Ancient History & Civilisation


The Expression of Terrestrial and Celestial Order in Ancient Mesopotamia

I. Introduction

The earliest examples of historical maps are traceable to the ancient Near East (fig. 1.1), where first Sumerians and Akkadians, and later Babylonians and Assyrians, developed the landscape of Mesopotamia by building houses, temples, palaces, cities, and states. Sumerian and Babylonian iconographic representation of features of the land and the built environment is known for the entire history of the cuneiform writing tradition, from the early third millennium to nearly the beginning of our era in Late Babylonian texts. Because of the diversity of the source material, mapmaking in ancient Mesopotamia has not been studied as a thematically unified corpus. The aim here is to afford a general overview and to make some comments on the various Mesopotamian cultural contexts for mapmaking. Completeness has not been a primary objective. Because of the sheer number of examples it is not possible to discuss each and every map on cuneiform tablets.¹ The following overview will move from the large (or local) scale to the small (or global) scale, beginning with evidence for plans of houses and other buildings, then proceeding to field surveys, city maps, regional maps, a map of the world, and finally the establishment of a spatial organization in the heavenly cosmos. The discussion will therefore not be strictly chronological, but each section will proceed from earlier to later examples.

Even in the prehistory of the Near East, a “map” from the Neolithic site Çatal Hüyük in central Anatolia attests to social awareness of the inhabited place and its relation to its geographic surrounds. Found in 1963 by the archaeologist James Mellaart during the excavation of Çatal Hüyük near Ankara, Turkey, this 3 m red-brown polychrome wall painting, radio carbon dated to approximately 6200,² appears to represent the town itself with eighty rectangular buildings of varying sizes clustered in a terraced town landscape (fig. 1.2 and plates 1a and 1b). Mellaart noted the similarity of the representation of the houses to the actual excavated structures found at the site, that is, rows of houses built one beside the other with no space between them. The wall painting shows an active double-peaked volcano rising over the town, likely to be the 3,200 m stratovolcano Mount Hasan, which is visible from Çatal Hüyük. Lava is depicted flowing down its slopes and exploding in the air above the town. A cloud of ash and smoke completes the scene.³

FIGURE 1.1 Map of the Ancient Near East. Adaptation by the Ancient World Mapping Center, University of North Carolina, Chapel Hill, of J. B. Harley and David Woodward, eds., The History of Cartography, vol. 1, Cartography in Prehistoric, Ancient, and Medieval Europe and the Mediterranean (Chicago: University of Chicago Press, 1987), 108, fig. 6.1.

While not a map in the sense of a surveyed and measured image of some part of the earth’s surface, the image of the town with its local mountain spewing molten rock is a representation of a phenomenon that one can well imagine was actually experienced. In the context of other prehistoric European maps, Catherine Delano-Smith (1982) speculated that the representation of this geographic scene most likely had a ritual function. The map was not meant as a projection of a landscape upon a measured framework or even as a lasting representation, as the walls of the dwellings at Çatal Hüyük were regularly replastered and painted over. As Delano-Smith put it (1982, 18), “If villages and fields were depicted in it, they would have been associated less with the ‘objective’ recording of a spatial distribution for purposes of direction or reference than with invoking the favours of the controlling forces of those aspects of life these topographical features were used to represent or with attempts to appease them.” Of course, such interpretations are conjectural in the absence of contemporary written evidence. As will be clear in this presentation, however, even in the presence of contemporary written evidence, the nature and motivations of mapmaking in the ancient Near East are not always easy to assess. The Çatal Hüyük map, while certainly not a map in the familiar sense, shares aspects with other ancient Near Eastern maps which reflect a desire to represent, for whatever reason, real aspects of the visible terrain of the immediate environment.

FIGURE 1.2 Part of a representation of a town in a Neolithic wall painting from Çatal Hüyük, Turkey, dated to the early seventh millennium BCE. Approximately 3 m in length. Image reproduced from J. Mellaart, “Excavations at Çatal Hüyük, 1963, Third Preliminary Report,” Anatolian Studies 14 (1964): 55 and pl. V. Photograph by James Mellaart, reproduced with permission.

One of the functions of ancient Near Eastern maps is the representation of architectural or topographical order imposed on the physical landscape, or the idea of the physical landscape, both terrestrial and celestial. Mesopotamian maps do not guide the traveler from one place to another, but rather represent those ordered features imposed by human beings on the natural world, either physically, such as the temple, the city, agricultural lands, roads, and canals, or philosophically, as in notions of a world inclusive of known and unknown regions shown in diagrammatic form on the one surviving Babylonian world map. They are maps in the sense that a map is, as defined by Denis Wood (1992, 122), “an icon, a visual analogue of a geographic landscape . . . the product of a number of deliberate, repetitive, symbolic gestures . . . formal items—the discrete elements of iconic coding . . . shaped within the space of the map . . . or preformed and imposed on the map, activating formal symbolism and formal metaphor as well.”

This definition of the map as an icon, and mapmaking as a process of developing iconic codes for representing topographical features, is useful for pulling together disparate sets of evidence that can be viewed together within a coherent framework of “mapmaking” in the ancient Near East. The features of the topography that eventually warranted representation, such as houses, temples, agricultural fields, and cities, distinguish undeveloped land from property. Mesopotamian maps were not exercises in the objective rendering of geography or the unbuilt landscape. One might say that land was not mapped; rather, property was. The reasons that made different kinds of property worthy of being mapped varied from economic to religious and cosmic.

Practical geography, on the other hand, is represented in literary form, in itineraries attested from Old Babylonian times up to Neo-Assyrian, that is, from the eighteenth to the seventh centuries. Their interpretation and their use for reconstructing ancient landscapes is fraught with difficulty, because the identification of toponyms and their relative distances from one another is often uncertain. The itineraries do often provide distances, given in months and days, or, in the case of the later examples, in the unit of distance and time called bēru (danna), which we translate as the “double-hour,” and which was also adopted for use in measuring stellar and other celestial distances in astronomical texts. Indeed, the bēru was the pervasive measure of both distance and time in celestial terms, as it was a one-twelfth subdivision of the nearly constant unit of the day (ud), the interval between successive sunsets. The celestial “double-hours” stem from the correlation made between the length measure of twelve terrestrial bēru and the fixed time measure of one day (ud). In view of the term’s Sumerian etymology—KASKAL.GÍD = danna, “long road”—the original length measure of the bēru referred to the distance traversable on foot in the time it would take for the sun to travel thirty degrees across the sky, or one-sixth the length of daylight (assuming an equinoctial day of twelve daylight hours). The distance of one bēru would then correspond roughly to a two hours’ march. Such a correlation between terrestrial distances and the division into twelfths of the “day,” or rotation of the sky, is, in a concrete sense, geodetic.

Not only on earth but in heaven too, structural order was imposed by the transposition and projection of terrestrial features of the built and cultivated environments from earth to heaven. The horizon was the “cattle pen,” and the heavenly bodies the cattle and sheep that followed their orderly, or, in the case of planets, somewhat less orderly, paths (Rochberg 2010). The religious text Enūma Eliš, also known as the Babylonian Creation Epic, uses this double metaphor for the gods as both celestial bodies and livestock when it says, “Let him (the creator god Marduk) assign the regular motions of the stars of heaven; let him herd all the gods like sheep” (Enūma Eliš VII 131). In addition, the regular paths followed by the stars across the sky from their risings to their settings were compared to furrows of a field. The fixed stars were also attached to specifically named “roads” serving to mark their directions, and the moon too moved against the starry background along what was called “The road of the moon” (harrān dSin).

Other units familiar from the texts of land surveyors, the cubit or “forearm” and finger, are also found in astronomical contexts where distance between celestial bodies is in question. Despite a dearth of iconographic representation of the celestial roads and the arrangement of constellations within them, there survives noniconographic textual evidence making reference to double-hours, cubits, and fingers (bēru, ammatu, and ubānu, respectively) within the heavenly roads or with respect to other devices for fixing celestial positions, such as the ecliptical stars. These references show that such imagery and its associated metrology had currency throughout the cuneiform tradition of astral sciences. Celestial mapmaking, as we would recognize it, is not well represented, despite the application of units of time-distance such as bērus and degrees to the measurement of relative positions of celestial objects. Although the surviving iconographic evidence for “mapping” the heavens is extremely limited (see III.2 below), a practical astronomy that systematically organized, schematized, and predicted celestial phenomena shows that the celestial landscape was quite well “mapped,” in fact.

Extant cuneiform maps cannot be considered to constitute a coherent tradition of cartography in which a continuous evolution of mapping techniques or even conceptions of the map itself are evident over time. Nonetheless, various aspects of the ancient Mesopotamian physical world, both terrestrial and celestial, are represented on cuneiform tablets over a considerable time span. These representations are eminently classifiable as cartographic, even according to the definition of professional cartography by the British Cartographic Society as “the science and technology of analyzing and interpreting geographic relationships, and communicating the results by means of maps” (Harley 2001, 151).

Mesopotamian maps have not always been readily incorporated into the history of cartography. In older scholarship resistance to including them can be found even with regard to the best-known example, the so-called Babylonian mappamundi from the later eighth or early seventh century. In his introduction to the 1959 reissue of Edward Bunbury’s 1883 History of Ancient Geography among the Greeks and Romans from the Earliest Ages till the Fall of the Roman Empire, William Stahl characterized this map as a crude representation (p. iii): “Before the time of the ancient Greeks, geography and cartography were in a primitive state. The early Babylonians, for example, had developed remarkable precision and skill in observing and predicting the orderly movements of celestial bodies, but their conceptions of the earth were what one might expect of a relatively isolated people. A Babylonian clay tablet . . . depicts the earth as a circular plane, bisected by the Euphrates River, with the capital city of Babylon located near the center and a few adjacent countries bordering upon an encircling ocean.” Stahl’s condescension takes it for granted that the map was intended as a direct and accurate rendering of its subject, and that its maker was a scientist with goals of precision and accuracy. However, such a position assumes criteria for mapmaking that will not be met by most ancient Near Eastern exemplars.

Recent cartographic historiography has revised the concept of what a map is (or what it is to make a map) and has consequently opened the way to consideration of maps as representations not only of the environment as a physical object, but also of notions of the environment, imagined, abstracted, or ideal realms of the world beyond the concrete features of the immediate experienced terrain (MacEachren 1995, 255–56). In particular, Brian Harley has articulated again and again the value of reassessing the history of maps and redefining cartography in accordance with evidence of its practice rather than by a priori modern criteria. As a result, he uncovered a deep level of humanity in maps, ancient and modern, and showed how maps invariably reflect culture and subjectivity: “As mediators between an inner mental world and an outer physical world, maps are fundamental tools helping the human mind make sense of its universe at various scales. Moreover, they are undoubtedly one of the oldest forms of human communication. There has probably always been a mapping impulse in human consciousness, and the mapping experience—involving the cognitive mapping of space—undoubtedly existed long before the physical artifacts we now call maps” (Harley and Woodward 1987, 1). What follows will summarize the evidence for maps and mapmaking in historical ancient Mesopotamia. Both maps that represent ideas of the landscape and those that represent its actual features are included; the line separating these is not always easy to define.

II. The Terrestrial Landscape


II.1.a. House and Plot Plans. At the most localized level, there survive tablets on which plans for houses and urban plots are drawn, accompanied by various kinds of inscription. House plans, showing walls and doorways, come from a wide chronological span, from the Old Akkadian and Ur III to Old Babylonian and Neo-Babylonian periods (ca. 2350–400). Early examples are also known from third-millennium Girsu, although it is difficult to identify with any certainty what kind of buildings they represent. A well-preserved house plan from Ur III Umma shows dimensions in cubits, probably in reference to the dimensions of the rooms themselves.¹⁰

The ground plans of houses etched into clay tablets show walls by parallel lines, and in what seems to be a real representation of the floor plan they indicate the placement of doorways (see fig. 1.3). Sometimes doors are indicated by a pair of cuneiform wedges as though to mark a break in the wall, or even by hatch marks perpendicular to the wall. One Old Babylonian building plan—dated to the reign of Abi-Ešuh and inscribed with the description “plan (literally ‘drawing’) of a house of Sippar-Jahrurum”—includes labels for the various rooms and parts of the building, such as doors (KÁ = bābum), under the stairs (šapal simmilti, written GIŠ.KUN4), a reception room (PA.PAH = papāhum), and a barbershop (É ŠU.I = bīt gallabim).¹¹ These indications would suggest that the building had some sort of public function, though it is apparently not a temple.¹² Such labeling of rooms, or parts of a building, is also to be seen in an earlier plan from third-millennium Girsu: it mentions a courtyard (KISAL = kisallu), reception room (PA.PAH = papāhum), and living quarters (KI.TUŠ = šubtu), and gives the length and width of rooms measured in cubits (KÙŠ) and ropes (NINDA.DU).¹³ An Ur III tablet from Nippur offers a further example of a building plan in which rooms are labeled: it represents a building containing a kitchen (É.MU), a washroom (É.LUH), a weaving room (É.UŠ.BAR.RA), and an archive, or “tablet room” (É.DUB).¹⁴

FIGURE 1.3 Ground plan of a residence in Umma with central courtyard, Ur III period. Vorderasiatisches Museum, Berlin, VAT 7031. Photo: Olaf M. Teßmer, Vorderasiatisches Museum—Staatliche Museen zu Berlin. Reproduced with permission.

Early house plans, such as from the Old Akkadian to the Old Babylonian periods, may well represent sketches, or scribal exercises, rather than true blueprints (see figs. 1.4 and 1.5). In the Neo-Babylonian period similar schematic house plans exist from the Nappāhu family archive (Baker 2004, 16). In addition to these documents, other Neo-Babylonian tablets, many of which are from Babylon during the reign of Darius I (522–486), preserve surveyed plans of urban plots in records of a land registry for both plots and fields, perhaps for the purpose of taxation. They do not show house plans, but merely the plot itself in relation to its surrounding streets or other properties;¹⁵ they provide measured boundaries and a total surface area of the property, a four-sided single plot with an occasional adjacent plot represented in addition. The houses so described in these plans are small (under 200 sq m) as compared against actually excavated houses. Various explanations are possible: they may have been additional rental properties, or the small-sized properties were only thus on the plan, whereas the actual area formed part of a larger house unit, not divided physically (Baker 2004, 62).

FIGURE 1.4 Modern redrawing of a Middle Babylonian plan of a house with courtyard. Original in Iraq Museum, Baghdad, IM 44036, 1. Reproduced from Joachim P. Heisel, Antike Bauzeichnungen (Darmstadt, Germany: Wissenschaftliche Buchgesellschaft, 1993), 35, with permission of the publisher.

The “maps” which document the delimitations of properties served a social and economic function of controlling real estate. Archival sources dealing with inheritance, sales, and rents confirm that the ownership of real property was a significant issue in Mesopotamian society over the course of its history. As Heather Baker (2004, 60) has pointed out, the house plans “are very similar in their basic configuration to the textual descriptions of houses, that is, they represent the sides of the property according to the points of the compass, labeled according to the neighboring property-owners and/or topographical features (usually streets) on each side, giving the measurements of the property’s boundaries plus the total surface area. Usually each plan consists of a single four-sided plot, but occasionally an adjacent plot is also depicted, according to the requirements of the survey.” From the point of view of function, then, the graphic documentation of houses and plots, as with the field plans (to be discussed below), played a role in the control of property ownership. From a cartographic viewpoint, the representation of houses and building plots and the use of certain iconic conventions for showing walls and doorways probably indicate a relative standardization effected by the training of the scribes who drew up the plans.¹⁶

FIGURE 1.5 Middle Babylonian house plan inscribed with dimensions of walls (?). The British Museum, BM 80083. Reproduced with permission.

II.1.b. Temple Plans. The conventions for house plans were also employed in representing the more elaborate houses of the gods. A number of temple plans on clay tablets are known, even though it is not possible to identify the names of either the deity or the temple in any instance. Such plans may be scribal exercises, or sketches, as is most likely the case in a tablet of uncertain provenance and date (possibly Middle Babylonian or later): it shows a building complex within which is an isolated smaller structure labeled “sanctuary” (aširtu) and comparable to northern temple architecture of the “bent axis” type.¹⁷

Several plans incised into clay tablets represent the stepped “ziqqurat” form of the traditional ancient Mesopotamian temple platform. One of these, from Nippur and of uncertain date (Old or Middle Babylonian), appears to present seven graduated levels in a square formation, and has inscribed on it a reference to an exorcistic purification ritual “to abolish (literally ‘smash’) what is evil” (hulu dúb).¹⁸ In the central portion of this plan, within what appear to be rooms surrounding the uppermost level are additional references to ritual objects: “pure (metal) hand,” “pure (metal) pedestal,” and a golden dagger and vessel.¹⁹ A Late Babylonian clay tablet with a drawing of a ziqqurat on both obverse and reverse does not permit a count of the levels, though on one side there are at least six.²⁰ Another Late Babylonian tablet, probably from Babylon, is a representation of a seven-stepped ziqqurat. The overall dimensions are such that the height, length, and breadth are all equal, measuring 42 cubits (21 m), and each level increases in equal steps by ½ GAR (3 m) per tier.²¹ Excavated ziqqurats, such as at Ur (62.5 × 43 m wide and high) and Dur-Kurigalzu (‘Aqar Quf) (69 × 69 m) are considerably higher, although the base (kigallu) of the E-temen-anki at Babylon was identical to this one, according to the E-sagil Tablet (AO 6555:19). While this ziqqurat plan could well be didactic or simply idealized, it certainly reinforces the likelihood that scribal training included plans of buildings and their mensuration (Wiseman 1972, 143–45).

From ca. 2100 we find the plan for a monumental temple drawn on a 93 cm high inscribed diorite statue of the ruler Gudea of Lagash (fig. 1.6).²² The king is portrayed as an architect or builder, sitting with his ground plan for the E-ninnu temple of Ningirsu, which was excavated by Ernest de Sarzec in 1880 in the city of Girsu (modern Tello).²³ The inscription identifies Gudea as one handpicked by the god Ningirsu “from among 216,000 (= 60³) persons” to shepherd his land (Edzard 1997, p. 32 iii 9–10). On ritually clean ground, he built the E-ninnu, which “raised its head to heaven,” an expression that would become a literary trope applied to other cities and temples, culminating perhaps in the very name of the temple of Marduk at Babylon itself, the E-sagil, “house whose head is raised (high).” The cedar timbers used for building E-ninnu were procured from distant forests of the Amanus mountain range in Syria, the road to which Gudea says Ningirsu “opened for him” (Edzard 1997, p. 33 v 21–27). The stone slabs for the stelae (monuments) set up in the temple courtyard were also brought from the western mountains; other rich materials, copper and gold, were brought in from long distances to the west and east (Elam, Meluhha). The statue’s inscription, therefore, tells not only of the building of E-ninnu, but also of the geographic knowledge of distant lands, as well as access to them and control of their resources. Moreover, on the board that rests on Gudea’s knees (fig. 1.7), beside the building plan, are the tools of the architect and mapmaker, the stylus and a graduated ruler, sometimes called the “cubit of Gudea.”²⁴ The measurements of this cubit have been the subject of discussion, with various values given, but all approximating 50 cm (Powell 1987–90, 462 sub §I.2.e).

FIGURE 1.6 Detail of life-size diorite statue of Ensi Gudea of Lagash with the building plan of the Ningirsu temple. Musée du Louvre, AO 2. Photo: Erich Lessing/Art Resource, NY. Reproduced with permission.

By the late third millennium, it was customary to draw preparatory plans for buildings on clay (Donald 1962, 190). Representations of temple plans on stone are in a different class, however. The fineness of the drawing and the presentation of the plan on Gudea’s lap—with not only the plan, but also the text describing the construction of the E-ninnu with its architectural details, carved in stone on the lower part of the body of the king—reflect an ideal. It is, so to speak, a representation of a representation, and never meant for “use” as a blueprint. This E-ninnu plan confines itself to the area enclosed by an external wall, with its merlons and crenels and six similarly crenellated gates appearing as battlements, giving the whole a castellated effect. It has been suggested by Wolfgang Heimpel that the six gates in the building plan are to be correlated with the six gates mentioned in Gudea’s Cylinder A describing the statues, standards, and figures that he placed at each one.²⁵ Representation of such monumental temple construction on clay can also be seen in a fragment of a tablet from Girsu.²⁶

FIGURE 1.7 Tools of the architect and mapmaker, the stylus and a graduated ruler, on the lap of Gudea of Lagash. Musée du Louvre, AO 4. Reproduced with permission.

Later texts that document the names and epithets of sacred buildings in the major Babylonian and Assyrian religious centers, Uruk, Ur, Kish, Nippur, Assur, and of course Babylon, can be viewed against the background of geographic and topographical Sumerian lexical texts as well as of hymns venerating cities and temples (George 1992, 1–2). Later, after the Old Babylonian period and the decline of the Sumerian literary genre of city and temple hymn, a similar purpose came to be served by scholarly tracts in list form such as the compilation “Tintir = Babylon.” These glorified a city’s cosmological centrality and sacred status by a comprehensive list of its epithets, temples, and shrines.²⁷

The two representations of temple plans in stone, one on the statue of Gudea, the other on a stele of Nebuchadnezzar II, both refer to significant architectural monuments, the E-ninnu and the E-sagil. These plans, unlike those preserved on clay, are an iconographic expression of glorification along the same lines as those identified by Andrew George for the scholarly lists. The E-ninnu plan is presented on the lap of Gudea in reverential demonstration of that ruler’s piety, and the plan of the E-sagil and the image of the ziqqurat E-temen-anki are shown atop a stele of the Neo-Babylonian ruler Nebuchadnezzar II, where he too is shown beside these images as well as below them in an attitude of devotion and dutifulness.²⁸

The Late Babylonian (third-century) “E-sagil Tablet” concerns the dimensions of the E-temen-anki using two different cubit standards, an early (Kassite and Early Neo-Babylonian) and a later (Standard Babylonian) one. The cubit, or “forearm,” was the Mesopotamian (Assyrian and Babylonian) standard linear unit of measure based on the distance from the elbow to the tip of the extended middle finger and was used from the third millennium (Old Akkadian period) onward.²⁹ It served to measure lots, fields and orchards, walls and other constructions, timber and beams, garments, and even the heights of people.³⁰ After determining the areas of two courtyards outside the Temple E-sagil proper, in the didactic form of a mathematical problem, the E-sagil text goes on to calculate the area of the base of the ziqqurat E-temen-anki from its length and width. The linear measure of length and width, using the later Neo-Babylonian cubit standard, is then converted to capacity-surface measure, that is, in terms of the volume of grain needed to seed such an area (in the standard equivalence 30 sila [or 1 ṣimdu] of seed to 1 ikû of land) (George 1992, 109–12). This conversion of linear to capacity measure is paralleled in the surveying of fields, first measured in cubits and then converted to a seed or other measurement (Nemet-Nejat 1982, 143–44).

Temple plans continue into the Neo-Babylonian period, as in the plan on a large clay tablet (23 × 31 cm) from Sippar that shows crenellations or buttresses in the outer wall, with bricks in a grid formation, and indicates the cardinal directions on the outer margins of two sides (fig. 1.8). Locations of doors are inscribed, as are other parts of the sanctuary; in the inside of the grid there is noted: “[n cubits, n] fingers (ubānu), the length of the storehouse of the northern gate.”

II.1.c. Field Plans and Cadastral Surveys. Drawing plans for real estate, including houses, plots, and fields, was part of professional scribal work, and it is possible that the same official, the šassuku (who assessed land and property for taxes), also measured, surveyed, and drew plans (Nemet-Nejat 1988). Actual surveying techniques are difficult to reconstruct for lack of evidence, but the third-millennium technical term gag . . . dù, “to drive in the peg,” in sale documents refers to holding a measuring line down by driving a nail or peg into the ground, a practice that may later have been reflected in the insertion of a clay nail into the wall of a house along with the sale document. There is again testimony to surveying in a group of mathematical texts which refer to a broken reed used in measuring a field.³¹ Literary evidence, too, confirms the connection between real estate and field surveying with the measuring line held by the peg. Thus, in a Sumerian disputation Enkimansi and Girinišag accuse one another: “You go to divide up an estate but are unable to divide up the estate. For when you go to survey the field, you cannot hold the measuring line. You cannot hold a nail in your hand” (Kramer 1963, 241).

Evidence for surveyed and measured parcels of land is to be found, for example, in Kassite-period (ca. 1400) boundary stones (kudurru texts), which legally document the granting of land by the king to a subordinate (ardu). Beginning in that period, kudurru texts show that fields are measured in “seed” (ŠE.NUMUN) and by the “great cubit” (ammatu rabītu) in accordance with a standard ratio of the amount of seed to the area of land (30 sila of seed to 1 ikû of land), having its origins in actual seeding practices of the Ur III period and continuing through the old Babylonian period, attested in evidence from Susa (Powell 1984, 35). In the context of fields, the capacity measure shows that the value of land was linked to yield capacity. The kudurru of Adad-zēr-iqīša notes the name of the surveyor of the field (māšihāni eqli, “the person who has surveyed the field”),³² as well as giving its location and size.

FIGURE 1.8 A,Temple plan on a Neo-Babylonian tablet from Sippar. British Museum, BM 68840—68845. Reproduced with permission. B, Modern redrawing.

It is difficult to reconstruct the topographical aspect of Mesopotamian agricultural lands, their precise features and dimensions. Archaeology can do only so much to realize macrostructural ancient landscapes by locating settlements, fields, and canals. Microstructural features are accessible through two types of documents relating to the shape and size of fields: plans, measured and drawn to scale, and cadastral surveys giving dimensions of the sides of fields oriented to the cardinal points. As is the case in other types of plans (house, plot, or temple), it is difficult to know whether the field plans and surveys have been produced in an actual administrative context as sketches for the preparation of land tax or sale documents, or whether they are only models for teaching surveying techniques. If the tablets with detailed plans of fields are models, not real cases, obviously they do not shed much light on ancient topography.

Thirty such field plans are extant for the Ur III period, between 2112 and 2004.³³ They demonstrate techniques of field surveying—specifically, the adaptation of irregularly shaped fields to regular rectangles with added or subtracted “appendages,” often triangular in shape, or with long narrow strips, as was apparently common in the layout of fields to meet the needs of irrigation and of plowing with a heavy plow that was difficult to turn; see figure 1.9.³⁴

The average Neo-Sumerian field was quite large, measuring 100 ikû, the equivalent of about 36 hectares, or nearly 90 acres. Earlier, in the Akkadian period, texts deal in much smaller-sized fields, between 4 and 10 ikû, perhaps because they were restricted to domestic rather than commercial use undertaken by large public (temple) institutions. In the administration of large commercial fields, also common were ½ fields (around 50 ikû), ¼ fields and ¾ fields, as well as “double fields” (around 200 ikû) and “triple fields” (300 ikû). Field text plans also reveal 1¼ (125ikû), 1½ (150 ikû), and 2½ (250 ikû) fields. Mario Liverani (1990, 157–58) concludes that these figures represent “a rounding up of the original measurements of the fields with reference to a ‘standard field’ of 100 ikû, which is also the area of ground cultivable by one apin-plough under the direction of one engar-farmer.”

FIGURE 1.9 Modern redrawing of an Ur III field plan. Original in Eski Sark Eserleri Müzesi, Istanbul, ES 1107. Reproduced from F. Thureau-Dangin, “Un cadastre chaldéen,” Revue d’assyriologie et d’archéologie orientale 4 (1897): 13.

The survey of one such agricultural tract, belonging to the city of Umma, is inscribed on a tablet that designates the field, measuring 4 BUR 2 iku (about 68 acres), as “the field of the irrigation ditch Guru” which belongs “to the goddess Ninurra.”³⁵ This field plan is dated to King Bur-Sin of the Third Dynasty of Ur. The drawing on the tablet, not to scale, may have represented notes taken down for the purpose of calculating the area within the field’s boundaries, which itself was necessary for figuring out how much seed grain was sufficient for planting such a field and how much yield to expect. As already indicated, such a conversion to surface-capacity measure points to the connection between the value of land and its economic productivity, as one would expect in an agrarian economy. Another field plan, from the city of Nippur and dated to the fifth year of Šu-Sin, shows a subdivided area of land allotted to several members of the temple personnel for their own subsistence. As was the case for irrigating the long fields of the Ur III period area of Lagash, this plan shows the strips of field aligned on the short sides to a canal, permitting irrigation to go down the line of furrows (Zettler 1989, text 6 NT 777).

FIGURE 1.10 Middle Babylonian (ca. 1500 BCE) Nippur Field Plan showing the curvature of the watercourse and its irrigation canals. University Museum, University of Pennsylvania, CBS 13885. Reproduced with permission.

Finally, in this connection, a map (fig. 1.10) from Kassite Nippur (ca. 1500) situates a centrally located field “between the canals, 8 NUMUN of arable land, the field of the palace” (Langdon 1916). On either side are fields for the upkeep of the bārû, or diviner priests. The neighboring towns of Kar-Nusku and Hamri, known from temple records at Nippur, are designated as well as the distribution of watercourses, both large and small canals, and a marshland to the north of the palace field. At the top of the map we find the Field of Marduk, the national god of Babylonia, probably a field owned by the temple of Marduk, E-sagil.


Examples of city maps, some quite fragmentary, are preserved for Gasur (later called Nuzi), Nippur,³⁶ Babylon,³⁷ Sippar,³⁸ and Uruk.³⁹ The ancient Mesopotamian city stands as the quintessential vehicle of self-identification in that fundamentally urban civilization (van de Mieroop 2004). Our knowledge of a Mesopotamian conception of “citizenship” is unfortunately quite poor, but a member of the community was identified as a “son of the city,”⁴⁰ and so the equivalent expression of the term “citizen,” or something perhaps similar to it, is tied to the concept and word for the city. We cannot determine the precise function of the city maps, and one hesitates to suggest the anachronistic notion that their production hints at something akin to self-representation on the social level. Even so, while perhaps not consciously self-representational, a cuneiform city map was in fact a representation of a social and topographical phenomenon that functioned not only as political and economic center but also as the residence of a god in the Sumero-Akkadian pantheon. This was the defining function of a Mesopotamian city, and so a number of identities related to community, government, and religion were in fact integral to the character of any given urban center (Stone 1995, 235).

II.2.a. Nippur. In the third millennium, Nippur was the single most important religious center of all Sumerian city-states. It was the earthly residence of the god Enlil, divine ruler of the pantheon, where his temple E-kur, “House Mountain,” was built. Not only had the city and its principal temple functioned as a sacred site for cult and ritual observance, but also ideologically Nippur represented the very center of the cosmos, as is clear from its epithet Dur-anki, “Bond of Heaven and Underworld.”⁴¹ In the fall of 1899, during the excavation at Nippur by the University of Pennsylvania, a 21 × 18 cm clay map was found (fig. 1.11).⁴² Published only in 1955, it was subsequently analyzed by Samuel Kramer and Cyrus Gordon in 1956 (Kramer 1956, 271–75). Scholarly consensus dates the map to the Kassite period (perhaps fourteenth to thirteenth century) (Gibson 1977, 1978). This marks a time of renewed vitality for Nippur, which had been abandoned for several centuries since the Old Babylonian period (reign of Samsuiluna, eighteenth century) (Gibson 1978, 119n9).

If in fact the map was first drawn by Kassite scribes, it may indicate the reconstitution of Nippur’s former glory with renewed building by Kassite kings. It represents, in the manner of house plans, the E-kur temple and its attached Kiur,⁴³ associated with the underworld. These buildings are shown with double lines for walls and with parallel crosshatches marking the positions of doorways, conventions employed in house plans. Also indicated are the An-niginna, a kind of enclosure, the Eš-mah, “Exalted Shrine,” a city park named “Central Park,” seven gates (including the Ur-facing gate and the Uruk, Gula, and Nergal gates), and two important canals, the Nunbirdu (Id-nunbir-tum) at the top of the map and the Center City Canal, appropriately named. Many features are given measurements in a standard Sumerian unit of length, the rod, or nindan (= 12 cubits).

FIGURE 1.11 Nippur city map drawn to scale. Frau Professor Hilbrecht Collection of Babylonian Antiquities, Friedrich Schiller University, Jena, HS 197. Photograph by Prof. Dr. Manfred Krebernik. Reproduced with permission.

With the aid of aerial photographs, and by study of archaeological remains and comparisons with the modern site plan of Nippur, Miguel Civil determined that the Nippur map was drawn to scale; he also found its orientation to be northwest to southeast. Not only is this map matched by the modern reconstruction of the site plan, but also when we read the Sumerian literary text “Enlil and Ninlil,” set in the city of Nippur, we are placed in the very topography represented on the map. The poem begins:

There was a city, there was a city—the one we live in. Nibru was the city, the one we live in. Dur-gišnimbar was the city, the one we live in. Id-sala is its holy river, Kar-geština is its quay. Kar-asar is its quay where boats make fast. Pu-lal is its freshwater well. Id-nunbir-tum is its branching canal, and if one measures from there, its cultivated land is 50 sar each way. Enlil was one of its young men, and Ninlil was one its young women. Nun-bar-še-gunu was one of its wise old women.

Enlil and Ninlil are described here as an adolescent boy and girl, and the girl’s mother, Nunbarshegunu, is there to advise. She says:

The river is holy, woman! The river is holy—don’t bathe in it! Ninlil, don’t walk along the bank of the Id-nunbir-tum! His eye is bright, the lord’s eye is bright, he will look at you! The Great Mountain, Father Enlil—his eye is bright, he will look at you! The shepherd who decides all destinies—his eye is bright, he will look at you! Straight away he will want to have intercourse, he will want to kiss! He will be happy to pour lusty semen into the womb, and then he will leave you to it!⁴⁴

II.2.b. Babylon. In the seventh century the city of Babylon was restored to former splendor by the Assyrian monarch Esarhaddon (680–669). He describes the temple of Marduk as “the palace of the gods, mirror image of the Apsû, counterpart of E-sharra, replica of the abode of the god Ea, counterpart of the constellation Field.”⁴⁵ The centrality of Marduk’s earthly residence was supported by a religious and political mythology accounting for the ascendancy of Babylon and its god, Marduk, over all other Mesopotamian cities and all other gods. The creation poem Enūma Eliš, composed to celebrate the rise of Marduk to kingship over all the gods, constructs a cosmic geography in which Marduk’s residence on earth is his city of Babylon, and this idea is further promoted in the theological text “Tintir” (George 1992, 6). Andrew George (1999, 69–70) discusses the fact that the religio-cosmic centrality of the city of Babylon lay in its role as the gathering place and home of the pantheon of all the gods of heaven and underworld. This ideology built upon a tradition already established in Nippur many centuries earlier: here Enlil’s temple was a cosmic center and Nippur functioned as the meeting ground for the entire assembly of the gods.

A small fragment of a Late Babylonian map of Babylon (BM 35385) marks the great Shamash Gate, and beside it the toponym Tubaki shown near a watercourse indicated by parallel lines infilled with wavy lines (fig. 1.12).⁴⁶ The topographical text “Tintir” makes mention of the city quarter of Tuba: “From the Navel of the Bow of the temple of Bēlet-Ninua to the river bank is called Bab-Lugalgirra; From the Šamaš Gate to the river is called Tuba.”⁴⁷ This text also describes Tuba as a city quarter of the “West Bank” (BAL.RI (balar) ereb šamši). Accordingly, George has reconstructed the map of Babylon from which the fragment came. A learned commentary on the topography of Babylon and the city’s dimensions occupies the obverse of the same tablet. Among the subjects it treats is the city wall (part of which is visible on the map) and the location of all gates; these George (1992, 135–37) was able to include in his reconstruction on the basis of the metrological data provided by the commentary. Notably, the little that is preserved of the Babylon map has been drawn as though with a ruler; the lines indicating the wall and the canal are straight, and the waterway is marked with the wavy lines that conventionally symbolize water.

FIGURE 1.12 Map of Babylon showing the city quarter of Tuba by a watercourse. The British Museum, BM 35385. Reproduced with permission.


II.3.a. Nuzi. The oldest of cuneiform maps was discovered by the Harvard-Baghdad School Expedition in 1930–31 16 km south west of modern Kirkuk at the site of Yorghan Tepe (Meek 1932). Here, during the late third millennium, the ancient site of Gasur was a city with a palace and a temple complex, which is the typical urban structure found in Mesopotamian cities of both the south and the north until the Neo-Assyrian period of the seventh century. Among the five thousand tablets found during the excavation of Gasur (later renamed Nuzi after a major influx of Hurrian-speaking people in the middle of the second millennium) was a 7.6 × 6.8 cm tablet with a depiction of settlements or estates and their environs, the latter comprising hills and watercourses or irrigation canals, one of them called Ra-hi-um, presumably within the local area. The settlements are marked as circles, inscribed with the name of the place. The only one of these completely preserved is MAŠ.GÁN. BÀD Eb-la, meaning “Fortress of Ebla.” At the center of the map there appears a circle without any inscription. To its left and right is an inscription indicating the size and ownership of the estate, specifically 354 ikû (approximately 128 hectares, or 318 acres) of cultivated land belonging to Azala, which could be the name of either a person or a place. Given its central location on the tablet, the natural inference must be that the map was produced specifically to clarify the location and size of this parcel belonging to Azala.

This earliest of cuneiform maps (fig. 1.13) is explicitly oriented, with the names of the winds written on the sides of the tablet. The top is identified as IM.KUR, “mountain wind,” or “east.” Correspondingly, the bottom is IM.MAR.TU, “west wind,” the left side is north (IM.MIR), and the right is south (the writing here is not preserved). Theophile Meek, the epigrapher for the Nuzi excavation, speculated that the map represented an area between the Zagros Mountains and the hills near Kirkuk. The watercourses indicated would then be identifiable as the Lower Zab, Radanu, Tigris, or perhaps just irrigation canals.

FIGURE 1.13 Akkadian-period (ca. 2400 BCE) regional map denoting cultivated land and its owner. Courtesy of the Semitic Museum, Harvard University, SMN 4172.

This Nuzi map is inscribed, and its synthesis of iconic symbols, topography, and script thus acts to convey its meaning. Two fundamental, if not almost universal, cartographic symbols occur on the map: the hill sign of overlapping humps, viewed from the side as it were, and the watercourse sign of parallel lines. It might be claimed that we recognize these signs so easily because they bear a visual likeness to the topographical features they represent. There is no guarantee, however, that hills will be drawn as rows of hummocks or rivers as parallel or wavy lines, and indeed such symbols taken out of context do not necessarily conjure up hills and rivers. At least it may be observed that these same icons bear resemblance to the abstracted and linear pictographic representations of hills and water seen in the early development of the cuneiform script.⁴⁸ Of course, by the date of the Nuzi map, these pictographic signs had already been long superseded by cuneiform ones. Nonetheless, the unified features of map icons and the precuneiform form of these word signs are noteworthy.

II.3.b. Nippur. During the third expedition of the University of Pennsylvania’s excavation of Nippur a tablet was found with a map of at least seven cities (some, but not all, written with the determinative URU, “city”) and their relative positions with respect to a canal called Tukulti-É-kur. This tablet (fig. 1.14) was published by Albert Clay (1905). He noted that several of the cities (or towns) mentioned on the map were identifiable with toponyms mentioned in temple inscriptions of the Kassite period (fourteenth century), specifically ones dealing with tax records, business documents showing transactions with income, receipts, and payrolls of temple and storehouse personnel.

FIGURE 1.14 Topographical map from Nippur. University Museum, University of Pennsylvania, CBS 10434. Reproduced with permission.

The canal, positioned down the center of the tablet, is represented by narrow parallel lines and is inscribed with its name, ÍDTu-kul-ti-É.KUR. Also more or less centrally located is a town of the same name, Tukulti-E-kur, marked as a small square incised with single straight lines. The other towns, too, are shown as squares. Near the bottom of the tablet, in what appears to be another canal, is the inscription “field of the town Similati” (eqil (A.ŠÀ) URUSi-mi-la-ti). This Similati is also shown on the side of the tablet, marked by a square. Clay conjectured that the writingeqil, “field of,” was a mistake for nār, “canal of,” though it could simply be a consequence of the inscription’s placement inside the lines representing the waterway instead of below them. In any event, the mention of a field here does not seem too incongruous. There are no measurements on this map, only relative positions. The reverse of the tablet does not seem to be inscribed, although a portion of the right-hand side is broken off. For what purpose this map was drawn is difficult to know, but it seems connected to the temple administrative documents making mention of the tax records from these same named districts outside Nippur, exploited for maintaining the E-kur temple during the Kassite period.


A Late Babylonian tablet (BM 92687), copied from an older original, contains a written description and a diagrammatic image of the world as conceived sometime after 900 (fig. 1.15).⁴⁹ Its provenance is uncertain, but its British Museum catalog number and the onomastics of the scribe who copied the tablet point to the city of Borsippa. The description written to accompany the map has an integral, but complicated, relation to the image (Horowitz 1988, 153). This world map is best matched with the category of medieval maps termed mappaemundi (see further chaps. 6 and 7below). It depicts a geographic totality that includes the local environment but is not limited to it. Circles represent cities (as in the Nuzi map), and parallel lines rivers. The largest demarcated area, shown as a rectangle on the upper central portion, is Babylon, the point of view from which the map was presumably made. Also on the map are the states of Urartu and Assyria, the latter written with the determinative KUR, “land.” The cities of Der and Susa and the territory of Bit Yakin are included. The regions, cities, and other geographic features such as the swamp and the water channel are all arranged inside a circle bounded by the waters of the ocean, designated as the “Bitter River” around the entire encircling band. Medieval mappaemundi typically show the world encircled by the “ocean sea” (Mare Oceanum) within a spherical world picture; equally, the “ocean river” (Alveus Oceanus) is sometimes shown as dividing the land zones (Woodward 1987, 300). For want of intermediary evidence, one can only note the similar cosmographical elements of the Babylonian world map without drawing conclusions about possible continuities.

Beyond the salt waters surrounding the known cities and lands are large triangular areas (only five survive of the original eight) representing lesser-known faraway places. They could be reached were one prepared to travel 7 bēru, but here the ordinary features of the known world are exaggerated or disturbed. One of these distant regions is described as a place where “a winged bird cannot safely complete its journey” (Rev 8′; Horowitz 1988, 150–51). Another is home to horned cattle. A reference to the “Four Corners” of the world (kibrāt erbetti) in the last section of the text is followed by the phrase “whose interior no one knows.” The inscribed portions of the map are lamentably damaged and incomplete. Even so, because they have a decidedly mythological literary character, with references to heroes and kings such as Sargon, Nur-Dagan of Purushhanda, and Utnapishtim—the only man to survive the great Flood and attain immortality, well known from the Epic of Gilgamesh (George 2003)—we are for certain in an imagined cosmical landscape.

FIGURE 1.15 Babylonian map of the known world and distant lands. The British Museum, BM 92687. Reproduced with permission.

The question of the map’s orientation has been discussed by Eckhard Unger.⁵⁰ As already noted, the third-millennium Nuzi map is clearly oriented to the east (IM.KUR, “mountain wind”). On the Babylonian mappamundi a mountain is drawn at the top of the circle of the world. From the vantage point of Babylon, the mountain could represent the east. If so, however, then other places are not arranged accordingly; Assyria, for example, is placed below and to the right of Babylon (in fact, it lies to the north), while Susa, shown at the bottom, should not be in the west (in fact, its location is southeast of Babylon). Altogether, places do not appear to be situated exactly in accordance with any given orientation of the map.

The historical and literary references in the description accompanying the map resonate with the historical and cosmical atmosphere of its iconographic representation of the world. That representation’s collection of toponymns indicates a date of composition not before 900, but the extant copy is Late Babylonian and seems to have a Babylonian provenance. The inclusion of Urartu and Bit Yakin points to Neo-Assyrian history. The territory of Bit Yakin, located in the southernmost part of Mesopotamia, was independent of Assyrian rule for a time during the eighth century (721–710) through the political and military efforts of Mardukapla-iddinna II.⁵¹ Urartu, Assyria’s northern neighbor, was a frequent target of attack during the ninth and eighth centuries by kings such as Assurnasirpal II (883–859) and Sargon II (722–705). It was very roughly at this juncture, early in the first millennium, that the ideology of Babylon as the new Nippur, the center of the world and linchpin of heaven and netherworld was formed (Maul 1997). In addition to historical references, the map incorporates cosmical features such as the “Bitter River” surrounding the known land, and the triangular “regions” beyond the sea and the limits of empirical knowledge.


Noniconographic topographies attested in cuneiform—such as “The Sargon Geography” describing “the totality of the land under heaven” conquered by the “King of the Universe” (Horowitz 1998, 67–95), or the five-tablet compilation entitled “Tintirki (=) Babylon, on Which Fame and Jubilation Are Bestowed”—provide detailed information on lands and regions as well as descriptions of temples, shrines, streets, walls, and rivers. Awareness of great distances and of the diversity of terrain in regions far from the cities of the Mesopotamian heartland is also preserved in the form of itineraries and surveys of the topography of places traversed by trading caravans or military campaigns (Millard 1987, 107–8). But the motivation for recording descriptions of faraway or even local places, or routes and distances between places, seems not to have been to produce a geographic “road map.” While these texts are not maps, they do perhaps reveal an affinity with them and with the claims to political, economic, or cosmological power from which they derive. Lands are known and then written about because they are either conquered or traversed for economic or military purposes; or, conversely, they are epitomized as the centers of power and of all creation.⁵²

Conquest, the establishment of overland and long-distance commerce, and even the elevation of a certain city to the role of “center” (as was the case first with Nippur then with Babylon), while all quite different phenomena, are nevertheless various manifestations of power relations between rulers (or traders) and the land. The dynamics of power result in the imposition of various forms of order or organization upon the land and its inhabitants. This imposition as confirmation of the exercise of power is the implicit substance of literary and iconographic representations of those conquered or otherwise controlled lands that merit such representation. Economic power had its direct and physical consequences in the creation of various hierarchies of wealth and control over resources. Such economic and political control shaped the land and the landscape; it in turn made possible the building of towns, cities, roads, and watercourses to irrigate fields for further development of economies. Consciousness of these processes is manifest in the geographic texts as well as in mapmaking.

A Larsa-period itinerary, dated by internal criteria to the reign of King Rim-Sin of Larsa (Hallo 1964, 85), a contemporary of Hammurabi of Babylon, seems to describe a historical “forced march” of nearly three months from the city of Larsa to the city of Emar on the Euphrates River in Syria, detailing layovers, delays, and detours. However, such travel was not accompanied by mapmaking, nor evidently did it require any. After all, a journey from point A to point B neither involved nor demanded a process of fixing certain ordered structures upon the landscape; hence, it did not lend itself to Wood’s “iconic coding.”

The “Sargon Geography,” a literary creation preserved in two tablets from the first millennium (one Neo-Assyrian, the other Late Babylonian), enumerates the lands conquered “three times” by the third-millennium ruler Sargon of Akkad and gives their relative sizes. Micellaneous ethnographic information about the peoples of conquered territories is included, such as the comments that the Amorites and the Lullubu “do not know construction,” and the people of Karzina “do not know burial,” but cremate their dead (Horowitz 1998, 91–92). The boundaries of the lands are defined, and the “circumferences” (talbātu) of major land areas measured (Marhaši, Tukriš, Elam, Akkad, Subartu, Amurru, Lullubi, Anšan). Lands beyond the Upper and Lower Seas are named; together with the measured territories they compose “the sum total of all the lands.” The literary character of the text makes it difficult to assess the historical geographic claims made about the extent of the empire. In general, the place-names are arranged in sensible geographic contiguities, with most places lying between the Euphrates and the Tigris, or east of the Tigris. What appears ostensibly geographic, however, is really more a declamation of the lands subject to Sargon’s power.

III. The Celestial Landscape

Images of the heavens projected onto the flat surface of a “map” and showing the totality of the starry skies were unknown in the ancient Near East. There was no method by which the arrangement of any celestial objects would be projected onto a flat plane. Stereographic projection requires the notion of sphericity, from which—using mathematical rules—points on a flat surface are plotted to correspond to points on a spherical one, be it earth or heaven. Our evidence argues rather forcibly against the notion that the spherical cosmos had any importance in Babylonian thinking. Thus any representation of stars or star groupings has other purposes.

As represented in Sumerian and Akkadian mythological poetry, heaven (divine AN) was paired with “earth” (divine KI= erṣetu), that is, everything that lay under heaven from the surface of the inhabited earth to the underworld, including the subterranean sweetwater (apsû) and the land of the dead. As such, AN and KI were the two principal and inseparable parts of physical space. Consistent with this idea of heaven and earth as counterparts, spatial relations were expressed by means of an extension of terrestrial metrology to the celestial. As mentioned above, the units for measuring time as linear distance were the bēru, “double-hour” (approximately 10 km), subdivided into 30 UŠ, “degrees” (each approximately 4 minutes of time). Marvin Powell (1987–90, 467) is surely right to explain the relation of terrestrial to celestial measurement as a consequence of the probably archaic tendency to measure long distances on earth by means of the perceived movement of heavenly bodies corresponding to the time taken by humans to travel so far.

In nonscientific contexts the bēru was often used simply as a means of denoting great distance, describing either overland or celestial spaces that were manifestly not in fact measured, as in this inscription of King Esarhaddon of Assyria: “Bâzu, a district located afar off, a desert stretch of alkali (lit., salt-earth), a thirsty region: 140 bēru of sand, thorn-brush and ‘gazelle-mouth’ stones, 20 bēru of serpents and scorpions, with which the plane (sic) was covered (lit., filled) as with ants, 20 bēru (through) Mt. Hazû, a mountain of saggilmut-stone, (these stretches) I left behind me as I advanced (thither, i.e., to Bâzu) where, since earliest days no king before me had come.”⁵³

The mythological tale of Etana, a man who ascended to heaven with an eagle, likewise describes travel in terms of bēru, as in the itinerary genre. Etana and the eagle fly 3 bēru up into heaven beyond the reach of the “Heaven of (the sky god) Anu,” passing through two gates to reach the house of Ištar. As they make their ascent, the land is viewed from above and described as a garden which also has agricultural features like an animal pen and an irrigation ditch. The sea is said to encircle the land, but when they have reached their goal, 3 bēru up, land and sea disappear from sight: “[He li]fted him up a third double-hour: Look my friend, the land, what is it [like?]. I am looking, but I do not see the land, and my eyes do not feast on the broad sea” (Horowitz 1998, 56–57, lines 38–41).

In traversing 1 bēru (= 30 NINDAN, “rods”) overland, therefore, the sun likewise traversed 1 bēru (= 30 UŠ, “degrees”), or one-sixth of the sky. Celestial movement—that is, distance in heaven—indicated time. The length-time system was based on an ideal day (UD) reckoned as 12 bēru. Powell (1987–90, 467–68) noted that the correspondence between the 12 bēru of the day and 360 UŠ on the one hand, and the twelve ideal months of 30 days each and 360 ideal days, on the other, was probably no coincidence. However, as he made clear, the link between geodetic and celestial length-time measure had no influence on the determination of the metrological standard equivalents for the terrestrial unit DANNA/bēru, which among other factors was considered to be the length of a field a team of oxen could plow in one day.

Measurement of time by means of celestial distances is seen in connection with a group of stars known as “zenith (ziqpu) stars” that “culminated”—that is, crossed the meridian—near to the zenith, around latitude 36° north (Nineveh). A Late Babylonian text (Textes cunéiformes du Louvre6 21) gives a list of successive culminations of a fixed sequence of such stars, beginning and ending with MUL.ŠUDUN, “The Yoke Star” (= Boötes). This list gives the intervals between the ziqpu stars in mana, “mina,” UŠ ina qaqqari, “degrees on the ground,” and bēru ina šamê, “celestialbēru.” The degrees “on the ground” seem to correspond to degrees of stellar motion. Because the fixed stars rise and set in arcs parallel to the celestial equator, the rising, setting, or culmination of equal arcs between these stars rise, set, or cross the meridian in equal times, and thereby make it possible to reckon the passage of time by means of those intervals, or stellar “distances.” The equivalence established between the units for the measurement of stellar distances in this system was 1 mana = 6 UŠ ina qaqqari = 10,800 bēru ina šamê.⁵⁴

Shorter apparent distances were sometimes designated by the cubit, subdivided into 30 fingers. The cubit had an astronomical application for measuring distances in the heavens between fixed stars and the meridian, for example, or between planets and ecliptical stars, as well as for measuring eclipse magnitude. The equivalence 1 cubit = 24 fingers = 2 UŠ (“degrees”) gives us 1 finger = 0;5° and 1° = 12 fingers. The cubit is used in two of the earliest observations recorded in the Almagest (9.7), from years –244 and –236. Ptolemy cites (1) Babylonian eclipse reports, giving time at which eclipse begins, statement of totality, time of mideclipse, and direction and magnitude of greatest obscuration in digits, in the manner of cuneiform eclipse reports, (2) distances in cubits from ecliptical norming stars (Normal stars) at dawn for Mercury, the dates for which are given in the Babylonian calendrical system of lunar months (translated into Macedonian month names) and years in the Seleucid era, and (3) the distance of Saturn in digits from a Normal star in the evening.⁵⁵

In Babylonian astronomy the concerns of celestial measurement had to do either with distances between the moon and fixed stars, or the moon and planets, in observational records, or with what we would call “longitudinal” distances along the ecliptic of planetary or lunar phenomena. In the latter case, however, longitude is not measured, but calculated in degrees of distance within the circle of the zodiac in accordance with algorithms for finding such distances. The ziqpu texts mentioned above, in which distances between fixed stars are quantified, also establish spatial relationships in the heavens. Even so, these interests in distances and relations between celestial positions fall within the boundaries of astronomy rather than what we would call star mapping.


The so-called astrolabes constitute a fundamental corpus of early Babylonian astronomical texts. The earliest exemplar stems from the Middle Assyrian period (the reign of Tiglath-Pileser I, 1115–1077). It is probably already a copy from an earlier, perhaps Old Babylonian source, and the text was still being copied in the Seleucid period (third century or later). The term “astrolabe” is a misnomer insofar as the cuneiform exemplars are not planispheric. However, they do map out fixed stars, constellations, and even planets in various parts of the sky for the twelve months of an ideal year (360 days, i.e., twelve months of 30 days each), either by listing in groups, with the stars of the three roads set alongside one another, or by arranging them in rings in a concentric circular diagrammatic form, three stars per month for a total of thirty-six stars.

The groups of stars are defined by their locations with respect to the horizon, and in a rough sort of distribution into three “roads” (harrānu), differentiated by declination.⁵⁶ The stars closest to the equator rise in their assigned month in the road of Anu, those to the north in that of Enlil, those to the south in that of Ea. These divisions of the sky, named for the three great cosmic deities, are also widely attested elsewhere, in the astronomical compendium MUL.APIN (discussed below), in prayers, scholia, and other star catalogs. The monthly rising of an astrolabe star represents the important reappearance of the star following its period of invisibility. The reemergence of the star in the sky is a seasonal event, with respect both to its date and its place of rising. The astrolabe text selects thirty-six heliacally rising stars (for fixed stars near to the ecliptic the heliacal rising is the first appearance in the morning following the seasonal period of invisibility when in conjunction with the sun), twelve in each of the “roads.”

The question of the purpose of the circular diagram of the astrolabe text arises in the context of celestial mapping. The same elements are presented in the circular diagram as in an alternative list form of the astrolabe. In general, the mapping function of these texts derives from their attention to the association of fixed stars and their “roads,” and from the fact that the entire sky over the course of a year is taken into account. As Johannes Koch has pointed out (1989, 120), the circular astrolabe was not an astronomical device, but rather an orientation guide and an aid to remembering which stars appear in which parts of the sky.


There survive two disk-shaped clay tablets with evidence for celestial topography: one is a seventh-century Nineveh library text showing constellations inside a circular arrangement of eight 45-degree segments (Koch 1989), and the other is a Neo-Babylonian tablet from Sippar showing the zenith, or ziqpu, stars in a circular arrangement of twelve 30-degree segments (Horowitz and al-Rawi 2001). The former has been dubbed the “planisphere” (K. 8538, King CT 33 10), and it alone provides iconographic representations of constellations. The only features common to both are that they are disk shaped, they divide the circle of the sky into equal parts, and they make use of dots to represent stars.

FIGURE 1.16 Redrawing of the seventh-century-BCE Nineveh planisphere, depicting constellations with their names. The British Museum, K. 8538. Reproduced from L. W. King, Cuneiform Texts from Babylonian Tablets in the British Museum, pt. 33 (London, 1912), plate 10.

The Nineveh planisphere (fig. 1.16) is slightly rounded on the reverse and inscribed only on the flatter side, or obverse, which has a slight rim or raised edge. It is not truly comparable to what is today considered a planisphere, a device that shows the sky in the different seasons of the year and different times of the night by the rotation of one disk that represents the horizon against another disk that projects the stars onto a circular plane. The Nineveh planisphere shows “connect the dot”–style figures in six of eight sectors (fig. 1.17). These figures represent constellations and are identified by name on the tablet. Beginning with sector 0, as indicated in the center of the diagram, and running counterclockwise, is the arrow figure inscribed as Ištar and Dumuzi. In his interpretation of the planisphere, Koch reasoned that because Ištar is associated with the Bow Star, Dumuzi must be the Arrow. In sector 1 is an image with two triangular figures, labeled muliku, “The Field,” and mul apin, “The Plow,” the combination of which makes sense; the Plow is given an epithet, “who goes in front of the stars in the road of Enlil.” Sector 2 has no star name preserved. In sector 3, containing an ellipse with two enclosed wedges, is written mulIs lê, “The Jaw of the Bull,” and mul MUL, “Pleiades,” as well as Sipazianna, “Orion.”

FIGURE 1.17 Reconstructed drawing of the Nineveh planisphere. Reproduced from Johannes Koch, Neue Untersuchungen zur Topographie des babylonischen Fixsternhimmels (Wiesbaden: Harrassowitz, 1989), p.112, with permission of the publisher.

Next, sector 4 contains a figure with three dots in a row and a long pointer with a wedge-shaped tip at the end, pointing directly to the middle of the circle. It gives the name of the twin stars known as “The Twins who stand in front of Orion.” They are known from a list of seven divine twins (7 mâšu)⁵⁷ that includes other sets of twins in the constellation Gemini, namely, MULMAŠ.TAB.BA GAL.GAL (who are the evil Meslamtaea/Nergal and Lugalgirra/Sin) and MULMAŠ.TAB.BA TUR.TUR. Koch has identified these with γ and ε Geminorum. Inscribed in this sector 4 too is the divine name Lātarāk, also a destructive figure, associated with the Asakku demon. Sector 5 mentions Regulus, “The King,” and Corvus, “The Crow,” written inside one of the small pointed figures. Inside another pointed figure is inscribed the sign BE, whose reading here is uncertain; Koch identifies this latter figure with mulAB.SÍN, “Libra.” Finally, in sector 6, a balance figure is inscribed with the name of Libra, written mulGI.GI (= mulZI.BA.AN.NA = zibānītu).

The other planispheric text comes from Sippar. It is inscribed on both sides.⁵⁸ The obverse of the tablet shows parts of twelve 30-degree segments indicated with straight lines radiating out from a central rosette. The names of ziqpu stars and arrangements of dots are preserved in six of the twelve segments. The ziqpu star names show that the disk is to be read in a clockwise direction. As labeled in the diagram (fig. 1.18), segment 1 is the “Hand of the Crook,” without dots preserved, and “The Twins” (Nabu and Nergal), shown with two dots. Segment 2 is “Crab” (ten dots) and “The Two Stars of the Head of the Lion” (two dots). Segment 3 has “The Four Stars of his Chest” (of the Lion, four dots) and “The Two Stars of his Tail” (two dots). Segment 4 gives “The Single Star of Its Tail” (one dot), “The Frond” (six dots), and “The Harness” (one dot). Segment 5 is “The Yoke” (two dots) and “The Rear Harness” (three dots). Segment 6 shows “The Circle” with a circle of dots. The circular arrangement of the ziqpu stars here on the obverse of the disk corresponds to the list of these stars on its reverse. The list there in turn seems to parallel the astrolabes’ presentation of stars both in lists and in circular diagrams.

FIGURE 1.18 Redrawing of a ziqpu star disk representing the number of stars in a group of constellations. Reproduced from W. Horowitz and F. N. H. al-Rawi, “Tablets from the Sippar Library IX: A Ziqpu-Star Planisphere,” Iraq 63 (2001): 177, with permission of the publisher.


An important astronomical compendium of the early first millennium is the two-tablet series entitled MUL.APIN, “The Plow Star” (Hunger and Pingree 1989). Like the geographic itineraries, its function is to map major celestial objects, their locations on certain “roads,” and their relative positions toward a kind of practical astronomy. By contrast to the geographic itineraries, however, interest in describing the landscape of the sky was not related to acts of political, economic, or military power. At the same time, the calendrical concerns underlying many of the astronomical subjects of interest in MUL.APIN demonstrate that its motivation was not merely disinterested description of “natural phenomena,” but rather an understanding of the environment together with some control of it. Further confirmation that a desire for understanding and control was at stake in the preparation of this astronomical compendium emerges from the sections of the text that provide rules for prediction relevant to a fixed calendar, such as the intercalation rules established by coordinating the date of the heliacal rising of the Pleiades with the first visibility of the moon, or indeed, the intercalation rule given with reference to the appearance of Sirius (Tablet II.i 22–24.). MUL.APIN deals with schemata and its calendrical underpinnings are schematic (the year of twelve 30-day months), as are its catalogs of stellar risings and settings, as well as its intercalation schemes. It presents a practical astronomy useful for celestial divination, whose general attitude about heavenly phenomena it shared, namely, that these were significant to know and understand: the gods had “produced” the appearances to signal events in the future, and the knowledge gained was to be used by human beings to enhance their security in the world. This belief made it desirable to identify periodic phenomena and to devise convenient ways of predicting them, such as by numerical schemes.⁵⁹

About sixty-six stars and constellations (including the five planets) are classified in this text in accordance with the arcs or pathways in the sky—termed “roads,” as mentioned above—which describe their risings and settings. Three such “roads” are demarcated, and named for the gods Anu, Enlil, and Ea. In modern terms, the Road of Anu is reckoned as the arc over the horizon, the “cattle pen,” where stars relatively close to the celestial equator, approximately 15 degrees ± declination are seen to rise; the Road of Enlil is to the north and actually includes the circumpolar stars; the Road of Ea is to the south. The Plow Star’s celestial itinerary designates the Road of Enlil as the head of the cattle pen, the Road of Ea as the foot. Given the knowledge that the fixed stars rise and set over fixed points along the horizon with seasonal regularity, the Plow Star also lists the simultaneous rising and setting of constellations and fixed stars in the three roads, season by season. Moreover, because it was understood that the planets do not appear in the same place with respect to the eastern or western horizons at their first and last visibilities, a rough estimate of the intervals of visibility and invisibility of planets was made. Another feature of the visible heaven included in MUL.APIN is a catalog of stars known as ziqpus, which are seen to cross the meridian and thereby are useful for telling time at night (Hunger and Pingree 1999, 84–90). MUL.APIN makes an association between twelve ziqpu’s crossing the meridian before sunrise in midmonth and the heliacal risings of certain constellations.⁶⁰

Without representing the observable features of the night sky iconographically, MUL.APIN described and presented in a systematic fashion the chief routes taken by stars and planets across the celestial landscape. As an “itinerary” of the heavenly region in the form of a practical astronomy, MUL.APIN is functionally parallel to the itineraries on the ground in the form of practical geographies.

IV. Conclusion

Although cuneiform maps may not be forerunners from which later Western maps originate, they share characteristics with other cartographic traditions in their graphic imaging of territorial, social, and cosmological space. Cuneiform texts provide several varieties of evidence for the ancient Mesopotamian efforts to express order by describing, delimiting, and measuring the heaven and earth of their experience, producing house, temple, plot, and field plans, city maps, and, with respect to the celestial landscape, diagrammatic depictions of stars. Various orders of power are implicit in the expression of these aspects of order in the environment. Administrative and economic powers support, or even require, the making of maps, as well as determining overtly the topographies that maps depict.

Where once such maps would not have been admitted within a general history of cartography, a new view of the meaning of the map can embrace them. The historiography of maps and cartography has emerged from criticisms similar in nature to those made against the modernist or presentist historiography of science, namely, that in reifying science or sciences such as cartography, false evolutionary histories are liable to be constructed. Some originating point is identified, such as the origins of science in Greece, or of mapmaking in Babylonia, from which a continuous history may be written from a presentist perspective, a tale of a discipline’s inexorable progress from its originating moment to the present. Critical cartographic history, however, has laid aside such ideas, and we no longer look to (in the words of Denis Wood), “a hero saga involving such men as Eratosthenes, Ptolemy, Mercator, and the Cassinis, that tracked cartographic progress from humble origins in Mesopotamia to the putative accomplishments of the Greeks and Romans” (Wood 1997, 549).

By no means do all ancient Near Eastern maps display metrological finesse or even the use of measurement, though some characteristically do, such as the agrarian field and urban plot cadastral surveys. Concern for orientation is attested in a number of maps, but not always in the same way, although with a tendency toward an oblique orientation northwest to southeast. Ancient Near Eastern maps may not have invariably been meant as exact or direct replications of territory, but there can be little doubt that they distinctively reflect the conceptual terrain of their social community and culture at large. The maps of buildings and fields focus on the urban and agricultural environment, matters of critical importance to whatever political and economic powers prevailed.

The maps of cities with their waterways and surrounding physical landscape combine cartography of sacred space, seen in the temple plans, with that of economic space, seen in the field surveys. The cities of Nippur and Babylon had a religious and cosmological function as well as a political and economic one. In the periods of their supremacy each was viewed as the center of the universe, as the meeting ground between heaven and the netherworld. The map of the principal temple in Babylon, E-sagil, which was the earthly abode of the national deity Marduk, represents the terrestrial counterpart to the celestial residence of the great god Enlil, designed, figuratively speaking, on the blueprint of the cosmic subterranean sweet watery region of the Apsû (Enūma Eliš IV 143–46).

The Babylonian world map is an attempt to encompass the totality of the earth’s surface iconographically: land, ocean, mountain, swamp, and distant uncharted “regions.” This said, it represents more of an understanding of what the world is from the viewpoint of historical imagination than an image of its topography against a measured framework. It offers a selective account of the relationship of Babylon to other places, including those that were at the furthest reach of knowledge.

The diversity of cultures that have sought to preserve their maps, putting them on clay, papyrus, parchment, and other writing media, points to a near universality of making maps in human culture. Cognitive psychologists claim that we come into our physical world mentally equipped to perceive and describe space and spatial relationships.⁶¹ The linguistic act of spatial description is perhaps a protomapmaking function of our very desire and attempt to place ourselves in relation to the physical world. By extension, we should not doubt that mapmaking too, in all its historical subjectivity, is a universal feature of human culture. The interest of the cuneiform maps lies in their rich articulation of such a feature, uniquely shaped by the particular social norms and forces that emerged and changed within ancient Mesopotamian history.



Antiquités Orientales


The Assyrian Dictionary of the Oriental Institute of the University of Chicago


Cuneiform Texts from Babylonian Tablets


Kuyunjik Catalogue, British Museum siglum

If you find an error please notify us in the comments. Thank you!