In his diachronic overview of the lexical tradition in the Mesopotamian textual record Niek Veldhuis referred to several Early Dynastic compositions that defy a clear classification as lexical texts. One among these texts sticks out, because of its comparatively long and complex history of transmission (Veldhuis 2014a: 128-129). This composition, which is given the preliminary designation Word List Z (henceforth WL-Z), is attested in sources from the middle of the third until the early second millennium BC. In this short contribution I attempt to discuss the various sources in more detail and deal with some of their characteristics, which hopefully shed some light onto this composition’s purpose and place in the late stages of transmitting third millennium scribal lore. A second aim is to incorporate some previous attempts to classify manuscripts of this composition, which aim at supplementing Veldhuis’ overview.
The classificatory designation “word list” is used here in a rather superficial manner. Word lists can be cursorily distinguished from sign lists, the latter usually grouping together similar sign forms or displaying graphemes in a determined sequence, which frequently concerns a grapheme’s complexity or certain grapheme features. Both categories of lists do not have clear boundaries. Word lists, in general, range from thematic or multi-thematic lists to vocabularies, both either with clearly marked categories (see graphemic classifiers) or containing entries which belong together because of certain features (e.g., an Emesal vocabulary). The various types of word lists are easily accessible in Veldhuis’ book, but from a methodological point of view it is important to note the ever-changing nomenclature. The term “word list” was mainly used by Giovanni Pettinato in his publication of the lexical texts found at Ebla (Pettinato 1981). Some of these designations, however, were partially abandoned and new designations were introduced in the secondary literature:
WL-Z has no archaic predecessor to my knowledge. According to the text witnesses which are currently available its origins can be traced back to the later Early Dynastic period in the scribal centers Fāra and Tell Abū Salābīkh. This of course does not exclude an earlier date of compilation, which will be briefly discussed further down. At our current stage of knowledge, this text is known from the following six sources (Q000292; see Veldhuis 2014a: 128):
Both the Fāra (A) and the Tell Abū Salābīkh (B) manuscripts follow the customs of the time. In each case the text is inscribed onto a large square or slightly rectangular tablet. It should be noted that none of the known sources introduces the entries by either the curvilinear horizontal wedge (1(ašc)) or the vertical wedge (1(diš)) in the later sources. If interpreted as a lexical text, the lack of any numerical notations is not per se clue enough for this composition being not part of the Early Dynastic lexical tradition. Tablets of approximately square shape are already attested in archaic Uruk at the end of the fourth millennium BC and this format is subsequently used until the end of the third millennium, as is now confirmed by the dated source of ED Plants (EDP) 6N-T 933 (see Civil 2013: 66, fig. 3). But this shape does not necessarily provide any hints for the content of a text, since at the various scribal centers of the Early Dynastic period this format was used to convey literary sources as well. A different medium is the multi-sided prism. Prisms are, however, absent from the Early Dynastic record and its use as medium for lexical texts can just be observed from the Old Akkadian period onwards (see also Taylor 2008: 206).
So far no traces of WL-Z can be found at Kish or beyond in Ebla, but it is an intriguing circumstance that the text was copied at Susa in neighboring Iran. The attestation of Mesopotamian lexical material is comparatively scarce at Susa, but this might as well be just a coincidence. MDP 14, 1 (C) is the large corner fragment of a multi-column tablet, which contained on its obverse a full copy of WL-Z. The lexical sources from Susa, which predominantly date to the Old Akkadian period, might appear to be chosen texts, for the contents of EDLu-A and ED Food (EDFo) demonstrate a rather close link to administrative processes, although their relevance for economic terminology of that time is negligible (Veldhuis 2014b: 245). But why did the text witness of WL-Z find its way to Susa? Is it mere coincidence? In any case, this text follows the Mesopotamian customs. The original tablet was of approximately square shape with its reverse left (mostly) blank.
All remaining sources are unfortunately unprovenanced, because they come from the antiquities market and their origin is therefore shrouded in obscurity. The seven-sided prism YOS 1, 11 (D) comes certainly next in the chronology of this composition. This prism (YBC 2126) arrived at the Yale Babylonian collection together with another, rather similar prism, a copy of the Standard Profession List Lu2 A (YBC 2125). Both prisms are said to originate from Nippur, but there is no real indication in the texts themselves which substantiate this claim. Regarding their scribal hand, both prisms can be dated to the Old Akkadian period at the earliest, although an Ur III date seems more likely. (The copy of EDLu-A bears a colophon on its top side; see also Veldhuis 2014a: 70, n. 115.)
A further prism containing a full copy of the text is currently unlocated, but was provisionally called “Ferrini prism” (E). Its date is uncertain, ranging from between the Ur III and the Old Babylonian period.
The most elusive source, however, can be dated to the Old Babylonian period. This unprovenanced manuscript is currently kept at the Cotsen Collection of Cuneiform Tablets (UCLA). This manuscript, PARS 12/01, 131 (henceforth Wilson 2008, 134; 14.2 × 11.5 × 4.9 cm) is in fact the fragment of a large compilation tablet originally containing copies of several compositions. The reasons for grouping together certain compositions on one tablet needs to be addressed separately for each instance. Often such reasons remain hidden. It is this text witness which will be discussed here in more detail.
Towards an understanding of the double rulings on Wilson 2008, 134
An intriguing feature on this tablet are double rulings, which are introduced in the text on various occasions. The preserved part of the text on the obverse contains one of these double rulings per column. (It seems that the scribe inserted in column iii' double rulings after each of two subsequent entries. This might have been caused by a partially erased entry in column ii', for which collation is needed.) In each case these double rulings appear approximately at the same height in the column. Double rulings are not an uncommon feature. In the Old Babylonian period they frequently appear at the end of a composition or extract before a rubric or catch line. But double rulings often mark the boundaries between conceptual units in a text as well. On the Old Babylonian literary text BM 15794, for instance, the boundaries between the various sections are rather clear judging by the respective content of each section (Wagensonner 2010: 219-242). These explanations, however, do not fit every instance and one must search for different solutions.
Double rulings represent an important feature in the tradition of the Standard Profession List Lu2 A (henceforth EDLu-A). Although this feature is particularly prominent on copies of the Old Babylonian period, it can already be observed in the ED IIIb sources originating from Ebla. The four exemplars known from this site can be divided into two groups, each consisting of a large square tablet according to the Mesopotamian lexical tradition and a small rounded tablet often referred to as “exercise” tablet. Each column on the two larger exemplars contains nineteen entries; in the manuscripts from Fāra and Tell Abū Salabikh the column lengths still vary slightly from source to source. The number nineteen is no coincidence, since the list totalling 129 entries fits perfectly in seven units à nineteen entries. The two smaller tablets in Ebla can only contain up to ten entries per column. However, each scribe marked the boundaries between every unit by a double ruling (Fig. 1). The most likely explanation for this is to see in each small tablet a (faithful) copy of the larger exemplars and not a preparatory version for drawing up the larger tablets. This interpretation is substantiated by the respective colophons. The colophon of the small rounded tablet MEE 3, 1 (TM.G.75.1312) refers to the scribe of the larger exemplar MEE 3, 2 + 5 (TM.G.75.2586+) (Archi 1992: 2-3)
The insertion of double rulings is also known from further lexical sources originating from Ebla. Alfonso Archi refers to several versions of ED Fish (EDFi), among which the two square manuscripts MEE 3, 28+ and MEE 3, 29+ contain columns of sixteen entries each. The list is further accessible in a non-orthographic version, where its content was divided into two halves of 48 entries each and inscribed onto two small tablets with rounded corners. Amid the small fragment MEE 3, 64 these tablets still await their publication. According to Archi both tablets insert double rulings after each group of sixteen entries (Archi 1992: 7). For EDLu-A nineteen entries per unit somewhat became a standardized length of columns in the centuries to follow. This is, in particular, apparent in the Old Babylonian sources, which are currently known. On SLT 112+ a copy of EDLu-A is preceded by a long list of personal names known as Ur-ki (see Peterson 2012: 252-253). Both the content of the personal names as well as the titles and professions list are subdivided into sections. In the latter case each section consists of nineteen entries. This evidence is not, as was discussed by Jon Taylor, coincidental. Other Old Babylonian sources of EDLu-A show the same feature as well. It is apparent on some annotated versions, such as the fragments UET 7, 86, which inserts a double ruling between entries 114 and 115 (i.e., 6 * 19). These double rulings therefore entered the stream of tradition. None of the known manuscripts of the Old Babylonian period belonging to EDLu-A diverts from the division of text into units of nineteen entries each. As in Ebla, we can most likely interpret this feature as a faithful transposition of a source text onto another medium such as a compilation tablet, rather than interpreting these artificial boundaries as separating conceptual units.
This feature is not restricted to EDLu-A and I would like to suggest a similar explanation for the double rulings on Wilson 2008, 134. Thanks to the earlier but parallel versions of WL-Z it is possible to calculate the original length of each column on this fragment. Column i' on the obverse starts with entry 13 of the text (according to text witness D). It is thus clear that WL-Z was not the first composition copied onto this tablet. The tablet’s profile supports two columns prior to the first preserved one. The calculated length of each column is 55 lines. Based on the placement of the double rulings, it seems quite likely to divide the copy of WL-Z into six sections of more or less equal length (Fig. 2).
The currently unlocated manuscript E (“Ferrini prism”) is the upper half of a hexagonal prism, which roughly dates to the end of the third or beginning of the second millennium BC. Face d starts with entry ir-nun. The same entry on the compilation tablet Wilson 2008, 134 is preceded by a double ruling. In the other instances the placement of double rulings roughly coincides with entries on top of each side of the “Ferrini prism” as well. This cannot be a coincidence and it appears rather likely to interpret the double rulings on Wilson 2008, 134 as the boundaries of columns on a source text such as a prism. Fig. 3 contains the representation of such a hypothetical source for the compilation tablet.
Double rulings occasionally occur at the bottom of columns or sides of prisms, such as in the case of the lexical text AD-GI4. At the bottom of each side on the hexagonal prism NMB 78411 the scribe inserted such a double ruling. For the Old Babylonian period the layout of this composition and therefore the length of its columns was widely standardized. This is supported by the text witness MS 3373, whose columns end with the exact same entries as on the aforementioned prism and contain double rulings as well.
Attempting an interpretation of WL-Z
Niek Veldhuis incorporated WL-Z into his diachronic overview of the lexical tradition in Mesopotamia and discussed this text as an unclear case. Not much can be done in order to clarify the many uncertainties. However, this text appears to be quite distinct from other lexical sources, in particular of the Early Dynastic period. One thing rests assured, the long tradition of this composition shows that its transmission throughout the second half of the third millennium and beyond at various places is certainly not an accident.
It is said that manuscript D probably originates from Nippur. This seven-sided prism received some attention in the past. It was first addressed by Jan van Dijk in the Vorberichte to the German excavations at Uruk. While comparing it to an ED I-II tablet found at Uruk, van Dijk discusses the prism as a possible literary text or a collection of literary expressions (van Dijk 1960: 59). Indeed this text contains quite a few short phrases and expressions and van Dijk’s suggestion, still supported by Niek Veldhuis (Veldhuis 2014a: 128), is still valid.
The first lines, which led van Dijk to this interpretation will not be repeated here (van Dijk 1960: 59 and Veldhuis 2014a: 128). The text appears to contain further entries, which allow for similar observations, most notably entries 129-131:
Interpreting WL-Z as lexical text is possible, although it lacks any obvious thematic organization. A feature that occurs several times is a method of arranging entries I like to refer to as “Cascade” (D.1.a). This pattern concerns a group of entries, which can be subdivided into smaller subgroups. These subgroups are linked together through entries, which contain a grapheme (group) used in the preceding entries and introduce a new grapheme (group) used in the ensuing entries. This well attested feature in early lexical texts can be demonstrated by the following example in WL-Z:
Configurations such as this example can reach quite a high level of sophistication, where the common denominators even concern certain graphical features of graphemes. This is best seen in the Ebla Sign List (ESL).
In his dicussion of the Early Dynastic Practical Vocabulary A (EDPV-A) Miguel Civil referred to WL-Z and argued that it might contain Semitic glosses (Civil 2008: 146). This observation was based on two subsequent entries, which according to Civil might refer to a term for “water pipe” (EDPV-A 378: gešE2×AŠtenû compared to gešE2 in WL-Z 104 and followed by a-la-lum), but there might be some other clues in the text in order to substantiate this possibility.
As for Example 1, it is quite striking that these two entries are preserved in the lexical text Nigga. Unfortunately the bilingual version of this lexical text (YBC 13524) is too damaged to verify the first entry (Civil et al. 1971: 118; collation: Elizabeth Payne). In entry 123 šu la2-la2 is rendered uṣṣulum in Akkadian. In WL-Z this entry is followed by ur-ge6, for which canonical Ura provides the rendering ṣalmu, “black”, therefore sharing some consonantal features. Example 2 might seem far-fetched, for the Semitic rendering for ir-nun derives from the Ebla Vocabulary. Be it as it may, it is not unlikely that there are several text layers present in WL-Z, which can be unlocked via such implicit links.
Single text witnesses of WL-Z received attention in various studies. Manuscript D was related to a prism and a cylinder kept in the Yale Babylonian Collection as well. Both YBC 2124 and NBC 11202 share many features with manuscript D (Cohen 1993a):
Cohen relates these texts to the lexical tradition of name lists. Recently Josef Bauer added to Cohen’s study further evidence and included manuscript B of WL-Z (Bauer 2014). Due to the date of publication Niek Veldhuis could not include Bauer’s study into his overview. Name lists have a long tradition reaching back at least to the ED IIIa period. The ED Names and Professions List (EDNPL) is attested in Tell Abū Salābīkh, Ebla and Fāra. At the last-mentioned site several other lists containing names and professions were compiled (see Fāra Names and Professions List; SF 29 with extracts SF 28 and SF 44 and possibly SF 63; see Pomponio 1984: 555). The two prisms published by Cohen date roughly to the Ur III period and are preceded by an Old Akkadian list containing personal names starting with the element nin possibly from Lagaš (Lambert 1988: 260-261). The Old Babylonian period sees the emergence of a wide variety of name lists, which are now easily accessible through the recent overview by Jeremiah Peterson (Peterson 2011).
Bauer could show that many entries in manuscripts B and D can be connected to personal names in administrative texts dating to the late Early Dynastic period. The parallels to the other two texts at Yale seem to prove that the various compilers had access to quite similar source material.
The reverse of Wilson 2008, 134
As mentioned above manuscript F of WL-Z is a compilation tablet. The reasoning for collecting various compositions onto one tablet is not always apparent. Already at Fāra certain thematic lists were combined. SF 9, for instance, contains ED Fish (EDFi) followed by ED Metal (EDM). SF 58 combines ED Plants (EDP) and ED Birds A (EDB-A). In both cases the texts combined into one tablet belong to the same genre, namely lexical texts. SF 18 was interpreted differently. This tablet contains a literary text written in UD.GAL.NUN orthography followed by a lexical text, which partially contains entries written in this orthography as well (Zand 2009: 190, text CUT 2; Krecher 1978: 156-157).
In the Old Babylonian period the use of compilation tablets is quite wide-spread. School texts such as the so-called Type II tablets contain on one side the extract of a lexical text and advanced exercises such as proverbs on the other (Veldhuis 1997: 35). The compilation tablet Wilson 2008, 134 was not such an exercise. It was written by an advanced scribe. This is implied by the careful script as well the even layout of both columns and lines. Wilson 2008, 134 was a fairly large tablet, which is usually referred to as Type I tablet (Veldhuis 1997: 30-31). As was demonstrated above, each column on this tablet had approximately 55 lines and we can infer from the tablet’s profile that each side had about seven to eight columns. This leaves quite a lot of space, which is now lost.
At first glance, the text preserved on the reverse appears to be quite different from the composition on the obverse. And indeed the text contains widely intelligible Sumerian phrases. This is quite different from the rather elusive content of Word List Z on the the tablet’s obverse. In his general book on scribal education Mark Wilson notes that this text contains several instances of the Sumerian negative clause nu-tuku, “not having” (Wilson 2008: 227). It is, in particular, concentrated at the end of column iv’. While looking for possible parallels I came across the small fragment OSP 1, 2 containing a couple of lines, which align quite neatly with the afore-mentioned passage on Wilson 2008, 134:
It was recognized just after Aage Westenholz published OSP that this fragment belongs to a composition identified as Early Dynastic Proverb Collection (henceforth EDPC 1). Bendt Alster collected the sources and published an edition of this composition, which is best known through sources from the sites of Fāra and Tell Abū Salabikh (Alster 1991/1992). It is one among a few Early Dynastic literary texts, which were still copied in the Old Babylonian period. Up until now – and this is also the case with new textual sources from the Schøyen collection – all known text witnesses of the early second millennium BC do not represent full copies, but just take small portions of the text into account. The sources, which were available to Alster concentrated on the beginning of the composition. The four-sided prism Ashm 1922-169 is in this respect most revealing, because the beginning lines of EDPC 1 dealing with insults against women were recycled and used to start the dialogue between two women.
To this list four further sources need to be added, which are currently in the Schøyen collection, which contains parts of EDPC 1. It is not the intention of this paper to treat EDPC 1 any further, but it must be emphasized that the reverse of Wilson 2008, 134 is the only source known so far, which contains a full copy of EDPC 1. Like the obverse, the reverse contains double rulings. There is one in column iv'. Having argued that the double rulings on the tablet’s obverse represent the boundaries between columns or sides on a source text, which had been faithfully copied onto the compilation tablet, it is not unlikely that the version of EDPC 1 on its reverse follows the layout of a source tablet as well.
The grouping of this text together with the Early Dynastic Proverb Collection on the same tablet does not necessarily give any clues towards an understanding of its content, although the obscure entries found in it might indeed have had some significance for this combination.
EDPC 1 is known from several bilingual versions, which cover the text partially. Both UM 29-15-174 and BT 9 are fragments of two-column tablets, which contain in the respective left column the Sumerian text including a pronunciation gloss in smaller script and an Akkadian equivalent to the right. Jacob Klein published BT 9 and argued that both fragments might belong to the same tablet or at least recension of the text. If both fragments are part of the same tablet, the text preserved on them might have had covered only a selection of entries of EDPC 1 (Klein 2003: 136-138).
Unfortunately the text covered by BT 9 is almost completely in the broken area of Wilson 2008, 134. Hence, both versions cannot be compared properly in terms of variants.
Amid the large missing portions of text on the compilation tablet Wilson 2008, 134 WL-Z was possibly conceived as a collection of sayings or short phrases suitable to be related to a collection of proverbs. Was WL-Z still conceived as a list of names at the beginning of the Old Babylonian period? It offers many vantage points for text analysis and interpretation, some of which are difficult to prove – such as a possible Semitic text layer. Niek Veldhuis emphasized the high amount of variation between the available sources (Veldhuis 2014a: 129). This adds to the issues in interpreting this text.
The reverse of Wilson 2008, 134 has its own significance, since it contains the most complete version of the Early Dynastic Proverb Collection on a tablet of the Old Babylonian period. Further study of this new source is necessary in order to relate it to the other known sources of EDPC 1. Besides the Keš Temple Hymn and the Instructions of Šuruppak EDPC 1 is the third major literary source of the later Early Dynastic period, which received ample attention in the scribal milieu of the Old Babylonian period.