Cuneiform Digital Library Notes
2015:17        «              »
The Gottstein System Implemented on a Digital Middle and Neo-Assyrian Palaeography*

Strahil V. Panayotov
BabMed Project, Berlin

The system for ordering cuneiform signs, which will be presented below, was developed by Dr. Norbert Gottstein. Sadly, on the 14 June 2014 he passed away. Dr. Norbert Gottstein was senior student in Assyriology at Heidelberg University. That was the first time when he was confronted with the cuneiform script. The “Gottstein system” arose out of his desperate need, as a beginner, to locate a single cuneiform sign in Borger’s sign lists, especially in the Mesopotamisches Zeichenlexikon (Borger 2010). Gottstein was convinced that the order of cuneiform signs, particularly for beginners, is useless because it is not stringent and a beginner can rarely find a single sign without going through multiple pages. On the contrary, there is a stringent system in chemistry to find around 40 million organic and inorganic chains (Gottstein 2012 and 2013). He used similar logic for his system of ordering cuneiform signs. The initial concept of his list was to make it a search tool, a ‘spotlight’ – named by the present author – for cuneiform signs (Gottstein and Panayotov 2014). Regrettably, he could not see the ‘spotlight’ published, as it was printed shortly after his death. The ‘spotlight’ was developed for beginners who want to find a single cuneiform sign according to its graphic parameters. It is not an extensive collection of signs but establishes initial order of the Middle Assyrian and Neo-Assyrian signs. It is based mainly on sign forms from Šēḫ Ḥamad, Tall Bderi, Tell Chuēra and the so-called ‘Tiglath-pileser I library’, roughly 13-11 centuries BC in comparison with 7th century BC Nineveh (see also illustrations 2 and 3).

What is different in Gottstein’s system if compared to the old one?

Current sign lists use an old-fashioned system for ordering cuneiform signs. It is based on the main five elements of cuneiform script:

(1) vertical wedgea
(2) horizontal wedgeb
(3) to the left falling wedgec
(4) to the right falling wedged
(5) Winkelhakene

This idea originates from the pioneer work of Westergaard from the first half of the nineteenth century (Borger 2010: xiii; Cathcart 2011: 2). All sign lists use this system, modified after the example of Delitzsch’s Assyrische Lesestücke (Delitzsch 1900: 3-40). Lately, Borger (2010: xiii) slightly modified the old system and renumbered the cuneiform signs. According to Borger (2010: xiii) it is virtually impossible to develop an ordered coherent system for sign forms except the Neo-Assyrian forms. In fact with Gottstein’s system all signs, except the earliest ‘linear’ signs, can be stringent ordered and searchable.

Gottstein’s system is based on the same five elements of cuneiform but ordered differently. The main criterion is the sum of the wedges (Fig. 1). The basic idea is to assign to each of these five elements its own name, namely to provide each element with a simple designation. The most simple and suitable names are the initial letters of the alphabet: a, b, c and d. Nevertheless, it is necessary that elements (3) and (5) are designated with the same letter c (see above), in order to eliminate the confusion, which occurs with these wedges. The elements in category a are the signs DIŠ, even if the vertical wedge turns slightly to the right, which is a matter of perspective, however. The next category b is the sign AŠ. The element c combines all kinds of Winkelhaken and wedges that are slanted left. Category d combines all the wedges that are slanted to the right.

Fig. 1

When the elements are defined like this, the building of a coherent system is a simple deduction. Still, only letters as building parameters are not enough for a precise specification of every single cuneiform sign. Therefore, it is necessary to provide each element with a number index. In order to do this, one shall count the elements within a single sign and then add numbers to the relevant designations. Then the sum of all the elements is the category of each sign. Accordingly, all signs with the same number of wedges are situated in the same category. In this way, signs can be ordered easily and then found in a matter of seconds. The sign EME consists of the following parameters a3, b5 and c1. This comprises category 9 (< 3+5+1). The sorting of the parameters follows the alphabet and is additionally combined with the indices (Fig. 2).

Fig. 2

Abbreviations: MA = Middle Assyrian, see Borger 1981 and Borger 2010

To find the EME sign in the paper version of the ‘Spotlight’ one shall go to category 9, noted in the right upper corner of the list and then to a3b5c1, which is situated after signs with parameters a3b4c2 (Fig. 2). If you search for it in the PDF version or in another electronic format, you shall simply type a3b5c1 in the search box and you will immediately find the signs consisting of these parameters. The order of the system follows the alphabet order and the decimal indexes. For instance, in category nine the sign MUŠ has the parameters a3b1c5, which will be followed by UZ3 with the parameters a3b2c2d2 (Fig. 2). Then come the signs with the parameters a3b2c4 then the signs with parameters a3b3c3, then another form of the sign UZ3 with the parameters a3b4c2 etc. Crucially, once the list has been established, one can insert and take out signs very easily and the system is always working and really flexible. In fact, the system cannot define each sign hundred percent precise. But, if the signs are structured in a list, this can be an advantage since it brings similar signs together. Thus, one can make comparisons between signs with the same parameters (Figs. 2 and 3).

Fig. 3

Abbreviations: MA = Middle Assyrian, see Borger 1981 and Borger 2010; SH = Šē̄ḫ Ḥamad, see Cancik-Kirschbaum 1996; TB = Tall Bderi, see Maul 1992; TC = Tell Chuēra, see Jakob 2009; TP = Tiglath-pileser I, see Weidner 1952/3.

In category four, the signs KAD2, BULUG3, DIN and MUNUS have the same parameters a1b1c2, as BA and NA do. The last two signs can be confused in some cases. The system provides the opportunity to check similar signs altogether, which was hardly possible with the other sign lists. Thus, a beginner can for instance compare EME and UKKIN (Fig. 2), or NA and BA, (Fig. 3). Then a beginner can decide for himself, which sign is more likely in the particular context, or easily understand that different signs sometimes look the same.

The number of wedges is decisive for the categorization of each sign. This is also reminiscent of the syllabaries, where different signs with the same number of wedges were occasionally listed next to each other. For instance, in Syllabary A some signs with the same number of wedges are combined together. The number in brackets on Fig. 4 refers to the number in Syllabary A (the sequence of signs follows Hallock 1955: 5-41).

Fig. 4

This illustrates that the sum of the wedges played some role in the ordering of signs in ancient syllabaries. However, this ordering principle is not the only criterion in the syllabaries, since phonetic values and different associations played an important role as well.

Yet, the more wedges there are, the more complicated Gottstein’s system will become. Therefore, it is especially convenient for periods with fewer wedges. For instance, the majority of the Middle and Neo-Assyrian signs span between categories 4 and 10, i.e., consisting of 4 to 10 wedges each. This system may be also particularly useful for palaeography research. Once the ‘spotlight’ is established it can be further developed into a palaeographic chart (Fig. 5).

Fig. 5

Abbreviations: Asb = Ashurbanipal; SH = Šē̄ḫ Ḥamad, see Cancik-Kirschbaum 1996; TB = Tall Bderi, see Maul 1992; TC = Tell Chuēra, see Jakob 2009; TP = Tiglath-pileser I, see Weidner 1952/3.

Such a palaeographic chart shall include more material, combined also in the initial order of the ‘spotlight’. While Van Buylaere (2009) established a palaeography of the Neo-Assyrian period, this is still not the case for the Middle Assyrian period (Freydank 2010: 252-259). For a comparison of Middle- and Neo-Assyrian sign forms, diverse colophons or eponym-dated texts shall be included. Essentially Royal Inscriptions, dated literary, or magical and medical texts shall be used as well. The signs shall come from clay tablets, clay cones, clay prisms etc. and not from monumental inscriptions on stone or metal, which employ specific shapes also due to the execution in the hard material. Only dated texts shall build up the chronological sequence for each single sign. Occasionally, sign forms seem to be chronologically distinguished as the sign TI shows older and newer forms in the 15th and 14th centuries BC (Freydank 1988: 73-84; Weeden 2012: 241), or as the AZ for the 13th century BC (Jakob 2005-2006: 326). Early Middle Assyrian period sign forms could be compared with clay tablets from the Egyptian capital Amarna, where letters from the Middle Assyrian King Aššur-uballiṭ I were unearthed (Kühne 1973: 77-83). Concerning Amarna, Kühne recognized parallels between sign forms in the Tušratta letters and the Assyrian sign forms (Kühne 1973: 44-45, note 209, 135-139). Other peripheral archives, such as Hattuša or Ugarit, also shed some light on the development of sign forms in the early Middle Assyrian period (Schwemer 1998: 8-39; van Soldt 2001: 429-444; Weeden 2012: 229-251). Sign forms from Šē̄ḫ Ḥamad: 13 century BC (Cancik-Kirschbaum 1996; Salah 2014), Tell Chuēra: 13th-12th centuries BC (Jakob 2009), Tall Bderi and Tall Ṭābān: 12th-11th centuries BC (Maul 1992 and 2005), and the so-called ‘Tiglath-pileser I library’ shall be further compared with the sign forms from Ashur (summary in Freydank 2010: 252-259), Sabi Abyad: 13th-12th centuries BC (in preparation by Frans Wiggermann), Kār-Tukultī-Ninurta: 13th century BC and others (see Freydank 2009: 16) or Dunnu-ša-uzibi: 11th century BC (Radner 2004: 54-61, 244-245). Tall Ṭābān signs are similar to Tall Bderi and are chronologically contemporary to the sign forms from the ‘library of Tiglath-pileser I’ (Maul 2005: 20, 83). Nevertheless, Daisuke Shibata traced some peculiar sign forms from Tall Ṭābān, possibly pointing to local differences (personal communication). Further texts from Tall Ṭābān are in preparation (Shibata and Yamada 2009: 87-109). Although, the signs from the so-called ‘Tiglath-pileser I library’, listed in Weidner (1952-1953: 201), are often mentioned throughout the literature, the context of the ‘library’ is not solely reserved for Tiglath-pileser I alone. The dating of the formally similar texts ranges from Tukultī-Ninurta I (late 13th century) to Aššur-bēl-kala at least (end of the first half of the 11th century BC). The existence of such a ‘library’ was already questioned by Lambert (literature and discussion in Pedersén 1985: 32, M2). Furthermore, some texts formally associated with the ‘Tiglath-pileser I library’ were used down to the Neo-Assyrian period until the 7th century BC (Pedersén 1985: 31, M2; Pedersén 1986: 12-14, 17 and N1). This casts doubt on their purported Middle Assyrian origin. Theoretically, some of them could also have been written in the Neo-Assyrian time, while sharing the formal similarities of the ‘Tiglath-pileser I library’ texts. For instance, VAT 10113 (Fig. 5) formally belongs to the ‘Tiglath-pileser I library’ (Weidner 1952-53: 207, no. 8) and is part of the so-called Middle Assyrian Coronation ritual (Müller 1937: 4, 47 dates the text in the late 13th until the late 12th century BC). Yet, the fragments belong to the archaeological context of the Neo-Assyrian period (Pedersén 1985: 31, N1: 70 and N1: 100 in Pedersén 1986). Therefore, the precise dating of the ‘Tiglath-pileser I library’ texts without colophons or eponyms is uncertain. Furthermore, the 10th century BC is still scarce on information and there are few Royal inscriptions on clay objects (Grayson 1991: 125-141).

For a comparison of the Middle Assyrian and Neo-Assyrian sign forms the archive in Kalhu is particularly important. In Kalhu, texts are datable from the 9th till the 7th century BC (Black 2008: 261-265, especially 263). Occasionally, texts from Kalhu were taken to Nineveh (e.g., Fincke 2003-2004: 115, note 24, 140; Black 2008: 264) and to Ashur as well (e.g., CTN 4, 58 (ND 5545) // KAR 147 (VAT 8780), reign of Ashurnasirpal II).

As shown by Weidner (1952-53: 201), the BA, ZU, SU group is abundant for comparisons of Middle and Neo-Assyrain. The BA sign will be used as an example (Fig. 6). BM 91033 is the Tiglath-pileser I prism from Nineveh, used for the ‘test case’ by the Royal Asiatic Society in 1857. VAT 8780 (Ashur) and ND 5545 (Kalhu) are hemerologies on clay ‘tablets with handle’ concerned with first eight days of Tishri and both date to the time of Ashurnasirpal II (Livingstone 2013: 177-189; Casaburi 2000: 13-29; Pedersén 1986: N4 (455). VAT 9026 is a dream ritual and dates to the eponym of Aššur-bāni ca. 713 BC. It stems from the house of incantation priests in Ashur (Pedersén 1986: N4 (87); Butler 1998: 249-312). In this case, sign forms under Ashurnasirpal II are closer to the ‘Tiglath-pileser I library’ then to Nineveh. The sum of the wedges illustrates the three basic forms of BA: a1b3 (mainly 8th-7th, but also 11th century BC) and a1b1c2 (mainly 13th-11th centuries BC), and a1c3 (13th century BC). The main criteria for the BA sign are the four wedges, always a1 with variations of b and c in front of it. It is obvious that regional differences played an important role and that it is rarely possible to date the text on palaeographic grounds only. In case of undated texts the palaeographic grounds, ductus (Weeden 2012: 229) and the general appearance of the cuneiform tablet shall be further considered in addition to the archaeological context and dialect (Pedersén 1985: 22)

Fig. 6

Abbreviations: Asb = Ashurbanipal; GR = Giricano, see Radner 2004; SH = Šē̄ḫ Ḥamad, see Cancik-Kirschbaum 1996; TB = Tall Bderi, see Maul 1992; TC = Tell Chuēra, see Jakob 2009; TP = Tiglath-pileser I, see Weidner 1952/3.

Moreover, not much development can be observed in some signs from the Middle Assyrian to Neo-Assyrian period (Figs. 5, 6). Yet, another palaeographical peculiarity is that even the scribes at Nineveh sometimes applied sign forms also used in Middle Assyrian texts. Compare, for instance, the different writings of IGI.SIG7.SIG7 in two consecutive lines of a tablet with Ashurbanipal colophon from Nineveh (Fig. 7).

Fig. 7

All this makes a comparison and assessment of the Middle and Neo-Assyrian signs fairly complex. It is necessary to establish a chronological sequence of dated texts for each sign. If the ‘spotlight’ is transposed to palaeographic charts, all signs can be traced in a matter of seconds on the computer, both in the ‘spotlight’ and in the palaeography chart as well. New signs can be easily inserted in the system, presently being in InDesign format. Thus, the ‘spotlight’ and the palaeographic chart can grow over time and notably always remain coherent.

Gottstein’s system offers an exact model, which could be learned in five minutes, providing an opportunity to define each sign, and to find each sign swiftly, which is especially essential for beginners. Furthermore, through this mathematical model one can easily describe each sign in cuneiform classes. It provides an opportunity to bring dynamic order to the signs, distinguishing the cuneiform signs into clear categories. In this respect the ‘spotlight’ is a supplementary list, supplying a summary of the existing similar signs. Furthermore, (digital) palaeographic charts can be made of this list.


Black, Jeremy A.
2008“The libraries of Kalhu”. In: J. Curtis, H. McCall, D. Collon and L. al-Gailani Werr (ed.) New Light on Nimrud (London):261-266.
Borger, Rykle
1981Assyrisch-babylonische Zeichenliste. Neukirchen-Vluyn.
2010Mesopotamisches Zeichenlexikon. Zweite, revidierte und aktualisierte Auflage, AOAT 305. Ugarit-Verlag: Münster.
Butler, S. A. L.
1998Mesopotamian Conception of Dreams and Dream Rituals, AOAT 258. Münster.
Cancik-Kirschbaum, Eva
1996Die Mittelassyrischen Briefe aus Tall Šēḫ Ḥamad, BATSH 4. Reimer: Berlin.
Casaburi, M. C.
2000“The Alleged Mesopotamian «LENT» The Hemerology for Tešritu”. SEL 17:13-29.
Cathcart, K. J.
2011“The Earliest Contributions to the Decipherment of Sumerian and Akkadian”. CDLJ 2011:1.
Delitzsch, F.
1900Assyrische Lesestücke mit grammatischen Tabellen und vollständigem Glossar. Leipzig.
Ebeling, E.
1915-1919Keilschrifttexte aus Assur religiösen Inhalts, WVDOG 28. Leipzig.
Fincke, J.
2003-2004“The Babylonian Texts of Niniveh”. AfO 50:111-149.
Freydank, H.
1988“Zur Paläographie der mittelassyrischen Urkunden”. In: P Vavroušek, V. & V. Souček (ed.) Šulmu, Papers on the Ancient Near East Presented at International Conference of Socialist Countries (Prague):73-84.
2009“Kār-Tukultī-Ninurta als Agrarprovinz”. AoF 36:16-84.
2010“Perspektiven einer mittelassyrischen paläografischen Liste”. AoF 37:252-259.
Gottstein, N.
2012“Ein stringentes Identifikations- und Suchsystem für Keilschriftzeichen.”. Material Text Culture Blog 2012:5.
2013“Ein stringentes Identifikations- und Suchsystem für Keilschriftzeichen”. MDOG 145:127-136.
Gottstein, N. & S. V. Panayotov
2014Cuneiform Spotlight of the Neo- and Middle Assyrian Signs. Dresden.
Grayson, A. K.
1991 Assyrian Rulers of the Early First Millennium BC I (1114-859 BC), RIMA 2. Toronto.
Hallock, R. T.
1955Syllabary A, MSL 3. Roma.
Jakob, S.
2005-2006“Review of H. Freydank-B. Feller, MARV 5”. AfO 51:324-329.
2009Die mittelassyrischen Texte aus Tell Chuēra in Nordost-Syrien. Wiesbaden.
Köcher, F.
1980Die babylonisch-assyrische Medizin in Texten und Untersuchungen VI. Berlin.
Kühne, C.
1973Die Chronologie der internationalen Korrespondenz von El-Amarna, AOAT 1. Neukirchen-Vluyn.
Livingstone, A.
2013Hemerologies of Assyrian and Babylonian scholars, CUSAS 25. Bethesda, Maryland.
Maul, S.
1992Die Inschriften von Tall Bderi, BBVOT 2. Berlin.
2005Die Inschriften von Tall Ṭābān (Grabungskampagnen 1997-1999). Die Könige von Ṭābētu und das Land Māri in mittelassyrischer Zeit, ASJS 2. Tokyo.
Müller, Karl F.
1937Das assyrische Ritual. Teil I. Texte zum assyrischen Königsritual, Mitteilungen der Vorderasiatisch-Ägyptischen Gesellscahft 41/3. J. C. Hinrichs Verlag: Leipzig.
Pedersén, Olof
1985Archives and Libraries in the City of Assur. A Survey of the Material from the German Excavations. Part I, Studia Semitica Upsaliensia 6. Almqvist & Wiksell: Uppsala.
1986Archives and Libraries in the City of Assur. A Survey of the Material from the German Excavations. Part II, Studia Semitica Upsaliensia 8. Almqvist & Wiksell: Uppsala.
Radner, K.
2004Das mittelassyrische Tontafelarchiv von Giricano/Dunnu-ša-uzibi, Subartu 14. Turnhout.
Salah, S.
2014Die mittelassyrischen Personen- und Rationenlisten aus Tall Šēḫ Ḥamad/Dūr-Katlimmu, BATSH 18. Wiesbaden.
Schibata, D. & S. Yamada
2009“The Cuneiform Texts From The 2007 Excavations At Tell Taban: A Preliminary Report”. In: H. Numoto (ed.) Excavations at Tell Taban, Hassake, Syria. Preliminary Report on the 2007 Season of Excavations, and the Study of Cuneiform Texts ().
Schwemer, D.
1998Akkadische Rituale aus Hattuša. Die Sammeltafel KBo XXXVI 29 und verwandte Fragmente. Heidelberg.
Van Buylaere, G.
2009Neuassyrische Paläographie. PhD Dissertation, Udine.
Van Soldt, W. H.
2001“Naḫiš-šalmu: an Assyrian scribe working in the ‘Southern Palace’ at Ugarit”. In: W. H. Soldt et al. (ed.) Veenhof Anniversary Volume: studies presented to Klaas R. Veenhof on the occasion of his sixty-fifth birthday (PIHANS 89; Leiden):429-444.
Von Soden, W. & W. Röllig
1991Das Akkadische Syllabar (4. durchgesehene und erweiterte Auflage), AnOr 42. Roma.
Weeden, M.
2012“Assyro-Mittanian or Middle Assyrian?”. In: E. Devecchi (ed.) Palaeography and Scribal Practices in Syro-Palestine and Anatolia in the Late Bronze Age. Papers read at a symposium in Leiden, 17-18 december 2009 (PIHANS 119; Leiden):229-251.
Weidner, E. F.
1952-1953“Die Bibliothek Tiglatpilesers I.”. AfO 16:197-215.
Wiseman, D. J. and J. A. Black
1996Literary Texts From The Temple Of Nabû, CTN 4. Oxford.

* I thank Jaume Llop, Luděk Vacín, Mark Weeden and Frans A.M. Wiggermann for help with the issues related to the article. During the writing of the article it was impossible to consult Van Buylaere 2009 and Salah 2014.
ISSN 1546-6566    © Cuneiform Digital Library Initiative | Archival: 0000-00-00