Ancient Iraq (5 page)

The Buried Cities of Iraq

To most tourists, the first contact with the ancient sites of Iraq comes as a surprise. They are taken to a hillock rising above the plain and they are told that this was once an ancient city. As they go nearer they may find such splendid monuments as the stage-tower of Ur or the Ishtar Gate of Babylon, but in most cases they are confronted with unsightly bits of brickwork and heaps of earth littered with broken pottery. Quite naturally they are puzzled and wonder how this happened.

To answer this question it should be first explained that these ancient towns were built of nothing but mud. Stone is rare in Iraq, whereas clay is everywhere at hand. In very early times houses were made of piled-up mud (
pisé
) or of shapeless lumps of clay pressed together (
adobe
), but as early as in the ninth millennium
B.C.
it was soon found preferable to mix clay with straw, gravel or potsherds, mould it into bricks, let these bricks dry in the sun and bind them together with a gypsum mortar. In that way, thicker, stronger and more regular walls could be built. Of course, kiln-baked bricks were much more resistant and durable, especially when they were jointed with bitumen, but this was a costly material, as wood fuel was rare and bitumen had often to be shipped from comparatively distant regions. Burnt bricks therefore were in general reserved for the houses of gods and kings, though this was by no means the rule,
²
and the vast majority of ancient Mesopotamian buildings were of simple mud bricks. The roofs were made of earth spread over a structure of reed mats and tree-trunks, and the floors of beaten earth sometimes with a coating of gypsum. A coat of mud plaster was also usually applied to the walls.

These houses with their thick walls were relatively comfortable, being cool in summer and warm in winter, but they required constant attention. Every summer it was necessary to put a new layer of clay on the roof in anticipation of the winter rains, and every now and then the floors had to be raised. The reason for this was that rubbish in antiquity was not collected for disposal but simply thrown into the street, so that the street level gradually rose higher than the floor level of the houses that bordered it, allowing the rain and the filth to seep in. Earth was therefore brought into the rooms, rammed over the old floors and covered with another coat of plaster. It is not infrequent for archaeologists to find two, three or more superimposed floors in one house. Provided these things were done, mud-brick buildings could last for a great many years. But then one day something happened. Whether it was war, fire, epidemic, earthquake, flood or change in river course, the result was the same: the town was partly or totally deserted. The roofs left unattended collapsed and the walls, now exposed to weather on both faces, crumbled down, filling up the rooms and sealing off the objects left behind by the householders. In the case of
war, the destruction was of course immediate, the victorious enemy usually setting fire to the city. These arsonists of yore unknowingly made modern ‘cuneiformists’ happy, since many sun-dried and therefore fragile tablets were baked by the fire and became almost indestructible.

After years or even centuries of abandonment, new settlers would perhaps reoccupy the site, attracted by such things as its strategically or commercially advantageous position, the abundance of its water supplies or, possibly, a lingering devotion to the god under whose aegis it had been built. Since they had no means of removing the enormous mass of debris, they levelled off the ruined walls and used them as foundations for their own building. This process was repeated several times in the course of years, and as ‘occupation levels’ succeeded one another the city gradually rose above the surrounding plain. Some sites, it is true, were abandoned early and for ever; others, like Erbil and Kirkuk, have been more or less continuously occupied from very ancient times until now; but the vast majority of them, after centuries or millennia of occupation, were deserted at some period or another of the long history of Iraq. It is not difficult to imagine what took place then: windborne sand and earth piled up against the remaining walls and filled in the streets and every hollow, while rainwater smoothed off the surface of the heaped-up ruins, spreading debris over a large area. Slowly but inexorably, the town took its present shape: that of a rounded, more or less regular ruin-mound or, as the Arabs say, using an old, pre-Islamic word, a ‘tell’.
³

The task of archaeologists is to dissect that closely woven fabric of standing or fallen walls and foundations, rubble, floors and earth-filling, to recover the plan of buildings, to collect and preserve the objects they may contain and to identify and date the successive ‘levels’ which constitute the tell. Depending upon the time and funds at their disposal, they use one of several methods.
⁴

The quickest and cheapest way of knowing roughly what is in a tell is to carry out a ‘sounding’. Several trenches are dug
into the surface of the mound at various angles. As the trenches are deepened, objects such as pottery are collected for dating purposes and a record is made of the floors and segments of walls encountered. This method is obviously imperfect and should only be used for preliminary surveys or for comparatively unimportant sites. A variety of sounding often applied to high and narrow tells consists of cutting a long trench, not on the surface but on the side of the mound from summit to base, just as one cuts into a Christmas pudding. An impressive series of occupation levels can be detected in this way, though it is practically impossible to circumscribe any building.

Another method, in theory perfect, is to divide the surface of the site into squares, dig up each square in turn until a certain depth is reached and start all over again for the second horizontal ‘slice’. The objects found in each square and in each layer are carefully numbered and plotted on maps. As the work goes on, monuments gradually take shape. This very slow and expensive method is rarely used. As a rule, archaeologists prefer what may be called ‘extended sounding’. A certain area is carefully selected on the surface of the tell and a trench dug, but as soon as walls are encountered, they are followed and denuded on both faces until the whole building is unearthed. Several areas are treated in the same way and may or may not join together. Whenever desirable, digging is pushed in depth underneath the uppermost and consequently more recent buildings, which are destroyed in order to bring older buildings to light. In one or more points a shaft or ‘test-pit’ may be sunk down to the virgin soil, giving a cross-section of the mound, a summary as it were of its various occupation levels. Some parts of the site remain, of necessity, untouched, but this is of little importance if the main monuments such as temples and palaces and a selection of private houses have been unearthed. Nimrud, Babylon, Uruk, Ur, Nippur and all the main sites of Iraq were or still are excavated by this method with, in the main, highly satisfactory results.

Dating the Past

Dating the monuments and objects discovered can be very easy or very difficult. Obviously, a building whose bricks are stamped with the inscription ‘Palace of Sargon, King of Assyria’ is
ipso facto
dated, provided we know when King Sargon reigned. But this is the exception. By far the majority of objects found in archaeological excavations – and of course the totality where prehistory is concerned – bear no inscription. In such cases, dating can only be approximate and ‘relative’, and is based on such criteria as shape, dimensions and style. The cumulative experience derived from the excavation of many a tell has taught archaeologists that bricks of a certain size, vases of certain shapes and decoration, weapons of a certain type, sculptures of a certain style, etc., are exclusively or predominantly found at a certain level and, grouped together, form what is called a ‘cultural horizon’ or ‘cultural stratum’. If only one of these objects is inscribed with a ‘date’, or if it is found in close and indisputable relation to a monument which is otherwise dated, then the whole cultural stratum easily falls in position within the scale of time. If not, attempts are made to correlate the period during which these objects were in use with more ancient and more recent periods. To take an example, in a number of southern Mesopotamian sites a certain category of painted vases (the so-called Jemdat Nasr pottery) appears immediately
below
a cultural stratum characterized, among other things, by ‘plano-convex’ bricks (i.e. bricks of which one side is flat and the other rounded) and immediately
above
a cultural stratum where plain, buff, dark or red ceramic predominates. Various inscriptions enable us to date the plano-convex bricks to the third millennium
B.C.
(Early Dynastic period:
c
. 2900-2334
B.C.
). The plain pottery is undated but forms part of the cultural horizon called ‘Uruk’ after the site where it was first identified. The Jemdat Nasr stratum can therefore be given a ‘relative’ date. It is intermediate in time between the Uruk period and the Early Dynastic period and ends about 2900
B.C
.
How long it lasted is another matter, but there are means of forming rough estimates.

When dealing with history it becomes necessary to express dates in figures, and it is not without interest to examine how these are obtained and to what extent we can trust them.

The ancient Greeks counted from the first Olympiad (776
B.C.
), the Romans from the foundation of Rome (753
B.C.
); the Moslems date from the
hijra
(
A.D.
622) and we have our own Christian era. The ancient Mesopotamians, however, had no such fixed chronological system until late in their history, when they adopted the Seleucid era (311
B.C.
). Before that time, they simply referred to the years of reign of their rulers. These could be expressed in three ways: (1) the years of the reign were given in plain figures, e.g.
12th year of Nabû-na'id (Nabonidus), King of Babylon
; (2) or within each reign each year was defined by some important event such as victories, royal weddings, construction of temples, etc. that had taken place in the
previous
year, e.g.
Year (when) Uruk and Isin were conquered
; (3) or each year of a king's reign was named after some high official of the kingdom (eponyms or, in Assyrian,
limmu
system). In Early Dynastic Sumer all three systems seem to have been used. Then the second system (year-names) was adopted in Babylonia and used until the Kassite period when it was replaced by the first system. In Assyria, however, the
limmu
system was kept throughout history.
⁵

These dating systems could only be of practical value for the Mesopotamians themselves if they possessed for each king a list of his year's names or a list of eponyms, for each dynasty a list of its kings with the duration of their reigns, and finally a list of the successive dynasties which ruled over the country. Such lists existed and several of them have fortunately survived. Here are some examples:

It will be seen from this list that the date quoted above is the seventh year of King Hammurabi.

The time-range of these lists varied. Some were restricted to one place and one dynasty. Others, like the king list B just quoted, included several dynasties which reigned – at least apparently – in succession. Others were even more ambitious and embraced very long periods and dynasties of several kingdoms. Such is the famous ‘Sumerian King List’ reconstructed by Th. Jacobsen, which ranges from the mythical rulers ‘before the Flood’ to Damiq-ilishu (1816 – 1794
B.C.
), last king of the First Dynasty of Isin.
⁹

To express such dates in terms of Christian chronology would have been impossible but for Claudius Ptolemeus (Ptolemy), a Greek from Alexandria who in the second century
A.D.
appended to one of his books a list of all the kings of Babylon and Persia from Nabonassar (747
B.C.
) to Alexander the Great (336 – 323
B.C.
). This list, known as ‘Ptolemy's Canon’, not only gives the length of each reign but the outstanding astronomical events that marked some of them. Now it so happens that by putting together data from several Assyrian tablets we can reconstruct a long, uninterrupted
limmu
-list covering the period between Adad-nirâri II (911 – 891
B.C.
) and Ashurbanipal (668 – 627
B.C.
), and this
limmu
-list also gives the main astronomical phenomena of these times. Between 747 and 631
B.C.
the limmu-list and Ptolemy's Canon coincide, and so do the eclipses, the movements of stars, etc. they mention. Moreover, astronomers have found that an eclipse of the sun, which in the limmu-list is said to have occurred in the month of
Sivan
(May – June) of King Ashur-dân's tenth year, actually took place on 15 June 763
B.C.
, and this is precisely the date arrived at by proceeding backwards and adding together on the list the years of each reign. The absolute chronology of Mesopotamia is therefore firmly established from 911
B.C.
onwards.
¹⁰
The chronology of early periods rests upon more fragile foundations. In theory, it should be possible to work it out from king lists and dynastic lists, but these have often proved to be misleading. Not only do they show significant differences, but they contain a number of gaps or scribal errors, or they give as successive dynasties which, in fact, partly overlapped or were contemporaneous.
One should not therefore be surprised to find different figures in different textbooks and occasional changes of opinions. For instance, the accession date of King Hammurabi of Babylon was given as 2394
B.C.
one hundred years ago (Oppert, 1888), 2003 after the First World War (Thureau-Dangin, 1927), and varies now between 1848 (Sidersky, 1940) and 1704 (Weidner, 1951), but most historians of the ancient Near East have pronounced in favour of the so-called ‘middle’ chronology according to which Hammurabi reigned from 1792 to 1750
B.C.
and this is the chronology that will be found in this book.
¹¹

We cannot leave this subject without mentioning the attempts made to put chronology on a more scientific basis by means of physical methods and, in particular, the Carbon 14 or Radiocarbon method developed in 1946 by Professor W. F. Libby of Chicago.
¹²
Its principle is briefly as follows: all living organisms contain ordinary carbon of atomic weight 12 and a radioactive isotope of carbon of atomic weight 14 which is formed in the upper layers of the atmosphere through the action of cosmic rays on nitrogen, falls upon earth and is absorbed by vegetation and ultimately by animals. The ratio of carbon 14 to carbon 12 remains fixed throughout life: one-billionth of a gram for every gram of ordinary carbon. After death, when no more carbon 14 is absorbed, that part of it which is in the organism decreases slowly and regularly by reverting to nitrogen. As the curve of disintegration, or ‘half-life’ curve, of carbon 14 is known (this is 5,568 years), it is possible to find the date at which the organism died, and consequently its age. This method can be applied to organic matter, such as bone, wood, charcoal, shells, reeds, etc., found in archaeological excavations, but its usefulness is limited by a number of factors (‘standard deviation’ inherent in the radiation counting technique, contamination by older or more recent material, variations in atmospheric carbon 14 concentrations with time) and recent attempts to ‘calibrate’ radiocarbon estimations by dendrochronology (the study of tree-rings) have met with problems. This means that radiocarbon dates must be taken with caution; they are of considerable help when
prehistory is concerned – since differences of a few hundred years matter little – but cannot be used for precise, historical chronology.