Citatio: B. Christiansen (ed.), hethiter.net/: HDivT (07-05-2026)

Extispicy (SU oracle)

1. Short Description

Extispicy, also known as SU or ‘flesh oracle,’ is a significant divinatory practice that involves examining the entrails and liver of slaughtered sheep. This technique, which was performed by the ‘diviner’, a male expert referred to by the Sumerogram ^LÚḪAL (^LÚAZU, Akkadian bārû), holds a prominent place within the extensive Hittite corpus of divinatory practices. Numerous Hurrian terms, some of which were adopted into Luwian, indicate that the Hittites learned this oracle technique from the Hurrians. This tradition resembles those found in Mesopotamia, where it developed into its definitive form during the Old Babylonian period. However, the Hittite SU oracles also show some differences to the Mesopotamian tradition. This is primarily evident in the terminology and the sequence of features observed.

2. Research History

The study of Hittite extispicy has always been embedded in the broader field of Mesopotamian extispicy, from which the Hittite SU oracle ultimately derives. The foundations for understanding the terminology and procedure of Babylonian extispicy were laid by Goetze and Nougayrol through their editions and discussions of numerous Old Babylonian texts (Nougayrol J. 1941; Nougayrol J. 1945–1946; Goetze A. 1947; Goetze A. 1957; Nougayrol J. 1966; Nougayrol J. 1967; Nougayrol J. 1971). Nougayrol’s significant contribution is the identification of a standard sequence for examining the various parts and features of the liver and other entrails, which proved fundamental for subsequent research (Nougayrol J. 1967, 232–233). Nougayrol was also the first to study the relation between Babylonian extispicy and the later Etruscian practice (Nougayrol J. 1955). Further advances in understanding the Babylonian terminology were achieved by I. Starr (see, e.g., Starr I. 1977; Starr I. 1978–1979; Starr I. 1985); and U. Jeyes (see, e.g. Jeyes U. 1978; Jeyes U. 1980; Jeyes U. 1989). Additionally, J.-W. Meyer’s analysis of clay liver models (Meyer J.-W. 1987) and R. Leiderer’s veterinary-anatomical study (Leiderer R. 1990) enhanced our understanding of extispicy terminology and practice from a medical perspective.

The first major comparative study of Akkadian and Hittite oracle texts was A. Boissier, “Mantique Babylonienne et Mantique Hittite” (Boissier A. 1935). Although now superseded in many respects, this monograph was foundational in establishing the close relationship between Mesopotamian and Hittite divinatory practice and in drawing attention to the Hittite extispicy corpus as a subject worthy of independent investigation.

The systematic study of Hittite extispicy terminology begins with E. Laroche’s two fundamental articles: “Elements d'haruspicine hittite” (Laroche E. 1952) and “Sur le vocabulaire de l'haruspicine hittite” (Laroche E. 1970). These studies assembled the key technical terms, identified many as Hurrian in origin, and established correspondences with the Mesopotamian omen vocabulary. His work remained the standard reference for more than two decades and is the explicit point of departure for all subsequent scholarship.

The most comprehensive and systematic analysis of Hittite SU oracle terminology to date is M. Schuol, “Die Terminologie des hethitischen SU-Orakels” (Schuol M. 1994). Working from the Middle Hittite text KBo 16.97 (dated to the beginning of the 14^th century BCE on palaeographic and linguistic grounds) and with systematic reference to the 13^th-century corpus and the Akkadian comparanda, Schuol investigated the technical vocabulary in unprecedented detail. Her study encompasses the chronological stratification of the corpus, the transmission history of technical terms (Hurrian loan-words, Luwian intermediary forms, Hittite loan-translations of Akkadian concepts), detailed anatomical and veterinary identifications, and the structure of oracle reports.

An important contribution to the understanding of Hurrian extispicy and terminology has been published by S. de Martino in 1992 (Martino S. de 1992a).

More recently, several overviews were published, each listing additional literature (see, e.g., Haas V. 2008a, 56–61; Beal R.H. 2002f, 59–64; Hout Th.P.J. van den 2003e). Particularly significant is Beal’s study, which provides a synthesis of the procedural logic behind the SU oracle reports, clarifying how individual signs were evaluated and how the final verdict was determined.

Another notable study is an article by Y. Cohen that focuses on the Hittite šašta-oracle, which involves observing the behavior of two rams and is always followed by a SU oracle. Building upon the work of Schuol and Beal, Cohen shows that in oracle inquiries utilizing both the šašta- and SU techniques, there is a correlation between the ram’s nocturnal behavior – specifically its sleeping positions, movements, and activity at the kamzuriti and pakmariti stations or procedures – and the results of the subsequent SU oracle.

3. Date and Findspots

The vast majority of oracle texts found at the Hittite capital Ḫattuša (modern Boğazköy) date from the second half of the 13^th century BCE, i.e., to the last generations of the Hittite Empire. The records of oracle inquiries conducted through extispicy, or a combination of extispicy and other techniques, primarily originate from buildings (or, the area of buildings) on Büyükkale (mostly A, E, C, and D). Another group of records comes from Temple 1 in the Lower City, and the so-called house at the slope („Haus am Hang’). Outside Ḫattuša, reports of oracle inquiries using this technique have been found at Alalaḫ (Tell Atchana), Kayalıpınar (Šamuḫa), and Kușaklı (Šarišša). This distribution shows that extispicy was practiced not only at the royal court in Hattuša but also in provincial towns.

The two oldest surviving extispicy reports – KBo 16.97+ and KBo 8.55+ – can be dated to the beginning of the 14^th century BCE on palaeographic and linguistic grounds (Schuol M. 1994a, 94–96; Hout Th.P.J. van den 2003e). KBo 16.97, the better-preserved of the two, was the basis for Schuol’s fundamental study of extispicy terminology (Schuol M. 1994a); it is also the Middle Hittite oracle text published as AhT 22 by Beckman in Beckman G. et al. 2011, 220–233. Its distinctive features – fully written-out Hurrian technical terms, archaic spellings, and references to persons and places datable to the period of Arnuwanda I – place it in the earliest stratum of Hittite extispicy practice (Schuol M. 1994a, 94–101). Both texts have been found in the debris of tablets (“Tablettenschutt”) in the area of building A of Büyükkale.

The existence of clay liver models in Old Script from Ḫattuša (ed. Mouton A. 2015b; De Vos A. 2013a; for a discussion, see also Schuol M. 1994a, 84–87) demonstrates that extispicy was known among the Hittites even earlier. A reference to the oracle technique can further be found in a prayer attributed to the priest Kantuzzili, dating to the late 15th century BCE, which derives from an Old Hittite text and incorporates elements borrowed from Sumerian-Akkadian incantations (see Schwemer D. 2015c with further literature). In this text, the priest asks the deity to disclose his transgressions either through a dream, through a seeress (^MUNUSENSI), or ‘through the liver,’ i.e., extispicy (KUB 30.10 obv. 24′–28′). Interestingly, the text refers to the technique by the term ^UZUNÍG.GIG, which also appears in the oracle reports from the Middle Hittite period, whereas the younger texts refer to it by the terms SU^MEŠ „flesh (signs)’ or TE^MEŠ, used as an abbreviation for SU^MEŠ. An interesting reference to extispicy, performed to obtain a divine decision during an oath ceremony, is found in the treaty between a Middle-Hittite ruler, likely Arnuwanda I, and Ḫuhazalma, king of Arzawa. The text runs as follows (KBo 16.47 obv. 15′–18′): 15′ anda=ma=kan UDU-un kuwāpi kuewēn nu linkiya 16′ [ka]ttan kiššan daiwen kuitman=wa ḫanneššar arḫa nāwi ariyaweni nu=wa ^DUTU-SI tuel KUR-i [Ú-UL] parḫzi zik=ma ŠA ^DUTU-ŠI [KU]R-i lē parḫzi: “Furthermore, while we killed a sheep, we placed it under oath as follows: ‘before we have not received a decision by an oracle inquiry, His Majesty will not attack your country, and you will not attack the country of His Majesty’” (ed. Martino S. de 1996a, 63–72). The text not only indicates that extispicy was performed during the middle Hittite period, but also that it was used in the context of treaty-making and oath-swearing with foreign rulers.

4. Topics

Like other Hittite oracle techniques, sheep extispicy was used to investigate the divine will in situations of crisis and at moments of major decision-making. The oracle functioned both as a diagnostic tool – “What is wrong? Which deity is angry? Why?” – and as a means of negotiating with the gods: “Do you, O deity, approve of this course of action?.” The topics investigated span the full range of royal concerns.

Military matters form a significant cluster: campaigns against enemy territories, route and winter-quarter choices, and questions about royal safety during operations. Matters of royal health – including the identification of the responsible deity and the required ritual remedy — are equally prominent; a single inquiry might proceed through many stages, gradually narrowing down the deity and the offense. Cultic omissions and their consequences are also frequently investigated: oracle texts ask whether a festival has been neglected, whether offerings have been improperly handled, or whether divine anger stems from some specific transgression. KBo 16.97 vividly illustrates this range of concerns, encompassing questions about military operations, cultic festivals, the welfare of the queen, and the identity of angry deities, all addressed in rapid succession within a single consultation (see Beckman in: Beckman G. − Bryce T. − Cline E. 2011a, 221–233). It is important to note that the preserved corpus tells us almost exclusively about royal and courtly divination: all SU oracle texts – as well as those using other techniques – stem from royal tablet collections and provide little information about the possible use of divinatory techniques among the broader population.

5. Experts

Sheep extispicy was performed by the diviner, whose title is written with the Sumerograms ^LÚḪAL or, rarely, ^LÚAZU, both corresponding to Akkadian bārû. The Hittite reading of this title is unknown. This specialist also performed various Hittite magical and purification rituals, reflecting the close connection between divinatory practices and apotropaic activities in Hittite religious life.

The diviner is mentioned by name only infrequently in the extispicy reports, unlike the augur, who regularly signs his name to bird-oracle observations. The diviner was also responsible for performing the ‘cave-bird oracle’ (MUŠEN ḪURRI), which appears to be a variant of extispicy involving a specific cave-dwelling bird rather than a sheep. Additionally, he conducted the šašta- or ‘bed oracle,’ which involved observing rams in a pen before they were slaughtered for the inspection of their entrails.

6. The Oracle Procedure

6.1 The Oracle Question

Before the sheep was sacrificed, a formal question was posed, which the gods could answer with either a ‘yes’ or ‘no.’ Afterward, the diviner specified the outcome he desired. If he sought the gods’ approval for a plan or wished to confirm that everything was in order and no danger was expected, he requested a favorable result (‘let the exta be favorable’). Conversely, if he suspected the gods were angry, disapproved of a plan, or if there was imminent danger, he would ask for an unfavorable outcome to reveal the truth (‘let the exta be unfavorable’). This situation also applied when a previous oracle yielded unfavorable results, prompting a second oracle as a cross-check (referred to as IGI-anda kallaranni, which means ‘as a countercheck for the misfortune’).

At times, two consecutive observations were conducted to eliminate any threats or to be absolutely sure that the gods were not displeased. In such instances, the standard formula is: ‘Let the first exta be favorable (SIG₅-ru , Hittite lazziyattaru) and the second unfavorable (NU.SIG₅-du, Hittite kallarešdu).’

In all cases, a match of the requested and actual outcome was interpreted as a ‘yes,’ whereas a mismatch meant a ‘no.’

6.2 The Canonical Inspection Sequence

The diviner inspected the liver and other organs in a fixed canonical sequence that broadly mirrors Mesopotamian practice, though with important differences in both order and terminology (Beal R.H. 2002f, 59; Schuol M. 1994a, 87–88; Christiansen B. forthcoming). For the standard order, see table 9.1.

6.3 The pars familiaris and pars hostilis Distinction

The fundamental spatial distinction underlying the evaluation system is between the right (ZAG, Hittite kunna-) and left (GÙB, (Hittite?/Luw. ipala-) sides of the liver. These sides correspond to what Roman haruspicy called the pars familiaris (right = favorable for the questioner) and the pars hostilis (left = favorable zone for the enemy, unfavorable for the questioner). A normal, healthy finding on the right side is a favorable sign for the questioner; a pathological alteration on the right side is unfavorable for them. Conversely, a pathological alteration on the left side is unfavorable for the questioner’s enemy and therefore favorable for the questioner. This applies to many standard observation terms: a feature ‘beaten’ (walhant-, Akk. MAḪIṢ) on the right was unfavorable, while ‘beaten on the left’ appears exclusively in favorable oracle outcomes (Schuol 1994b, 292; Beal R.H. 2002f, 60).

Within each side, a further distinction is made between the upper (šer, UGU) and lower (kattan, GAM) zones. The direction of bending, or the orientation, of a feature is also regularly noted and evaluated.

6.4 How the Final Verdict Was Reached: The ‘Single Bad Sign’ Rule

The interpretive logic of the SU oracle is not a simple arithmetic sum of positive and negative signs. Conversely, the system operated according to a ‘single bad sign’ principle: a single unfavorable anomaly was sufficient to render the entire oracle result unfavorable. In favorable answers, all meaningful deviations recorded are, without exception, signs that appear only in favorable outcomes elsewhere in the corpus. In unfavorable answers, it was typical for the scribe to record the unfavorable anomaly immediately and stop. This is why unfavorable answers in the texts are characteristically much shorter than favorable ones (Beal R.H. 2002f, 63–64).

Cohen Y. 2024a, 53 summarizes these principles as follows: “The result of the flesh oracle will be determined by the presence (positive) or absence (negative) of the nipašuri, šintaḫi, tanani, and keldi organs; the right or left position of the ^gištukul or the Throne; the appearance of the zizaḫi (usually positive), the position of níg (positive); the appearance of the Path (positive), a gall-bladder which is ḫilipšiman (negative); the appearance of the enti to the left (positive); the right position of the mazeri (positive); the presence of the irkepelli (negative); the number of coils (10 or 12) of the intestines (positive); the presence of the sag.me (negative); and the presence of the šuri (a positive sign, as well as a „joker’, which can overturn the result, like the nipḫu in Babylonian extispicy).’

6.5 Inconclusive Results: IZKIM NU.KIN

On rare occasions, no meaningful sign was found. Such cases were recorded with the formula IZKIM NU.KIN ‘no decision, no (divine) sign.’ A particularly important case is when the nipašuri- was absent (ni UL ešzi or NU.GÁL); this sometimes, though not always, led to the termination of the inspection, paralleling the role of the Mesopotamian manzāzu. Cf., e.g., KUB 5.1 rev. IV 81: ni UL e-eš-zi NU.SIG₅ IZKIM NU.KIN ‘ni(pašuri)- does not exist. No (divine) sign.’ Cases where the oracle procedure was continued can be found in KUB 16.29 obv. 10: IGI-zi SU^MEŠ ni KIN ši ta EGIR SU^MES ni NU.KIN ir-liš ‘The first exta: ni(pašuri-) exists, ši(ntaḫi-), ta(nani-). The second exta: ni(pašuri-) does not exist, irliš-.’ (for further attestations, see Schuol M. 1994a, 249). This clearly indicates that the term cannot refer to the liver as a whole, as suggested by Laroche E. 1970b, 131 (see also the criticism by Schuol M. 1994a, 249).

6.6 Cross-Checking with Other Oracle Types

Frequently, the SU oracle was combined with other oracle techniques which either preceded it, like the šašta- oracle or followed it. After an oracle was completed on a given question with one technique, one or two further oracle types were applied to verify or falsify the result. The formulaic introduction of a cross-check with a different technique is IŠ-TU [expert] ERTUM QA-TAM-MA-pát/KI.MIN ‘through the [expert], the very same/ditto.’ A countercheck using the same technique, or a technique performed by the same expert, was introduced by the formula IGI-anda SAG.KI-(an)za DAB-annaš, i.e., menaḫḫanda ḫanza appannaš (see Sakuma 2025a).

7. Terminology

The technical vocabulary of the Hittite sheep extispicy is strikingly multilingual and reflects its complex transmission history. The overarching designation SU^MEŠ is Sumerian; which can also be replaced by TE^MEŠ (deriving from têrtu ‘sign’, pl. têrētu „sign’). This term is to be explained as a short-hand writing of SU^MEŠ, since the sign TE only differs from the sign SU by ending with only one vertical instead of two.

Many of the names for features of the liver and other organs, as well as anomalies, are Hurrian loan-words, some of which were further modified through Cuneiform Luwian before entering Hittite usage; a few terms are loan-translations of Akkadian terms; and a small number appear to be native Hittite (Schuol M. 1994a, 87–88). An important chronological feature is that Middle Hittite texts (notably KBo 16.97+ and KBo 8.55+) write all technical terms fully in syllabic spelling, while the later 13^th-century tablets progressively replace these with abbreviated forms or logographic writings (Schuol M. 1994a, 98–101). For the list of terms, their meanings, and Akkadian counterparts, see section 9.

8. Difficulties

The study of Hittite extispicy encounters several interconnected challenges. One major issue is preservation: most oracle tablets from Ḫattuša are only partially intact, and the locations where many of these tablets were found remain unknown. The smaller tablets, which were likely produced during or immediately after the oracle sessions, are often written in a messy script and are usually in poor condition. In contrast, larger two-column tablets, which were compiled later from working notes, tend to be better preserved. However, they often compress the data significantly.

One additional challenge is terminology. Despite Schuol’s comprehensive analysis (Schuol M. 1994a), many technical terms still lack clear definitions or can only be estimated. The exact anatomical locations of liver features and the definitions of terms for anomalies are still a subject of debate. This uncertainty arises partly because the Hittite evidence does not always align with known Babylonian parallels, and partly because the meanings of the Hurrian terms are unclear.

9. Glossary of Technical Terms in Hittite SU Oracle Texts

The following table lists the major technical terms attested in Hittite SU oracle reports, with their Hittite (abbreviated) forms, language of origin, probable anatomical identification, Akkadian counterparts, and main English translations used in scholarship. Entries are arranged in the standard order of inspection, followed in Hittite extispicy reports.

9.1 Hittite Liver Divination Terms (in the sequence of their observation)

9.2 Observation terms (natural features and anomalies)

10. Selected Literature

• Beal R.H. 2002f. Hittite Oracles. In: L. Ciraolo and J. Seidel (eds.), Magic and Divination in the Ancient World (Ancient Magic and Divination 2). Leiden/Boston/Koln: Brill, 57-81.
• Boissier, A. 1935. Mantique babylonienne et mantique hittite. Paris: Librairie Orientaliste Paul Geuthner.
• Cohen, Y. 2024a. “The Hittite Šašta Oracles: A Presentation of the Sources and An Analysis of the Šašta Oracle Procedure.” In Y. Cohen et al. (eds.), “Let the Tabarna, the King, Be Dear to the Gods,” IOS Annual 24. Leiden/Boston, 28–61.
• Goetze A. 1947. Old Babylonian Omen Texts (YOS 10). New Haven.
• Goetze A. 1957. Reports on Acts of Extispicy from Old Babylonian and Kassite Times, JCS 11, 89-105.
• Hout Th.P.J. van den 2003e. Orakel (Oracle). B. Bei den Hethitern. Reallexikon der Assyriologie 10, 118-124.
• Jeyes U. 1978. The “Palace Gate” of the Liver. A Study of Terminology and Methods in Babylonian Extispicy, Journal of Cuneiform Studies 30, 209-233.
• Jeyes U. 1980. The Act of Extispicy in Ancient Mesopotamia: An Outline, in: B. Alster (ed.) Assyriological Miscellanies 1, Copenhagen, 13–32.
• Jeyes, U. 1989. Old Babylonian Extispicy: Omen Texts in the British Museum. Istanbul.
• Laroche E. 1952. Éléments d’haruspicine hittite. Revue Assyriologique 54, 19-48.
• Laroche E. 1970. Sur le vocabulaire de l'haruspicine hittite. Revue Assyriologique 64, 127-139.
• Leiderer R. 1990. Anatomie der Schafsleber im babylonischen Leberorakel. Munich.
• Meyer J.-W. 1987. Untersuchungen zu den Tonlebermodellen aus dem Alten Orient (Alter Orient und Altes Testament 39), Kevelaer/Neukirchen-Vluyn.
• Nougayrol J. 1941. Textes hépatoscopiques d’époques anciennes conservés au Musée du Louvre (I), Revue Assyriologique 38, 67-88.
• Nougayrol J. 1945–1946. Textes hépatoscopiques d’époques anciennes conservés au Musée du Louvre (II), Revue Assyriologique 41, 56-97.
• Nougayrol J. 1950. Textes hépatoscopiques d’époques anciennes conservés au Musée du Louvre (III), Revue Assyriologique 44, 12-14.
• Nougayrol J. 1968. Le foie “d'orientation” BM 50494, Revue Assyriologique 62, 31-50
• Nougayrol J. 1967. Rapports paléo-babyloniennes d’haruspices, Journal of Cuneiform Studies 21, 219-235
• Nougayrol J. 1969. Nouveaux textes sur le ziḫḫu (I), Revue Assyriologique 63, 149–157.
• Nougayrol J. 1971. Nouveaux textes sur le ziḫḫu (II), Revue Assyriologique 65, 67-84.
• Nougayrol J. 1955. Les rapports des haruspicines étrusques et assyro-babyloniennes, Comptes rendus des séances de l'Académie des inscriptions et belles-lettres, Paris 1955, 509-517.
• Schuol, M. 1994a. Die Terminologie des hethitischen SU-Orakels. Eine Untersuchung auf der Grundlage des mittelhethitischen Textes KBo XVI 97 unter vergleichender Berücksichtigung akkadischer Orakeltexte und Lebermodelle, Altorientalische Forschungen 21, 73-124 (part 1); 219-320 (part 2).
• Starr I. 1977. Notes on Some Published and Unpublished Historical Omens, Journal of Cuneiform Studies 29, 157-166.
• Starr I. 1978–1979. Omen Texts Concerning Holes in the Liver, Archiv für Orientforschung 26, 45-55.
• Starr I. 1985. Historical Omens Concerning Ashurbanipal’s War against Elam, Archiv für Orientforschung 32, 60-67.