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

The Hittite šašta- Oracle

1. Short Description

The šašta- oracle, along with the snake oracle (MUŠ), represents one of the rarest types of oracle found in the Hittite corpus. Currently, only 22 fragmentary tablets are known, which document this oracle type that was performed by the diviner (^LÚḪAL), who also conducted the so-called ‘flesh’ or SU oracle (sheep extispicy), and the so-called ‘cave-bird’ or MUŠEN ḪURRI oracle. The šašta oracle involved observing the behavior of two rams before they were slaughtered for extispicy. Unlike extispicy, which can also be a standalone oracular practice, the šašta- oracle is always linked to the subsequent examination of the rams’ entrails and is considered a single unit with it. This is further evident from the fact that the final divine verdict is based on both the findings of the šašta- oracle and those of the SU oracle.

2. Research History

This oracle type was first identified by E. Laroche as a distinct oracular practice which he called “clinomancie” and listed under number CTH 576 in his “Catalogue des textes hittites” (Laroche E. 1971c, 99). The term is a compound of the Greek word for ‘bed’ (klinē) and the suffix -mancie, which derives from the Greek word manteia, meaning ‘divination’ and draws on other French terms for oracular practices such as cléromancie (‘lot oracle’) and lécanomancie, which derives from Greek lekanê ‘bowl, pool’ and involves the observation of oil or water poured into a bowl or on a plate for divinatory purposes. The title for this oracle type is based on the crucial role of the bed or sleeping place (Hittite šašta-) in this particular type of oracle.

Laroche identified only three texts in this category: KUB 16.40; KUB 18.11 (erroneously classified as KUB 18.1; see Berman H. 1982b, 124), which he considered duplicates, and KUB 6.27, which he regarded as a parallel to the other two texts. In the following years, other scholars identified other texts or text passages that report oracle procedures carried out with this technique. Kammenhuber A. 1976c, 13, provided an incomplete list of texts, to which Berman H. 1982b, 124–126, added several further texts. Recently, Cohen Y 2024a, 32 updated the list, which now contains 21 entries. The only additional entry for this list is KUB 18.45, which Berman had already included but Cohen had mistakenly left out.

It is important to note that all of these texts document oracle inquiries that incorporate multiple techniques. Thus, the šašta- oracle is always followed by extispicy, performed with the same rams whose behavior had been observed before slaughter. Additionally, both procedures are often combined with KIN oracles, ‘cave bird’ oracles (MUŠEN ḪURRI), and bird oracles.

The texts were first studied in detail by H. Hoffner (Hoffner H.A. 1993a) and Y. Cohen (Cohen Y. 2024a). Hoffner focuses on the terminology and origin of the text, arguing that the technique ultimately derives from the Old-Babylonian šumma immeru omen tradition and reached the Hittites via the Hurrians. Yohen’s study encompasses editions of six texts and provides a detailed discussion of the oracle type, particularly its connection to the SU oracles and the grounds on which the experts arrived at the final divine verdict.

3. Date and Findspot of the Texts

All reports of oracle inquiries that involved the šašta- technique have been found in the Hittite capital Ḫattuša (modern Boğazköy). Many of the tablets lack information about their original storage locations. This is partly because the exact sites of discovery were not documented during excavations before 1931, and later finds were often found in the debris of earlier excavations. Additionally, many oracle tablets that came to light in later excavations were not found in archival contexts but in debris between buildings, suggesting that they had already been discarded by the Hittites themselves. The tablets with a known findspot come from Büyükkale Buildings A (KBo 34.139) and F (KBo 46.71), the House on the Slope (‘Haus am Hang’; KBo 49.18), and Temple 1 (L/19 debris; KBo 57.125 + KUB 18.11).

Based on palaeographic features, all texts can be dated to the New Hittite or Late Hittite periods (from the middle of the 14^th to the end of the 13^th century). Dating based on content is difficult and often impossible due to fragmentary preservation and limited or lacking historical context. Considering that oracle texts were typically not preserved over several generations, the majority of the texts likely date back to the second half of the 13^th century, specifically the reigns of Ḫattušili III and Tutḫaliya IV.

4. Topics

In terms of content, the texts address topics that are also found in other oracle reports. They include military campaigns (KUB 50.90; KUB 16.40; KUB 6.19+; KUB 6.14), the celebration of important cultic festivals, such as the AN.DAḪ.ŠUM festival and the purulli- festival (KBo 57.125+), and potential dangers to the king, such as threats during campaigns (KUB 50.90), divine anger and illness (KUB 6.14; KBo 34.139), and unspecified evil and ominous signs (KBo 46.71).

5. Expert and technique

The šašta- oracle was conducted by the diviner (^LÚḪAL), who also conducted sheep extispicy (also known as ‘flesh or SU oracles’) and the ‘cave bird’ oracle (MUŠEN ḪURRI). This expert submitted, on behalf of his client, a specific yes/no question to the gods, which was referred to by the Akkadogram ERTUM, an abbreviation of Akkadian erištu(m) ‘wish, request.’ The procedure was conducted with two rams, referred to in logographic writing as UDU.ŠIR-iš or in its abbreviated form UDU-iš. These spellings indicate that the Hittite word, whose pronunciation remains unknown, is an i-stem, whereas the word for sheep is a u-stem (see Hoffner H.A. 1993a, 117).

The oracle procedure unfolded in a fixed sequence governed by a strict internal logic.

Step 1) Posing the question: The diviner formulates the oracle question – either afresh or by repeating the previous session’s query or the query conducted with another technique (IŠ-TU ^LÚḪAL ER-TUM QA-TAM-MA-pat “the same question through the diviner”) – and announces the expected result. In most cases, the expert asks the first ram to give a favorable and the second one an unfavorable result (see, e.g., KUB 50.90 obv. 3′–4′ IGI-ziš UDU-iš SIG₅-ru EGIR-ma NU.SIG₅-du “Let the first ram be favorable and the second be unfavorable”). Sometimes, both rams are asked to produce a favorable result (see, e.g., KUB 50.90 11′ 2 UDU-uš SIG₅-ru “Let the two rams be favorable’). There is also one inquiry that asks the first ram to yield an unfavorable result and the second a favorable result (KUB 6.19+ l. col. 16′ [… IGI-ziš] UDU-iš NU.SIG₅-du EGIR-ma SIG₅-ru “Let the [first] ram be unfavorable and the second one favorable”).

Step 2) Observing the rams: The rams are observed at night in their sleeping place (šašta-). The observation if further described by the terms IGI-zi/ḫantezzi ḫāli and EGIR-zi-/appezzi ḫāli, which might either be translated as ‘(in/during) the first nightwatch’ and ‘(in/during) the second nightwatch’ or ‘in the first pen’ and ‘in the second pen.’ Whereas HW² III: Ḫ, 30 translates ḫāli as ‘pen’, Cohen Y 2024a holds the meaning ‘nightwatch’ more likely. Choosing between the two is challenging, especially since, except for one instance (KUB 16.29+ 19–20), the rams' behavior is documented only in or during the first ḫāli. At their first sleeping place (IGI-zi-/ḫantezzi- šašta-), the rams lie down either on their right or left side – the key datum that the diviner records. Movement thereafter is equally important: a ram may shift from right to left, left to right, stay on the same side, or remain still. These movements and positions constitute the primary ‘sign’ of the procedure.

Step 3) kamzuriti and pakmariti: After observing the sleeping positions, the diviner records what happens at two additional stations or processes called kamzuriti and pakmariti. The precise meaning of these terms, which appear exclusively in šašta- oracles, is unknown. A non-Hittite, presumably Hurrian, origin for the two terms is reasonable, given their likely path of transmission. This is further supported by a Glossenkeil that precedes kamzuriti in KBo 39.56 l. col. 6. In the vast majority of cases, the ram does nothing (UL kuiški DÙ-at) at these two stations. Occasionally, exceptional actions are recorded: the ram turns itself over, sticks out its tongue and bites it, gnashes its teeth, or hits its side with its leg.

These descriptions suggest that the two terms designate specific stages or processes in the killing or sacrificial procedure (Hoffner H.A. 1993a, 119). According to Cohen, the terms might refer to offerings of different kinds given in preparation for the following extispicy procedure (Cohen Y. 2024a, 52).

Step 4) Slaughter and extispicy (SU): As the šašta- procedure is typically followed by a SU oracle, and reports do not document a result from the šašta- procedure, it is reasonable to assume that the oracle’s result is derived from the combination of both techniques. By comparing the oracle findings with the final verdict, Cohen Y. 2024a, 53 identified the following rules used by the experts to arrive at their conclusion.

Favorable verdicts are associated with the ram moving during the observations and ending on the right side, and with any action at pakmariti (as opposed to ‘doing nothing’); unfavorable verdicts are associated with the ram ending on the left side, being inactive throughout, and with any action at kamzuriti (as opposed to ‘doing nothing’).

The result of the SU oracle – determined by the presence or absence of specific key liver parts or marks, the location of certain features, particularly in relation to the right and left side, the condition of the gallbladder (with ḫilipšiman interpreted as a negative sign), and the number of coils of the intestines – provides the definitive positive or negative answer. The šašta- procedure and the flesh oracle are, according to Cohen, perfectly correlated: a ram whose behavior was positive will produce favorable exta, and vice versa.

The underlying logic is that the ram’s behavior at night anticipates the appearance of its inner organs. Observing the living animal in its sleeping place was considered a preparatory step for the post-mortem examination.

6. Origin

As demonstrated by Hoffner H.A. 1993a and Cohen Y. 2024a, the Hittite šašta- technique shows similarities to the Old-Babylonian šumma immeru omens. Phraseological parallels in specific sheep behavior descriptions suggest that the Hittite šašta- oracle ultimately derived from the Old-Babylonian tradition and reached the Hittites via Hurrian transmission. However, there are clear differences between the Mesopotamian and Hittite evidence. The Akkadian šumma immeru omens record similar animal actions but are structured as omen texts with protasis and apodosis of the type ‘if … then …’, whereas the Hittite šašta- oracles are reports of oracle inquiries. Furthermore, the Mesopotamian šumma immeru omens are not linked to extispicy omens, although there is some evidence suggesting that, in Mesopotamian divination practice, both oracle methods might have been linked as well.

As Cohen Y. 2024a highlights, the Hittite šašta- oracles bear an even closer similarity to a specific set of Akkadian omens from Tigunānum, located on the Upper Euphrates and dating to around the 17^th century BCE. Unlike the Mesopotamian šumma immeru omens, the omens from Tigunānum reference oracle practices and describe how the oracle expert would lock an ewe in a pen, seal the door, and inspect the animal the following day. In contrast to the Hittite šašta- oracles, however, the behavior of the ewe in these omens is articulated in protasis and apodosis, similar to the Mesopotamian tradition.

7. Key Terminology