PHYSIOGNOMY

Cuneiform signs are also mentioned in treatises on physiognomy, the intellectual discipline that explains how to infer the qualities and future prospects of human beings from physical features of their body, especially the face.

A conspicuous aspect of the passage, briefly discussed by Bilbija but not fully investi­gated, is that quite a few of the entries mention not just one but two or even three signs, all of them apparently holding the same ominous significance. In the following overview of the passage, which is based on Bock’s edition (2000a: 92-97), I discuss both the potential links between protases and apodoses and the connections between these variant signs. To facili­tate referencing, the numbering of the examples continues that of the extispicy omens in the preceding section. The first entry of the text provides the protasis in full, while the later ones present it in an abbreviated version.

56) [DIS a]lam-dim-me-e SAG.KI NA ina SAG.KI NA AN SUB NA BI HUL (line 76) [Concerning] the appearance of the forehead of a man: (If the grapheme) AN appears on the forehead of a man, this man will experience misfortune.^[69]

No etymographical link; the prediction is negative.

57) [DIS] SID NA BI SU LUGAL KUR-dd (line 77)

[If] (there is the grapheme) SID, the hand of the king will reach this man.

Bilbija (2008: 22-23) suggests the apodosis is based on the fact that sarru “king” is semantically related to the word issakku “city ruler,” one of the readings of SID. While not impossible, this explanation remains doubtful since the two words are otherwise clearly distinguished; SID does not occur among the numerous logograms listed in lexical texts as representing sarru (see CAD S/2, 76-78). The prediction is negative.

58) [DIS] BA NA BI HUL IGI : GE₆ IGI IGI-mar (line 78)

[If] (there is the grapheme) BA, this man will face misfortune; var.: he will face rage (sulum pani).

No etymographical link; the prediction is negative.

59) [DIS] ZI DUMU.MES E AD-su-nu i-za-aq-qa-pu (line 79)

[If] (there is the grapheme) ZI, the sons will raise the house of their father.

As seen by Bilbija (2008: 23), the apodosis could be based on the fact that ZI corre­sponds to Akkadian tebu “to arise,” which is semantically related to zaqapu “to raise.” The prediction is positive.

60) [DIS] MU : BI DUMU.MES E AD-su-nu ZAH.MES (line 80)

[If] (there is the grapheme) MU (or) BI, the sons will ruin the house of their father.

The signs MU and BI, semantically unrelated, look rather similar, especially in Old Babylonian cursive script. No etymographical link; the prediction is negative.

61) [DIS B]U^? MU-jm NU GAL-si (line 81)

[If] (there is the grapheme) BU, he will not maintain his name.

Reading of the sign uncertain, no obvious etymographical link; the prediction is nega­tive.

62) [DIS] 1x1 DUMU.MUNUS.MES E AD-si-na i-za-aq-qa-pa (line 82)

[If] (there is the grapheme) x, the daughters will raise the house of their fa­ther.

Compare no. 59; the prediction is positive.

63) DIS [x] DUMU.MUNUS.MES E AD-si-na i-kab-ba-sa : ZAH.ME (line 83)

[If] (there is the grapheme) [x], the daughters will tread down, var.: they will ruin the house of their father.

Compare no. 60; the prediction is negative.

64) DIS 1GIS : US± EGIR E LU GAL-si (line 84)

If (there is the grapheme) GIS (or) US, the legacy of the house of the man will remain.

In the Old Babylonian cursive (but not in later Babylonian or Assyrian script), the — semantically unrelated — signs GIS and US look quite similar. No obvious etymo­graphical link;⁶⁷ the prediction is positive.

65) DIS TAB : PA EGIR E LU ZAH (line 85)

If (there is the grapheme) TAB (or) PA, the legacy of the house of the man will perish.

TAB and PA, semantically unrelated, have similar shapes throughout the history of cuneiform writing. Bilbija (2008: 24) argues that “the apodosis... can be linked to the sign TAB if it is read as hamatum ‘to burn (up),' and the head carrying the sign is interpreted as the man's house,” but this explanation seems rather far-fetched to me. The prediction is negative.

66) DIS EN RI HU LU BI be-en-nu 1x± [...] (line 86)

If (there is the grapheme) EN, RI, (or) HU, this man [(...)] epilepsy [(.)].⁶⁸

The three graphemes, semantically unrelated, have similar shapes throughout the his­tory of cuneiform writing. No obvious etymographical link;⁶⁹ the prediction is prob­ably negative.

67) DIS UR : IB SU.BI.AS.A[M] (line 87)

If (there is the grapheme) UR (or) IB, the same.

UR and IB, semantically unrelated, have similar shapes throughout the history of cu­neiform writing. Their shapes also resemble to some extent those of the graphemes from the preceding entry, which has the same apodosis. No etymographical link; the prediction is probably negative.

68) DIS GAN : UD LU BI sa ^d30 i-ma-[at^?] (line 88)

If (there is the grapheme) GAN (or) UD, this man will die (...) of Sin.⁷⁰

In the Old Babylonian cursive (but not in later Babylonian or Assyrian script), the — semantically unrelated — signs GAN and UD look quite similar. No obvious etymo­graphical link;⁷¹ the prediction is negative.

69) DIS MA : LU : KU NA BI US hi-bil-ti : [... imat] (line 89)

If (there is the grapheme) MA, LU, (or) KU, this man [will die] violently, var. [.].

In the Old Babylonian cursive, MA and KU can look very much alike, and LU has a similar shape; in other periods of cuneiform writing, the similarities are less pro­nounced. No etymographical link; the prediction is negative.

70) DIS KI US SA HUL UG₇ : US hi-ti [(...) imat] (line 90)

If (there is the grapheme) KI, he will die of grief, var.: [he will die] in a sinful way [(.)].

No etymographical link; the prediction is negative.

71) DIS KA NU mit-gur-ti ina E NA GA[L-si] (line 91)

If (there is the grapheme) KA, there will be discord in the house of the man.

According to Bilbija (2008: 23-24), this protasis-apodosis string may be based on the widely attested readings of KA as gu = sasu “to shout” and rigmu “voice, noise,” possibly indicative of loud altercations. This interpretation is ingenious, but not com­pletely compelling. One could also argue that a reading du₁₁(KA)-du₁₁(KA) = dababu “to litigate” is behind the entry. Perhaps, there is, in fact, no etymographical link at all. The prediction is negative.

72) DIS AB : UM mus-ke-nu i-sar-[ru] (line 92)

If (there is the grapheme) AB (or) UM, the poor man will become rich.

In the Old Babylonian cursive, but usually not in other periods, AB and UM can have the same shape. No etymographical link; the prediction is negative.

73) DIS AD : IL bi-ir-ta u-sa-kal sa bu-tuq-qe-1 e± [...] (line 93)

If (there is the grapheme) AD (or) IL, he will provision the fortress, of the losses [.].

The signs AD and IL look quite different in all periods of Mesopotamian writing, but in Old Babylonian, there is a certain similarity between them (see Kraus 1935: 22). No etymographical link. The prediction seems to be positive, but its meaning is not completely certain.

74) DIS BI : GA SUB IBILA NA DAM NA 1x1 [...] (line 94)

If (there is the grapheme) BI (or) GA, the man’s heir will fall, the man’s wife [.].

In Babylonian script, but not in Assyrian, BI and GA look rather similar. No etymo- graphical link; the prediction is negative.

75) DIS UL DINGIR KI LU BI SILIM [.] (line 95)

If (there is the grapheme) UL, the god will make peace with this man [...].

No etymographical link;^[72] the prediction is negative.

76) DIS NA DUMU.MES-su UG₇.UG₇ [...] (line 96)

If (there is the grapheme) NA, his sons will die [.].

No obvious etymographical link;^[73] the prediction is negative.

77) DIS TAB : UB NU U.TU U.TU NU SI.SA SI.[SA] (line 97)

If (there is the grapheme) TAB (or) UB, an infertile woman will have a child, a woman having difficulties in childbirth will easily give birth.

In Babylonian script, but not in Assyrian, TAB and UB have similar shapes. No etymo- graphical link; the prediction is negative.

78) DIS URU : GUR US [imat] (line 98)

If (there is the grapheme) URU (or) GUR, [he will die] through a crossbeam.

The shapes of URU and GUR are similar throughout the history of Babylonian and As­syrian cuneiform writing. As for a possible link between the protasis and the apodosis, one could point to the readings Ri of URU and RI of DAL, but this remains speculation. The prediction is negative.

79) DIS N1 : IR BA.UG₇ KIMIN MUNUS ina hi-ti LU Ixl [...] (line 99)

If (there is the grapheme) N1 (or) IR, he will die, ditto, a woman, through a crime [(...)] the man [(...)].

The shapes of N1 and IR are similar throughout the history of Babylonian and Assyrian cuneiform writing. No obvious etymographical link;^{74 75 76} the prediction is negative.

80) DIS IGI DU₁₀-ub lib-[bi] (line 100)

If (there is the grapheme) IGI, there will be happiness.

For the same apodosis, see examples no. 17 (grapheme: AN) and 86 (graphemes: SE and PI, both similar to IGI); compare also no. 122. No obvious etymographical link. The prediction is positive.

81) DIS KI E LU IZI [ikkal] (line 101)

If (there is the grapheme) KI, a fire [will devour] the house of the man.

No etymographical link; the prediction is negative.

82) DIS LA : SU dan-na-tu⁷ LU BI i-ra-am-[mi] (line 102)

If (there is the grapheme) LA (or) SU, this man will dwell in a fortress.

The shapes of LA and SU are similar throughout most of the history of Babylonian and Assyrian cuneiform writing. No obvious etymographical link. The prediction is appar­ently negative.

83) DIS AL US KI.HUL [imat] (line 103)

If (there is the grapheme) AL, [he will die] through mourning.

No etymographical link; the prediction is negative.

84) DIS SAG US su-ub-ti U[G₇] (line 104)

If (there is the grapheme) SAG, he will die in (his) dwelling.¹⁶

No etymographical link; the exact meaning of the prediction is unclear.

85) DIS U US a-si-i [imat] (line 105)

If (there is the grapheme) U, [he will die] through the asu-illness.

Bilbija argues the entry is based on the fact that U is read sammu “plant” in Akkadian, and that the plant used to cure the asu-illness was called sammi asi; this explanation, however, seems rather far-fetched. The prediction is negative.

86) DIS SE : PI DU₁₀-ub lib-[bi] (line 106)

If (there is the grapheme) SE (or) PI, there will be happiness.

There is a certain similarity between the two graphemes from the Old Babylonian period onward. For the same apodosis, see examples no. 17 (grapheme AN) and 80 (grapheme IGI, similar to SE and PI), cf. also no. 122. No obvious etymographical link. The prediction is positive.

87) DIS SA ba-la-at SA [amzli(?)^[77]] (line 107)

If (there is the grapheme) SA, a healthy life (lit., life of the heart) [(is in store) for the man].

There is an obvious link between protasis and apodosis, as pointed out by Bilbija (2008: 22, n. 12): both include the sign SA. The prediction is positive.

88) DIS DA SE ina la sa-at-ti S[U^?...] (line 108)

If (there is the grapheme) DA (or) SE, [he will]... [...] in the wrong year.

The two graphemes do not resemble each other. Unlike other variant signs, they are not divided by separating cola, and one wonders if the ancient scribe (or one of his predecessors) may have copied the beginning of the line incorrectly. Alternatively, one could suppose that SE introduces the apodosis, and translate: “If (there is the grapheme) DA, the barley [will...]... outside the season [...]” (see CAD S/2, 206a). No etymographical link; the prediction is probably negative.

89) [DIS] DAR LU ina hi-ti [...] (line 109)

[If] (there is the grapheme) DAR, the man [will...] through a crime.

Too broken for an analysis. The prediction is probably negative.

90) [DIS A]L^? ra-bu E LU i-b[a^?...] (line 110)

[If] (there is the grapheme) AL, a magnate will [...] the house of the man [.].

Too broken for an analysis.

91) [DIS] 1x1 LU BI ^giSGI ina¹ 1x1 [...] (line 111)

[If] (there is the grapheme) x, this man [...] a reed [...].

Too broken for an analysis.

92) DIS [x] KIMIN ne ne^[78] ina SA 1x± [...] (line 112)

If (there is the grapheme) [x], ditto,. in the heart. [.].

Too broken for an analysis.

93) DIS MI US hi-it-nu-[qi imat] (line 113)

If (there is the grapheme) MI, [he will die] through strangulation.

Bilbija (2008: 23) argues that MI, read GE₆ = salamu “to become dark,” could “de­scribe the effects of strangulation,” but this is again a rather speculative idea. The prediction is negative.

94) DIS GAN/KAM TES LU [...] (line 114)

If (there is the grapheme) GAN/KAM, the potency of the man [will...].

Too broken for an analysis.

95) DIS U sal-tu ZI.GA [...] (line 115)

If (there is the grapheme) U, there will be quarrel, loss [...].

No obvious etymographical link. The prediction is negative.

96) DIS HAR : AH US si-il-la-ti [U]G₇ (line 116)

If (there is the grapheme) HAR (or) AH, he will die a death (caused by) blas­phemy.

The graphemes resemble each other in Babylonian, but not in Middle and Neo-Assyri- an script. No etymographical link; the prediction is negative.

97) DIS AZ : LUGAL US sar-ri US bu-ri UG7 (line 117)

If (there is the grapheme) AZ (or) LUGAL, he will die a death (caused by) the king (or) a death (caused by) a well/a calf/hunger.

The two graphemes resemble each other most closely in the Old Babylonian cursive. The reference to the king in the apodosis is clearly motivated by the occurrence of LUGAL in the protasis. The prediction is negative.

98) DIS LI : TU US ID US ha-am-ta UG7 (line 118)

If (there is the grapheme) LI (or) TU, he will die a death (caused by) the river (or) a speedy death.

The two graphemes resemble each other throughout much of the history of Babylo­nian and Assyrian writing, but most closely in the Old Babylonian cursive. No etymo- graphical link; the prediction is negative.

99) DIS ZA US su(A, B: su)-um-me(A, B: me)-e UG7 (line 119)

If (there is the grapheme) ZA, he will die from thirst.

No etymographical link; the prediction is negative.

100) DIS BAD GIG.MES LU DAB.MES (line 120)

If (there is the grapheme) BAD, diseases/wounds will seize the man.

Compare example no. 4. BAD, read us, means matu “to die” in Akkadian, but one wonders if this really explains the (negative) prediction.

101) DIS U : LU LU BI ina E TUS-ab (line 121)

If (there is the grapheme) U (or) LU, this man will live in a house.

The two graphemes do not resemble each other. The reference to the man in the (posi­tive) apodosis could be motivated by the occurrence of LU in the protasis (see Bilbija 2008: 21), but, obviously, most of the predictions deal with a “man.”

102) DIS TAR : GAM i kur id ZI.GA LU NA UG₇ (line 122)

If (there is the grapheme) TAR (or) GAM,... loss for the man, the man will die.

The two graphemes resemble each other in Babylonian writing, but not so much in Assyrian. Bilbija (2008: 21) argues that the last two apodoses are based on readings of GAM as pilsu “breach” and matu “to die.” TAR/KUD, with its reading parasu “to cut off,” is semantically not too far off, but this may be simply by chance. The prediction is negative.

103) DIS NU i kur id ZI.GA LU E (line 123)

If (there is the grapheme) NU,... the man will experience loss.

Especially in Old Babylonian, NU looks quite similar to TAR and GAM, the signs featured in the preceding entry, which has a similar apodosis. NU means la “not,” and this negative connotation could have inspired the prediction, but if it really did remains doubtful.

104) DIS UD u-la-lu-tam(B, A: lu^? :^?1 i-la-lu-tam) LU GIN (line 124)

If (there is the grapheme) UD, the man will become helpless.

No etymographical link; the prediction is negative.

105) DIS NINDA LU NINDA i-be-ru (line 125)

If (there is the grapheme) NINDA, the man will hunger for bread.

The reference to the bread (NINDA, akalu) in the apodosis is clearly motivated by the occurrence of NINDA in the protasis. The prediction is negative.

106) DIS GIR : UG (A, B: AZ) : BAN (A, B: GIM) US re-i-ib-ti LU UG7 (line 126) If (there is the grapheme) GIR, UG (A) / AZ (B), or BAN (A) / GIM (B), the man will die from the re °ibtu-disease.

GIR, UG, and AZ look similar in Old Babylonian, but not so much in later phases of cuneiform writing. BAN and GIM are similar to each other throughout most of the his­tory of Babylonian and Assyrian cuneiform, and in Old Babylonian, the signs also look to some extent similar to the other three characters. No etymographical link; the prediction is negative.

107) DIS TI : IM ba-la-at SA(B, A adds -bi) NA(A, B: LU) (line 127)

If (there is the grapheme) TI (or) IM, a healthy life (lit., life of the heart) (is in store) for the man.

Compare no. 87. There is a certain, even though somewhat superficial, similarity between the shapes of TI and IM. TI is often rendered as balatu “life” in Akkadian, which explains the reference to balatu in the apodosis (see Bilbija 2008: 22). The prediction is positive.

108)DIS ES ZI.GA SU NA (line 128)

If (there is the grapheme) ES, there will be losses for the hand of the man.

ES consists of three “Winkelhakens” (U), and it is interesting that in example no. 95, the grapheme U indicates losses (ZI.GA) as well. No etymographical link; the predic­tion is negative.

109)DIS MES SU DINGIR LU DAB-bat : KUR-ad (line 129)

If (there is the grapheme) MES, the hand of the god will seize, var.: reach the

man.

No etymographical link; the prediction is negative.

110)DIS A na-mar E LU ana sa-a-tim (line 130)

If (there is the grapheme) A, the man’s house will be bright forever.

No etymographical link; the prediction is positive.

111)DIS MAN bu-tuq-ti(A, B: tum) E LU GAR-an (line 131)

If (there is the grapheme) MAN, a breach¹⁹ will be made in the man’s house.

No etymographical link; the prediction is negative.

112)DIS NA(A, B: BA) : MA li-i°-bu E(B, A om.) LU i-la-ib (line 132)

If (there is the grapheme) NA (A) / BA (B) (or) MA, the (household of) the man will suffer from the li °bu-disease.

While BA and MA look similar in Babylonian script, NA does not. However, NA does look similar to BA in Assyrian script, suggesting that the reading NA in ms. A goes back to a mistake made by an Assyrian scribe copying an Assyrian manuscript. No etymographical link; the prediction is negative.

113)DIS he-pi DAM.MES LU(A, B om.?) UG₇.MES (line 133)

If (there is the grapheme) — broken —, the wives (of the man) will die.

The prediction is negative.

Another passage referring to cuneiform characters observed on the body of a man occurs in the Assur text KAR 395, edited by Bock (2000a: 290-95).^{[79] [80] This is the second tablet of a series, but not the canonical Alamdimmu series as we know it from Nineveh. As in the case of Alamdimmu III, the section on the signs occurs toward the end of the tablet. Its beginning is lost, and it is not completely clear which body part it describes. Most probably, though,}

ECKART FRAHM

the section deals with cuneiform characters on the cheek. Kraus (1935: 52-53), pointing out that the catchline of KAR 395 refers to the usukku, or upper cheek, suggested, quite convinc­ingly, that this word may also occur in rev. iv 2', which is followed by the section on the graphemes.⁸¹ The passage includes the following omens, all referring to one grapheme only (the line numbering follows Bock 2000a):

114-16) DIS NU [...] / DIS KUR Fx± [...] / DIS NE LÜ B[I...] (lines 69-71)

If (there is the grapheme) NU, [...]. / If (there is the grapheme) KUR,... [...].

/ If there is the grapheme NE, this man [...].

Too broken for analysis.

117)DIS IGI IGI^II.BI [...] (line 72)

If (there is the grapheme) IGI, his eyes [...].

The occurrence of IGI (= inu “eye”) in the protasis is mirrored by the reference to eyes in the apodosis.

118)DIS GAG ina-kud KUR US_n SU Ixl [...] (line 73)

If (there is the grapheme) GAG, he will become anxious, (there will be) an attack through sorcery, the hand [...].

No etymographical link; the prediction is negative.

119)DIS NIGIN me-si-ru DAB-su (line 74)

If (there is the grapheme) NIGIN, confinement/hardship will befall him.⁸²

The reference in the apodosis to mesiru “confinement” seems to be based on the well- established reading of NIGIN as eseru “to confine” (but cf. the discussion below). The prediction is negative.

120)DIS SAG AD₆^! KUR-su (line 75)

If (there is the grapheme) SAG, a corpse will reach him.

No etymological link; the prediction is negative.

121)DIS LAL al-ma-nu-tam GIN-ak (line 76)

If (there is the grapheme) LAL, he will become a widower.

The sign LAL is associated with notions of poverty and dearth; it can be read matu “to become little” and qalälu “to become weak.” These connotations might have inspired the apodosis, but this is not certain.

122)DIS UD SA.BI DU₁₀.GA (line 77)

If (there is the grapheme) UD, he will be happy.

Compare example nos. 80 and 86, where the analogous apodosis tub libbi “happiness” is preceded by references to the signs IGI, SE, and PI, all similar to UD. No etymo- graphical link; the prediction is positive.

123)DIS BAR ina la-li-su BA.UG₇ (line 78)

If (there is the grapheme) BAR, he will die in his prime.

No etymographical link (but the cross-like shape of the sign may have played a role); the prediction is negative.

124)DIS PA SU DINGIR KUR-.su (line 79)

If (there is the grapheme) PA, the hand of a god will reach him.

No obvious etymographical link; the prediction is negative.

125)DIS RA ϋ§ sa-ga-as-ti BA.UG7 (line 80)

If (there is the grapheme) RA, he will die through murder.

One could speculate that a reading of RA as mahasu “to beat, smite” influenced the negative apodosis, but this remains uncertain.

126)DIS BA^[83] U4.MES-.su TIL.MES (line 81)

If (there is the grapheme) BA, his days will come to an end.

No obvious etymographical link; the prediction is negative.

127)DIS ZU ra-ga-am DINGIR ana NA (line 82)

If (there is the grapheme) ZU, there will be divine prosecution against the man.

No etymographical link; the prediction is negative.

128)DIS GAN^? SU LUGAL KUR-su (line 83)

If (there is the grapheme) GAN, the hand of the king will reach him.

No etymographical link; the prediction is negative.

The two texts presented here mention the following graphemes (in alphabetical order): AB (72), AD (73), AH (96), AL (83, 90[?]), AN (56), AZ (97, 106), BA (58, 112, 126), BAD (100), BAR (123) BI (60, 74), BU (61?), DA (88), DAR (89), EN (66), ES (108), GA (74), GAG (118), GAN (128), GAN (68), GIM (106), GIR (106), GIS (64), GUR (78), HAR (96), HU (66), IB (67), IGI (80, 117), IL (73), IM (107), IR (79), KA (71), KI (70, 81), KU (69), KUR (115), LA (82), LAL (121), LI (98), LU (69), Lύ (101), LUGAL (97), MA (69, 112), MAN (Ill), MES (109), MI (93), MU (60), NA (76, 112 [scribal mistake]), NE (116), N1 (79), NIGIN (119), NU (103, 114), PA (65, 124), PI (86), RA (125), RI (66), SAG (84, 120), SA (87), SE (86, 88[?]), SID (57), SU (82), TAB (65, 77), TAR (102), TI (107), TU (98), U (95), ύ (85), ϋ (101), UB (77), UD (68, 104, 122), UG (106), UL (75), UM (72), UR (67), URU (78), US (64), ZA (99), ZI (59), ZU (127). Forty-nine of the apodoses are inauspicious, fourteen are auspicious, and nine remain unclear.

The preceding overview clarifies a number of issues. First, it is obvious that the signs ana­lyzed in the physiognomic texts differ substantially from those of the extispicy treatises. In the latter, the number of different graphemes observed on the exta is fairly small, with the same characters reoccurring again and again, apparently because of their similarity with certain le­sions and grooves typically found on the liver and other organs. In the case of Alamdimmu III, the author/compiler of the text was interested in the analysis of a much larger sample of signs. His goal was to point out with regard to each of them what its specific meaning was when it occurred, most probably in the form of wrinkles, on a man’s forehead. Only a few signs are mentioned two or three times.

What governs the sequence of the signs investigated in Alamdimmu III remains unclear — no lexical list seems to have provided the model. In a few instances, the entries seem to be organized according to acrophonic principles reminiscent of the Old Babylonian tu : ta : ti lists,⁸⁴ but these principles are not applied with any consequence. The same holds true for the rare cases in which sign sequences mirror those of Proto-Ea.⁸⁵ There is no question, however, that the bulk of the text’s section on graphemes goes back to Old Babylonian times. As out­lined in my notes (and already recognized in Kraus 1935: 22, but not taken into account by Bilbija 2008), the many variant signs mentioned in the omens resemble one another, almost without exception, in the Old Babylonian cursive script of the time of Hammurapi and his successors, but not necessarily in other periods of Babylonian writing, and even less so in the Neo-Assyrian script used in the two Nineveh manuscripts that preserve the passage.^{86 *} This in­sight, unfortunately, does not settle the question of when the variant signs were actually added. Theoretically, they could already have been part of the original Old Babylonian version of the passage, with a scribe assuming that similarly shaped graphemes observed on the forehead all had the same import. It is also possible, however, that a later redactor of the text, perhaps even the famous scholar Esagil-kin-apli, who according to Mesopotamian tradition edited the canonical series Alamdimmu in the eleventh century B.C. (see Finkel 1988), provided the vari­ants. Working with older manuscripts, the redactor in question may no longer have been able to establish the exact nature of the decontextualized graphemes, and this uncertainty may have prompted him to give every possible reading of them in his new compilation. The truth could also lie somewhere in between, with some variants being old and some of a later date.⁸⁷

Whatever the exact editorial history of Alamdimmu III, the fact is that the variant signs mentioned in many of its entries are grouped together because of their shape, and not be­cause their logographic or phonetic readings share some tertium comparationis. This strongly mitigates against the idea that “etymography” is to be regarded as the main rationale behind the protasis-apodosis strings of the various entries. To be sure, there are a few cases where etymography does seem to play a role. In example nos. 87, 97, 101, 105, 107, and 117, the grapheme of the protasis is either repeated or rendered syllabically in the apodosis, and in example nos. 59, 102(?), 119, 121(?), and 125(?), somewhat more subtle links seem to ex­ist.⁸⁸ But these are only eleven out of seventy-three entries (some, admittedly, badly broken), representing exceptions rather than the rule.

Unfortunately, what is the rule, in the other cases, remains difficult to establish. Apparently the sign’s shape, in the Old Babylonian cursive, played a major role; yet why, for instance, the shape of the KI sign, in no. 81, points to a future conflagration remains obscure to the present writer.

Here and there, however, some vague patterns seem to emerge. Nos. 59 and 60, for ex­ample, provide very similar apodoses, one positive and the other negative, and it is notewor­thy that the graphemes adduced in these entries, ZI and MU, resemble each other. In the Old Babylonian cursive, ZI looks like a MU supplied with two additional vertical wedges. Could ZI therefore symbolize the “raising” (zaqapu) of the house mentioned in the particular apo­dosis,⁸⁹ while MU signifies the exact reverse? Example nos. 64-65 provide a comparable pair of omens with opposite predictions, and again, the signs, GIS and US in no. 64 and TAB and PA in no. 65, have similar shapes; yet TAB and PA, unlike GIS and US, are “open” on the right side, a feature that might have indicated to the ancient experts that the legacy of the house dealt with in the omen entry was about to “flow out” and perish.^{90 91} Another reference to TAB, in no. 77, is followed by a positive prediction: a woman having difficulties in childbirth will easily give birth (sutesuru). Could it be that in this case, the two parallel wedges of the TAB sign signaled a smooth delivery? Example no. 119 is also of interest. The link between the sign NIGIN in the protasis and the word mesiru “confinement” in the apodosis could be based on etymography, as argued above, but also on the shape of the sign, a square formed by four wedges “confining” an empty space in the center. And finally, it is noteworthy that the rather similar signs IGI, SE, PI, and UD in nos. 80, 86, and 122, for whatever exact reason, all refer to happiness (tub libbi)⁹

There is one more physiognomic text that needs to be taken into account here: the highly unusual Nineveh manuscript K. 2087(+?)K. 2088, copied by Kraus (1939: pls. 35-36; see also figs. 7.2-3 below), and edited by Bock (2000a: 258-61). Its section on cuneiform graphemes

differs from the corresponding passages in Alamdimmu III and KAR 395 in several respects. First and most conspicuously, while otherwise written in Neo-Assyrian script, the tablet pres­ents the graphemes it discusses in forms that seem to be based on an attempt to reconstruct the earliest, essentially pictographic stages of cuneiform writing, even though closer inspec­tion reveals them to be artificial concoctions of a younger age that do not match the real sign forms of the late fourth millennium. Second, while some entries seem to have the usual omen format, others do not. And third, quite a few of the entries display very clear examples of “etymographical” thinking.

The section on graphemes is introduced, in K. 2087, rev. i'(?), “III'”,⁹² by the heading alam-dim-me-e SAG.KI N[A...] “(Concerning) the appearance of the forehead of a man [...],” a line highly reminiscent of the introduction to the analogous passage in Alamdimmu III (see above, no. 56). Then, in IV", follows the entry (129)⁹³ DIS ina SAG.KI NA BAD SUB U₄.MES- [su.] / ϋ§ Μί?1-Γχ1 [...] / EN(adi?) kim-ti-[su^!...] “If (the grapheme) BAD appears on the forehead of a man, [his] days [will be short (...)], death through... [...] together with [his] family⁹⁴ [...].” A drawing of a BAD that resembles an arrow accompanies the entry.⁹⁵ The reference to death (BAD = ϋ§ = m„tu) seems to be based on etymography. Entry no. V' (130) reads: DIS KIMIN SIG₇ SUB [...] “If ditto (the grapheme) SIG₇ appears [...].” Entry no. VI^õ (131) refers to the sign GISIMMAR, and entry nos. VII' and VIII' (132-33) to signs mostly broken away. The apodoses of these last entries are lost, and of the drawings only modest traces remain.

K. 2087 rev. i' breaks off at this point. The text seems to continue, after a gap, with K. 2088, a fragment with remains of one column, probably the last of the reverse. Entry I' of this piece (134) is mostly lost. Entry II' (135) deals with the sign TUK, presented both in an archaizing and in its Assyrian form. The short text passage accompanying these sign forms is badly damaged and largely unintelligible, but it includes the logographic writing NIG.TU[KU] = isarru “he will become rich,” indicating that there is an etymographical link between the sign and the text passage. It is also clear that the passage, like the ones in the following en­tries, does not have the omen format found in the entries in rev. i'. We cannot be absolutely sure, therefore, even though it seems likely, that we are still dealing with signs observed on a man’s forehead. Entry no. III' (136) provides an archaizing drawing of a sign interpreted by Bock as KUM, with an inscribed smaller sign resembling a monumental Babylonian NIG and another, badly broken sign on the right. The accompanying short text — ku-um-ma' / ib-ta-ni “He built a shrine (kummu)” — is clearly linked to the sign through paronomasia.⁹⁶ Entry no. III' is followed by a subscript (IV') explaining that the preceding section presented “four cuneiform signs from a second liginnu-tablet” (4 G&.SUM sd KA 2-ti IM.GI[D.DA]).⁹⁷ The

following entry, V' (137), presents another, unidentifiable archaizing sign form, and a short text too badly damaged to make much sense of it.

K. 2088 breaks off at this point. After what may have been an extremely small gap, the left column of the tablet continues with K. 2087 rev. ii'.⁹⁸ Entry no. I' of this section (138) reads: tu-kul-ta-su / ^dasari-ma / i-da-as HUL.GAL.BI / HUL.HUL.BI “He treated (or: he will treat) the god Asari (i.e., Marduk), who supports him, with disregard — misfortune for him, evil for him.” The archaizing sign accompanying this sentence is tentatively identified by Bock as UB, but Kraus’s suggestion (1935: 50) to read it as HUL would provide a better etymographi­cal link. Entry no. II' (139) shows a stylized palm tree that is supposed to represent the sign GISIMMAR, whose Neo-Assyrian form is given as well. A short text on the left reads: dum-qa / u-sat-lim-su / ύ-kin-su / tak-li-me “He provided him with good things, established for him the taklimu-offering.” Since damaqu can be written with the GISIMMAR sign, read sa₆, there is again an obvious link between text and grapheme.^{99 *} Entry no. III' (140) offers an archaizing and a Neo-Assyrian version of the sign DU, accompanied by the phrase al-la-ku / sa ur-hi / i-du-us-su / i-ba-a° “A traveler went (or: will go) at his side”; it is linked to the sign through the well-established reading of DU as alaku “to go.” This part of the text comes to an end with yet another subscript (IV'), which states that the “four graphemes” treated in the preceding lines were taken “from the third liginnu-tablet.”

The last preserved section of the fragment seems to contain nothing but drawings of pseudo-archaic signs and their Neo-Assyrian equivalents. Entry no. V' (141) presents the Neo-Assyrian form of the sign MAH and a drawing that looks like a hill, perhaps because MAH = siru means “exalted, high(-ranking).” No. VI' (142) offers the sign RAD/SITA and two hori­zontal lines possibly symbolizing an irrigation channel (note that the sign represents the word ratu “water-channel”). No. VII' (143) has yet again GISIMMAR, this time accompanied by a drawing of a half-circle, and VIII' (144) has SA, in its Neo-Assyrian form, a fairly realistic archaic version, and the pseudo-archaic shape, probably based on the latter, of a triangular structure. The remaining entries are mostly damaged and obscure. IX' (145) presents GA and a drawing made up of horizontal wedges, X' (146) TUK(?) and IL with two small stars in between, XI'¹⁰⁰ (147) a DU inscribed in a rectangular configuration, together with NIGIN(?) KIB(?) written on the right, XII' (148) GIM with a drawing of a pseudo-archaic form of the sign, and XIII' (149) LIL ύ KLJR(?), together with a drawing that is mostly lost. After another — badly damaged — subscript (XIV') probably stating that the preceding section included “nine cuneiform signs from the forth liginnu-tablet,” K. 2087 rev. ii' breaks off.

The graphemes mentioned in K. 2087(+) are (in alphabetical order): BAD (129), DU (140, 147), GA (145), GIM (148), GISIMMAR (131, 139, 143), HUL(?) (138), IL (146), KIB(?) (147), KUM(?) (136), κύΕ(?) (149), LIL (149), MAH (141), NIGIN(?) (147), RAD (142), SA (144), SIG₇ (130), TUK (135, 146(?)), and ύ (149). Omen no. 129 has a negative prediction, while of the intelligible short texts of the left column nos. 135, 136, 139, and 140 seem to be positive and no. 138 negative.¹⁰¹

The highly archaizing sign forms listed in K. 2087(+) seem to indicate, at first glance, that K. 2087(+) represents a tradition that precedes Alamdimmu III with its Old Babylonian background. But in reality, the text probably originates from a later period. Quite a few of the signs analyzed in it were usually employed as logograms or CVC signs, grapheme types more widely used in post-Old Babylonian Akkadian writing. The text’s focus on “etymography” points to a later stage of cuneiform culture as well. And finally, attempts by Mesopotamian scholars to systematically reconstruct the original forms of cuneiform graphemes are otherwise known only from first-millennium sources, most prominently from a number of Neo-Assyrian and Late Babylonian syllabaries with added columns featuring what the scribes apparently believed were those forms,¹⁰² but also from a small fragment from Kalhu inscribed with what appears to be a first-millennium historical text written in extremely archaic characters.^{103 *} The text represented by K. 2087(+) was probably composed by scribes who, aware of the tradi­tion of analyzing “Old Babylonian” sign forms on the face of human beings, felt motivated to replace them with even older forms, which they believed were closer to the beginnings of all wisdom.