or another, almost anywhere from Bologna to Mbanza Congo. Just as gods (or God) are not seen as bound by morality – since only a principle existing beyond good and evil could have created good and evil to begin with – so ‘divine kings’ cannot be judged in human terms; behaving in arbitrarily violent ways to anyone around them is itself proof of their transcendent status. Yet at the same time, they are expected to be creators and enforcers of systems of justice.

The problem with this sort of power (at least, from the sovereign’s vantage point) is that it tends to be intensely personal. It is almost impossible to delegate.

Even if one does develop an administrative apparatus (as they of course did), there is the additional problem of how to get the administrators actually to do what they’re told – and, by the same token, how to get anyone to tell you if they aren’t.

Sovereignty always represents itself as a symbolic break with the moral order; this is why kings so often commit some kind of outrage to establish themselves, massacring their brothers, marrying their sisters, desecrating the bones of their ancestors or, in some documented cases, literally standing outside their palace and gunning down random passers-by.62 Yet that very act establishes the king as potential lawmaker and high tribunal, in much the same way that ‘High Gods’ are so often represented as both throwing random bolts of lightning, and standing in judgment over the moral acts of human beings.

insofar as such a figure does manage to establish themselves as genuinely standing above the law – in other words, as sacred or set apart – another apparently

universal principle kicks in: in order to keep him apart from the muck and mire of ordinary human life, that same figure becomes surrounded with restrictions. Violent men generally insist on tokens of respect, but tokens of respect taken to the cosmological level – ‘not to touch the earth’, ‘not to see the sun’ – tend to become severe limits on one’s freedom to act, violently or indeed in most other ways.63 For most of history, this was the internal dynamic of sovereignty.

Lastly, then, we should consider the impact of such first-order regimes on our third basic form of freedom: the freedom to shift and renegotiate social relations, either seasonally or permanently. This is, of course, the hardest to assess. Certainly, most of these new forms of power had a decidedly seasonal element. During certain times of year, as with the makers of Stonehenge, the entire social apparatus of authority would dissolve away and effectively cease to exist. What seems most difficult to comprehend is how these strikingly new institutional arrangements, and the physical infrastructure that sustained them, came into being in the first place.

When sovereignty first expands to become the general organizing principle of a society, it is by turning violence into kinship. The early, spectacular phase of mass killing in both China and Egypt, whatever else it may be doing, appears to be intended to lay the foundations of what Max Weber referred to as a ‘patrimonial system’: that is, one in which all the kings’ subjects are imagined as members of the royal household, at least to the degree that they are all working to care for the king. Turning erstwhile strangers into part of the royal household, or denying them their own ancestors, are thereby ultimately two sides of the same coin.78 Or to put things another way, a ritual designed to produce kinship becomes a method of producing kingship.

In the case of Egypt, it seems, ‘state formation’ began with some kind of Natchez or Shilluk-like principle of individual sovereignty, bursting out of its ritual cages precisely through the vehicle of the sovereign’s demise in such a way that royal death ultimately became the basis for reorganizing much of human life along the length of the Nile.

As we’ve seen, modern Nilotic peoples, and particularly the Shilluk, show how relatively mobile societies that place great value on individual freedom might, nonetheless, prefer an arbitrary despot – who could eventually be got rid of – to any more systematic or pervasive form of rule. This is especially true if, like so many peoples whose ancestors organized their lives around livestock, they tend toward patriarchal forms of organization.

What preceded the First Dynasty, then, was not so much a lack of sovereign power as a superfluity of it: a surfeit of tiny kingdoms and miniature courts, always with a core of blood relatives and a motley collection of henchmen, wives, servants and assorted hangers-on. Some of these courts appear to have been quite magnificent in their own way,

These kings give every sign of making grandiose, absolute, cosmological claims; but little sign of maintaining administrative or military control over their respective territories.

How do we get from here to the massive agrarian bureaucracy of later, dynastic times in Egypt?

a kind of extended argument or debate about the responsibilities of the living to the dead. Do dead kings, like live ones, still need us to take care of them? Is this care different from the care accorded ordinary ancestors? Do ancestors get hungry? And if so, what exactly do they eat?

It is no coincidence that arable wheat-farming – though long familiar in the valley and delta of the Nile – was only refined and intensified around this time,

The two processes – agronomic and ceremonial – were mutually reinforcing, and the social effects epochal. In effect, they led to the creation of what might be considered the world’s first peasantry. As in so many parts of the world initially favoured by Neolithic populations, the periodic flooding of the Nile had at first made permanent division of lands difficult; quite likely, it was not ecological circumstances but the social requirement to provide bread and beer on ceremonial occasions that allowed such divisions to become entrenched. This was not just a matter of access to sufficient quantities of arable land, but also the means to maintain ploughs and oxen – another introduction of the late fourth millennium BC. Families who found themselves unable to command such resources had to obtain beer and loaves elsewhere, creating networks of obligation and debt. Hence important class distinctions and dependencies did, in fact, begin to emerge,88 as a sizeable sector of Egypt’s population found itself deprived of the means to care independently for ancestors.

seasonal work units (or phyles, as Egyptologists call them) seem to have been made up only of men who passed through special age-grade rituals, and who modelled themselves on the organization of a boat’s crew.

interesting parallels to explore here with what happened in the Industrial Revolution, when techniques of discipline, transforming crews of people into clock-like machines, were first pioneered on sailing ships and only later transferred to the factory floor. Were ancient Egyptian boat crews the model for what have been called the world’s first production-line techniques, creating vast monuments, far more impressive than anything the world had yet seen, by dividing tasks into an endless variety of simple, mechanical components: cutting, dragging, hoisting, polishing? This is how the pyramids were actually built: by rendering subjects into great social machines, afterwards celebrated by mass conviviality.95

Perhaps this is what a state actually is: a combination of exceptional violence and the creation of a complex social machine, all ostensibly devoted to acts of care and devotion.

today refer to as ‘states’ really do have any common features, one must certainly be a tendency to displace this caring impulse on to abstractions; today this is usually ‘the nation’, however broadly or narrowly defined.

most of human activity was directed upwards, either towards tending rulers (living and dead) or assisting them with their own task of feeding and caring for the gods.97 All this activity was seen as generating a downward flow of divine blessings and protection, which occasionally took material form in the great feasts of the workers’ towns.

The problems come when we try to take this paradigm and apply it almost anywhere else.

other cases referred to in the literature as ‘early states’ were entirely different.

The kingdom of Egypt and the Inca Empire demonstrate what can happen when the principle of sovereignty arms itself with a bureaucracy and manages to extend itself uniformly across a territory. As a result, they are very often invoked as primordial examples of state formation, even though they are dramatically separated in time and space. Almost none of the other canonical ‘early states’ appear to have taken this approach.

Early Dynastic Mesopotamia,

The cities they ostensibly ruled over had been around for centuries: commercial hubs with strong traditions of self-governance, each with its own city gods who presided over local systems of temple administration. Kings, in this case, almost never claimed to be gods themselves, but rather the gods’ vicegerents, and sometimes heroic defenders on earth: in short, delegates of sovereign power that resided properly in heaven.

The Maya lowlands were different again.

Classic-period rulers lacked a sophisticated administrative apparatus, but they imagined the cosmos as itself a kind of administrative hierarchy, governed by predictable laws:

So do these ‘early states’ have any common features at all? Obviously, some basic generalizations can be made. All deployed spectacular violence at the pinnacle of the system; all ultimately depended on and to some degree mimicked the patriarchal organization of households. In every case, the apparatus of government stood on top of some kind of division of society into classes. But as we’ve seen in earlier chapters, these elements could just as well exist without or prior to the creation of central government – and even when such government was established, they could take very different forms.

Looking at China only seems to complicate things even further.

But in most important ways, we find little similarity between the Shang and either Old Kingdom Egypt or Inca Peru. For one thing, Shang rulers did not claim sovereignty over an extended area.

At this point it should be clear that what we are really talking about, in all these cases, is not the ‘birth of the state’ in the sense of the emergence, in embryonic form, of a new and unprecedented institution that would grow and evolve into modern forms of government. We are speaking instead of broad regional systems;

In terms of the specific theory we’ve been developing here, where the three elementary forms of domination – control of violence, control of knowledge, and charismatic power – can each crystallize into its own institutional form (sovereignty, administration and heroic politics), almost all these ‘early states’ could be more accurately described as ‘second-order’ regimes of domination. First-order regimes like the Olmec, Chavín or Natchez each developed only one part of the triad. But in the typically far more violent arrangements of second-order regimes, two of the three principles of domination were brought together in some spectacular, unprecedented way. Which two it was seems to have varied from case to case. Egypt’s early rulers combined sovereignty and administration; Mesopotamian kings mixed administration and heroic politics; Classic Maya ajaws fused heroic politics with sovereignty.

the third form of domination was largely pushed out of the realm of human affairs altogether and displaced on to the non-human cosmos (as with divine sovereignty in Early Dynastic Mesopotamia, or the cosmic bureaucracy of the Classic Maya). Keeping all this in mind, let’s return briefly to Egypt to clarify some remaining points.

Egyptian kingship was, however, Janus-faced. Its inner visage was that of supreme patriarch, standing guard over a vastly extended family – a Great House (the literal meaning of ‘pharaoh’). Its outer face is shown in depictions of the king as a war leader or hunt leader asserting control over the country’s wild frontiers; all were fair game when the king turned his violence upon them.111 This is very different, however, to heroic violence. In a way, it’s the opposite. In a heroic order, the warrior’s honour is based on the fact that he might lose; his reputation means so much to him that he is willing to stake his life, dignity and freedom to defend it. Egyptian rulers, in these early periods, never represent themselves as heroic figures in this sense. They could not, conceivably, lose. As a result, wars are not represented as ‘political’ contests, which imply a match between potential equals. Instead, combat and the chase alike were assertions of ownership, endless rehearsals of the same sovereignty the king exercised over his people and which ultimately derived from his kinship with the gods.

Here, too, all state officials had to be in some sense appendages of the king’s own person.

Everything remained confined to life at court. This is abundantly clear in the ‘tomb biographies’ of Old Kingdom officials, which describe their life achievements almost exclusively in terms of their relationship to and their care for the king, rather than any personal qualities or attainments.112

Dramatic public contests of any sort, political or otherwise, were well-nigh non-existent. There is nothing in the official sources of the Egyptian Old Kingdom (nor much in later periods of ancient Egyptian history) that is remotely reminiscent of, say, Roman chariot-racing or Olmec or Zapotec ball games.

Dead kings, perhaps, compete with one another; but by the time sovereignty comes down to the domain of mortals, matters have already been settled.

if we are trying to understand the appeal of monarchy as a form of government – and it cannot be denied that for much of recorded human history it was a very popular one – then likely it has something to do with its ability to mobilize sentiments of a caring nature and abject terror at the same time.

It is also worth observing that monarchy is probably the only prominent system of government we know of in which children are crucial players, since everything depends on the monarch’s ability to continue the dynastic line. The dead can be worshipped under any regime – even the United States, which frames itself as a beacon of democracy, creates temples to its Founding Fathers and carves portraits of dead presidents into the sides of mountains – but infants, pure objects of love and nurture, are only politically important in kingdoms and empires.

whenever state sovereignty broke down, heroic politics returned – with charismatic figures just as vainglorious and competitive, perhaps, as those we know from ancient epics, but far less bloodthirsty.