2.0 A TRANSMITTED LITERATURE? EVIDENCE FOR EXTISPICAL TEXTS PRIOR TO THE OLD BABYLONIAN PERIOD

The understanding of extispicy as a transmitted literature fundamentally depends on the existence of forerunners² to the three forms of technical literature we see arising in the Old Babylonian period: liver-omen models, compendia, and reports.

It has been a problem of many analyses of these three text-types that they freely compare terms and features of texts from different times, places, and text-types on a presumption of fixed meanings and direct transmission. Dispensing with a historically critical approach, this presumption does not reduce, but rather increases, the possibility of creating anachronisms and contextual noncomparabilities. Lexical and semantic understandings in extispical texts are often reconstructed by referring between Mariote, northern and southern Babylonian cor­pora, between the three text types mentioned above, and/or between Old Babylonian and Neo-Assyrian attestations. So eager are we to know what the “Comb” of the lung is — to resolve definitional problems through intertextual references — that we have ignored large problems of synchronic and diachronic comparability: the terms of compendia rarely appear in the reports (and vice-versa); the omens of Larsa do not show up in Sippar; the proportion of hapaxes is through the roof; and so forth.

If a unitary and accumulated literature existed, it should be demonstrable in some mea­sure — but what evidence for a pre-Old Babylonian literature do we have? Eight categories of evidence will be discussed relative to arguments supporting the existence of extispical literature prior to the nineteenth century:⁵

2.1 The appearance of diviners in third-millennium professional lists

2.2 Third-millennium references to animal omens

2.3 Third-millennium references to liver divination

2.4 Purported examples of third-millennium extispical texts

2.5 The appearance of third-millennium kings in Old Babylonian “historical” omens

2.6 The size, extent, and comprehensiveness of the Old Babylonian extispical com­pendia

2.7 Later references to extispicy’s antiquity

2.8 Procedural dissimilarities to scientific method with respect to observationalism

2.1 The Appearance of Diviners in Third-millennium Professional Lists

The earliest evidence for extispicy is the appearance of diviners in Early Dynastic pro- fessonal lists, in the entries lu.mas.su.gid (“one who reaches the hand (in)to the goat”) in Lu E from Ebla⁶ and Lu C from Fara and Abu Salabikh.⁷ Yet while these entries attest to an identifiable class of ritualist at this early stage, they tell us nothing of the apparatus of ritual itself. If anything, Lu C, which displays some apparent groupings of professional types, lists the lu.mas.su.gid together with persons working with animals, not with professions more likely to have been working within a scribal or cult tradition.⁸

It also bears observation that, despite the early appearance of the professional name, it does not appear again until the middle of the Ur III period, when once again the documentation is strictly concerned with the administration of animal management, not with cult or ritual practice as such.^{9 *} It is also not possible to locate diviners within rosters of cultic personnel at major temples.¹⁰ Whatever the ritual functions of the professional baru in the third millen­nium, we cannot point to any instance in which he functioned in a cultic or literate context with or within the institutional households where textual traditions were most prominently supported.

2.2 Third-millennium References to Animal Omens

A host of third-millennium references to omens procured through small livestock are often cited as evidence for early extispicy.

Gudea’s Cylinder A is commonly cited as providing evidence for third-millennium ex­tispicy (as indeed it does: see section 2.3). What is commonly overlooked, however, is that in this one composition alone several other kinds of animal omens are also mentioned. In one instance, a goat is led to the brick shed to identify the pure brick for building.¹¹ Elsewhere, Gudea leads two sheep and a kid to lie down on animal skins to induce an omen in an incense ritual.¹² Within his initial dream, Gudea recalls seeing a donkey pawing the ground, a sign of his own eagerness to build Eninnu.¹³ All these animal omen techniques also appear alongside several forms of non-animal divination used by Gudea: dream, grain, and kledon omens.¹⁴ The existence of multiple formal procedures for procuring omens from animals should warn us away from a “presumption of extispicy” when extispicy is not specified (as Gudea elsewhere does): there clearly were a number of ways to get an omen out of a goat.

This in turn must cast some doubt on just what procedures were meant in the large number of southern Mesopotamian year-names from the Akkad, Lagas, Ur, Isin, and Larsa dynasties referring to omens.¹⁵ At least twenty-eight year-names — from Naram-Sin’s years “o” and “ll” (ca. 2250 B.C.),¹⁶ as late as Damiq-ilisu of Isin’s year 4 (1812 B.C.) — refer to sheep omens identifying cult officials to be appointed in temples, using the following formulae (see Appendix 1 for a complete listing):

Naram-Sin “o”:... mas.e ib.dab₅.ba

Naram-Sin “ll”:... mas.e ib.dab₅.ba

Lagas: Ur-Ningirsu I “a”:...

Ur III: Ur-Namma “b”:... mas.e ba.pa.da

Isin/Larsa: Isbi-Erra 13:... mas.e ba.pa.da

These year-names differ only in the type of priest and deity named,¹⁸ the expression for goat (mas/mas), and the verb (Akkadian dynasty only: dab₅; thereafter: pad), all meaning “Year in which NN-priest(ess) was named (dab₅: installed) by (means of) a goat.”¹⁹

First, a literal-minded translation of these formulae must take note of the fact that ex- tispicy per se (i.e., some variation on su... gid) is not mentioned, though we know that the verbal formulation was in use at this time (see sections 2.1 and 2.3). A further question is raised by the ambiguities of the verb pad “to call,” which is most heavily employed in contexts which imply speaking (or, perhaps, bleating), though of course the semantic range of “calling” (both in Sumerian and English) affords the idea of “naming.”²⁰ The meaning is thus unclear, and dab₅ is even more obscure as regards the ominous method. We must remain sensible that the “calling” in question is no more likely to have involved reading the entrails of a dead goat than of the other procedures illustrated in the Gudea cylinder.

Further problems arise that make this more than a matter of raising a reasonable doubt about the nature of the ominous procedure. Three disconnects — geographic, temporal, and functional — must be established between this class of year-names and the later technical literature; these disconnects substantially separate the nominative year-names from the later technical extispicy. First, the practice of choosing priests “by means of a goat” was restricted to southern Mesoptamian cities (Nippur, Isin, Uruk, Larsa, Lagas, and Ur), which were not, with the exception of Larsa,²¹ the cities from which the later technical literature is attested (Larsa, Mari, Sippar, Babylon, Esnunna). Second, the technical literature post-dates the year- names with very little overlap.²² Our latest-known sheep omen year-name is for Damiq-ilisu’s year 4, 1812 B.C.; the earliest exemplars of technical literature probably date to nineteenth century Mari (see section 2.5); the earliest securely datable technical document for liver divi­nation is now the omen for the accession of Dadusa of Esnunna, ca.

Third, the apodoses of the technical literature are virtually silent about the concerns (so far as we know) of the third-millennium sheep omens, the appointment of priestly personnel or the identification of temple sites. Indeed, the classes of officials in the two sets of literature show almost no overlap: third-millennium references to extispicy address the choosing of cult figures such as the en, nin.dingir, lu.mah, gudu₄,²⁴ and isib; the later compendious texts (e.g.,

YOS 10 and OBE) are concerned with non-temple officers such as the sukkal, sakkanakku, sipru, zabardab, nun, and lugal. This dichotomy is not without exceptions. Sulgi, for instance, boasts of using extispicy to determine not only cultic matters, but also military action (Sulgi Hymn B, lines 131-49).²⁵ On the Old Babylonian side, a very few omens from the technical literature do take priestly personnel as their subject matter in various contexts — but only three, out of perhaps three thousand omens,²⁶ for their selection or appointment. In all three cases, these omens are about ugbabtu-priestesses,^{2 *} who are not among the personnel appear­ing within the nominative year-names.

Divination of the mas... dab₅/pad type should be understood as older, southern, and cultic, while literature of the mas su... gid omens should be seen as newer, northern, and em­phatically statist and non-cultic. The year-names and the Old Babylonian omens are mutually exclusive in terms of time, space, and subject, two fundamentally different sets of practices, neither precursor nor finished form.

2.3 THIRD-MILLENNIUM REFERENCES TO LIVER DIVINATION

Notwithstanding, there is no doubt that extispicy was practiced in the third millennium. Yet written references to the practice of the extispical craft cannot be regarded as evidence for a technical literature of liver divination. In fact, the undoubted antiquity of references to practice then makes the millennium-long absence of procedural and reference materials all the more remarkable, underscoring the nature of that practice as a craft.

The very diversity of contexts for these references (administrative documents, literary works, year-names — but see section 2.2, above) has been distracting. Two Early Dynastic pieces of evidence are, together with the appearance of diviners in Lu C and E, the oldest specific mentions of extispicy. The first is an enigmatic Sumerian proverb “The songs of a city are its omens (uru^kl en-du-bi mas-su-gid-gid-bi-im),”²⁸ which suggests only perhaps a likeness at the level of orality. The second is the cultic text OIP 99 114;²⁹ this composition probably names rituals for the reader to perform, but contains no information about method or

procedure; though its contents are “obscure,” as Alster notes, it is not itself a technical text.³⁰ What can be said about the proverb and the cultic text is that they point away from written technical instruments, and towards oral performance.

Throughout the third millennium, a host of literary compositions make reference to divi­nation with varying emphasis. The most well-known extispicies among these are the two by Naram-Sin in “The Cursing of Akkade,”³¹ at least two performed by Gudea (Cyl. A xii 16-17; xx 5),^{32 *} and those boasted of by Sulgi (Richardson 2006). In all these instances, the verbal formulae mas/màs su... gid is used to designate the method used to gain an omen. In none of these instances is there any reference to textuality, nor could the passages themselves conceivably constitute any kind of observational document drawn on by future readers. The one exception to this state of affairs has long seemed to be a crucial passage in Sulgi’s Hymn B (“I am the very Nintud of the omen collections (gìr-gin-na)”), which supported the idea that a fully serialized library of omens existed at least by Sulgi’s time. My recent argument (Richardson 2006) that gìr-gin-na should be translated as “procedures” rather than “omen col­lections” considerably alters this picture. A mistaken conflation of Sulgi’s learned skills with his innate ones in the secondary literature masked the emphasis on extipicy as a natural and intutive art, not a “book-learned” technical skill, nor an observational and documentary one.

Categorical errors about what skills and practices lay behind extispicy have been magni­fied by a definitional drift in which ominous procedures of all kinds have often loosely been translated simply as “omens.” The tertiary effect has been for students, scholars, and editors to sometimes interpolate extispicies where other kinds of omens were actually meant (see Appendix 2 for the effects of this problem in a particularly influential set of translations in Jacobsen 1987). The image produced has thus been one in which extispicy was practiced more often than it was and stood in some clearly preeminent position vis-à-vis other divina- tory systems. It did not.

However: even were these instances all to be understood as liver divinations, what they have in common is that — though they would certify that extispicy was in use — none of them mentions or suggests the use of texts. In “The Cursing of Akkade,” the evidence is equivocal on this point, since Naram-Sin is simply said to “perform” extispicies. But for Gudea, the contexts point more toward an intuitive or memorized craft than a scholarly one. The omen of Cylinder A xii 16-17 is said to reveal Ningirsu’s intention (sà-^dnin-gir-su-ka) which “stands out as clear as daylight” (u₄-dam mu-na-è), and that the revelation was due to Gudea’s (repeat­edly, emphatically) proclaimed qualities of “great knowing” (gal mu-zu) and “great carrying­out” (gal ì-ga-tum-mu), epithets which suggest an unmediated and untutored access to divine knowledge based on innate gnostic ability — not on learned knowledge.³³

2.4 Purported Examples of Third-millennium Extispical Texts

Two very different letters — one from Ebla, one a school text from the “royal correspon­dence” of Ur — have either been proposed as or pretend to be extispical texts dating to the third millennium. The first, however, is not an omen (though it is from the third millennium), and the second is not from the third millennium (though it is an omen). A third direct claim in Sulgi’s hymns to have produced serial literature for extispicy is evidence I have disputed elsewhere on the grounds that the crucial Sumerian term gir-gin-na, often translated as “col­lections” or “library” (of omens, in this case), should rather be understood as “procedures,” relevant to extispicy’s unwritten and performative protocols (Richardson 2006).

In the first case, the Ebla text TM.76.G.86 has been published as an “extispicy report.”³⁴ Strictly speaking, it is a letter which refers to an extispicy. Coser asserts that a “structural analysis” reveals that the letter contains both protases and apodoses.^{35 *} This is not the case. The letter refers to two inspections of sacrifices (no specific animal is mentioned) in II.2-3 (wa hul, “and (the omen result) was bad”), and III.7-IV.1 (wa igi.gar gu-sum ug₇ ds-dag igi.gar, “and, when he observed the victim, he saw death by your side”). An extispical procedure is discussed, but the relevant passages fall short of the operative criterion of extispical literature: to record a specific observation (a protasis, indispensible in reports, as Coser herself notes) in order to read a specific result (an apodosis, often, though not always, present in reports), reproducible conditions which can be consulted in the future. No sign or mark is recorded in the Ebla letter: there are no protases, and there are no technical terms of any kind.³⁶ No obser­vation, as such, is recorded in TM.76.G.86: nothing from the document could be reproduced as an omen. The letter talks about an omen, but doesn’t contain one.

A different case presents itself with the Old Babylonian school letter, in which an “omen” appears embedded within long and short versions of a putative royal letter of Ibbi-Sin of Ur:

Enlil has looked upon me with grace and has taken my prayer to his holy heart; he established for me in my omens the favorable parts. Furthermore, he fashioned the right side for him, and the left side for me. He beautifully set there the Weapon on my favorable side with a straight flank; the Weapon on his unfavorable side was present and (looked over) to the other side, bound steadfast to the filament. (This means:) “My enemy will be delivered over to me and killed.”³⁷

The passage (in contrast to the Ebla letter) refers to specific signs, employing a technical terminology, connected to specific results. Yet although it is genuinely extispical, it is not genuinely third millennium: eight copies of the long and short versions of the letter are known, and they all date closely to the latter half of the reign of Samsuiluna in the late eighteenth century B.C., and not to the twenty-first century B.C. reign of Ibbi-Sin. As Michalowski argues, the “omen” is one of a variety of Old Babylonian scribal exercises inserted within a school text, written in the “highly baroque style” of the Larsa court, an insertion fully consistent with the wider program of archaizing elements of the “royal correspondence.”³⁸

2.5 The Appearance of Third-millennium Kings in Old Babylonian

“Historical” Omens

Some of the very first written omens have been understood to require written third-millen­nium sources: these are the so-called “historical” omens, which mention the long-dead kings of Akkad and Ur, among others.³⁹ These omens themselves give not a hint of any contemporary conviction, however, that the observations had their origins in histoire evenementielle. The “information” they provide better reflects scribal-scholarly interests in paronomasia (e.g., the Naram-Sin omen about Apisal) and the historiographic tradition of Heils/Unheilsherrschaft, a dualistic scheme which fit well into the interpretive matrix of extispicy. It is also no ac­cident that the kings of the historical omens were often the same famous kings who were the subjects of other literary compositions then popular in the Old Babylonian school curriculum — Gilgames, the Akkadian kings, Ibbi-Sin, etc.^{40 *} — and visible in the statuary at Nippur. Though these ominous significations clearly referred to the past — as did literary tales of those kings — there was no claim that the texts (or even the omens) themselves had come from the past — as with the colophons that scrupulously marked the copying of original documents of other kinds, such as royal inscriptions.

From the start, the problematic datation of these historical omens has confused the histori- cal/historiographic issue. The liver models found at Mari (the earliest-known documents to be inscribed with historical, indeed any, omens) were written in the sakkanakku script which does not clearly distinguish the century of their composition. It has thus been possible to suppose that the omens so inscribed had been composed contemporaneous with their subject matter. The pivotal historical figure here is Isbi-Erra:⁴¹ his appearance among these omens has been used to argue that he marks a teminus ante quem for the liver omens, that is, that they had

all been composed between the time of Sargon and the death of the first king of Isin.⁴² Gelb, to the contrary, argued that the liver models could not have been inscribed before the reign of Isbi-Erra — that the rebel king was simply the least venerable in the company of “histori­cal” kings.^{43 *} I feel this is the more sensible explanation: a group of texts, found together, all mentioning past events in similar form and script, are more likely to have been composed or compiled together soon after the latest recorded event among them, not from the earliest one and over a period of four centuries.⁴⁴

Historians of these texts have asked why they were first composed. But given the above, we should perhaps invert the question: if “historical” omens were observational, why did they ever stop? If the scribes believed in the authenticity of observational omens, why were there never again recorded liver omens about any Old Babylonian kings who reigned during the time when the technical texts were actually being produced?⁴⁵ The “historical” liver-model omens of the twentieth/nineteenth century B.C.⁴⁶ have the highest comparability among the Old Babylonian technical texts⁴⁷ and are thus the strongest evidence that extispical literatures drew on common-culture sources. Liver models are also the first apparatus appearing among the technical types, with compendia surfacing only in the later nineteenth and early eighteenth centuries, and reports in the eighteenth and seventeenth centuries. Yet the kings who were treated as “historical,” and whose significance was broadly similar from text to text, was limited: they reflected the Old Babylonian idea of what constituted history, that is, the events of the Akkad and Ur III dynasties. By contrast, when compendious texts were still in produc­tion in the seventeenth century, at a time when onomastica still reflected pious veneration of Hammurabi and Samsuiluna and the kings still traced their lineage through them, we never find any new ominous signs named for these or any other more recent kings. Thus, not only are the “historical omens” poor historical sources for those kings (as Cooper 1980 stated so succinctly), their temporal restriction to the pre-compositive phase of the literature also speaks

against any ongoing interest in observable phenomena. The “historical” kings were chosen in a later period precisely for their historical veneer.

2.6 The Size, Extent, and Comprehensiveness of the Old Babylonian Extispical Compendia

The impression that forerunners for extispical literature existed is also suggested by the dramatic appearance of the extensive compendious literature; without antecedent materials, how could such a corpus be formed ex novo? The massive series from southern (nineteenth­century Larsa) and northern (eighteenth-seventeenth-century Babylon and Sippar) Babylonia are the earliest compendia known, yet these already display a series numbering nearly 10,000 omens. The texts appear to us so fully formed that it is hard to believe they were not the outcome of a long process of scholarly redaction or compilation. This seeming impossibility induces assumptions that earlier texts, though not yet found, nevertheless must have existed prior to the nineteenth century B.C. This is, of course, precisely the interpretation which the scattershot of earlier secondary references would seem to favor (but see sections 2.2-5, above).

And, indeed, some aspects of the internal, formal organization of the compendious series could be taken to mean that a few of the Old Babylonian texts known to us cannot be “first­generation” documents. Goetze (1947a) long ago pointed out the existence of both duplicates and variants, possibly indicating the existence of earlier recensions (cf. section 3.2),⁴⁸ and the arrangement of the compendia by the zones of the liver has encouraged an association of complex organization with antiquity. These have occasionally led to speculations about writ­ten⁴⁹ and oral⁵⁰ sources for the compilation of such texts.

Yet the hard fact remains that, while Assyriologists have been studying liver-omen litera­ture for over a century, in this time no technical texts dating earlier than the Old Babylonian period have emerged.⁵¹ Despite the propensity of third-millennium scribes to compile lists

and collections of many kinds, and the avidity of Old Babylonian scribes for copying them, no earlier lists of omens have emerged. Despite the antiquarianism abroad in the Old Babylonian period (especially at Nippur), we find little reflection in extispical texts (absent at Nippur) of the topoi which concerned the Sumerian literati (e.g., fertility, mortality, purity, cosmic order; see also sections 2.1-2); it is hard to see that the Sumerian Weltanschauung is reflected in the extispical corpus. In contrast to the wide variety of Old Babylonian texts coming out of a true scholastic tradition of copying (epics, hymns, prayers, commentaries, mathematical documents, lexical and other lists), no known Old Babylonian extispical text bears a colophon indicating it to have been copied from another source, nor is there any reasonable expectation that such sources will emerge.^{52 *} This is especially strange when we consider the degree to which Old Babylonian omens were accurately transmitted to Neo-Assyrian Barutu: are we to understand that a great textual tradition, maintained and transmitted with a high degree of reliability in all periods when it is visible, is to be constructed where it is not visible?

One of the other great bodies of serialized Mesopotamian literature was similarly not preceded by materials identifiable as “forerunners”: the Early Dynastic proverb collections were sizeable, extensive, and comprehensively organized, yet seem unlikely to have been compiled from any antecedent literature. The only “smaller” materials for the Early Dynastic corpus are a handful of school texts that are not earlier than the collections themselves, and perhaps later in date. Only the Old Babylonian proverb collections are accompanied by great numbers of excerpts and school tablets, that is, long after the collections themselves were well established.⁵³ While it has been debated whether or not Early Dynastic proverbs were collected from genuine phrases and sayings or were compiled for purely academic purposes, it is clear that they do not reflect other parts of the scribal curriculum — despite having been composed by scribes.⁵⁴ In the cases of both the omens and proverbs, there is no “primitive” literate background to these massive, well-organized corpora. This absence suggests that, while the compilations may have been genuine in the sense of collecting existing knowledge based on oral tradition, they did not emerge from a scholastic tradition over time, gathered from multiple sources.⁵⁵

2.7 Later References to Extispicy’s Antiquity

Potency and legitimacy were accorded to Mesopotamian cultural forms for their vener- ability, and extispicy was indeed viewed as an ancient art — but only in the first millennium, when it was already more than a thousand years old. The claim of antiquity was advanced for the first and only time in a text from Assurbanipal’s seventh-century B.C. library, that the antediluvian king Enmeduranki was taught the art by the god Samas, the king then passing his knowledge on to wise scholars.⁵⁶ “Enmeduranki” is a slight corruption of the Enmeduranna known from the Sumerian King List. Yet though the Sumerian King List dates to at least the twenty-first century B.C., it mentions no wisdom traditions of any kind — only that Enmeduranna was a king ruling at Sippar for 21,000 years.⁵⁷ A third and final reference to Enmeduranna is in the King List compiled by Berossus in the third century B.C., but here again we find no reference to liver divination.⁵⁸

As Pongratz-Leisten argues,^{59 *} the Assurbanipal-era claim has little value as historical evidence. The ancient pedigree of knowledge texts was part of a wider royal claim to hold independent access to divine will by privileging the past as a site of original knowledge production, such as with Assurbanipal’s famous boast to have “read tablets from before the Flood.”⁶⁰ Earlier ages had in fact emphasized the antiquity of knowledge to a lesser degree. Neither within the Old Babylonian technical literature or in secondary references to liver divination are there any references to its antiquity, nor even to its general origins (see section 2.5 regarding the absence of colophons).⁶¹ Old Babylonian scribes, like Neo-Assyrian ones, embraced antiquarian learning, but there is nothing to suggest that they looked on extispicy as an especially ancient tradition. This is reflected in the Old Babylonian use of the terms baru and barutu: though we know of plenty of barus in the Old Babylonian period, the term barutu was little used.⁶² We know the names of hundreds of Old Babylonian “diviners,” but almost no abstract concept of “divination”; the Old Babylonian craft was still too heterodox (or newly orthodox) to admit abstraction.

2.8 PROCEDURAL DISSIMILARITIES TO SCIENTIFIC METHOD WITH RESPECT TO OBSERVATIONALISM

Finally, a theoretical problem: divination’s similarity to scientific procedure, and the implication that observationalism was its underlying mechanism, has lent weight to the idea that its process was documentary in nature. The analogy to “science” is partly welcome. It sets liver divination at a distance from the semantic fields of “temple religion” and “magic.” It is justifiably pinned on both a) science’s similar status in modernity as an irreducible form of knowledge, and b) divination’s likeness to the scientific method in its systematic organization of phenomena, causal association to other repeatable phenomena, the creation of extensible theoretical categories, and (apparently) in the employment of observation.⁶³ But the analogy is limited: absent are the critical methodologies which also characterize modern science: ex- perimentalism, problematization, falsification, disproof.⁶⁴

Observations of livers have been presumed to be the means by which the first omens were transferred to their place in the texts (e.g., “If X is observed, then Y”), but that process is not visible in the textual precipitate.⁶⁵ The presumption that a gradual process of accumula­tion and compilation retrojects observationalism into extispicy’s genetic development.⁶⁶ A historicist point of view, however, looking at the concentration of early evidence into the century ca. 1850-1750 B.C., sees this idea as dubious: the absence of a documentary trail (as discussed above) itself militates against the existence of either an observational procedure or a principle of causation whose mechanism did not require the heavy framing of both scribe and specialist.⁶⁷

It has been almost fifty years since Thomas Kuhn (1962) first critiqued the presump­tion of cumulative observationalism as the mode of progress in the sciences (The Structure of Scientific Revolutions). Kuhn argued that change in scientific knowledge is characterized by sudden crises in thought that demarcate otherwise long periods of quiescent paradigm.

Although this stance has not been adopted uncritically by intellectual communities,^[122] one of Kuhn’s most long-lasting and widely subscribed ideas is that observation has never been — can never be — free of theoretical framing. In these points — the punctative nature of scien­tific development on the one hand, and the rejection of some root of “pure” observationalism on the other^[123] — extispical literature deserves the analysis of its entextualization, of its texts as a literature with a history, not as a unitary form that presumably existed from time imme­morial. Someone created it, and for a reason.

2.9 Synthesis

This part of the discussion has argued against the existence of any scholarly tradition for liver divination prior to the nineteenth century B.C. In so doing, it refutes no particular opinions to the contrary, but counters a scholarly discourse too accepting of certain very mod­ern premises about the observational origin of the practices, amplified by some tendentious claims of later antiquity. Thus, though little that I have argued above has not been considered in some fashion elsewhere, it is my hope that there is a particular value in bringing all these strands of evidence together in a systematic fashion. It is not my purpose to destroy a “straw man”: the next section turns its attention to the entextualization of liver divination, to think­ing about the reasons why it came into being when and as it did. Central to the discussion is the coincidence of the rise of the extispical literature with the 150-year period during which Mesopotamia descended into intra-regional war.