Mereological themes in cuneiform worldmaking

Introduction

It is a well-known characteristic of Babylonian astronomy that not only did its predictive methods function independently of a physical spatial framework, but its epistemic goals did not include understanding how the world worked as a physical or mechanistic system. If neither a physical structure for the celestial phenomena nor the aim to grasp some structural unity is present in astronomical cuneiform texts, the question of a world conception readily presents itself.

This characteristic of the late Babylonian mathematical astronomy of the second half of the first millennium B.C.E. differentiates the tradition fundamentally from that of its closest historical relative, Hellenistic Greek astronomy, which developed as a direct result of contact with Babylonia. The Greek cinematic form of astronomy is intrinsically tied to a cosmological framework, namely, the spherical geocentric arrangement of the planetary spheres. Babylonian astronomy did not embed its mathematical methods within a cosmos thought to govern the very phenomena the mathematical methods sought to describe. Consequently, we are unable to deduce a cosmological framework from cuneiform astronomical texts.

In order to take account of ideas about a world structure in cuneiform texts, it is necessary to disconnect from the astronomical corpus and focus instead on other sorts of texts, e.g. literary narrative texts, divinatory compendia, divine epithets, names of temples, and royal inscriptions. These afford a picture, if fragmented and partial, about how the world had divine order and organization. These sources do not afford easy access to a unified cosmology either, but in their various descriptions of parts of a world, they offer clues to a way of worldmaking.

In adopting an approach to the notions of world and worldmaking for the cuneiform context, I am borrowing directly from Nelson Goodman's Ways of Worldmaking (1978). One of the more interesting, and puzzling, aspects of the cuneiform evidence to be examined here is that rarely is reference made to the world as an integral whole, or a place with explicit boundaries. Not only does this problematize the question of a spatial framework in the Babylonian conception, but also the very question of what ancient Assyrian or Babylonian reference the modern word ‘world’ really has.

If cuneiform culture's way of worldmaking comes under the rubric ‘science in the past,’ it underscores the discontinuity between its native tradition and that of the European natural sciences that repudiated and eliminated supernatural phenomena as not belonging to the world science investigates. The idea of the ‘external world,’ which came to be tied closely to science, has no counterpart in the knowledge tradition of the cuneiform texts. In ancient Assyria and Babylonia a long-standing intellectual tradition aimed to collect and organize knowledge of the phenomenal world. To this tradition belong divinatory texts as well as astronomy, incantation series as well as therapeutic medicine. Where we often go wrong is in equating the phenomenal world of interest to this tradition with phenomena of ‘the external world’ conceived and explained by modern science, or that the phenomena belonging to the system of reference in cuneiform knowledge should be evaluated in terms of phenomena of an external ‘reality’ of nature. Instead, the phenomena referred to in cuneiform knowledge comprise an inclusive category of things not limited to natural objects or events. If science is implicated in the basic process of finding order and intelligibility in the phenomenal world of experience and imagination, then the present paper, focusing on the relationship between people and their worlds, their immediate tangible as well as imagined surrounds, concerns itself with science.

As far as the ‘science of the forest’ is concerned, the relationship between peoples and their worlds, between world and worldmaking, is obviously a fundamental of anthropological interest. Joanna Overing (1990), in an inquiry into the Amazonian shaman as a maker of worlds, embraced Goodman's sense of worldmaking as a productive mode of anthropological understanding. She said,

the processes of worldmaking … followed in the West and in the jungle are much akin to one another. The scientist, artist, myth teller or historian, and shaman-curer are ‘doing much the same thing’ in their construction of versions of worlds. However, while the thought processes for constructing worlds are in many ways similar, the facts of which these worlds are made are very different indeed. For instance, in the jungle a world version may comprise angry creator gods and translucent streams of madness, rather than force fields of energy or atoms and molecules. (Overing 1990, 603)

Following Goodman's claim that, ‘our universe, so to speak, consists of these ways [of describing] rather than of a world or of worlds,’ I propose to begin with descriptions of the world attested in scholarly cuneiform texts, principally from mythologies of creation, hymnic texts in praise of deities, and royal inscriptions from Neo-Assyrian kings (7th century BCE). By focusing on their ways of describing it becomes possible to discern a complex of themes that point to relationships established between major interrelated parts of the world, such as the regions of the heavens, earth, the netherworld, and the abyss (Apsû).

Prominent among these interrelated themes are complementarity, correspondence, and the relation of counterparts. Meronyms, that is, names for parts of the one whole (we would say ‘world,’ or ‘universe’), and the part-whole relationships they indicate, are of importance to this inquiry into worldmaking that is based on words, terms and descriptions and the things/world parts to which they refer. In looking at meronyms, or names for parts of the whole, it is interesting to note that the expected holonym meaning ‘world’ in fact does not emerge. Instead we see a variety of words for ‘totality, all, or entirety (of parts).’

In all cognizance of the variation in descriptions and representations, with especially wide variation diachronically, this paper pursues a thematic analysis from Akkadian meronyms (names for parts) and meronoms (the parts themselves) in order to understand how mereological world-parts, such as they are described, were set in relation to a whole. It further raises the question of how, or in what way, out of the many terms for ‘all of x (parts),’ there was a notion of the whole, as that is an essential part of meronomy, which deals with part-whole relations in contexts where there is a set of parts forming a whole.

Worldmaking in cuneiform

The present investigation of worldmaking in the cuneiform culture of the first millennium B.C.E. proceeds from descriptions found in literary and scholarly texts. In Goodman's words (1978, 2–3): ‘If I ask about the world, you can offer to tell me how it is under one or more frames of reference; but if I insist that you tell me how it is apart from all frames, what can you say? We are confined to ways of describing whatever is described.’ By virtue of such descriptions we establish a relationship between frames of reference and systems of description. How the first millennium Babylonian literati understood the structure and composition of their world is described in mythological narratives (see Horowitz 1998; Katz 2003; Lambert 2013). The description of the world, or its parts, found in these sources construct a world version true to its frame of reference.

The literary sources for world-description are paramount, but what is not so obvious is how far those conceptions of the world and its component parts extend into non-literary sources. The question, then, of the relationship of world-versions is problematic. Mythological narratives and god-lists of cuneiform culture have been approached as representative of the rudimentary beginnings of the history of cosmology, from the standpoint that historical systems can be assessed by comparing them to the ‘right rendering’ of the world of modern science. In the case of ancient Mesopotamia, the world-version attested in the stories of gods and the Creation was once viewed as going nowhere, as in the following statement by W.G. Lambert (1975, 49):

In many ways it is a disappointment that the civilisation which produced so much in the collection of data and in the abstract sciences did not develop its cosmological ideas during the 3,000 years of its existence, and the reason for this can be sought. Two main factors were at work. One was the old, prehistoric mythological thinking, which, despite the fictional element in its products, was properly based on sound observation of the universe. This could have provided the basis for a scientific cosmology if conditions had been right. The other was the personification of natural forces into anthropomorphic gods and goddesses, and this led away from the realities of nature into theological fantasy. Unfortunately the growth of civilisation resulted in a diminution of the first of these two factors and an increase in the second.

The imagery of world-parts such as the watery Apsû or ‘Abyss’ (home to the god Ea) and the imagined Ešarra (home to the god Enlil) was not of interest to mathematical (or observational) astronomical texts, which as mentioned before, operated independently of a spatial cosmology. References to world-parts are found in texts that relate directly to the gods, such as in a late Assyrian prayer to the god Marduk, where the major constituent parts of the world are named as entities that should ‘witness’ that god's deeds:

May all the gods and goddesses, Anu, [Enlil], the constellations (lumāšu), the Abyss, the netherworld … behold the deeds of the lord of the gods, Marduk. (līmurū epšet bēl ilī Marduk … lumāši apsû daninnu Livingstone 1989, 2:36–37).

When it comes to the question of world-structure (Lambert 1975, 2013; Horowitz 1998; Geller 1999; Katz 2003), the key text is the poem Enūma Eliš ‘When Above.’ The text was recovered in 1849 when Austen Henry Layard found the clay tablets containing Enūma Eliš in the ruins of Nineveh. In 1876 George Smith brought out an English edition, and in 1890 Peter Jensen's Die Kosmologie der Babylonier set the text within the wider scope of cosmology. At the start of the twentieth century, the British Museum published L.W. King's cuneiform copies of the text in volume 13 of Cuneiform Texts from Babylonian Tablets in the British Museum (1901). King's edition and translation, titled The Seven Tablets of Creation, or the Babylonian and Assyrian Legends concerning the creation of the world and of mankind, appeared in rapid succession in 1902. The authoritative edition appeared in Lambert 2013. The poem, ‘When Above,’ is perhaps the finest example of cuneiform literature and scribal scholarship, not only in terms of its high poetic style, but also and notably its scholarly mastery of cuneiform orthographies and hermeneutics. It belongs at the centre of any discussion of the Babylonian conception of world structures according to first millennium scribes.

When we look at the narrative of creation from Enūma Eliš Tablet IV, we see that the story is principally a moral one, and yet there are the world-structural components explained by the narrative. Marduk, the hero, slays the goddess Tiamat, whose body will become the heavens. In killing Tiamat, Marduk vanquishes evil, treachery, and improper divine rule. Enūma Eliš accuses Tiamat of having contempt for mercy, for appointing her own spouse, Qingu, to ‘the rank of Anuship’ (the highest divine station) though he had no right to that office. She is further accused of giving Qingu the Tablet of Destinies, 1 though he had no right to hold it. She is charged with stirring up trouble and perpetuating evil against the gods. In the battle Marduk is aided by an incantation for undoing evil (Akkadian tû, Sumerian TU₆) as he approaches his foe: (Enūma Eliš IV 60-62) ‘He set his face toward the raging Tiamat. In his lips he held a spell (tû) and in his hand he held a plant to counteract (Tiamat's spell?).’ Tiamat as well, when she found she was up against the redoubtable Marduk, first turned into a raving maniac (Enūma Eliš IV 88 mahhûtiš ītemi ušanni ṭēnša ‘like one insane she lost her mind’) and then repeatedly recited an incantation to protect herself (Enūma Eliš IV 91). The moral victory was complete when Marduk rounded up Tiamat's divine helpers, smashed their weapons, threw them all in a net and confined them to prison (Enūma Eliš IV 111-114). The culminating triumph was when the hero Marduk wrested the Tablet of Destinies away from Qingu, whose possession of it had been an infraction against divine propriety (Enūma Eliš IV 121).

When after the Babylonian Chaoskampf the creation of the world is finally mentioned in Enūma Eliš IV 135-136, the god rests and surveys ‘the corpse’ of Tiamat, which is also referred to as a kūbu ‘lump’ (of flesh, usually in reference to a stillborn or premature fetus, or a monstrosity of some kind). Having then fashioned the heavens from half of her tightly stretched body, and keeping her waters from escaping by installing watchmen, he proceeded to make heaven as a counterpart measured to the size of the subterranean watery Apsû. Finally, he rectified the order of divine propriety in setting up the shrines where the high gods were to have their residences (Enūma Eliš IV 141-146). The major parts of the world that are created in order to house these deities and set the world aright are the heavens (home to Anu), Earth (as Marduk's new residence), the Apsû (home to Ea), also called Ešgalla, and Ešarra (literally ‘House of the All’) for Enlil.

Enūma Eliš provides an authoritative picture of the creation of the world, but variations in world-structural imagery are also attested (Rochberg 2020), just as in the Classical and Greco-Roman worlds more than one cosmology took shape (Pythagorean, Milesian, Platonic, Atomist, Aristotelian, Stoic, Ptolemaic). Before the various philosophical traditions pursued inquiries into the nature, structure, and material constituents of the cosmos, notional underpinnings of the word kosmos (κόσμος), as Jaan Puhvel's etymological inquiries revealed, are derivable from Homer as well as from Cretan inscriptions (Puhvel 1976, 154–57). These contexts, he said (1976, 154), ‘point to a notion of ordering, arraying, arranging, and structuring discrete units or parts into a whole which is ‘proper’ in either practical, moral, or esthetic ways.’ The semantics of order, ritual procedures, and cultic correctness are similarly reflected in the Akkadian word parṣu, used by the scribes to translate Sumerian GARZA, a word frequently associated with Sumerian ME, whose uses go back into third millennium sources. ME's semantic range covers concrete to abstract things, from divine attributes and offices, to rites and norms, but also including drawings or emblems, presumably for the abstractions ME can reference (Hrůša 2015, 35–38). 2 Akkadian parṣu can also be used to mean emblems, symbols, and insignia of abstract things such as divine authority, the power of kingship or other office (CAD parṣu mng. 4).

The significance of this complex of words and meanings is that before kosmos took on the meaning ‘world, universe’ with ontological implications, there was a more active sense of the order and harmony established by correct ritual procedure and the rightful carrying out of divine (or royal) offices. Certainly in cuneiform literature, the order and stability of the world was expressed in terms of its functioning rather than its material form.

System of world description: Meronyms and meronomies

In the fields of semantics and pragmatics, according to Alan Cruse, a meronomy is a ‘type of branching hierarchy’ of the ‘part-whole type’ (Cruse 1986, 157). The prototype for meronomies is the division into parts of the human body, where the whole in that case is the ‘body’ occupying the apex of the hierarchy and underneath are parts such as ‘head,’ ‘arm,’ etc. These, then, have their own parts that branch underneath, e.g. ‘nose,’ ‘hand,’ and so on. Meronymy, on the other hand, considers, as Cruse put it (1986, 159), ‘the lexical items used to designate parts and wholes, and the semantic relations between them.’ Expressions for part-whole relations might, then, point to ways that the ancient cuneiform scribes considered questions such as, what it means for a thing to have parts, for something to exist in the world, how existing things and their parts are classified, and the mereological sums, or wholes, that result from such thinking about what we now call parthood.

In one important respect the cuneiform evidence makes such analyses extremely difficult. The first and most formidable stumbling block, in my view, is that the very idea of ‘matter’ is not within the sights of the scribes. The kinds of levels of existence we assume, such as the physical and the non-physical, not to mention the idea of matter itself, or the idea of a combination of formal and material properties (e.g. Aristotelian hylomorphism), are a world away. 3 Thus the ontological implication for ideas about material things, and what makes them divisible into parts, or composed of such parts, will be quite different from what generally comes of more standard Western philosophical mereologies, even of the ‘Neo-Aristotelian’ kind (Koslicki 2007).

It could rightly be asked, as a result, what kind of ontological system can be derived from such limited structural terminology as is available in cuneiform texts. These terms reflect ideas about how world-parts are arranged and structured, but not how the ‘nature of matter’ or of ‘material things’ determines that they should be composed of parts at all, or in what way. We find no talk of essences or essential properties because there was no interest in matter or material existence such as would give priority to physical things or their ability to cause other physical things to occur or change in the material world.

The focus on meronymy/meronomy, or the names of parts and the parts themselves, makes possible a comparison, or counterpoint, with more familiar approaches to understanding the cosmos from the Greek world. Even in otherwise quite divergent cosmological schemes, the part-whole relation is clear in the notion of ‘cosmos’ that underpins the particular Greek way of worldmaking exemplified in the one all-encompassing world that persisted into the later Christian and Ptolemaic version. 4 A case in point is Aristotle's definition, or rather three definitions (De caelo 1.9 278b9–21), of ouranos from the work On the Heavens. These are (1) the substance of the outermost revolution of the universe of the world and the seat of divinity; (2) the body containing the moon, sun, and some of the stars; (3) the body enclosed by the outermost revolution, i.e. the entire world, that is, the universe, ouranos. (Leggatt 1995, 88–89) The passage concludes with the remark that ‘the whole world which is enclosed by the outermost circumference must of necessity be composed of the whole sum of natural perceptible body, for the reason that there is not, nor ever could be, any body outside the heaven’ (Aristotle, On the Heavens, Bk A, Ch. IX, 278b, 10-25).

Another locus for the idea of a unified cosmos is found in Stoicism. Brooke Holmes (2019, 245) notes, ‘The conspicuous fact of cosmic sympathy compels us in turn, the Stoics insist, to accept the reality of less obvious things, such as the unity of the body of the cosmos.’ 5 She also observed (2019, 240) that the Stoics’

theorization of the cosmic organism enacts a scenario in which the idea of Nature as a totalized whole is taken to its absolute limit … In Stoic sympathy, we find a pious commitment to the unity, completeness, and perfection of the world together with a tarrying with its parts and its particulars.

A number of Akkadian words mean ‘all,’ ‘totality,’ or ‘entirety,’ but in each case the word is constructed with another noun in the genitive for a sense of ‘all of’ some set of things included within it or counted as parts of it. For example, a word meaning ‘all’ or ‘totality’ (kiššatu) is used precisely in the sense of ‘all of x,’ or ‘the totality of x.’ In these constructions kiššatu is found together with some other noun in the genitive, as in ‘all (of the) lands’ (kiššat mātāte), ‘the entire sky’ (kiššat šamê), ‘all mankind’ (kiššat nisī), or ‘all the gods’ (kiššat ilī). 6 Alone the term kiššatu connoted ‘the entire inhabited world,’ as in the well-known royal title ‘king of the universe’ (šar kiššati), but there kiššatu certainly has the sense of all known inhabited places, not ‘all’ in the sense of all that exists. Divine epithets for Marduk and Nabû use the phrase ‘all heaven and earth/netherworld’ (kiššat šamê u erseti) as a designation of worldwide power, but the meaning is conveyed by the construction of ‘all of’ with the merism of ‘heaven and earth/netherworld.’ 7

A notable aspect of the term kiššatu is that it can be written with the logogram ŠÁR, 8 which, as the word for the number thirty-six hundred, is further used to mean an indefinite large number, and thus in a certain sense, ‘totality.’ 9 Sometimes its meaning is ‘countless,’ 10 and so by extension ‘all.’ The latter, however, is limited to the expression adi šāri to convey a sense of ‘for all time’ or ‘everywhere,’ mostly in bilingual literary contexts. These expressions convey a sense of temporal as well as spatial ‘totality,’ as in the passage that refers to the radiance (melammû) of kingship that overwhelms ‘everywhere’ (adi šāri, see CAD s.v. šār mng 2).

In a note on the origin of the Greek term saros, Otto Neugebauer (1969, 141) said that Sumerian ŠÁR meant ‘universe or the like’ as well as the number thirty-six hundred. He explained the crooked path by which the number 3600 came to give its (Sumerian) name to the eclipse relation 223 synodic months = 242 anomalistic months, which is the basis for prediction of eclipses every approximately 18 years, or 6585 and 1/3 days as a standard value. From the point of view of usage, and available attestations, the Akkadian loan šār, however, does not fill the gap in terminology for ‘universe.’ On the other hand, the Sumerian word ŠÁR, represented in its earliest form graphically as a circle, may indeed have connoted something of an encompassing ‘all.’ As mentioned before, the name Ešarra (É.ŠÁR.RA), literally ‘house of the 3600,’ or ‘house of the all’ becomes a cosmic location as the abode of the Sumerian creator god, Enlil.

Another word (kullatu), similarly, means ‘all,’ or ‘totality,’ and is frequently constructed with a genitive of parts that can be counted, e.g. ‘all the gods’ (kullat ilī), ‘all peoples’ (kullat nišī), or ‘all cities’ (kullat māhāzi). In extended meaning there is ‘the whole of all wisdom’ (kullat nagbi nēmeqi), and, perhaps most general in semantic force is the construction with the idiom ‘everything, anything,’ literally ‘whatever its name’ (mimma šumšu), to mean ‘all of everything.’ Esarhaddon says of the god Ea, lord of wisdom, that he was ‘the one who fashioned everything, whatever its name’ (pātiq kullat mimma šumšu, Leichty 2011, 104:4). In this case, ‘whatever its name’ (mimma šumšu) is an abstract way of designating ‘all parts,’ and so is similar to the previous examples where kullat + gods, lands, or peoples is the way to express ‘all of x.’ Finally, from a prayer to Marduk by Assurbanipal, a construction with another lexeme meaning ‘totality’ (hammatu, from hamāmu ‘to gather’) gives us ‘possessing the whole of all wisdom, the totality of all strength’ (hammata kullat nēmeqi gamir emūqi [gašrāti(?)], see Livingstone 1989, 7:4).

The substantive kalu similarly means ‘whole, entirety, all,’ and is sometimes constructed following a noun referring to things that can be counted, as in ‘signs, all of them’ (ittātim kalašina), or ‘prediction from birds, all (kinds) of them’ (purussû iṣṣūrū kalašunu). The extended meaning of the word napharu ‘all, whole, totality,’ stems from its basic meaning ‘sum, total (of things counted or added),’ which occurs in the context of a total tally of items in a list, or the sums of numbers. 11 It is sometimes found in absolute use, 12 e.g. ‘the totality of people, all of them’ ([te]nēšetu ša naphari kališunu), or to describe rulership over (ana malikūti) everything, or one who is appointed to be in charge (pāqid) of everything. More often it is found in construction with kullatu, as in ‘the totality of everything’ (kullat naphari), also preceding a genitive such as ‘all of Assyria’ (naphar māt Aššur), ‘all of Chaldea’ (naphar māt Kaldi), or ‘the totality of gods’ (naphar ilī), and ‘all the black headed (people) as many as exist’ (naphar ṣalmat qaqqadi mala bašû). Finally, a poetic word nagbu means ‘totality’ or ‘all,’ in the same way as its synonymns kullatu, napharu, and kiššatu. 13 Just as in the case of its synonyms, nagbu refers to all ‘of something,’ constructed with what is being totaled in the genitive, such as ‘all the Aramaeans’ (nagab Ahlamê), ‘all his enemies’ (nagab zā’erīšu), ‘all the swamps’ (nagab berate), ‘all living beings’ (nagab napišti). Already mentioned is also ‘the whole of all wisdom’ (kullat nagbi nēmeqi). 14

Somewhat different is the use of the term ‘regions’ (kibrātu), often constructed in the expression ‘the four quarters’ (kibrāt erbetti), which then represented ‘the entire inhabited world.’ 15 In one of its usages, the phrase ‘the four regions’ (kibrāt erbetti) has a political rather than a cosmological connotation, used in royal titulary and in inscriptions describing either hostilities in or domination over ‘the four regions.’ The notion that the four regions represent the entire world reflects a royal ideology in which divine appointment to kingship is in fact to rule over the entire world, as in the claim to be the one ‘whom Assur appointed to rule the entire world’ (ša Aššur … ana mu'irrūt kibrāt arba'i šumšu … išquru). 16

The last terms to consider here, gimru and gimrētu, from gamāru, whose basic meaning ranges from ‘to finish,’ ‘encompass,’ ‘complete,’ 17 present a somewhat different sense of totality as compared against the previously discussed terminology. The key passage is in the scene in Enūma Eliš where the assembly of the gods give Marduk kingship ‘over the entirety of the whole of everything’ (kiššat kal gimrēti), which he then demonstrates by annihilating and then recreating a star (or constellation, using the term lumāšu, Enūma Eliš l IV 14). Another such passage is from the Akītu, or New Year's, ritual: ‘You (Marduk) survey everything with your eyes, you watch over all omens/oracles/decrees/extispicies (têrtu) through the decrees (têrtu) you give’ (ina īnēku tabarri gimrētu [ina] HAR.BAD.MEŠ-ka tahâṭu HAR.BAD.MEŠ 18 ). Another reference to gimrētu, or gimrātu is in a lexical synonym list where gi-im-ra-tum = MIN (=[pu-uh-ru]), 19 and puhru stems from pahāru meaning to gather things together. In another passage for puhru, the plurality of things making up ‘the whole’ is clear: ‘Anu created all (meaning the collection of all) the heavens, Anu created (the collection of) all the earths’ (Anu puhur samê Anu puhur erṣeti … ibtani, see Köcher BAM 538 ii 52).

The foregoing examples underscore the difference between the concept of a totality expressed by ‘all xs,’ and the concept of a whole that functions within a lexical part-whole hierarchy, or meronomy (Cruse 1986, 157). Among the terms for ‘all of x (parts)’ discussed above, none offers a true holonym for the whole of what the parts comprise. Nor is it possible to construct a branching hierarchy from the world parts as belonging to the whole of which each is an autonomous part. Given that Cruse (1986, 171) said, ‘there are no meronomies of unnamed wholes,’ it is, therefore, difficult to know how to analyse the various world parts (heaven(s), earth, netherworld, Apsû) as meronoms (parts) or the words for them as meronyms (names of parts) of something that has no word or name (holonym) for itself. In the absence of a straightforward meronomy, then, let us consider the themata that relate the parts to each other, and, perhaps then, indirectly, indicate a notion of the whole.

Themata: complementarity, counterparts, and correspondence

It is clear that the passages containing the many Akkadian terms for ‘all’ (kullatu, kiššatu, kalu, napharu, nagbu, gimru, gimrētu and puhru) do not focus these descriptions on the internal structure or material substance or even overall shape of the world as a whole, but rather as the sum of its many parts or constituent things. It is my sense that a meronomic branching hierarchy cannot be reconstructed. More common are references to parts of the world in various non-hierarchical binary relations to one other, most basic of which are the regions of heaven(s) and earth(s). The idea that certain parts of the world, such as ‘above’ and ‘below,’ complement one another, forming by virtue of their relation one whole, is an essential ingredient of the cuneiform conception of the world.

The use of merisms, comprised of terms for contrasting or complementary parts, e.g. ‘heaven and earth,’ to express what is conceived of, presumably, as a whole, is evident in the cuneiform sources as it is commonly in other languages (Lloyd 1966; Wasserman 2003). Parts that complement one another, perfecting and making a whole from those parts, are often at the same time viewed as counterparts of one another. These two themata can therefore be considered together, as indeed the complementary parts often constitute counterparts.

The complementarity of two principal parts seems to be the oldest and most basic of conceptions of ‘world.’ The idea is preserved in the terms AN.TA.MEŠ (Akkadian elâtu) ‘the upper region(s)’ and KI.TA.MEŠ (Akkadian šaplâtu) ‘the lower region(s),’ and the idea of upper and lower halves of the world reflects an original mythology of creation by the separation of AN ‘above’ from KI ‘below’ (Lambert 2013, 169–171). The complementary parts heaven and earth, taken as a pair in Sumerian as AN.KI ‘heaven and earth’ (as well as the Sumerian AN.ŠÁR and KI. ŠÁR or ‘all heaven’ and ‘all earth’) and in Akkadian as šamû u erṣetu ‘heaven and earth (or netherworld)’ remain throughout cuneiform cultural history as the most basic structural duality for the world as a whole. 20 Wasserman (2003, 82) analyses the common merism of ‘heaven and earth’ as referring to a ‘vertical totality of space,’ which is a wholly different concept from one of ‘universe.’ 21

The notion of ‘above and below’ is, moreover, basic to the way the world is made through celestial divination, that signs above are equivalent to signs below. They are counterparts, as expressed in a singular text known as the ‘Diviner's Manual,’ which puts it as follows: ‘the signs on earth as well as of the sky bear (lit. ‘wear,’ našû) omens 22 for us; heaven and earth alike bring us omens; they are not separate from one another; heaven and earth are interconnected. A sign that is evil in heaven is evil on earth; that which is evil on earth is evil in heaven.’ 23 This describes a world wherein the counterparts of good and evil are manifested in the two complementary realms of the world, construed as above and below. 24

The complementary status of heaven and earth, however, is complicated by the fact that the ‘below’ of this equation (the lexical equation KI = erṣetu) denotes either earth or the netherworld, depending on context. A few sources confirm the identification of the erṣetu as the place where netherworld deities resided; thus erṣetu means ‘netherworld’ in such contexts. In other contexts erṣetu clearly means the earth as dwelling place of the living (Horowitz 1998, 273–4, and see the passage from Atra-Hasīs, ‘I cannot [set my feet on] the earth of Enlil,’ Lambert, Millard, and Civil 1999(1969), 91:48). Noegel (2017) suggests that in the Bible (Gen 1:1), as well as in Enūma Eliš, the merism should be understood as heaven and netherworld, not heaven and earth. This, he argues, is the result of the use of the term erṣetu for both earth and netherworld, and that the structure that makes sense of the terminology is a five-level organization of the world into heavens (šamû) – starry sky (šamû) – earth (erṣetu) – fresh water abyss (Apsû) – netherworld (erṣetu), wherein the words šamû and erṣetu each have two references. (Noegel 2017, 124)

The complementarity of heaven and earth/netherworld (erṣetu) is found in a motif often repeated in temple hymns and royal inscriptions concerning the building of ziggurats and temples, whose tops were so high as to ‘rival heaven’ and whose foundations were so deep as to reach into the netherworld (erṣetu). As an example, ‘he (Sargon) established its (the temple Eanna's) foundation in the depth of the netherworld like a mountain.’ (temmenšu ina irat kigalla ušaršid šadū’aiš YOS 1 38 I 40) It was also a way of expressing great magnitude, as in the description of a mountain encountered by Sargon II (721-705 bce) on campaign, ‘whose summit above leans against the heavens, and whose base, below, is firmly rooted in the netherworld’ (ša eliš rēšāša šamāmi endāma šaplānu šuršūša šuršudu qereb aralli TCL 3, p.6: 19.) The motif was also applied to mythological mountains in the Gilgamesh Epic (George 1999, 71 IX 38–41) that reached up to the base of heaven and down to the netherworld. It is also found in the Erra Epic, where the sacred tree, whose wood was suitable for constructing a statue of the supreme god Marduk, is similarly described as having roots in the netherworld and a crown that touches the highest heaven (Foster 1996, 765).

The conception of the ultimate heights and depths of a sacred tree, a mountain, or a temple had a variant where earth/netherworld is replaced by the depths of the abyss itself, the Apsû. The metaphor is traceable to the early third millennium bce in an archaic hymn: ‘Great, true temple, reaching the sky, temple, great crown, reaching the sky, temple, rainbow, reaching the sky, temple, whose platform(?) is suspended from the midst of the sky, whose foundation fills the Apsû.’ (Biggs 1971, 201). The Apsû was the place of the watery abyss, not the salt sea, rather the sweet fresh waters of the springs (nagbu 25 ), as we know from Sumerian and Akkadian mythology, divine epithets, and hymnic literature. The unceasing, seemingly eternal, fresh water springs and source of the rivers no doubt accounts for the inclusion of the Apsû as one of the cosmic regions. Thus in the Hymn to Šamaš, about the man who does good deeds, it is said ‘[his] descendants will go on forever like the eternal spring waters (nagbu).’ 26

The innermost part of the fresh water abyss was beyond the reach of human perception. No one had plumbed the depths of the Apsû nor seen its interior (Horowitz 1998, 317). The depth of the Apsû was an unknown region, as described in a broken line that compares something (in the broken beginning of the line) to the unknowable inner regions of the Apsû, saying ‘that, like the inside of the distant Apsû, not even a god can find (it).’ 27 Although a part of the world largely inaccessible to human perception, the Apsû's sweet waters in rivers and springs that bubble up from underground were places where the Apsû reached the earth's surface.

Apsû's watery depths were associated with the creation, abode, and kingdom of the god Enki (Sumerian)/Ea (Akkadian), so closely associated with him that Enki/Ea's son, Marduk, was known as ‘first-born son of the Apsû.’ Because of Enki/Ea's association with wisdom, magic, and incantations, the Apsû was the fount of wisdom and source of the secret knowledge of incantations. The temple of Enki in the oldest Sumerian city of Eridu was called the E-Abzu ‘House of the Abyss.’ In Enūma Eliš, Marduk's temple Esagil in Babylon was said to be the counterpart (mihirtu) of Apsû (Enūma Eliš VI 62). 28 As we have already seen, Enūma Eliš IV 142 places Ešarra, the dwelling place of Enlil, as a counterpart (tamšīlu) of Ešgallu, the ‘great shrine’ or Apsû. And the ziqqurat foundation, Etemenanki (its name itself testimony to the idea of the complementarity of above and below) is said to be a counterpart (GABA.RI) of Esharra in an inscription of Nabopolassar (VAB IV 62 iii 19; cited Lambert 2013, 200).

One of the great sources for the notion of counterparts is Enūma Eliš, when Marduk demonstrates his authority in a wholesale reorganization of the world and the settling of gods in it. Equally as important, the poem accounts for the location of the city of Babylon, and site of Marduk's own temple Esagila there, as a new cosmic centre. Marduk's residence in the Temple Esagila effectively placed Babylon at the centre of the world. Marduk's ‘house,’ the Esagil, is said to be ‘the equivalent’ or ‘counterpart’ to the great abyss, the Apsû, where Ea dwelled: Enūma Eliš VI 62 ‘They raised the peak of Esagil, a replica of the Apsû’ using the term mihirtu ‘equivalent, counterpart.’ The importance of the idea of measured counterparts, equivalents, is also clear in Enūma Eliš IV 141–46 ‘He (Marduk) crossed the heavens, surveyed the sky, and he adjusted them as the equivalent of the Apsû, Nudimmud's abode. Bel measured the shape of the Apsû and set up Esarra, the counterpart to Esgalla. In Ešgalla, (in) Ešarra which he had built, and (in) the heavens, he settled Anu, Enlil, and Ea in their shrines.’

A further indication that the idea of counterparts played a thematic role in a cuneiform world description is the use of the term maṭṭalātu from a verb (naṭālu) meaning ‘to look at, face, or point toward.’ The word is relatively rare, occurring only in first millennium scholarly or literary contexts. The clearest usage is no doubt the one found as the incipit of Tablet 16 of the liver omen series (Barûtu ‘the art of inspection [of the ominous entrails]’), which is: šumma amūtu maṭṭalāt šamê ‘If the liver is an image/counterpart of heaven.’ (CT 20 1:31) Another attestation of maṭṭalātu occurs in the context of the account of rebuilding of the temple Esagil by Esarhaddon. In his royal inscription, in a deliberate reference to the Babylonian Enūma Eliš, Esarhaddon says the temple Esagil is the counterpart of the Apsû (maṭṭalāt apsî) and the equivalent, or likeness (tamšīlu) of Ešarra (Leichty, 2011, p. 198 Esarhaddon 104 iii 41b-iv 1, and p.206 Esarhaddon 105 iv 37b-v 15). The passage is worth quoting in its entirety:

[In] a favorable month, on a propitious day, I laid its (Esagil's) foundation platform over its previous foundations (and) in exact accordance with its earlier plan I did not diminish (it) by one cubit nor increase (it) by half a cubit. I built and completed Esagil, the palace of the gods, an image (maṭlāt) of the Apsû, a replica (tamšīl) of Ešarra, a likeness (mehret) of the abode of the god Ea, (and) a replica of (the square of the constellation) The Field; I had (Esagil) ingen[iously] built (and) I laid out (its) square (mithartu). 29

In the context of world regions, the thematic notion of the counterpart refers to temples, mountains, trees, and the liver, and it connects the divine, or heavenly, world with something important for human life. When the temple of Ea, the Ešarra, is set up as the replica or counterpart (mihirtu) of the Field constellation, or the Square of Pegasus, the celestial quadrant marked by the stars is established as the model in heaven for the earthly shape and foundation for the Ešarra temple in Babylon. The best indication that the meaning of mithurtu is ‘opposition,’ in the sense of complementary counterpart, is in the writing of the Sumerian word that is translated as mithurtu, namely HA.MUN, meaning ‘harmony (of opposites),’ which is spelled with the Sumeriogram NAGA.NAGA (where the second NAGA is written upside down).

Continuing the binary nature of themata attested in many world descriptions from cuneiform scribal scholarship is that of correspondence. The theme of complementarity between celestial and terrestrial applies in divinatory interpretation, where evil above is evil below, and so on, but correspondence forms a different relationship constructed of signs and their consequents, 30 the protases and apodoses of omen texts. Every omen then, in making a relationship between phenomena from one domain to those in another, instantiates the thema of correspondence, so this is perhaps the best attested of all themata in cuneiform knowledge.

Typical examples of correspondence can be given where deities are set in corresponding relationships with phenomena, for example with the four winds: South (corresponds to) Ea, the father of the gods; East (corresponds to) Enlil, the lord of all; North (corresponds to) Ninlil, lord of phantoms; West (corresponds to) Anu, father of heaven. (George 1992, 152–3) The parts of the liver correspond to specific deities, and the months to zodiacal signs, as in this example from a late text from Uruk:

The Presence (corresponds to) Enlil, the month Nisan, [the Hired Man (Aries)]. … The Path (corresponds to) Shamash, the month Ajaru, The Bull of Heaven (Taurus) … The Pleasing Word (corresponds to) Nusku, the month Simanu, True Shepherd of Anu (Orion) … […] The Strength (corresponds to) Ninurta, the month Du'uzu, the Crab (Cancer) … and so on. (Koch-Westenholz 2000, 24–25)

Conclusion

The themata characterizing the descriptions of world parts, namely complementarity, counterparts and correspondences, are non-hierarchical binary relationships. The fact that there is no Akkadian counterpart to our notion of ‘world,’ but rather identification of a totality is always as the sum of parts, as in ‘all of x (things),’ raises the question of whether these elements of world description are effected through meronomy. Meronomies are typically based on part-whole relationships, but the nature of those relationships are not narrowly defined, either with respect to spatial, structural, material or functional relations of various parts to the whole. What they do have are distinct ideas of those part-whole relations, necessitating a designated name (holonym) for the whole in relation to the parts. In the passages from cuneiform texts in which world-parts are described, the interest appears to be less about the part-whole relationship and more about the part-part relationship, specifically about binary relations in the form of counterparts, complements, and correspondences.

On no level do we find examples of branching hierarchies of parts in their relation to an overall whole. Nor can we find evidence that such wholes were ontologically prior to its parts. Totalities of parts constituted wholes, as lands were sums of all inhabited cities. Another kind of totality is found in the sum of many things, e.g. all wisdom or ‘the entirety of all of everything,’ both of which connote sums rather than unitary wholes.

The exploration of mereological themes in cuneiform worldmaking shows a focus on non-hierarchical binary relationships and the tendency to disregard questions of substance and composition in favour of correspondence, complementarity, and correlation. In view of Goodman's statement (1984, 14, emphasis in the original) that ‘there is no one correct way of describing or perceiving “the world,”’ it is important to be able to account for the particularity of the cuneiform ideas through their own way of worldmaking, and also to consider that different traditions within the cuneiform corpus refer to different ‘worlds.’ We are then able to leave the idea of a physical external world out of that picture, to focus instead on what constituted the ancients’ right rendering, or right description, of the world in its own context and within its own frame of reference. Goodman's view (1978, 3–4) that ‘“the world” depends upon rightness,’ advocates for a multiplicity of ‘rightness of renderings of all sorts.’ If the anthropologist of historical sciences can establish from the evidence ‘what makes a version right and a world well-built’ (Goodman 1984, 29) the ‘science of the past’ can open onto more diverse and multiple kinds of knowledge both within and across cultures.