CCP 6.6 - Neo-Babylonian Grammatical Text (9)

Catalogue information
Free Library of Philadelphia
FLP unn72
Nippur(Nippur) (?)
AfO 24 79 and 81 fig. 1-2

Leichty, 1973E. Leichty, Two Late Commentaries, Archiv für Orientforschung, vol. 24, pp. 78-86, 1973.: 79-82

LexicalGrammatical texts


Base text: 
Neo-Babylonian Grammatical Text (9)
Tablet information
obv 28, rev 18
Achaemenid (5th cent - 331 BCE) (Uruk, Anu-ikṣur / Nippur / Babylon)
Ninurta-[...] (?) s. Andul-d[...] (?)

Frahm, 2011E. Frahm, Babylonian and Assyrian Text Commentaries. Origins of Interpretation. Ugarit-Verlag, 2011.: 40, 75

Gabbay, 2016U. Gabbay, The Exegetical Terminology of Akkadian Commentaries. Brill, 2016.: 24 (5, 17), 119 (2), 144 (14), 213 (6), 119, 151 (7), 199, 287 (24)

Gabbay & Jiménez, forthcomingU. Gabbay and Jiménez, E. , From Nippur to Uruk: The Tablets of the Gimil-Sîn Family.
[On the colophon]

Leichty, 1973E. Leichty, Two Late Commentaries, Archiv für Orientforschung, vol. 24, pp. 78-86, 1973.
: 79-82

Oelsner, 1986J. Oelsner, Materialien zur Babylonischen Gesellschaft und Kultur in Hellenistischer Zeit. Eötvös Loránd Tudományegyetem, 1986.
[vorhellenistisches Datum]
: 469-470 fn. 882

Frazer, 10/2016 (Transliteration)
Frazer, 10/2016 (Translation)
Frazer, 10/2016 (Introduction)
Frahm & Jiménez, 10/2016 (Suggestions)
Veldhuis, 04/2017 (Suggestions)
By Mary Frazer |
Cite this edition
Frazer, M., “Commentary on Neo-Babylonian Grammatical Text (9) (CCP no. 6.6),” Cuneiform Commentaries Project (2017), at (accessed October 24, 2017)
Make a correction or suggestion

This portrait-oriented tablet, FLP unn72, contains the only attested commentary on the Neo-Babylonian Grammatical Texts (henceforth NBGT).1 NBGT, which was edited by R. Hallock and B. Landsberger in MSL 4 pp. 129-202, and studied by J. Black,2 is the modern title given to several lists in which morphemes of the Sumerian prefix and suffix chains are equated with Akkadian words. The apparent aim of these equations is to explain different elements of Sumerian grammar to a non-Sumerian speaking audience. In antiquity, the work seems to have been divided up into 9 tablets, but only the incipit of the first tablet, ù = anâku, is preserved. These lists are attested on tablets from Assyria and Babylonia that date mostly to the first millennium BCE.3

As Leichty observes in his editio princeps of FLP unn72, several of the grammatical terms or components that it explains are attested in NBGT Tablet 9, and so the commentary seems to be at least partially concerned with this tablet.4 Other grammatical terms and components that appear in the commentary are attested in Tablets 1 and 2 of the work, but since there seems to a certain degree of repetition from Tablet to Tablet, it cannot be excluded that these terms and components were also originally in NBGT 9. Indeed, one of the greatest challenges this commentary poses to modern attempts to understand it is that NBGT is itself in a very fragmentary state of reconstruction, as well as being difficult to understand.

Enough of the colophon survives to indicate that the manuscript was produced at Nippur: the name of the tablet’s owner seems to begin with the theonym Ninurta, which is a frequent theophoric element in names of Nippureans in the Late Babylonian period; furthermore, the colophon concludes with an imprecation that invokes Enlil, the patron deity of Nippur. Although a Nippur origin for the tablet seems very likely, it is possible that it was found at Uruk.5 The transfer of other Nippurean manuscripts of text commentaries to Uruk at some point after their production is well attested.6

Although FLP unn72 was probably produced at Nippur, it cannot be dated precisely. It may well date to the Achaemenid period, since other commentaries have been excavated at Nippur in Achaemenid-period contexts. In view of the lack of evidence for FLP unn72’s precise date of production, it is worth noting that a manuscript of NBGT that was produced at Nippur (AO 17602, the only known source for Tablet 1 of NBGT) is dated to Year 17 of Darius (i.e., in either 505 or 407 BCE, depending on whether Darius I or II is meant), as we know from its surviving colophon (BAK no. 121).7

Beyond FLP unn72 indicating that NBGT was a focus of commentators in first millennium Babylonia, the importance of this commentary lies in the fact that it is one indicator of the level to which Sumerian was studied at Nippur in the Achaemenid period, and the fact that it is another piece of evidence for the importance of Sumerian as a facet of Nippurean identity. In this respect, it is highly interesting that the text features a previously unidentified quotation from a Sumerian literary text, Enlil and Sud, which is used to illustrate the commentator’s contention that the Sumerian verb ĝen can sometimes be used to introduce direct speech (l. o 13).

A final interesting feature of the commentary is the fact that it contains the paratextual remark ul ašme, “I did not hear it/ have not heard of it” (o 10 and 17), which is sometimes encountered in cuneiform text commentaries. This remark is intriguing for two reasons. First, it may indicate that the commentary was initially written down by means of dictation.8 Second, since the text contains no references to an earlier manuscript of the commentary (i.e., there are no ḫepi-glosses in the commentary or reference to a Vorlage in the colophon), the presence of the remark ul ašme may suggest that FLP unn72 is an autograph, i.e., the original manuscript of the commentary.9


Powered by Oracc
(Base textCommentaryQuotations from other texts)


FLP unn72 (unpublished unassigned ?) [commentary on NBGT]

o 1 o 1

[x (x)] x x KU : dur : MURUB₄ú : ki dili : it⸣-te--⸢šú? [...] 1

[...] ... dur means “infix”; (the phrase) ki.dili means “onc[e” ...].

o 2 2

[šu?-bi?]-⸢gin₇? : ki-ma qa-ti-šu-ma : šá-niš ki-ma qa-ti maḫ*⸣-[ri-i ...] 2

[šubig]in means “like its hand”; alternatively, (it means) “like an ear[lier] hand [...]

o 3 3

GI : maḫ-ru-ú : ḫa! mi-ḫi-il- an-ni-tu₄ : x [...] 3

gi means “first”; this sign is HA; ... [...]

o 4 4

ú-nu a-na e-na i-na : áŋ-kúš : šu-ús-ḫur-tu₄ ⸢:* [...] 4

(The prefixes) “un-, an-, en-, in-” (belong to the category of) ág.kúš, which means “non-indicative”; [...]

o 5 5

li-la-a ul áš-me : níŋ til-ri- : ga-mer-tu₄ ul x [...] 5

(The word) lilâ I did not hear it. (The phrase) níg.til.ri.iš means “completeness” ... [...]

o 6 6

šá a-na ru-qu ŠUBú : -luḫ : mu-gal-lit-tum ⸢:* [...] 6

(which means) “that which is thrown far away.” (The phrase) dù.luḫ means “trouble maker.” ... [...]

o 7 7

áš-šú ra-qa-tu₄ : GIŠ.PA.A : ki-ma qa-ti-⸢šu*-ma* [...] 7

concerning “thin ones (fp).” giš.pa.a means “Like its hand” [...]

o 8 8

níŋ dili-me : e-de-e-ti : ša u ù : šá mu-⸢x [...] 8

níĝ means “single things.” (The prefixes) ša- and ù- mean “that which ... [...]

o 9 9

e-su*-ub : lu-ta-mi-ka : i-qa-at-⸢ta [...] 9

esub means “may I conjure you!” (The word) “It comes to an end” [...]

o 10 10

si dur : e-lu-u MURUB₄ú : íb x [...] 10

(The words) si dur mean “prefixed” (and) “infixed.” (The word) ... [...]

o 11 11

ra-i : e lu-ú i-⸢x a-lak.MEŠ : a-⸢la-ku* [...] 11

(The adverb) “definitely” means ē (or) . ... (The word) “a-lak.MEŠ” (is derived from the verb) “to g[o” ...]

o 12 12

ša-át- ù át-tu-nu :? ša-at-ta u at-⸢tu⸣-[nu ...] 12

“Him/her/it” and “you (2 mp) mean (the pronouns) “him/her/it” and “yo[u (2 mp) ...]

o 13 13

ŋen : ma-a AN.TA : ŋen ga-ra-ab-dug₄ inim mu-un-⸢da⸣-[bala-e/en ŋidlam-ŋu₁₀ ḫé-me-en] 13

(The word) ĝen means “thus” (i.e., it is used to introduce direct speech) (it is) a prefix” (as in) “Come, I want to speak to you, I will have a talk with you, [please be my wife” (quotation from Enlil and Sud 24)].

o 14 14

3 GI : mi-ḫi-il- : ḫa àm sukud-⸢gin₇* áš-šú x [...] 14

“3 gi means “cuneiform sign/wedge”; ... as above ... [...]

o 15 15

ú-ka ap? li is mu ti-ma : tu⸣-šá-ma : ki*⸣-[šá-ma (...)] 15

(The adverb) uka ... (the adverb) “seemingly” means ... [...]

o 16 16

lu-man : i-na-an-na : a-nu-um-ma : a-na i-[rat? ...] 16

(The adverbial expression) “if only” means “now.” (The adverb) “herewith” means “to the b[reast” ...]

o 17 17

U₄.TA : al-la-pak ul áš-me : ú-ka x [...] 17

u.ta means allapak I did not hear. The adverb “uka” ... [...]

o 18 18

    šá-niš al-lu-ú-a 18

alternatively, (it allapak should be read as) allū’a.

o 19 19

Á-ka i-šar : a-na i-di-⸢ka? [...] 19

“Your arm is straight” means “to/for your arm” [...]

o 20 20

a-sir-tum-ma : su-uSU₈ ⸢: x [...] 20

(The word) “captive woman” means “to bring”;... [...]

o 21 21

saŋ u₄-sakar nu-èš bala gub [...] 21

At the beginning of the month, the nêšakku-priest, the man who stands ready to perform the bala duty [...]

o 22 22

ar-ḫi šá ina É.KUR i-na x [...] 22

The month that in Ekur ... [...]

o 23 23

bar-NUN : BAR.NUNta : SAG.⸢KI? [...] 23

The word bar-nun means “diagonal.” The word [...]

o 24 24

im-me la tuš-tab-bal : [...] 24

Do not consider the im-me! [...]

o 25 25

šá ina a-la-ku u a-ma-ru ⸢:?[...]

He/she/they who when going and seeing [...]

o 26 26

ZIK KI ZI SAL NIM x [...] 25


o 27 27

NIM.GÍR GE₆ : bi-ri-⸢iq? [...]

“NIM.GIR₂ GE₆” means “lightning [at night]

o 28 28

AN AN ti? SAG šá tum? [...]

r 1' r 1'

x [...]

... [...]

r 2' 2'

x [...]

... [...]

r 3' 3'

x [...]

... [...]

r 4' 4'

šá [...]

... [...]

r 5' 5'

ZA* x [...]

The exorcist ... [...]

r 6' 6'

MAŠ.MAŠ x [...]

... [...]

r 7' 7'

ina še-er-tu x [...]

At dawn [...]

r 8' 8'

BAD x x [...]

... [...]

r 9' 9'

šá ú-⸢ṣi?⸣-[...]

... [...]

r 10' 10'

šá ur₅-⸢ra [...] 26

that which ur₅.ra [...]

r 11' 11'

sa saŋ-ki-⸢a-ni x [...] 27

“Sinew, temple ... [...]

r 12' 12'


Cut the sinews ... stones [...]

r 13' 13'

ri : ku-ú x [...] 28

ri means “your(s) ... [...]

r 14' 14'

    ka-ta x [...]

You (sg.) ... [...]

r 15' 15'

    at-ta NU x [...]

You (ms) ... [...]

r 16' 16'

im--da mdnin⸣-[urta-...] 29

Long tablet of Nin[urta-...]

r 17' 17'

A šá mdDÙL-⸢x [...] 30

son of ... [...]

r 18' 18'

pa-liḫ délal dUTU? [...] 31

He who reveres Enlil, Shamash [and Marduk must not take it (i.e., the tablet) away!]

1As noted by Leichty, the equation DUR = qablû, which occurs again in l. 10, probably relates to the term for Sumerian infix, MURUB₄, which is equated with qablītu in Examenstext A 16. Leichty proposes that the equation of MURUB₄úwith DUR may be due to the rare equation DUR = abunnatu (“navel, umbilical cord”), which is attested in the lexical list Aa A = nâqu (Tablet 8, 1:76 = MSL 14 p. 491). With respect to the following entry, Leichty notes that itteššu is probably a biform of ištiššu, “once,” which is treated in NBGT 4: 34, in which case the equation seems to be based on synonymous equivalence. The poorly attested logogram KI.DILI otherwise appears on a liver model, BM 50494 (82-3-23, 1485), written by one Mušēzib-Marduk, where, however, the term is also poorly understood: see the discussion by J. Nougayrol RA 62 (1968) pp. 47-48. itt- is not attested for ištīššu, but it is attested for ištēštu (e.g. Gilg V 103, it-til-ti). The equation with ištēššu may derive from a notarikon analysis of itteššu as itti(KI) + (DILI).

2The phrase kīma qātīšu, which occurs again in l. 7, is otherwise attested in Old Babylonian letters: see the brief discussion by W. van Soldt ad AbB 13 8: 11, where he translates the phrase as “in accordance with his share.” The phrase may be related to the term kīma šumīšu, which means “ditto” (see the discussion by M. Rutz JCS 58 (2006), pp. 87-88. However, according to U. Gabbay CHANE 82 (2016), p. 119 n. 176, the phrase kīma qātīšu is “not necessarily a hermeneutical term.”

3As Leichty notes, the equation GI = mahrû is not attested elsewhere, but GI = ma-ḫa-ru is found in the lexical list Lanu B iii 6 (for an edition, see dcclt/P385927). Leichty also notes that, since mahrû is equated in lexical lists both with AN.TA.GI and with IGI (for specific references see CAD M/I 108b), GI in both cases may be a shortened form of IGI. In the present line, if the reading of the first sign of l. 2 is correct, then GI would probably explain GIM. With respect to the second entry in the line, Leichty observes: “If this is a correct translation, the scribe must be commenting on a reading of a sign. Perhaps the sign was written in an archaic form or was damaged.” Leichty also notes that ḪA occurs in NBGT 9: 32 and 59, but also in 1: 407 and 2: 44. The sign ḪA is written in its Old Babylonian and Assyrian form, rather than in its Late Babylonian counterpart.

4The interpretation of the first part of this line follows Leichty, who suggests that it contains writings of the Sumerian verbal prefixes un-, an-, en- and in-, which are treated in NBGT 2: 86-89, 99-102 and passim in NBGT 1. Alternatively, as suggested by C. Wilcke to Leichty (private communication), the first part of the line might refer to NBGT 9: 43-44, which treats the Akkadian prepositions ina and ana. However, if this were the case, the commentary in this line would be reversing the order in which the prepositions appear in the base text (first ina then ana), and the meanings of the other halves of the two equations would be unclear; it therefore seems a less likely possibility. As noted by Leichty, the word ÁG.KÚŠ occurs in NBGT 9: 39. The Akkadian word also occurs passim in NBGT 1 and 2, as well as in Examenstext A 16, where it is equated with NIGIN. See J. Black Sumerian Grammar in Babylonian Theory (1984), pp. 92-95 for the proposal that it means “non-indicative verbal form.”

5As noted by Leichty, the word li-la-a also occurs in NBGT 9: 50 (in the MSL 4 edition, it has been misread as li ú a). Leichty, albeit reluctantly, understands this word as the word for lilû-demon, it seems more likely that it is an otherwise unattested grammatical term. The second entry of the line comments on a phrase attested in NBGT 9: 55 (MSL 4: 172). The proposal by Frahm GMTR 5 (2011), pp. 254-255 n. 1207, that NÍG.TIL.RI., should be read as Akkadian šá be-ri-, “that which is distant,” is not followed here. As noted by N. Veldhuis (private communication), níĝ-til-la also appears in ‘In Praise of the Scribal Art’ (see, a text that contains abundant grammatical terminology.

6The proposal by Frahm GMTR 5 (2011), pp. 254-255 n. 1207 that šá ... ŠUBú is a hermeneutic term (meaning “that which pertains to ...”) is viewed skeptically by U. Gabbay CHANE 82 (2016) p. 213 n. 51, on the grounds that “all the elements of the phrase šá a-na ru-qu ŠUBú correspond to the elements NÍG.TIL.RI. that they interpret.” Accordingly, the phrase šá ... ŠUBú is not translated as a hermeneutic term here. However, until the meaning of the entry in l. 5 to which these words refer is clarified, certainty on how to interpret both this phrase and the signs NÍG.TIL.RI. in the preceding line is impossible. With respect to the second entry in this line, the equation .LUḪ = mugallitu is otherwise attested in NBGT 9: 57 (in the MSL 4 edition, it has been misread as -ú), as well as in the lexical list Antagal B 209 (MSL 17 p. 193). As with li-la-a in the preceding line, the phrase .LUḪ seems more likely to refer to a grammatical term than to a demon.

7The word ra-qa-tum occurs in NBGT 9 345, but its specific meaning in a grammatical context is unclear (compare the use of marû, lit. “fat, slow”, to describe Sumerian verbal forms). The translation offered here of the hermeneutic term aššu in this passage is based on the suggestion of Gabbay CHANE 82 (2016) p. 151, who argues that aššu can sometimes function as “a linguistic indicator in a lexical contextualization.”

8As pointed out by Leichty, the second entry refers to the prefixes ŠA- and Ù- in Sumerian.

9As Leichty suggests, the first entry is connected with Malku = šarru 3 118 (e-su*-ub = la te-me-ek-ki (p. 367 in the 2010 edition of I. Hrůša). The reading e-su-ub in this commentary indicates that those mss of Malku = šarru 3 broken in this line should be read es-su-ub instead of e-zu-ub. Leichty suggests: “Either the scribe was badly confused or lu-ta-mi-ka is some sort of strange dialectal Akkadian.” Leichty points out that NBGT 1 435, where ŠUB.BA is equated with e-zib₄ may also be relevant (if zib₄ is the correct reading of KA: on which issue see the cautious words of R. Borger MesZL P. 255). Note that the following entry in NBGT 1 is quoted verbatim in l. 13 of the commentary.

10The equation SI = elû is otherwise unattested but may be motivated by the learned writing dSI for the god Enlil, on which see Frahm GMTR 5 (2011), p. 140 n. 692. For the equation DUR = MURUB₄ú see ad l. 1.

11As noted by Leichty, the optative Akkadian particle raʾi is attested in NBGT 9 119. The reading of the signs differs from Leichty’s, who considered e-lu to mean “prefix.” The equation of the adverb with the particles ay (sometimes written “ē”) and makes better sense.

12The commentator explains two words as unusually written syllabic spellings of two common Akkadian words. Leichty regards this explanation as unlikely.

13The first three words correspond exactly with the entry NBGT 1 436. As noted by E. Jiménez (private communication), the passage cited is l.24 of the Sumerian literary work Enlil and Sud, which is attested in first millennium manuscripts (M. Civil, “Enlil and Ninlil: The Marriage of Sud,” JAOS 103 [1983], pp. 43-66. See E. Jiménez The Babylonian Disputation Poems (CHANE 87, 2017), p. 91 for further discussion of the passage in question.

14It is unclear to what “3 GI” refers in this context. The meaning of the signs following mi-ḫi-il-tim is unclear too. As noted by E. Frahm (private communication), the expression SUKUD.GIM is also attested in the Esangil-kīn-apli colophon (Finkel, Fs Sachs [1998], p. 147 A 37 and 43). Note that ḪA.ḪU and ÀM are equated with in the grammatical text HS 1610 (dcclt/P229745). In the same text, a sign is followed by the word “sign/wedge” tum GU.SUM (= mihiltu) in r iii′ 33′.

15The adverb uka, which is also cited in l. 17 of this commentary, is otherwise attested only in three Old Babylonian letters from Mari (CAD U/W 54, it is perhaps related to the Neo-Assyrian particle muk). As noted by Leichty, the Akkadian adverb tušāma, “seemingly,” appears in NBGT 3, 4 and 9 (l. 268 and perhaps Appendix 2:2 C). It also appears in the so-called Reisner Vocabulary, where it is one of many Akkadian adverbs equated with the Sumerian adverb i-gi₄-in-zu. Unfortunately, neither context clarifies how to read the preceding signs in the present line. However, since tušāma is equated with kīam and kīšāma in Malku = šarru 3 109 and 8 114, the restoration of kīšāma seems likely.

16As noted by Leichty, both lumān and anumma appear several times in NBGT, but in view of the preponderance of correspondences with entries from NBGT 9, the context NBGT 9 276-277 seems most likely to be the source of the present line. The tentative restoration of the phrase following anumma is based on the entry NBGT 9 279, as Leichty suggested it might be. Note, however, that in the only preserved manuscript of NBGT 9 279, the phrase is instead written an-na i-rat.

17As Leichty noted, the equation cited occurs in NBGT 9 280 as well as elsewhere in NBGT. The alleged attestation of the adverb allapak in Malku = šarru mentioned in CAD A/1 356a is incorrect: see Hrůša AOAT 50 (2010) p. 368.

18As noted by Leichty, the commentary’s explanation in this line is based on a reading of the final sign in al-la-pak as ḪU not pak.

19As Leichty observed, the association of idu and išaru in this line is reminiscent of the Akkadian part of Nabnītu Tablet R 288 (i-du i-ša-ru). The various components of the line also appear in the grammatical text in the Hilprecht Sammlung edited by Veldhuis at http:/].

20The word asirtu is very rare. It is otherwise attested in SB Ālu (and is commented on in a late commentary on this text, CCP 3.5.49). The signs could also be Sumerian.

21The context of this line is obscure. As noted by Leichty, the Sumerian phrase BALA.GUB occurs in MSL 12 40: 212b and 131: 75, where it is equated with ša manzalti, “office-holder.”

22The context of this line is also obscure.

23Neither BAR.NUN, nor its Akkadian reading ṣiliptu, nor SAG.KI is attested in NBGT, but, interestingly, BAR.NUN is attested in LB mathematical texts with the meaning “diagonal”: see Friberg RlA 7 (1987/1990) p. 553b. Note, however, that SAG.KI is attested in r 9 of the present tablet.

24The translation of this line is tentative, due to the difficulty of translating IM.ME, which may a quotation of Sumerian.

25This line is currently unable to be understood. In view of the occurrence of NIM.GIR₂ in the following line, the NIM sign may be the beginning of NIM.GIR₂. Unfortunately, the traces of the following sign are inconclusive.

26As suggested by Leichty, these signs seem to refer to the lexical list UR₅.RA = hubullu.

27Leichty suggests that both this and the following line refer to diagnostic omina. But compare perhaps Muššu’u 2 38: sa saŋ-ki-a-ni sa ši-[...] (reference courtesy of E. Jiménez).

28Leichty suggests that this and the following two lines contain the catchline (i.e., the first lines of the next tablet). He also notes that, though this is the only occurrence of the specific equation RI = , it is an unsurprising equation because RI is attested equated with other forms of the 2nd person pronoun. As noted by N. Veldhuis (private communication), the equation ri = ku-ú and the next two lines appears in HS 1610 o ii 24′-26′.

29The restoration Nin[urta-...] follows Gabbay and Jiménez: Ninurta is a common theophoric element in names of Nippureans. As Gabbay and Jiménez note, the person in question is not Ninurta-ušabši son of Enlil-kēšir, the owner of at least one other commentary (on Šumma ālu = CCP no. 3.5.45.A), because the traces of the father’s name in the following line do not fit the reading Enlil-kēšir. See U. Gabbay and E. Jiménez, “Cultural Imports and Local Products in the Commentaries from Uruk: The Case of the Gimil-Sîn Family,” forthcoming, no. 3.

30As pointed out by Gabbay and Jiménez, the personal name could be interpreted either as mAN.DÙL = andullu or ṣulūlu (MZL, top of p. 250), or as a cryptographic writing of a theonym (Šamaš or Enlil). Note that the relatively rare sign DÙL is attested in the l. 4 of the commentary.

31On dŠID (ŠID read here as élal) as a cryptographic writing for Enlil, see Frahm GMTR 5 (2011) p. 198 n. 936 and p. 298 n. 1424. The restoration of this line follows Frahm (ibid. p. 299), who bases it on the parallel in the colophon of the SA-GIG commentary CCP 4.1.21 (also housed in the Free Library of Philadelphia), which was produced in Nippur by a member of the Rīš-Gula branch of the Gimil-Sîn family. For discussion of the commentaries produced by members of the Gimil-Sîn family, see now U. Gabbay and E. Jiménez, “Cultural Imports and Local Products in the Commentaries from Uruk: The Case of the Gimil-Sîn Family,” forthcoming.

Photos by Mary Frazer

© Free Library of Philadelphia