Akkadian and Akkadian Texts in Hittite Anatolia

F. Giusfredi; V. Pisaniello

Jump to Content

Chapter 8 Akkadian and Akkadian Texts in Hittite Anatolia

In: Contacts of Languages and Peoples in the Hittite and Post-Hittite World

Authors:

F. Giusfredi

F. Giusfredi
Search for other papers by F. Giusfredi in
Current site
Google Scholar
PubMed

and

V. Pisaniello

V. Pisaniello
Search for other papers by V. Pisaniello in
Current site
Google Scholar
PubMed

Type:: Chapter

Pages:: 206–241

DOI:: https://doi.org/10.1163/9789004548633_009

Open Access

Download PDF

1 Previous Studies on the Akkadian of the Hattuša Archives

Akkadian is one of the famous acht Sprachen of the Hittite archives that were identified by Forrer (1922). Its role, however, is by the very nature of the ancient Mesopotamian scholarship different from that of most of the other idioms that are represented in the Hattuša archives. Indeed, while Palaic, Luwian, and Hattian have a specific local significance in the mixed cultural environment of Bronze Age Anatolia, and Hurrian and ‘Boğazköy Indo-Aryan’ are introduced through a historically specific and well-defined pattern of interactions with the Mittanian world, both Akkadian and Sumerian are fundamental elements of the cuneiform koiné that constitutes the core of the historical object that we generally call the ancient Near East.

The original introduction of Akkadian (and, in general, of the Sumero-Akkadian Mesopotamian cultural tradition) in Anatolia has been discussed in Chapter 7. But after the first cultural ‘quasi colonization’ during the Middle Bronze Age phase of the Old Assyrian markets and the complex and still foggy circumstances of a reintroduction at the origin of the history of Hatti, the role of Akkadian in the Hittite archives does not appear to be monolithic and homogenous over time. Of course, Akkadian does have features that appear consistently in the different types of documents and diachronic stages (cf. Wilhelmi 2011:260–273). These include the inverted order of genitive and noun, with the reduplication of the marker of possession, which seems to calque very closely the structure of Hittite, and some peculiarities that emerge in different varieties of peripheral Akkadian, such as the confusion of case endings in nouns and especially pronouns, mistakes in the use of dual number and gender, imperfect agreement in noun phrases, and inconsistencies in the use of the subjunctive and ventive in verbal forms. Although these commonalities may give the impression that a single grapholect of Akkadian existed, they are relatively trivial features. Some are defined sociolinguistically by the non-native competence of the scribes or authors of the texts and others arise from very general mechanisms of imperfect learning.

1.1 Boğazköy Akkadian and Peripheral Akkadian

The study of the Akkadian of Boğazköy began relatively early.¹ Only a few decades and years after the attention of the scholars was drawn to the existence of peripheral Akkadian archives from Nuzi and Ugarit, respectively, and before those of Mari and Alalah were even unearthed, R. Labat (1932) devoted his doctoral dissertation to the Akkadian texts coming from the Hittite archives. His work represents a solid effort to describe the writing conventions, phonology (wisely separated from the previous category), morphology, and syntax of the texts. For nearly forty years after this outstanding seminal effort, the works on ‘peripheral Akkadian’ were mostly focused on the Syrian archives, and studies of Boğazköy Akkadian were chiefly philological. One should, of course, mention F. Sommer and A. Falkenstein’s edition of the Political Testament in 1938. In the same year, G. Meier published an Akkadian healing ritual (KUB 29.58, Meier 1938), thereby inaugurating the study of the non-bilingual Akkadian documents that are more directly linked to a Mesopotamian tradition. The 37th volume of the Keilschrifturkunden aus Boghazköy by F. Köcher (1953) was influential in introducing this meta-genre to scholars. Lexical lists have been known since the 1910s and were discussed in some early volumes of the Materialien zum sumerischen Lexikon (MSL), but they were generally and reasonably examined from an Assyriological perspective. Although linguistic studies of Boğazköy Akkadian were not entirely absent, they tended to be unsystematic. Two very important contributions were E.H. Sturtevant’s 1938 study of the Sumerian and Akkadian words used in Hittite and W. von Soden’s 1973 comparative analysis of the iterative verbal forms in Akkadian and Hittite. However, comprehensive studies were not attempted.

The second extensive work on the Akkadian of the Hittite capital as a grapholectal phenomenon (even if the term ‘grapholect’ is a bit anachronistic for that stage, it is theoretically appropriate) was once again a dissertation. It appeared in 1976 and was authored by John W. Durham. While the dissertations by Labat and Durham are superficially similar in structure, Durham’s work is significantly longer and much more detailed than Labat’s, but, at a closer look, it is less structured from a theoretical standpoint. Almost 450 pages are devoted to graphemics and phonology, which is hardly unexpected: not only is the structure of the syllabary the first barrier when one tries to examine an extinguished corpus language, but also the relationship between graphemic and linguistic competence in the culture of the cuneiform world is highly complex. In comparison, Durham devotes only 60 pages to the grammatical aspects of Akkadian. These pages consist of an overview of nominals and nouns (verbs are treated earlier in the book, in a section dedicated to the graphophonemic treatment of vowels and semivowels). Half a century later, Durham’s work still represents a precious source of information and examples; however, its biggest shortcoming is the lack of a general theorization. It contains a short introduction and some brief ‘comparative notes’ at the end of significant sections, but no proper ‘synthetic’ conclusions. References to other corpora are frequent, but there is no obvious attempt to cross-reference them to identify tendencies, patterns of influence, or similarities between central and peripheral scribal praxes.

Entirely different is the work that M. Marazzi (1986) dedicated to the use of Akkadian in the Old Hittite texts. Rather than considering all of the available examples, even for the earlier phases, the author concentrates on only three texts: the Political Testament, the Siege of Uršu, and the Annals of Hattušili I. Marazzi’s work is scientifically more daring than Durham’s and does not neglect theorization: it investigates interference-driven mistakes both from the linguistic and philological perspective, including the linguistic identity of the authors/scribes and the direction of translation (in the cases in which a translation may or must have occurred).

In 1976, when Durham completed his dissertation, 43 volumes of the KUB series and 23 volumes of the KBo series had been published. More volumes were to come, and, from the 1980s, more and more Akkadian texts were published in hand copies in the main series of Hittite epigraphy. The KBo volumes 28 (1985, by H.M. Kümmel) and 36 (1991, by G. Wilhelm) contain non-literary and literary texts from Mesopotamian traditions. Texts of different types appeared in other volumes of both series, as well as in other minor series of cuneiform hand copies. In this phase, research in the field became more specialized. By then it was clear that the different genres of Akkadian documents had to be addressed in different ways.

Finally, in recent years a doctoral dissertation was dedicated to the problem of Akkadian in the Hittite scribal world. L. Wilhelmi’s work (2011) is still unpublished but was kindly made available to us by the author, whom we wish to thank. It concentrates mostly on those documents that are certainly local productions (political documents, historical materials, and, above all, diplomatic texts such as letters and treaties).²

It would be unproductive here to sketch a history of the publications and editions of the different monolingual or bilingual documents in Akkadian: an attempt at cataloging the documents based on their genre will be made in § 2. What should not be neglected, however, is the presence of Akkadian texts produced from the Hittite scribal schools that have been found in archives outside Hatti. These texts exhibit some features in common with those of the central Hittite archives, whereas other features appear to be local. The texts from Karkemiš and Ugarit were the object of Huehnergard’s 1979 doctoral dissertation, which was the basis for his 1989 monograph on Ugarit Akkadian. This was followed, some twelve years later, by a book on the same subject by van Soldt (2001). An important, although often overlooked, follow-up article was authored in 1992 by H.J.M. van Deventer and P.J.J. Huÿssteen. This 1992 article has the merit of providing comparative information about several peripheral varieties of cuneiform Akkadian and a useful overview of the types of material.

A few general observations are in order. First, many of the main studies that are dedicated to Boğazköy Akkadian (Labat 1932; Durham 1976; Huehnergard 1979 and 1989; van Deventer and Huÿssteen 1992) focus on orthography and morphophonology, as well as the differences that emerge in the different peripheral traditions. Huehnergard places more emphasis on the peculiarities of the syntax: in his dissertation, after comparing some features that differ in Karkemiš and Ugarit, the author speaks of dialects, implicitly assuming that the western peripheral varieties are to be regarded as (admittedly elusive) dialectal varieties of the main language. In his 1989 work on Ugarit Akkadian, however, the sociolinguistic perspective is much more refined, and the concepts of local substrata and superstrata become much more relevant in the theoretical framework. As regards Hatti, reflections on the local peripheral varieties of Akkadian are also frequent in Durham’s dissertation, with references to corpora such as those from Alalah and Nuzi.

In general, neither a purely philological/historical nor purely linguistic approach can be expected to provide a fair representation of the Akkadian production in Boğazköy. In the next sections, we will discuss the problems inherent in categorization and then propose a categorization of the texts that compose the corpus (or corpora). After that, we will outline the characteristic features of the different Akkadian traditions.

1.2 Problems of Categorization

Given the large number of Akkadian documents in the Hittite archives, a presentation of the material cannot be attempted without attempting a categorization. Categorizing the Akkadian documents in the Hittite archives, however, is no trivial task. The simple fact of belonging to the Hittite scribal production would produce a large set of ‘local’ texts and very few documents that may have been composed elsewhere, a distinction that, due to the overwhelming majority of the former group, is not conducive to a comfortable presentation.³ The bilingual vs. monolingual criterion is more promising but is essentially epigraphic and weaker than other options, especially because the presence of more languages is directly correlated with specific types of compositions, as will be evident in the next section of the present chapter.

As for the paleographic criterion, it poses as many concerning issues as promising features. The oldest Hittite documents display a rather consistent type of cuneiform that mostly patterns with the ductus of Old Babylonian Alalah (Alalah VII).⁴ The evolution toward the Late Script (or New Script) ductus is paralleled beginning in the late 15th or early 14th century by the emergence of documents written in an Assyrian or Mittanian ductus. Furthermore, non-Hittite sign types that correlate with the Upper Mesopotamian and Syrian Mittanian traditions contribute to the syllabary of the latest version of the Late Hittite ductus (the so-called ductus IIIc).⁵ These ducti distribute in a fashion that still appears irregular with respect to text types. For instance, in a few cases a Middle Assyrian ductus emerges in the Hittite-Assyrian correspondence but does not prevail and is not limited to this subcorpus. Similarly, the very rare documents that present a Middle Babylonian ductus carry texts from the Mesopotamian literary and cultural tradition, but not all of the literary and cultural Mesopotamian texts found in Hattuša exhibit this kind of ductus.

All in all, research on the paleography of the Hattuša cuneiform has progressed, but the correlations among language, text, and script are still partly obscure, especially for the documents in foreign languages. Therefore, in the next section of this chapter, we will attempt a categorization of the Akkadian texts from the Hittite archives based on textual macro-categories, providing linguistic and, occasionally, paleographic information as necessary.

2 The Akkadian Texts from Boğazköy: A Categorization

The Sumerian and Akkadian texts present in the Hattuša archives, or found in the Hattuša excavations, represent several different texts and collections. Many studies are dedicated to the Mesopotamian literature found in Boğazköy but, with very few exceptions (e.g., Viano 2016 on the Sumerian texts), most are unsystematic in discussing the sources and significance of the texts. Furthermore, not all of the texts in Akkadian belong to a Mesopotamian literary tradition. Instead, most are locally produced historical, political, or diplomatic documents.

The Akkadian texts from the Hittite archives correspond to the following textual categories:

Political and administrative Akkadian:
1. Old Hittite political texts
2. Diplomatic texts
3. Landschenkungsurkunden
Akkadian of the cultural tradition:
1. ‘Scholastic’ texts belonging to the scribal curriculum
2. Texts concerning magic and medical technical knowledge
Royal inscriptions on votive objects

The first group includes the Akkadian version of the Annals of Hattušili I (CTH 4.I),⁶ the Political Testament of Hattušili I (CTH 6), the Siege of Uršu text (CTH 7),⁷ and the Akkadian version of the Edict of Telipinu (CTH 19.I).⁸ These are obviously local productions and are all, to different degrees and in different ways, bilingual. The Political Testament of Hattušili I is recorded on a bilingual tablet, with columns i and iv containing the Akkadian version and columns ii and iii the Hittite version. In contrast, the Hittite and Akkadian versions of the Annals of Hattušili I and Edict of Telipinu are recorded on separate tablets. Finally, the Siege of Uršu text is an Akkadian narrative with only some portions in Hittite; their relationship with the Akkadian text is one of complementarity rather than correspondence. It is a matter of debate whether the Siege of Uršu is a translation of a Hittite original (Kempinski 1983:34; Beckman 1995:27; Archi 2010:40–41) or rather one of the first attempts to adapt the cuneiform script to the Hittite language (van den Hout 2009a:92).

Two diplomatic texts dating to the Old Kingdom—the letter sent by Hattušili I to Tunip-Teššub of Tikunani (CTH 187), probably coming from Tikunani,⁹ and the treaty between Telipinu and Išputahšu of Kizzuwatna (CTH 21.I), of which three Hittite versions are found in the archives of Boğazköy—can also be placed in this group.

Akkadian was used as a lingua franca for diplomacy. Diplomatic texts written in Akkadian (particularly a peripheral Middle Babylonian dialect) include treaties with non-Anatolian partners and international correspondence. The Akkadian treaties that were found in Boğazköy are the treaty with Paddatiššu of Kizzuwatna (CTH 26, variously dated to the reigns of Alluwamna, Hantili II, or Huzziya II),¹⁰ as well as the treaties between Tahurwaili and Eheya of Kizzuwatna (CTH 29), Tuthaliya I and Šunaššura of Kizzuwatna (CTH 41.I), Tuthaliya I (?) and Tunip (CTH 135), Šuppiluliuma I and Aziru of Amurru (CTH 49.I), Šuppiluliuma I and Šattiwaza of Mittani (CTH 51.I), Šattiwaza of Mittani and Šuppiluliuma I (CTH 52.I), Šuppiluliuma I and Tette of Nuhašše (CTH 53), Muršili II and Tuppi-Teššub of Amurru (CTH 62.AA), Muwattalli II and Talmi-Šarruma of Aleppo (CTH 75), Hattušili III and Ramses II (CTH 91), and Hattušili III and Bentešina of Amurru (CTH 92).

We can add to this list some documents found at Ugarit, such as the treaty between Šuppiluliuma I and Niqmaddu II (CTH 46) and that between Muršili II and Niqmepa (CTH 66), as well as some royal verdicts or decrees on several local issues.

Akkadian treaties were recorded on monolingual tablets, but corresponding Hittite versions sometimes exist on separate tablets.¹¹ Letters were also monolingual, although Hittite copies (drafts or archival copies) are found sporadically. Akkadian international correspondence has been found both in Hittite and non-Hittite (especially Ugarit and Amarna) archives. It includes letters exchanged by Hittite kings or dignitaries and foreign addressees (both kings and vassals). The texts in this category mainly include correspondence with Ugarit (CTH 77, 110, 112–114), Egypt (CTH 153–170), Babylon and Assyria (CTH 172–174), Mittani (CTH 179), and Amurru (CTH 193).

Royal land grants or Landschenkungsurkunden (CTH 222) are paleographically Middle Hittite documents dating from the reign of Telipinu to that of Arnuwanda I. Those issued from the reigns of Telipinu to Muwattalli I are written in Akkadian, specifically, in an Old Babylonian dialect (with sporadic Hittite lexemes, which are mostly technical terms and topographical indications). The only land grant composed under Arnuwanda I and Ašmunikkal (KBo 5.7) maintains a formulaic Akkadian framework but contains more details in Hittite.¹² All of the land grants are original documents bearing a cuneiform seal of the king and display a fixed set of formulas: 1) an introductory clause mentioning the seal of the tabarna; 2) the grant formula with the name of the recipient; 3) a vindication clause; 4) a curse formula; and 5) a final clause indicating the place where the document was written, the name of the scribe, and the list of witnesses. Such a structure represents a Hittite re-elaboration of Syrian models: the verb našûm in the grant formula probably calques the Hittite šarā dā-,¹³ and the curse formula is only attested at Boğazköy.¹⁴

Among the Akkadian texts belonging to the cultural tradition, two major subgroups can be singled out. The first includes texts that were probably mainly (but perhaps not exclusively) used by Hittite scribes to learn Sumerian, Akkadian, and the cuneiform script, to a more or less extent degree. This subgroup includes lexical lists, hymns and prayers, epic narratives related to the Sargonic kings, mythological narratives (of both Mesopotamian and Hurrian origin), wisdom literature, and, tentatively, omen series, although the technical knowledge they contained might also have been relevant. These texts usually have translations in Hittite, either alongside Akkadian (and when present, Sumerian) on multilingual tablets (as is the case with lexical lists and many omen series) or on a separate tablet (as occurs with mythological narratives). In some cases, only the Hittite version is preserved, although an Akkadian model should be assumed. Furthermore, some of these compositions—particularly hymns and prayers—were probably used mainly as models from which new original texts in Hittite were drafted.

The second major subgroup consists of magic and medical texts, including rituals and incantations, medical treatises, pharmaceutical texts, etc., which we believe were mainly employed for the transmission of specific technical knowledge rather than for mastering the cuneiform script and Akkadian language (which, however, cannot be entirely excluded). Some of the rituals were also probably performed on specific occasions. Indeed, these compositions were only sporadically provided with Hittite translations, although they sometimes provided models for the composition of new texts in Hittite.

The compositions included in the two subgroups of the cultural tradition were generally foreign in origin although often drafted locally by Hittite scribes. Nevertheless, some imported tablets have been found. Therefore, these documents do not form a consistent and homogeneous collection, from the perspective of chronology, provenance, and paths of transmission; each document would require a separate discussion.

The last group of Akkadian documents includes brief royal inscriptions on votive objects, which are generally short and bear limited information relevant to this research. Only three documents belong to this category: the inscriptions on the dagger of Anitta, axe of Ammuna, and sword of Tuthaliya I. The dagger of Anitta, found at Kültepe, only bears the short Akkadian inscription É.GAL Anita ruba’im “the palace of Anitta, the king.”¹⁵ The inscription on the bronze axe of Ammuna (findspot unknown) shows the same ductus found on the Old and Middle Hittite seals, and the curse formula against forgery is the same (barring small variants) as the formula found on the tabarna seals on the Landschenkungsurkunden.¹⁶ Finally, the bronze sword of Tuthaliya I, found at Hattuša, was dedicated by the Hittite king to the Storm god after the victory over Aššuwa, as stated by the text of the inscription: inuma ^mDuthaliya LUGAL.GAL KUR ^URUAššuwa uhalliq GÍR^HI.A annûtim ana ^dIŠKUR bēlišu ušeli “When Tuthaliya, the Great King, destroyed the land of Aššuwa, he dedicated these swords to the Storm god, his lord.”¹⁷

Akkadian texts	Origin	Hittite versions
Old Hittite political texts	Local	Yes
Dipl. texts A: treaties	Local	Yes
Dipl. texts B: letters	Local/foreign	Drafts/archival copies
Land grants	Local	No
Scholastic texts A: Scribal	Foreign	Yes
Scholastic texts B: Magic	Foreign	Sporadic
Votive inscriptions	Local	No

3 The Akkadian of Politics and Administration

3.1 Old Hittite Political Texts

Old Hittite political documents, written in Akkadian or presenting an Akkadian version are, as previously discussed, a heterogeneous set. Most resist being ascribed to one of the typical genres of the later Hittite tradition. Telipinu’s text (CTH 19, MS), despite being labeled an edict, is a complex historical narrative; the Uršu text (CTH 7) and Hattušili’s political testament (CTH 6, NS) are unica, and the annals of the first Hittite king (CTH 4, NS) are the only decently preserved specimen of an Old Hittite annalistic tradition. However, despite their unique textual features, they are linguistically consistent, at least if we consider those that derive from the scribal activities of Hattuša (thus excluding the Tikunani letter, which was in all likelihood composed by a Syrian scribe).¹⁸ Old Hittite ‘political Akkadian’ is a variety of Old Babylonian that exhibits few or no significant traces of Middle Babylonian or Middle Assyrian traits.¹⁹

One of the oldest documents found in Hattuša is probably the Uršu text. It presents paleographic and graphemic similarities to the Alalah VII tablets (Klinger 1998:370–372; Archi 2010:40), while its language has been described as a variety of Old Babylonian with some forms that are attested in the archives of Mari (e.g., Beckman 1995b:28–29 on the form emqet for emqēta, neti for niati). The language of the Uršu text supports the important role of the Syrian cultural interface for the setup of the Hittite scribal praxis after the hiatus that separates the first historical Hittite dynasts from the Kārum and post-Kārum ages. As is well known, however, while its language could be labeled a form of North Syrian Old Babylonian, the Uršu text contains some traces of Hittite interference. These are not very indicative of the use of a linguistic mixed code by the speakers but belong to the easiest kind of contact phenomena identifiable in corpus languages. Most of the interference consists of lexical insertions of Hittite words in the Akkadian context. They should be described not as ‘loanwords’ but as ‘foreignisms,’ which are not integrated into the target language but rather represent artificial cases of code-switching. This pattern is consistent with a translation from Hittite to Akkadian by a Syrian scribe—an interpretation that combines the observations of Beckman (1995b) and Archi (2010). The fact that only one case of syntactic interference can be detected (obv. 18 ša šunūti ṭemšunu, calquing a Hittite double-marked possessive structure) is also consistent with this scenario as a single instance points to an occasional mistake rather than the linguistic predisposition of a native Hittite scribe.

With its atypical paleographic, graphemic, and linguistic profile, the Uršu text may have represented an exception to the usual relationship between Hittite and Akkadian in the Old Hittite documents. The most difficult problem, however, is trying to describe this relationship in the other, less peculiar political documents of this early phase. What complicates the issue is the impossibility of dating the available linguistic material precisely. Indeed, leaving aside the Uršu text, the extant manuscripts of the other Old Hittite political documents are generally relatively late. The Akkadian version of Telipinu’s edict may predate the Hittite ones but does not seem to have originated any earlier than the Middle Hittite paleographic facies.²⁰ The Akkadian versions of the Annals of Hattušili I and that of his Political Testament are both preserved in New Hittite copies. The available fragments of the Akkadian version of the Išputahšu Treaty (CTH 21) are also older than the Hittite fragments. They appear to exhibit a Middle Script ductus but according to Devecchi (2015:64) may be regarded as original documents from the age of Telipinu. However, the fragments are too small to permit a linguistic analysis.²¹ This complex situation implies that some features of the available documents do not stem from original Akkadian versions (setting aside the problem of the dating of the first Hittite versions, which is discussed in Chapter 6).

However, some features that must derive from later stages of the tradition can be identified confidently. For example, the use of writing conventions that are typically late, such as the syllabograms labeled by Devecchi (2005:86–87) as ‘non-babilonesi,’²² and the Middle Babylonian use of ŠÚ to write the third person singular pronoun can be regarded as modifications that occurred when Hattušili’s annals were copied in the pre-imperial and imperial ages. Other traits connect the language and writing of the Akkadian of the OH political bilinguals to earlier or peripheral Akkadian traditions (Devecchi 2005:84–109). Although Hittitisms are rare in the syntax and morphosyntax, they do exist (e.g., double-marked genitival appositions, following both a Hittite and non-Hittite word order).²³ However, Hittite loanwords or foreignisms (comparable to those in the Uršu text) are nowhere to be found.

Formulaic calques from a Hittite cultural environment do exist: one may consider, for instance, the form pahru ibbašu, which is used to render the Hittite taruppanteš ešer in Telipinu (e.g., KBo 3.1+KUB 11.1 i 4, 15, 25–26), or the expression ana sunišu iškunšu in the annals (KBo 10.1 obv. 13), which occurs only in the Old Assyrian corpora from Anatolia and therefore probably has a Hittite or other local, non-Mesopotamian origin. However, some formulas are typically Akkadian, as are some topoi that were collected and discussed by Steiner (1999) and Devecchi (2005:111–116). These include the use of the LUGAL GAL title (obviously Mesopotamian, at least formally), the epithet ‘beloved by the god’ (Akkadian narām DN, maintained as an Akkadogram in the Hittite Version KBo 10.2 i 27), the expression ‘my city GN’ (URU-ya GN in KBo 10.1 obv. 6), the topoi of the enemy king who is tied to the chariot (KBo 10.1 rev. 24–25), the god running before the king on the battlefield (KBo 10.1 rev. 13–14), and the designation of the king as a ‘lion’ (KBo 10.1 obv. 35).²⁴ It must be stressed, however, that the presence of formulas from Mesopotamian, Syrian, or local traditions testifies to the fact that the OH annalistic production was derivative from the literary point of view but says nothing about the linguistic profile of the scribe and does not provide conclusive evidence about the original language of the OH political texts.

Our view is that it is impossible to attempt to identify a single Anatolian or Mesopotamian tradition and original language for the compositions; instead, each of the main OH political documents may have a unique philological history. The task of identifying an ‘original version’ is made even harder by the fact that specific expressions or phrasemes present in one version of a text can be drastically modified or reduced in another, which qualifies the production of the text as something more complex than a literal translation.²⁵

However, a description of the type of Akkadian employed in these documents is feasible as the code seems to be relatively consistent within the subcorpus. If elements that belong to later stages of the tradition are eliminated, OH political Akkadian can be seen as a form of Old Babylonian that is characterized by influences from the Syrian area. The Annals of Hattušili I, for instance, employ the adjective kalû instead of the gabbu that prevails in Middle Babylonian. They also contain the formula ana warduti târu ‘to become a subject again’ (KBo 10.1 obv. 22), which, according to Devecchi (2005:66–67) is also known from the Syrian area (as does the writing ARAD^MEŠ for the abstract wardutu). Another formula of Syrian origin is, again according to Devecchi (2015:73), pānam u bābam ul išû ‘to have neither front nor gate’ (KBo 10.1 obv. 36).

In contrast, few features of the language of the OH political texts relate to the local Old Assyrian traditions. Some of these features pertain to the vocalism of inflected verbs (e.g., itarab from erēbu in KBo 10.1 obv. 48),²⁶ but could also have come from other peripheral traditions. A more striking case, syntactic in nature, is the use of raṭābu with the infinitive, meaning ‘to begin to,’ which is attested in Old Assyrian and present in both the Annals (KBo 10.1 obv. 14) and the Political Testament (KUB 1.11 obv. 16–17; cf. Devecchi 2005:62); however, this case, rather than indicating a link to the Old Assyrian scribal production, may have represented a calque of an Anatolian serial construction in both traditions. We can only suggest that between the penetration of Old Assyrian in Anatolia in the Middle Bronze Age and the introduction of Middle Babylonian cultural elements beginning in the 16th or 15th century, an intermediate phase must have existed, during which Mesopotamian cultural and linguistic material typical of the North Syrian Old Babylonian period was adopted in the Hittite scribal world, possibly in a gradual fashion.

3.2 The Landschenkungsurkunden

The land grants—often referred to by the German technical term Landschenkungsurkunden or LSU—represent a very peculiar type of Akkadian documents produced by Hittite scribes. These documents are unique, firstly, for their contents: since they record land assignments from the crown to specific noble families of the kingdom, they represent the main known corpus of administrative tablets that were produced in the Hittite world. According to the material presented by Rüster and Wilhelm (2012, pc. 72), land grants were written in Akkadian until the time of Muwattalli I,²⁷ a king who reigned in the 15th century BCE. They are chronologically close to some diplomatic documents, in particular, the Kizzuwatna treaties, which likewise date to the 15th century except for the treaty between Tuthaliya I and Šunaššura. Apart from the complex problem of defining a Middle Script ductus (Rüster and Wilhelm 2012:60–64), the paleography of the documents does not contradict that dating.

However, the language and the formulas employed in the LSU are peculiar. While alleged Middle Babylonian traits occur in the diplomatic documents as early as the Paddatiššu treaty (CTH 26),²⁸ the grants seem to adhere to Old Babylonian principles and forms. To some extent, this may indicate the maintenance of a linguistic tradition closer to that of the Akkadian version of the Old Hittite political documents that would perhaps dwell in the praxis of scribal schools that issued internal official documents. However, it is surprising that a synchronic change in the use of Akkadian between internal grants and international documents was not matched by a clear change in the shape of the signs as the Middle Hittite scribal phase is mostly characterized by a gradual shift from the older ductus to the imperial one. This fact certainly complicates the description of the role of a new penetration of the Mesopotamian cultural elements in the Hittite world during the pre-imperial phase. If it happened, it does not seem to have greatly affected the scribal ductus, which continued its gradual and uneven evolution until the 13th century, while the linguistic influence seems more prominent in the international documents than on the internal administrative tablets.

Occasional influences of the Hittite language on the Akkadian are sometimes detectable: besides formulas that seem to derive from a Syrian tradition, occasional inflected Hittite loans emerge, and some calques of the Hittite structures appear. Rüster and Wilhelm (2012:73) explicitly refer to the formula (the land, the king) išši-ma (the grantee(s)) iddin, which would reflect a Hittite construction;²⁹ however, as no grammatical mistake occurs in Akkadian, this can hardly be considered a form of proper interference and is merely a local formula.

3.3 The Akkadian of Diplomacy

As already mentioned, diplomatic texts written in Akkadian include treaties with non-Anatolian partners, international correspondence, and decrees or verdicts of the Hittite kings concerning matters outside Hatti. Akkadian treaties between Hittite kings and their international partners have been found in Hattuša and other important centers, especially Ugarit. The language of these treaties is the so-called ‘diplomatic’ Akkadian, that is, a peripheral Middle Babylonian dialect used as lingua franca, mostly by non-native Akkadian speakers, across the ancient Near East in the Late Bronze Age.³⁰ Diplomatic Akkadian was a rather artificial language: texts were written using a standard formulary and display several surprising features. Most of these features can be explained as interference phenomena, although the path of interference was not always linear and unidirectional.

At present, we do not have many original treaties, that is, the official copies that were sealed and deposited in temples. Only three diplomatic documents can be regarded as original, all written in Hittite: the Bronze Tablet Bo 86/299, containing the treaty between Tuthaliya IV and Kurunta of Tarhuntašša (CTH 106.I.1); the MS tablet KUB 31.103, containing the treaty with Pahhuwa (CTH 212.1); and the fragment 544/f, recording a declaration of Kurunta of Tarhuntašša (CTH 96).³¹ The only alleged original treaty in the Akkadian language found so far, namely, the MS tablet CTH 29.A (a treaty between Tahurwaili and Eheya of Kizzuwatna), may not be original: the fragment with the seal of the Hittite king cannot be joined to any of the other fragments of the treaty, and the material characteristics of the tablet make it possible that the seal belongs to a different document.³² Therefore, most of the tablets we have are archival copies, drafts, or scribal exercises.

A different problem that concerns bilingual treaties specifically is the issue of the original text, not to be confused with the problem of original tablets dealt with above. When multilingual documents are involved, most scholarly efforts are directed toward determining which is the ‘original’ text—presumably that written in the language of its author or scribe—and which the translation. Linguistic idiosyncrasies or idiomatic phrases may reveal which is the secondary version.

The treaty between Telipinu and Išputahšu of Kizzuwatna (CTH 21) is very fragmentarily attested by one tablet in Akkadian in MH script and three fragments in Hittite (one in OS). The tablet KUB 34.1+, recording an Akkadian treaty between an unknown Hittite king³³ and Paddatiššu of Kizzuwatna (CTH 26), was probably drafted by a scribe who did not belong to the Hittite scribal tradition as shown by the shape of some diagnostic signs³⁴ and other writing features,³⁵ although the syntax shows Hittite interferences.³⁶ The text of the treaty between Tahurwaili and Eheya of Kizzuwatna (CTH 29), attested only in Akkadian on two different tablets (one of them allegedly sealed and therefore original; see above), seems to parallel CTH 26, at least in part, but it is too fragmentary to be fully evaluated.³⁷

The treaty with Šunaššura of Kizzuwatna (CTH 41), issued by a Hittite king most commonly identified with Tuthaliya I, is attested by five tablets that preserve the Akkadian version and two that contain the Hittite one. The redactional history of the text seems to have been complex because the Akkadian versions present significant divergences and the Hittite and Akkadian manuscripts do not have passages in common. The final version of the treaty seems to be the one recorded in the MS Akkadian copy KBo 1.5 (CTH 41.I.2.A), whereas the MS Akkadian tablet KBo 28.110+ (CTH 41.I.1) possibly consists of the collection of diplomatic documents exchanged by the two partners before the conclusion of the treaty.³⁸ Some Hittite influences can be detected in the Akkadian versions,—for example, the occurrence of the endyadic phrase ÉRIN^MEŠ ANŠE.KUR.RA^HI.A ‘infantry (and) horses (= chariotry)’ (e.g., KBo 1.5 i 20–21), which denotes the whole army and only emerges in MH (probably starting from the reign of Tuthaliya I) and is later restricted to texts in Hittite language. In contrast, Akkadian texts consistently have ÉRIN^MEŠ (u) ^GIŠGIGIR^MEŠ ‘infantry (and) chariotry’ (cf. Giusfredi, Merlin, and Pisaniello [forthcoming], and see below). Another specifically ‘Hittite’ phrase is NAM.RA^HI.A GU₄^HI.A UDU^HI.A ‘people, cattle, sheep’ (KBo 1.5 i 22), with the semantic narrowing of NAM.RA from whatever is taken as booty (Akkadian šallatu) to deported people only (cf. Watkins 1979). In all other occurrences in this text, the booty is expressed by the phrase maršit URU^KI NAM.RA^HI.A ‘the property of the city (and) the NAM.RA people’, which also has clear parallels in the OH tradition. Compare, for example, ^URUHalpaš NAM.RA^MEŠ āššu=ššet [^URU]Hattuši udaš “he brought the NAM.RA people and the goods of Aleppo to Hattuša” in the Edict of Telipinu (§ 9).³⁹ A likely calque found in this treaty is also the use of etēqu Gt in the phrase ‘transgress the oath’ (cf., e.g., nīš DINGIR^MEŠ ītetiq in i 25), based on the Hittite -kan šarra-. As remarked by Del Monte (1986a:72) with regard to the Syrian treaties, where the same phenomenon can be found (see below), the Akkadian verb usually occurs without the -ta- infix in similar phrases outside the Hittite world. In our view, however, the choice of a Gt form may not depend on its separative meaning, as suggested by Del Monte, but instead be a matter of diathesis: the Hittite šarra- ‘transgress (an oath)’ is usually (and originally) middle (cf. CHD Š:237–238; EDHIL:727–729), although secondary active forms are sometimes found. Therefore, an Akkadian Gt form, with its reflexive and reciprocal meanings,⁴⁰ may have appeared to be the best match for the Hittite middle. Other likely examples of calques on Hittite are EGIR paṭaru in KBo 1.5 iii 56, 62 (< Hittite āppa lā- ‘release from an obligation’) and šapal nīš ilī in KBo 1.5 iii 60, iv 10 (< Hittite linkiya kattan ‘under oath’).⁴¹ Furthermore, Hittite influence may explain the mistaken use of the gen./acc.pl. awati in place of nom.pl. awatū in KBo 1.5 iv 34, 36 because it may have been determined by the fact that the Hittite uttar ‘word, fact’ was a neuter noun.⁴²

The treaty between Šuppiluliuma I and Šattiwaza of Mittani (CTH 51) is preserved in three distinct Akkadian versions (A = KBo 1.1+; B = KBo 1.2; C = KUB 3.1a(+)) and one Hittite manuscript (KUB 21.18(+)).⁴³ From a paleographic and orthographic point of view, the Akkadian manuscripts A and C probably belong to the Hittite scribal tradition. B was surely drafted by a Mittanian scribe and differs somewhat in content from A and B, which argues that it was a Mittanian version of the treaty rather than a copy of the Hittite version written by a Mittanian scribe.⁴⁴ Some phenomena of interference depending on Hittite can be detected, at least in the versions drafted by Hittite scribes. For example, in KBo 1.1+ obv. 15 we find a relative clause with resumption of the antecedent in the main clause by a demonstrative and the noun repeated, which is a typical Hittite structure (cf. GrHL:424) but uncommon in Akkadian (anīna ÉRIN^MEŠ ša ina ŠU-ya išhiṭu ÉRIN^MEŠ annû ina KUR ^URUIšuwa īterub “then, the people who escaped from my hand, those people have entered the land of Išuwa”).

CTH 52 is complementary to CTH 51 since it seems to represent an edition of the treaty with Šuppiluliuma I issued by Šattiwaza, although the normative section is absent. The text is preserved both in Akkadian and Hittite, with the Akkadian manuscript apparently being a copy by a Hittite scribe of the original document issued by the Mittanian chancellery. However, although it is presented as a Mittanian edition, Hittite linguistic influence sometimes emerges, especially in the phraseology, alongside typical Mittani Akkadian features. For example, the sentence found in KBo 1.3+ obv. 41, u DINGIR^MEŠ ša LUGAL GAL LUGAL KUR [^URUH]atti ittalku ana pa[nini] “and the gods of the Great King, King of [H]atti, walked bef[ore us]” is a patent calque on the common Hittite phrase peran huwai- + dat. ‘run before someone’, referring to assistance in battle.⁴⁵ In contrast, the phrase ana mārutti epēšu ‘to adopt as son’ (obv. 24) is largely attested at Nuzi, as is the term tertennuttu ‘status of hereditary prince’ (obv. 29) with e-vocalism (vs. tartennūtu in the Hittite tradition).⁴⁶

Five treaties between Hittites and Syrian partners form a structurally consistent group: those between Šuppiluliuma I and Aziru of Amurru (CTH 49; Akkadian and Hittite), Šuppiluliuma I and Tette of Nuhašše (CTH 53; only Akkadian), Muršili II and Tuppi-Teššub of Amurru (CTH 62; Akkadian and Hittite), Muršili II and Niqmepa of Ugarit (CTH 66; only Akkadian), and Hattušili III and Bentešina of Amurru (CTH 92; only Akkadian). Del Monte (1980, 1986a) has shown that the Akkadian versions of these treaties present some unusual features as well as calques that typically point to an original Hittite version or Hittite as the language in which the text was conceived, regardless of whether a version in Hittite existed. However, the situation is not always straightforward because we have Hittite versions that seem to be based on Akkadian versions (perhaps back-translations?) as well as manuscripts that seem to involve languages other than Akkadian and Hittite.

An obvious phenomenon of interference in these texts is that almost every sentence begins with the Akkadian conjunction u, calquing the Hittite nu. However, less trivial phenomena can be found. In the treaty between Muršili II and Niqmepa (CTH 66), found at Ugarit and preserved only in Akkadian, we find a construction with a double possessive pronoun, where an independent possessive pronoun precedes the head noun, which is followed by a clitic possessive: RS 17.338+ obv. 2 (restored after RS 21.53 obv. 1′) attuka ša ^mNiqmepa a[na Š(EŠ^MEŠ=ka)] “to your brothers, Niqmepa” (lit. “of you, of Niqmepa, to your brothers”). As shown by Del Monte (1986a:44–45), constructions with resumptive clitic pronouns are rarely attested in Akkadian texts issued outside the Hittite world.⁴⁷ Because such forms correspond to the OH split genitive,⁴⁸ they seem to be calques on this typical Hittite structure. Furthermore, in these treaties the use of etēqu Gt is found in the phrase ‘transgress the oath,’ probably calquing the Hittite -kan šarra- (see above).

Another clear phenomenon of interference can be found in the treaty between Šuppiluliuma I and Tette of Nuhašše (CTH 53): in KBo 1.4+ i 9, the Hittite army is referred to with the endyadic phrase ÉRIN^MEŠ ANŠE.KUR.RA^HI.A ‘infantry (and) chariotry’ (lit. ‘infantry and horses’). As shown in Giusfredi, Merlin, and Pisaniello (forthcoming), such a sumerographic phrase in which the chariotry is metonymically expressed by horses is specifically Hittite and generally only occurs in texts in the Hittite language. The corresponding phrase consistently found in Akkadian texts from Boğazköy is ÉRIN^MEŠ (u) ^GIŠGIGIR^MEŠ ‘infantry (and) chariotry’,⁴⁹ which is also found in all of the other passages of the treaty with Tette. Therefore, a single occurrence of the ‘Hittite’ phrase in the Akkadian version of this treaty can be explained as a scribal mistake deriving from either a Hittite version used as a model (although such a model is yet to be found) or the fact that the text was originally conceived in Hittite, regardless of whether a Hittite version existed.

Other features of these treaties seem to have resulted from the intervention of other languages or different scribal traditions. For example, in the incipit of the treaty between Muršili II and Niqmepa of Ugarit, the Hittite king is referred to with the title ^dUTU (RS 17.338+ obv. 1) instead of the expected ^dUTU-ši. According to Del Monte (1986a:38–39), this could be explained as the direct intervention of an Ugaritic scribe because the corresponding Ugaritic form špš is usually employed without the possessive clitic pronoun.

Although the interference phenomena found in these texts mostly point to Akkadian versions depending on Hittite, KBo 10.12+, currently the only Hittite version of the treaty between Šuppiluliuma I and Aziru of Amurru (CTH 49.II), probably derived from an earlier Akkadian text that is preserved on six tablets. Several idiosyncrasies show that the Hittite text should be regarded as a (back-)translation of an Akkadian version: 1) literal translations of Akkadian phrases resulting in uncommon Hittite expressions, such as Akkadian ištu ṣābē narkabāti namāšu ‘mobilize with infantry (and) chariotry’, translated with the two nouns in the instrumental case and a middle form of the verb ninink- (cf., e.g., [IŠTU ÉRIN^M]^EŠ ANŠE.KUR.RA^MEŠ ŪL neniktari in KBo 10.12+ ii 30′), whereas a transitive construction with ÉRIN^MEŠ ANŠE.KUR.RA^MEŠ as the direct object of the active verb ninink- would typically be used in Hittite (cf. Del Monte 1986a:65–66);⁵⁰ 2) some mistaken translations; see particularly the thorough discussion by Del Monte (1980:105–107) on the use of the Hittite walh- (vs. expected zahhiye-) to translate the Akkadian mithuṣu in KBo 10.12+ ii 31′; and 3) the use of ‘Akkadian’ rather than ‘Hittite’ language-specific writings—for example, ÉRIN^{MEŠ GIŠ}GIGIR^MEŠ in KBo 10.12+ ii 26′ rather than the typical ‘Hittite’ ÉRIN^MEŠ ANŠE.KUR.RA^MEŠ that is used elsewhere in the text (see above for the opposite phenomenon in the Akkadian treaty with Tette).⁵¹ Furthermore, the manuscript does not record the full text of the treaty because the divine witness list, curses, and benedictions seem to have been intentionally omitted by the scribe (cf. Devecchi 2015:202).

Besides treaties, diplomatic correspondence was also exchanged with non-Anatolian states in Akkadian. Several Akkadian letters have been found in the Hittite archives, including those sent by foreign kings to the kings of Hatti and copies of those sent by the Hittite kings. Some letters are written in Hittite: they should probably be regarded as either archival copies or drafts after which the Akkadian versions were prepared.⁵² The language of the letters is the same diplomatic Akkadian used for the treaties and, since they are products of the Hittite chancellery, they present interference phenomena similar to those described above.

The correspondence with the Assyrian kings is limited to two letters in Akkadian sent by the Hittites and several letters—drafts or archival copies—in Hittite. The two Akkadian letters were probably drafted by Hittite scribes who were capable of using a different ductus from that employed in the Hittite text; this is suggested by the occurrence of typically Hittite signs alongside non-Hittite sign shapes (Mittanian or Assyro-Mittanian).⁵³ These two letters in Akkadian feature some Assyrianisms, but also interference phenomena indicating that they were either conceived in Hittite or translated from Hittite drafts. In the Akkadian letter KBo 1.14 obv. 7′, the use of the preposition ultu in [ultu ammak]a … ultu annaka—if correctly restored—is unexpected. Mora and Giorgieri (2004:69) suggest that it is a calque on Hittite kēz … kēzziya or kēz … apēz ‘on the one hand … on the other hand’. Even more striking is the case of KUB 3.125, in which we read ma-a-an šum-ma LU[GAL …] (rev. 11), where mān probably represents the Hittite hypothetical conjunction, mistakenly written by the scribe before the corresponding Akkadian conjunction šumma (cf. Mora and Giorgieri 2004:149).

4 The Akkadian of the Cultural Tradition

Unlike the Akkadian documents dealt with in the preceding sections, which were compositions produced locally by the Hittites in Akkadian (historical narratives, edicts, treaties, land grants, etc.), the Akkadian texts belonging to the cultural tradition mostly include original foreign Akkadian compositions that were written elsewhere and imported into the Hittite capital city or were composed elsewhere but copied onto tablets locally by Hittite scribes and sometimes provided with a Hittite translation. This heterogeneous group includes texts of different genres, reflecting both the stages of the scribal curriculum and the technical knowledge coming from Mesopotamia. We have tentatively defined these two subgroups as follows:

texts belonging to the scribal curriculum, including lexical lists, hymns, mythological narratives, etc., that were probably employed chiefly for mastering the cuneiform script and the Sumerian and Akkadian languages.⁵⁴ They are generally provided with translations in Hittite, and some were used as models for the composition of Hittite original texts;
magic and medical texts, including rituals, incantations, medical treatises, etc., perhaps used primarily to master technical disciplines other than the scribal art and thus only sporadically provided with Hittite translations.

Such a distinction is not clear-cut. At least some of the compositions included in the first subgroup could have been used outside the scribal schools. Similarly, ritual and medical texts could have been employed for mastering cuneiform and Mesopotamian languages, and surely some were used as models for drafting original compositions in Hittite.

The lexical lists that were found in the Hittite archives (CTH 299–309) are monolingual (Sumerian), bilingual (Sumerian-Akkadian), or trilingual (Sumerian-Akkadian-Hittite) tablets.⁵⁵ Most date to the Empire period, although some may be older or have been copied from earlier manuscripts.⁵⁶ The majority are copies of standard Mesopotamian lists. The fragments identified so far belong to the following lists: S^a, Diri, Erimhuš, S^o, Ura, Izi, Kagal, Sag, Lú = ša, Lú.ázlag = ašlāku, Ea, and perhaps also An.⁵⁷ Besides canonical vocabularies, the Hittite archives also contained lexical lists that do not have a direct Mesopotamian model, perhaps being local productions.⁵⁸ Some of the lists belonging to the Mesopotamian scribal curriculum, especially the most elementary ones, seem to be almost entirely missing—for example, the so-called tu-ta-ti lists, of which an example only occurs as a filler in the Akkadian column of an Erimhuš manuscript (KBo 26.20 ii 39–41).⁵⁹ Possibly only the advanced lexical lists belonging to the second stage of scribal education were kept as archival materials to be reused, and the more elementary ones were discarded.⁶⁰

The issue of how and when these lexical materials reached Hattuša is very complex. A thorough analysis of the manuscripts found so far suggests multiple directions of transmission.⁶¹ Some orthographic features and the possible sporadic inclusion of West Semitic lexical material and peripheral Akkadian words point to a Syrian path for some of the lists (cf. Weeden 2011:103, 131). An Assyrian stage can be possibly assumed for the Ura 20 fragment KUB 37.145(+) (Weeden 2011:129).

As remarked by Scheucher (2012:228), “Akkadian and Syllabic-Sumerian columns unilaterally refer to the Sumerian column, and the Hittite column unilaterally refers to the Akkadian column” in lexical lists from Boğazköy. This is shown by several mistakes in Hittite translations that reflect misunderstandings of the respective Akkadian entries⁶² and by the fact that Hittite translations vary when different Akkadian words correspond to the same, repeated Sumerian entry, but this does not happen when different Sumerian entries are matched by a single Akkadian translation (cf. Scheucher 2012:275).⁶³

Lexical lists represented the first stages of the scribal curriculum. The students then continued by copying more elaborate and difficult texts such as proverb collections, wisdom literature, epics, mythological compositions, and omens. This advanced stage is witnessed in the archives of Hattuša by compositions in Akkadian in different textual genres. Most are provided with Hittite translations. The Akkadian fragmentary MS five-sided prism KBo 19.98 and perhaps also the four-sided prism KBo 19.99, both listed under CTH 819, preserve the MB recension of an epic of Narām-Sîn, the so-called Cuthean Legend. We also have several tablets of an epic of Narām-Sîn in Asia Minor in Hittite (CTH 311), but their relationships with these Akkadian fragments remain unclear. The scribe of KBo 19.99, Hanikuili, is indicated as the son of Anu-šar-ilāni (probably a Babylonian scribe residing at Hattuša) who is qualified as BAL.BI, which could be an abbreviation for EME.BAL.BI ‘its translator’, so a direct link between this Akkadian prism and one of the Hittite compositions mentioning Narām-Sîn is possible.⁶⁴ The text is also attested by two OB manuscripts from Sippar and a late recension preserved on six NA copies from Nineveh, one NA tablet from Sultantepe, and one NB tablet from Kiš.⁶⁵

In the archives of Boğazköy, the composition known as šar tamhāri ‘King of Battle’ (CTH 310), which describes the military campaign of Sargon of Akkad in Anatolia against the city of Purušhanda, is preserved on seven fragmentary tablets in Hittite dating to the imperial period. The text represents a free adaptation of an Akkadian composition attested elsewhere by four manuscripts: a tablet from Amarna written in a western peripheral dialect of Akkadian (EA 359), a fragment from Aššur dating to the NA period that preserves the same recension as the Amarna tablet (VAT 10290), a fragment from Nineveh with a different recension (K. 13228), and another small fragment from Amarna that cannot be fully evaluated (EA 375).⁶⁶ The material characteristics of the Amarna tablet, as well as orthography and some linguistic features, suggest that it belongs to the Hittite scribal tradition and was possibly written at Boğazköy in the Middle Hittite period (cf. Westenholz 1997:105–107).

The bilingual Akkadian-Hittite tablet KUB 4.97 (CTH 315.C) contains the OB composition known as ‘The Message of Lu-diĝira to his Mother,’ while part of the Sumerian version is recorded on a different fragment (KUB 4.2).⁶⁷ In this composition, Lu-diĝira asks a royal courier to bring a letter to his mother, Šāt-Ištar, who is at Nippur, and provides five ‘signs’ (i.e., a poetic description) that will allow the courtier to recognize her. The same composition is attested on the trilingual Sumerian-Akkadian-Hittite tablet RS 25.421, found at Ugarit. The latter tablet was drafted by a Hittite scribe and imported from Boğazköy.⁶⁸

The wisdom literature includes two bilingual Akkadian-Hittite texts listed under CTH 316—the so-called Instructions of Šūpê-amēli (or Šimâ Milka ‘hear the advice’)⁶⁹ and the proverb collection KBo 12.128 (a four-sided prism), of which the Akkadian text is almost completely lost⁷⁰—and the Akkadian prism KUB 4.40 that also contains a collection of proverbs (CTH 814).⁷¹ Šimâ Milka consists of proverbs concerning several topics in the narrative framework of a dialogue between a father (Šūpê-amēli) and his son. The composition is also attested by manuscripts from Ugarit and Emar. Although no tablet containing it has been found so far in Mesopotamia, it is mentioned in an OB catalog of literary texts (AUAM 73.2402 obv. 15), which proves the existence of an earlier Mesopotamian forerunner.

The two fragmentary proverb collections from Hattuša are probably copies of lost Babylonian compilations. Some of the individual proverbs that they contain are known from other sources. For example, the proverb in KBo 12.128:5′, preserved in an incomplete Hittite translation (hantezzin pahhuenanza karapi “the fire devours the first one” [the rest of the paragraph was left blank]), is attested in Sumerian and Akkadian in earlier collections.⁷²

Turning to the Akkadian hymns and prayers, they were probably used by Hittite scribes as literary models for drafting Hittite original compositions. Whether they were also part of the scribal curriculum is controversial. However, there are hints that some of these texts might have been used in scribal schools as texts for the study of cuneiform, Sumerian, and Akkadian (see, e.g., the discussion on KBo 1.12 below).

The hymns and prayers include Sumerian-Akkadian texts provided with Hittite translations and texts for which Hittite translations have not been found. The first group includes the OB hymn to Ištar (CTH 312.I), with a Hittite translation on a separate tablet;⁷³ the trilingual hymn to Iškur-Adad (CTH 314), which likely dates to the Kassite period,⁷⁴ and probably the OB(?)/MB(?) hymn to Šamaš KBo 1.12 (CTH 792.1; with a partial duplicate found at Aššur, KAR 1.19), although the Hittite translation in the right-hand columns is almost completely lost.⁷⁵ The latter tablet includes a section dealing with the treatment of slaves (KBo 1.12 rev. 7′–16′) that is seemingly unrelated to the hymn and possibly represents an excerpt from an otherwise unknown wisdom composition. The composite nature of the tablet could perhaps suggest that it was a scribal exercise (cf. Ebeling 1954:210). Also note that KUB 37.36+, the Sammeltafel preserving the Akkadian version of the hymn to Ištar (CTH 312.I), was perhaps a school exercise because it contains several mistakes that involve the division of the verses. The Hittite version written on the LNS tablet KUB 31.41 (CTH 312.II) seems to have been produced using a different and better copy, but the scribe was unable to translate some parts of the Akkadian text.⁷⁶

The texts that do not have a Hittite translation are the Akkadian hymn to Šamaš KBo 9.44 (CTH 792.2), a Sumerian-Akkadian invocation to Šamaš (CTH 793), the Sumerian-Akkadian hymn and prayer CTH 794 (in an Assyro-Mittanian ductus), and other fragments that are listed under CTH 795. An OB hymn to Adad is preserved in Hittite translation only on the MS tablet KBo 3.21 (CTH 313), which was probably drafted by a Babylonian scribe living in Hattuša (cf. DUB.SAR pabilili in iv 12′); its original Babylonian model is unknown.⁷⁷ The MH hymn and prayer to the Sun deity that is commonly known as the Prayer of a mortal (CTH 372), as well as the parallel Prayer of Kantuzzili (CTH 373) and Prayer of a king (CTH 374), also had Akkadian forerunners (perhaps OB), although they are original Hittite reworkings.⁷⁸

As for mythological narratives, we found some Akkadian fragments of the Epic of Gilgameš (CTH 341.I) in the Hittite archives, as well as Hittite and Hurrian versions. The Akkadian fragments belong to three different recensions: one distributed on four MS tablets found in Temple XVI and not written in the typical Boğazköy Akkadian (CTH 341.I.1.A–D), the second dated to the Empire period and written in the local peripheral Akkadian variant (CTH 341.I.2), and the third represented by the very fragmentary KUB 37.128 (CTH 341.I.3).⁷⁹

By comparing the text preserved in the Boğazköy archives with all of the other sources of the Epic of Gilgameš, Beckman (2003, 2019b) showed that the Hittite recension seems to be closer in content to the SB Twelve Tablet Edition ascribed to Sin-leqi-unnini rather than the OB texts. Therefore, it probably belongs to the stage of re-elaboration of the Kassite period that is poorly known from Mesopotamia,⁸⁰ when the epic was revised before developing into the canonical version of Sin-leqi-unnini. A different problem concerns the relationship between the Hittite and the Hurrian versions of the poem. Judging from the seven very fragmentary Hurrian manuscripts belonging to this composition, the Hurrian version was a reworking of the Akkadian Gilgameš to accommodate the Hurrian religious world rather than a straightforward translation.⁸¹ Some hints in the Hittite version and similarities with passages from other Hurrian poems argue that the Hittite text was translated from Hurrian, although a direct match between the preserved Hittite and Hurrian fragments cannot be established.⁸²

Given the total absence of references to the figure of Gilgameš in the Hittite world outside of the tablets that describe his deeds, it is probable that the Akkadian text was only or primarily used in the Hittite capital city for training scribes.⁸³ Nevertheless, comparison of the Hittite Gilgameš manuscripts with OB and MB sources reveals adaptations in the narrative. These were probably made to accommodate the Hittite mindset⁸⁴ and allow the text to be recited at the royal court. However, some of the changes could have stemmed from Hurrian intermediation since, as mentioned above, the Hurrian recension was probably the direct model for the Hittite Gilgameš.

A small fragment of the Akkadian Atramhasis (KBo 36.26, CTH 347.1), probably written by a Hittite scribe based on the ductus,⁸⁵ was also found at Boğazköy along with some versions in Hittite and two fragments of a still unidentified Akkadian mythological narrative (CTH 796).⁸⁶ One of the Atramhasis fragments in Hittite, KUB 36.74, which preserves a few lines of the third column, might have belonged to a bilingual tablet because the colophon seems to run along the entire length of columns iii and iv.⁸⁷ Another fragment, KUB 8.63+, probably represents the Hittite translation of a Hurrian version.⁸⁸

Finally, we provisionally include omen series and oracle texts (CTH 531–560) in the subgroup of the texts belonging to the scribal curriculum,⁸⁹ although they cannot be regarded as an entirely homogeneous group because the series differ in how and when they reached the Hittite capital. There may also have been case-by-case differences in why they were copied. Some (almost certainly the liver models, but perhaps also some omen series) might have been of greater interest for the technical knowledge they contained than for their teaching value in training scribes.⁹⁰

Omen series were mostly recorded on monolingual tablets (sometimes with Hittite translations on separate tablets),⁹¹ although bilingual tablets are sporadically found.⁹² Some collections are now only preserved in Hittite translation, but the absence of their Akkadian models is probably due to chance. Many series probably reached the Hittite capital directly from Mesopotamia, in some cases by the Old and Middle Hittite period, especially the birth omens, liver models, and tirānu oracles. Others could have been transmitted through Hurrian intermediation (e.g., the MH liver oracle KBo 16.97+).⁹³ The picture is further complicated by the difficulty of identifying and assigning to standard Mesopotamian omen series those tablets found at Boğazköy whose fragmentary condition allows multiple solutions. As remarked by Rutz (2012:174), “Identifying the transmission of a given textual tradition relies on being able to observe the distinctiveness and the stability of a text over time.” In many cases, individual entries in the series from Hattuša find their exact parallels in series from other places and times, but the vertical dimension, namely, the sequence of entries, seems to be much more elusive.⁹⁴

In the Catalogue of Hittite Texts, the first omen series—the largest group of tablets—are represented by astronomical omens whose Hittite translations can be dated linguistically to the Empire period, later than the Hittite translations of the liver omens (cf. Riemschneider 2004:xli).

CTH 531 only contains a small fragment of the Hittite translation of the introduction to the canonical series Enūma Anu Enlil (KUB 34.12). CTH 532 records omens related to the lunar eclipse: only two Akkadian fragments are preserved, but there are several Hittite translations. The Akkadian tablet KUB 4.64+ (CTH 532.I.1) seems to be an OB original.⁹⁵

The lunar omens are listed under CTH 533. Among these texts, KUB 29.11+ (CTH 533.3.B; NS) is an Akkadian-Hittite bilingual tablet (with a parallel text from Emar, Msk 731041). It was drafted by Pikku, son of Tatta, according to the colophon. The Akkadian text includes some mistakes in word division. For example, aš-na-an mu-ri-šu occurs instead of the correct ina(AŠ) na-an-mu-ri-šu lit. “(if the moon) on its being sighted” (i 14, 16, 18). This is rightly paraphrased in Hittite by takku ^d30 autti ‘if you see the moon’, probably showing that the scribe of the tablet, although not able to understand ina nanmurišu, was copying from a model drafted by a translator who worked from a correct Akkadian text and understood it.⁹⁶

CTH 534 includes sun omens. The Akkadian manuscript KUB 4.63 (CTH 534.I.1.A) probably is an imported OB tablet⁹⁷ as shown by the ductus (different from the ductus of Boğazköy) and the single column divider. The text was then copied several times at Hattuša.⁹⁸ Hittite translations also exist, some of them displaying a MS ductus. The star omens cataloged under CTH 535 are only preserved in Hittite translation, although Akkadian models probably existed.

The only assured Akkadian fragment of the terrestrial omen series šumma ālu found at Boğazköy is written on the reverse of KBo 36.47, whose obverse is occupied by a recension of the šumma immeru (see below). However, some Hittite adaptations are attested (CTH 536). A second Akkadian fragment of šumma ālu may be represented by the reverse of KUB 4.53 (cf. Rutz 2012), currently listed under CTH 537 (medical omens).

Medical omens (CTH 537, for which the closest parallel is provided by the later canonical SA.GIG series) were used to interpret medical signs. Most come from Büyükkale and were drafted by scribes belonging to the Boğazköy scribal tradition of the Empire period (or perhaps even by a single scribe). Only four tablets were written by Mittanian scribes; one of the latter, KUB 4.53 (CTH 537.I.15), whose inclusion in this group of omens remains questionable, was written by a scribe with a Hurrian name, Agi-Teššub and seems to be a Schülertafel, meaning that it was probably drafted locally rather than imported.⁹⁹ The language of the texts is MB with Assyrian influence, as in the kingdom of Mittani and northern Syria, which suggests a transmission from Babylonia to Hattuša through a Syro-Mittanian path. Only two small fragments of medical omens in Hittite are known so far, KBo 13.32 and KBo 13.33 (cf. Burde 1974:48), although Hittite-Luwian glosses sporadically occur in Akkadian manuscripts—for example, :taršiyai (KUB 37.193+ obv. 2), :tarpalli[š] (KUB 37.193+ obv. 5), :paptartanzi dankuwaeš (KUB 37.193+ rev. 13′), :GIM-an GIG-anza arha dalāi (KUB 37.190 obv. 4′), and :irmananza (KUB 37.190 obv. 6′). As discussed in Pisaniello and Giusfredi (2021), such glosses rarely translate the Akkadian entries in which they occur. Instead, they add further symptoms, possibly based on different entries of the Akkadian text. Given the existence of fragments of medical omens in Hittite, Akkadian manuscripts that include interpolations by the Anatolian scribes in Hittite and Luwian could be regarded as intermediate materials that preceded the production of a complete version in Hittite.

CTH 538–540 contains šumma izbu omens. The Akkadian manuscripts belonging to this group derive from Babylonian models and show MB features.¹⁰⁰ Their Hittite adaptations, although surely based on the same models, display the OH ductus. Moreover, many include linguistically archaic features that point to an earlier transmission, maybe via Hurrian, given the existence of a bilingual Akkadian-Hurrian fragment.¹⁰¹ The striking similarities between the Akkadian šumma izbu omens found at Boğazköy and the standard version found in Assurbanipal’s library at Nineveh suggest that the Boğazköy recensions belong to the MB period during which the texts were being systematized into what would be the standard versions of the first millennium.

The earthquake omens listed under CTH 541, which were part of the iqqur īpuš series,¹⁰² only include three Akkadian fragments: KUB 37.163 and KUB 37.164, which may have belonged to the same tablet, and KBo 36.36, which is a later direct copy of KUB 37.163 (cf. Fincke 2010b). A Hittite version is preserved on the obverse of KUB 8.28 (CTH 535.4; dupl. KBo 47.62), which contains star omens on the reverse. Characteristic of the Hittite translations of these omens is the occurrence of the deity Ninga, elsewhere unattested in the Hittite corpus. This deity was likely created ad hoc to match the Akkadian figura etymologica rību īrub ‘an earthquake quakes’ (= the Hittite ^dNingaš ninikzi).¹⁰³

KUB 37.198(+) is the only manuscript found at Boğazköy containing oil omens (CTH 542) and the only omen compendium of this type found outside of Babylonia.¹⁰⁴ All of the other oil omens date from the OB period. The Boğazköy tablet shows MA script, with sporadic OB monumental signs, perhaps taken from the copy that served as a model;¹⁰⁵ no peripheral spellings are found. Therefore, it was probably imported at Hattuša during the reign of Šuppiluliuma I or was drafted locally by a foreign scribe. Since it is a single copy and no Hittite translations have been found, we may provisionally assume that it was probably not used as a scribal didactic tool.

Physiognomic omens are listed under CTH 543, which includes three Akkadian fragments and two Hittite versions. The only animal omens (CTH 544) are NS Hittite versions from Bk. A and KBo 13.29, the latter being a MS collection of different omens from the House on the Slope. The birth omens listed under CTH 545 may belong to the omen series iqqur īpuš. They include one Akkadian fragment and two Hittite tablets: KBo 25.2+ is in OS and KUB 8.35 in NS. The latter fragment was probably drafted by a non-Hittite scribe and contains only one paragraph that matches one of those preserved on the Akkadian fragment KUB 37.118, showing that the translator probably had other Akkadian versions available.¹⁰⁶

Some Akkadian tablets of hemerologies (CTH 546), which contained rules and prohibitions guiding actions that could be taken on different days of the year, are also attested at Boğazköy. The tablet KUB 43.1(+), which records rules for ‘crying out laments’ (šigû šasû) and ‘cleansing of his clothes’ (ṣubāssu ubbubu), shows orthographic and paleographic features that suggest that it is a local product written by a Hittite scribe based on a Babylonian or Assyrian model, probably during the MH period.¹⁰⁷ Hittite translations have not been found yet, but the Hittite ritual CTH 432, which has a Babylonian background (see below) contains a hemerology with a šigû-lament (in Hittite, duddu halzai-). Five other tablets contain ‘Offering bread hemerologies’ (Fincke 2010a); one of them, KUB 4.45, belongs to a tradition that seems to have no parallel elsewhere.¹⁰⁸

The liver models (CTH 547) include both monolingual Akkadian models and bilingual Akkadian-Hittite ones. The latter group is represented by four models with complementary texts in the two languages: the protasis in Akkadian, and the apodosis translated into Hittite. Akkadian models are linguistically OB and display an archaic (or archaizing) ductus,¹⁰⁹ whereas some of the bilingual ones are in OS. The liver models may be connected with the northern Syrian or southeastern Anatolian areas because most liver models have been found in the west periphery rather than in Mesopotamia.¹¹⁰

Liver models were pedagogic tools to teach hepatoscopy¹¹¹ and may have been connected to the presence of Babylonian haruspices in the Hittite capital. The analysis of the script reveals no Hurrian influence: they belong to the Babylonian tradition, with some Syrian or northern influence. The technical terms occurring in these texts are always written in Akkadian or as Sumerograms, whereas in the Hittite liver oracle reports—which were archival documents and did not have a pedagogic function—they are written consistently in Hurrian, proving Hurrian intermediation in the transmission of the practice of hepatoscopy.¹¹²

It should be noted that the Hittite apodoses in the bilingual liver models were not originated by Hittite scribes, as is shown, for example, by KUB 4.72 rev. 6–7, in which ÉRIN^MEŠ ITTI DINGIR hingani wekzi almost perfectly calques the OB apodosis in YOS 10, 46 iii 41, ummānī itti DINGIR-lim ana dâkim eršet “with the god (i.e., divine approval) my army was demanded to death”, mechanically matching ana dâkim with a dative and converting the stative form eršet into the active present wekzi, resulting in a syntactically odd Hittite sentence.¹¹³ According to De Vos (2013:80), the translations were made by Akkadian speakers,¹¹⁴ which would explain why only the apodoses were translated and account for mistranslations and unusually structured Hittite sentences. This is possible, but other solutions can be envisaged. For instance, Cohen (2015:124) posits that “the technical language of the protases was intentionally left un-translated, as much as Hurrian terminology was retained in Hittite SU-oracles and the šašta-omens: the basic keys of interpretation, like in spells or incantations, were intentionally left untouched.”

Besides liver models, Akkadian and Hittite omen series treating signs of the different parts of the liver are also attested at Boğazköy. CTH 548, relating to gall bladder omens (ZÉ, martu), includes two Akkadian fragments, KBo 7.4 and KUB 37.180. CTH 549, about signs of the ‘position’ (KI.GUB), contains various fragments in Akkadian and Hittite, as well as two bilingual tablets, KUB 8.34+ (perhaps MS)¹¹⁵ and KBo 34.133(+). CTH 550, concerning the ‘yoke’ (nīru), includes the Akkadian tablet KUB 4.66, which was possibly imported (Riemschneider 2004:58). Two Akkadian tablets are listed under CTH 553, signs of ‘well-being’ (šulmu), and four Akkadian fragments are included in CTH 554, omens of the ‘weapon’ (kakku). CTH 555 includes two Akkadian fragments that treat the signs of the ‘palace gate’ (bāb ekalli); they may belong to the same tablet, which was written by Tarhuntaziti under the supervision of Anuwanza. Finally, diverse fragments of Akkadian liver omens are listed under CTH 556.

Two Akkadian tablets and one Hittite fragment, showing ancient ductus (OS or MS), concern entrail omens (CTH 551) that involve observation of the coils of the intestines (tīrānu). Kidney omens (CTH 552) are preserved in an Akkadian-Hittite bilingual recension in the Sammeltafel KUB 4.1, together with the ritual text CTH 422 (Incantation at the enemy’s border).¹¹⁶

Finally, CTH 560 includes several Hittite and Akkadian omen fragments. Notably, KBo 36.47 has been recognized as an Akkadian fragment of the šumma immeru omen series, with a recension of šumma ālu on the reverse (see above). The recension of šumma immeru is identical to that found at Emar (Emar 698), indicating that the two manuscripts belong to the same stage of elaboration and transmission of the text. Both show traces of an OB stage, especially in spelling conventions, although some post-OB linguistic features can be identified; therefore, they were probably created during the MB stage when this material was being standardized. Hittite translations have been not identified so far, but, as recognized by Hoffner (1993), the Hittite šašt(a)- oracles (CTH 576) can be traced back to the Akkadian šumma immeru, although the presence of Hurrian technical terms points to an intermediated transmission, perhaps through older forerunners.¹¹⁷

Unlike the scholastic texts belonging to the scribal curriculum, which were mainly although not exclusively used for mastering the cuneiform script and the Sumerian and Akkadian languages, magic texts—including rituals, incantations, and medical and pharmacological texts—were probably employed primarily for the study and transmission of technical medical and ritual knowledge and the execution of ritual performances. That they were found in the Hittite archives is probably related to the presence of Babylonian experts (physicians, exorcists, etc.) at the royal court.¹¹⁸

It appears that Sumerian and Akkadian magic texts were rarely translated into Hittite. Aside from the medical omens listed under CTH 537, which had Hittite translations (see above), only two Hittite rituals, CTH 432 (see below) and the medical text CTH 461.L, could be regarded as translations of lost Akkadian originals.¹¹⁹ Furthermore, Hittite substitution rituals for the king (CTH 419–421) depend on Babylonian models (possibly with Hurrian intermediation), although they are not direct translations of Akkadian ritual texts.¹²⁰

The ritual against depression (CTH 432)¹²¹ and the ‘babilili ritual’ for Ištar-Pirinkir (CTH 718),¹²² preserved in several copies, are bilingual texts in which more or less extensive Akkadian recitations are embedded in a Hittite ritual framework. The Akkadian passages of CTH 432 are in good Middle Babylonian (but with a west peripheral syllabary) and show a more sophisticated vocabulary in comparison to the other Akkadian texts found at Boğazköy,¹²³ whereas the ductus is typically New Script. Beckman (2007:81), observing that the Hittite instructions are fluent and do not seem to be translated from an Akkadian original, suggested that “this text represents the collaboration of a Hittite student and his foreign teacher, a Babylonian scribe resident at the Hittite capital.”¹²⁴ Conversely, as mentioned previously, Schwemer (2013:158–159) regards this ritual as a Hittite translation of a lost Akkadian original text. The Akkadian of CTH 718 is also different from the common Boğazköy Akkadian of diplomatic texts, despite its west peripheral syllabary. It is generally more correct in the use of verbal forms and feminine pronouns, but there are also several Assyrianisms, as well as elements that seem to point to an Old Babylonian dialect.¹²⁵ The fact that these texts are characterized by Hittite instructions and Akkadian recitations—thus being structurally similar to the Hittite rituals with Luwian, Palaic, and Hurrian formulas—strongly points to their use in ritual practice rather than as mere tools for mastering cuneiform and Mesopotamian languages.

Conversely, other rituals and incantations are original Mesopotamian compositions without any Hittite translations or adaptations. These include the Sumerian-Akkadian incantations listed under CTH 801, the ritual against impotence listed under CTH 802 (type ŠÀ.ZI.GA), the Akkadian incantations šumma amīlu kašip (CTH 803) and ana pišerti kišpī (CTH 804),¹²⁶ the Sumerian-Akkadian bilingual incantation UDUG.HUL.A.MEŠ (CTH 805), and the Sumerian-Akkadian incantations mentioning the deity Asalluhi (CTH 806).¹²⁷

KUB 37.1, listed under CTH 808, is an Akkadian medical text that concerns the application of poultices. Because of the poor quality of the script and the presence of several Hittite and Luwian glosses, it was formerly identified by Köcher (1952–1953) as a student exercise¹²⁸ but is now more precisely interpreted as an exercise of a student in medicine (Giusfredi 2012). Other Akkadian medical texts include seven copies of an ophthalmological treatise (CTH 809), some imported and others copied locally;¹²⁹ the incantation known as ‘the Moon-god and the cow’ (CTH 810), in non-Hittite ductus; two non-identical copies of a ritual and prescriptions against fever (CTH 811), both written in non-Hittite script;¹³⁰ and other fragments belonging to different compositions (CTH 812).¹³¹

5 Concluding Remarks

Because Akkadian was used widely in the Hittite scribal world, many texts share local similarities, especially interference with local Hittite or more generally Anatolian substrata that resulted in features such as doubly marked genitival chains with a head noun following the modifiers or confusion in the use of the Semitic gender. However, some differences exist between different Akkadian grapholects. The Old Hittite political documents seem to pattern with a peripheral variety of Old Babylonian that was open to Marisms and to the preservation of sparse formulaic material in common with Old Assyrian. Old Babylonian features with Syrian influences also appear to emerge in the land grants, whereas the Akkadian of diplomacy, from the MH stage onwards, appears to pattern more with Middle Babylonian and standardized international Akkadian. In contrast, most of the literary and technical texts of the cultural tradition seem to have been created during a mature phase of the Hittite scribal history and tend to remain closer to the Mesopotamian cultural and linguistic milieu of origin even when they are almost certainly copies produced locally.

We are referring here only to the Hittite archives from Boğazköy; the Old Assyrian phase in Anatolia was discussed in Chapter 4. Cf. also Chapter 6 for the problem of the reintroduction of Cuneiform in Anatolia.

After Wilhelmi’s dissertation, two more important works appeared. Beckman (2021) dedicated an article to a general presentation of the problem of Boğazköy Akkadian. The contribution is informative, although very synthetic. Middle Babylonian features are observed in the corpus, and the author hints at differences related to document age and type but fails to carry the discussion further. Wilhelmi (2022), on the other hand, summarizes several aspects of the contextualization of Boğazköy Akkadian as a specific grapholect.

Bilingual documents in Akkadian and Hittite generally form a relatively small subset of the first group of A-type texts according to the categorization proposed by van den Hout (2005:286–287), and date to an early phase.

See van den Hout 2009c.

On the origin and main features of the Hittite IIIc ductus, see Klinger 1996:32–29 and, more recently, Weeden 2016.

See Devecchi 2005 for an edition.

Edition by Beckman 1995.

Edition by Hoffmann 1984:58–62.

See edition and discussion in Salvini 1994.

Cf. G. Wilhelm (ed.), hethiter.net/: CTH 26 (INTR 2014-02-25).

CTH 41, 49, 51, 52, 62.

Cf. Rüster and Wilhelm 2012:72–73.

Cf. Rüster and Wilhelm 2012:36.

awāt tabarna LUGAL.GAL ša AN.BAR ša lā nadīam ša lā šebērim ša ušpahhu SAG.DU-su inakkisū “the word of the tabarna, the great king, (is) of iron, (it is) not to be rejected, not to be broken. Whoever changes (it), his head will be cut off.” See also KBo 1.6 obv. 6–7 (CTH 75.A; Treaty between Muwattalli II and Talmi-Šarruma of Aleppo) and KBo 6.28+ rev. 28–29 (CTH 88; Decree of Hattušili III regarding the exemption of the hekur of Pirwa from taxation). See also WIlhelmi 2011:107–108.

Cf. Özgüç 1956.

Cf. Salvini 1993. The inscription says: tabarna Ammuna LUGAL.GAL ša išar [I]NIM ušpahu BA.ÚŠ “Tabarna Ammuna, Great King. He who forges the right word will die!”

Cf. Ünal et al. 1990–1991; Ertekin and Ediz 2003; Ünal 2003.

We fully support the interpretation offered by Archi (2010:39). Therefore, the Tikunani letter cannot be considered a document issued by a Hittite archive. It may, of course, offer material for comparison regarding the origin of the scribal praxis in Hattuša. These issues, however, are dealt with in Chapter 6 of the present book.

For the Middle Babylonian features, we refer to Aro 1955. On its use in Akkadian magical texts, cf. also Schwemer 1998:47–50. Middle Babylonian elements are also found in the Akkadian of the international treaties, cf. Beckman 1996:2 and below. Middle Assyrian elements seem to be more circumscribed in the Hattuša production, with certain examples emerging in two letters exchanged with the Assyrian court (Mora and Giorgieri 2004:40, 57–75, 145–149). Here, however, other elements also emerge, including writings that are reminiscent of other peripheral Akkadian corpora (especially from Syria) as well as regular Old Babylonian or Babylonian forms. This variability seems to indicate that the Akkadian letters from the Hittite court to Assyria represent linguistically unique products that were probably quite artificial and represented an effort to adapt the language to the audience. The mixed ductus that is employed for KBo 1.14 would confirm the unusual nature of these linguistically isolated scribal products.

Some of the Akkadian versions (KBo 7.15+ and 19.96+) show a mixture of typically Old Script and typically New Script signs (in some cases, the two variants of the same sign, e.g., IG, co-occur).

We have no clear examples of how the Hittite cuneiform ductus must have looked like at the time of Telipinu, apart from the mixed ductus of the edict and the two fragments of the treaty. Other texts that might go back to this phase (CTH 20 and 22) are only preserved in later copies.

It deals with the variants BÁ, PÁ, DÁ, TÁ, KÀ, and QÁ for sylllables containing a stop and a vowel. They seem to have originated in an Old Assyrian and Old Akkadian tradition and do not belong to the graphemic inventory of the local cuneiform system used for Hittite. In the Akkadian version of the Annals of Hattušili I, they represent minority variants with respect to the signs BA, PA, DA, TA, KA, and QA that are expected to prevail both in Old Babylonian and the Akkadian of Hattuša.

Devecchi considers the construction ša GÉME^MEŠ-šu ŠU^MEŠ-šina “of the slaves their hands” a calque of a Hittite syntactic pattern. In contrast, the construction ^GIŠGIGIR^MEŠ-šu ša KUR ^URUAbbaia “its chariots of the land of Abbaya” employs a double-marking but does not follow a Hittite word order and may be a rare emphatic structure of core Akkadian.

Other topoi listed in Devecchi (2005:114–116), such as the presence of a general rebellion, the designation of an exploit of the king as unprecedented, or his representation as someone who frees the peoples oppressed by the enemy are, we believe, mere anthropological universals.

Cf., on the Annals of Hattušili I, the discussion by Melchert (1978), who argues for a true bilingual text deriving from a Hittite original. More recently Giusfredi (2013) challenged the idea that an ‘original version’ can be reconstructed.

It should be observed that the form occurs in a very disturbed context in which the Hittite and Akkadian versions seem to differ (cf. also Melchert 1978:19).

In the corpus collected by Rüster and Wilhelm, documents in Hittite exist for earlier rulers. These are texts that list personal names and toponyms without other indications of a Hittite underlying dictation except in the case of a late text, no. 91 (by Tuthaliya I and Ašmunikkal), in which some sentence particles point to proper Hittite syntax.

The treaty, dating to the reign of an unidentified 15th-century ruler (cf. Devecchi 2015:70–71), presents a use of mimation limited to the CVm-syllabograms and some instances of non-assimilated nasal consonants (KUB 34.1+ obv. 29, 33). However, other features appear purely Old Babylonian, e.g., the uncontracted -ia- diphthong rendered by Ci-a.

The Hittite construction mentioned by Rüster and Wilhelm (2012) would involve the verbal form sara dai- followed by another verb. As no clear comparanda exist due to the lack of proper Hittite administrative jargon, the reason for this proposal must lie in the lack of comparable formulae in the peripheral Akkadian one.

On the spread of ‘diplomatic’ Akkadian, see especially van Soldt 2011.

Cf. Devecchi 2015:53–54.

Cf. Wilhelm 2013:348–349.

Cf. Devecchi 2015:70, including the references.

See G. Wilhelm (ed.), hethiter.net/: CTH 26 (INTR 2014-02-25) and Wilhelmi 2011:114.

Cf., e.g., the inconsistent use of the determinative KI after geographical names or the way ^URUKi-iz- / :wa-ta-ni is written in lo. e. 22′–22′a, i.e., split and partly placed on a new line (with a Glossenkeil indicating that wa-ta-ni belongs to the word that begins on the preceding line).

Cf. Durham 1976:70.

On the treaties with Kizzuwatna, cf. Del Monte 1981.

Cf. Beckman 1996:13–14, Devecchi 2015:73–75, and the introduction of the online edition by G. Wilhelm (ed.), hethiter.net/: CTH 41.I.1 (INTR 2011-12-20).

According to Watkins (1979:274), the semantic narrowing of NAM.RA^(MEŠ) also involves this phrase because we would have an opposition between NAM.RA [+ human] and āššu-, the latter probably [– animate], thus only referring to inanimate booty (as in the Anitta text and in the Annals of Hattušili I) rather than [± animate] and including livestock (cf. ŠA ^URUArziya āššu[=ššet^? QADU NAM.RA^MEŠ] GU₄^MEŠ UDU^MEŠ in the Deeds of Šuppiluliuma I, KUB 14.22 obv. 9′–10′, if the restoration is correct). However, it is theoretically possible that NAM.RA retains in this phrase the original feature [± human], thus including livestock, only later expressed by GU₄ UDU. KBo 1.5 may reflect a fluctuating situation between the still predominant OH usage and a less common use that became more prevalent later.

Cf. Huehnergard 2000:393.

Cf. Wilhelmi 2011:129.

See Wilhelmi 2011:124, where also other examples of the same phenomenon are listed, for which, however, an explanation in terms of Hittite interference is less straightforward.

The Akkadian fragment KBo 68.190, currently listed in the CTH as a separate tablet, may belong to C.

Cf. Beckman 1993 and especially Devecchi 2018 for a more detailed analysis.

Cf. Del Monte 1986b:62; Devecchi 2018:77.

See Devecchi 2018:78 for other possible examples of Hurrian influence.

Cf., e.g., the structurally identical tuel ŠA ^mKupanta-^dLAMMA DUMU^MEŠ=KA “your children, Kupanta-Kurunta” in the Hittite treaty between Muršili II and Kupanta-Kurunta of Mira and Kuwaliya (KBo 5.13 ii 11′; CTH 68.C).

See GrHL:251 (cf., e.g., ammel tuēggaš=miēš “my members” in VBoT 58 i 24).

Incidentally, such a complementary distribution makes it clear that ANŠE.KUR.RA^MEŠ/HI.A in the ‘Hittite’ phrase meant ‘chariotry’ rather than the more generic ‘horse troops’ (pace Beal 1992:6 with fn. 24).

See also išhiulaš lenkiyaš in iii 24′ translating Akkadian ša riksi u ša māmīti, while in Hittite context išhiul- and lingai- usually occur independently from each other (Del Monte 1986a:69).

The ‘Akkadian’ phrase also occurs in a Hittite treaty(?) fragment from Oylum Höyük (Oy. 12–401 obv. 11; cf. Ünal 2015) that is too fragmentary to be fully evaluated.

Some hints to determine whether they are drafts or archival copies may come from the greeting formulas that usually open the letters and are sometimes absent in these texts. Letters without a greeting formula can be regarded as preparatory drafts, while those including a greeting formula are probably copies or translations of the letters that were sent. The same may apply to letters in Akkadian: they may be copies of the letters sent by the Hittite king, preparatory drafts, or letters that were ready but never sent (see the discussion in Mora and Giorgieri 2004:43–45).

Cf. Mora and Giorgieri 2004:37–38. On the problems connected with the label Assyro-Mittanian, see however the important observations by Weeden (2012a).

Cf., e.g., Beckman 1983b.

Two lists, KBo 26.56 and KBo 26.5+, are on prisms.

The oldest manuscript is the late MH Ura fragment Or. 95/3, found at Ortaköy/Šapinuwa (cf. Süel and Soysal 2003).

For a presentation of the lexical lists, see the doctoral dissertation by Scheucher (2012).

Cf., e.g., KUB 37.122, currently under CTH 815 (Giusfredi and Pisaniello 2019a).

Cf. Weeden 2011:91–92.

Cf. Cohen 2013:68.

Cf. Weeden 2011:91–131 and Scheucher 2012.

Cf., e.g., KBo 1.31 obv. 11′, where the Akkadian entry qa-a-tù ‘finish’ is translated by ŠU-[aš] ‘hand’ in Hittite (= Akkadian qātu). However, as noted by Veldhuis (2014:275), these mistakes should not be simplistically regarded as signs of scribal ignorance: “rather than attribute those wholesale to incompetence, we may well look for signs that indicate that the Hattuša scribes tried their best at decipherment and made an intellectual investment that went beyond mere copying.”

However, sporadic examples exist in which the Hittite translation seems to refer to the Sumerian rather than the Akkadian entry (see Scheucher 2012:276).

Cf. Beckman 1983b:103–104 with fn. 37.

Cf. Westenholz 1997:263–368.

Cf. Westenholz 1997:102–139. See Rieken 2001 for the Hittite version.

Cf. Civil 1964 and Klinger 2010:324–328.

Cf. Nougayrol 1968:310 and Viano 2015:382 fn. 8.

KBo 12.70+, cf. Cohen 2013:81–128.

Cf. Cohen 2013:201–206.

Cf. Cohen 2013:199–201.

Cf. Cohen 2013:204–205.

Cf. Reiner and Güterbock 1967. This composition also has a later NB version that includes some additions.

Cf. Laroche 1964 and Cooper 1971:8–9.

Cf. Ebeling 1954.

See Reiner and Güterbock 1967:256, 265.

Cf. Archi 1983.

See Schwemer 2015 for a detailed analysis of this group of texts.

Cf. George 2003:306–326. See also Beckman 2003:42.

Only two small MB Gilgameš manuscripts have been found so far in Mesopotamia (at Ur and Nippur), while other MB tablets come from Megiddo, Emar, and Ugarit.

Cf. Beckman 2019b:23.

Cf. Archi 2007:187–188.

Cf. Beckman 2003:37–38 and Beckman 2019b:1.

Cf. Beckman 2019:5–6.

Cf. Haas 2006:278.

Also note that an Akkadian version of the Tale of the hunter Kešše and his wife (CTH 361.III) has been found at Amarna, while in Boğazköy archives there are only Hittite and Hurrian versions.

Cf. Beckman 2019b:67.

Cf. Archi 2007:186.

For a comprehensive edition, cf. Riemschneider 2004.

Cf. Koch Westenholz 1993:237–240. Also consider that celestial omens were stored in Hittite archives in multiple copies, while non-celestial ones were usually kept in single copies (van den Hout 2002:872).

CTH 532.I, 533.I, 534.I, 537.I (some of them include Hittite glosses in the Akkadian text), 538.I, 540.I, 541, 542, 543.I, 545.I, 546, 547.I, 548, 549.a, 550, 551, 553, 554, 555, 556, 560.I.

Cf. CTH 533, 547, 549.b, 552.

Cf. Beckman 1983b:101–103. Note however that different opinions exist about the path through which the omen series reached the Hittite capital (according to Kammenhuber 1976, for example, Hurrian intermediation is always implied).

Cf. Rutz 2012:176.

Cf. Koch Westenholz 1993:235.

Cf. Güterbock 1997b:168.

Koch Westenholz (1993:235) tentatively ventilates the possibility that this text (and perhaps others) were carried from Babylon to Hattuša by Muršili I.

Cf. Riemschneider 2004:xxxii, 46; Fincke 2009a.

Thus Wilhelm 1994:6–9. More recently, Rutz (2012) proposed that KUB 4.53 was a Sammeltafel containing a hymn or prayer to Šamaš on the obverse and, on the reverse, the omen series šumma nūru ša rēš marṣi, later incorporated in the šumma ālu series but possibly already included in sakikkû.

100

Cf. Riemschneider 1970:3–4. Only two OB manuscripts of this omen series exist, YOS 10, 56 and CUSAS 18, 12, which are orthographically different from those found at Boğazköy.

101

KUB 29.12. The Akkadian text in the right column does not seem to match the Hurrian omens in the left column so that it cannot regarded as properly bilingual unless we imagine that the Hurrian column referred to a missing Akkadian column on the left and the Akkadian column had a corresponding Hurrian column on its right (cf. KUB 29:v, with fn. 1). See also Cohen 2007 and Cohen 2017:16. Also note that textual correspondences between Akkadian šumma izbu tablets and Hittite ones can only be suggested in a single case (KUB 34.18 ii 9–11, matching KUB 4.67 ii 2′–7′; cf. Riemschneider 1970:70–71) and no bilingual tablets have been found (KUB 37.184 contains šumma izbu omens on the reverse and an unrelated Old Hittite composition on the obverse; KBo 36.46+ has Akkadian omens on the obverse and a Hittite text on the reverse). The Hurrian fragment KBo 27.215 (CTH 774), containing šumma izbu omens, has been proven to be an exact parallel of the Hittite text KUB 8.83 (CTH 538.II.1; MS), although they are probably independent translations of the same post-OB source (cf. Cohen 2017).

102

Cf. Riemschneider 2004:133 fn. 1 and Fincke 2010b:10.

103

Cf. Riemschneider 2004:246.

104

Cf. Anor and Cohen 2018:200.

105

Cf. Anor and Cohen 2018:206.

106

Cf. Fincke 2004:238.

107

Cf. Fincke 2009b:115–117.

108

Cf. Fincke 2010a:143.

109

However, they cannot be dated as a group because cuneiform signs employed are not consistent. According to De Vos (2013:80–108), their dates range from the 16th to the end of the 15th century. It is generally assumed that the bilingual liver models in OS predate the monolingual Akkadian ones, but De Vos (2013:105–106) has suggested that monolingual Akkadian models were transmitted first and used to draft the bilingual models.

110

Cf. Mouton 2015b:232.

111

See Mouton 2015:229, including the references. Note that liver models were probably stored in libraries; they were not archival documents (cf., e.g., Mouton 2015b:215 on KUB 37.218).

112

Cf. Mouton 2015b:207, 230. Note that Hurrian translations of hepatoscopy treatises exist.

113

Cf. Güterbock 1997a:159 and Riemschneider 2004:280–281.

114

See already Riemschneider 2004:xli.

115

But ‘alter Duktus’ according to Riemschneider 2004:99.

116

Dupl. Bo 3476, with the same two compositions.

117

As mentioned, this may also be the case for the šumma izbu omens (see above).

118

On Akkadian magic texts at Boğazköy, cf. especially Schwemer 2013.

119

Cf. Schwemer 2013:158–162. The other medical rituals in Hittite found in the archives of the Hittite capital city, published by Burde (1974), seem to be local products (cf. Beckman 1990:630).

120

Cf. Schwemer 2013:162–164.

Cf. Beckman 2007.

Cf. Beckman 2014.

Cf. Beckman 2007:79.

Beckman 2007:81.

Cf. Beckman 2014:5–6.

126

Cf. Abusch and Schwemer 2011:27–64.

127

Cf. Zomer 2019 for the prism KBo 1.18.

128

See also Beckman 1990:630: “the product of an Anatolian student under an Assyrian master.”

129

Cf. Beckman 1990:630.

130

Cf. Meier 1939 and Schwemer 2013:155.

131

Some of them edited by Schwemer 1998.

Citation Info

Save
Cite
Email this content

Share link with colleague or librarian

You can email a link to this page to a colleague or librarian:
Email this content
or copy the link directly:

https://brill.com/display/book/9789004548633/BP000008.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Collapse
Expand

Contacts of Languages and Peoples in the Hittite and Post-Hittite World

Volume 1, The Bronze Age and Hatti

Series: Ancient Languages and Civilizations, Volume: 4

Cover Contacts of Languages and Peoples in the Hittite and Post-Hittite World

E-Book ISBN:: 9789004548633

Publisher:: Brill

Print Publication Date:: 23 Jun 2023

Subjects

Front Matter

Part 1 The Theoretical and Historical Setting and the Earlier Phases

Part 2 The Foreign Languages of the Hittite Archives and Textual Evidence for Interference

Part 3 Contact Phenomena in Late Bronze Age Anatolia

Back Matter

Metrics

	All Time	Past 365 days	Past 30 Days
Abstract Views	0	0	0
Full Text Views	1893	872	232
PDF Views & Downloads	897	274	12

African Studies	Education	Media Studies
American Studies	History	Middle East and Islamic Studies
Ancient Near East and Egypt	Human Rights and Humanitarian Law	Musicology
Art History	International Law	Philosophy
Asian Studies	International Relations	Religious Studies
Biblical Studies	Jewish Studies	Slavic and Eurasian Studies
Biology	Languages and Linguistics	Social Sciences
Book History and Cartography	Life Sciences	Theology and World Christianity
Classical Studies	Literature and Cultural Studies

Subjects

Authors

Open Access

Product Information

Company

Contact

Chapter 8 Akkadian and Akkadian Texts in Hittite Anatolia

1 Previous Studies on the Akkadian of the Hattuša Archives

1.1 Boğazköy Akkadian and Peripheral Akkadian

1.2 Problems of Categorization

2 The Akkadian Texts from Boğazköy: A Categorization

3 The Akkadian of Politics and Administration

3.1 Old Hittite Political Texts

3.2 The Landschenkungsurkunden

3.3 The Akkadian of Diplomacy

4 The Akkadian of the Cultural Tradition

5 Concluding Remarks

Citation Info

Share link with colleague or librarian

Contacts of Languages and Peoples in the Hittite and Post-Hittite World

Volume 1, The Bronze Age and Hatti

Table of Contents

Metrics

Metrics