This prefix has two allomorphs: i) /i/and ii) /a(l)/-; and the second allomorph has again two allomorphs: a) /al/-when there is no other prefix between it and the verbal stem, see ex. (203) below; b) /a/-in all other environments, see, e.g., ex. (206) below. The /i/-prefix assimilates to the vowel of the following syllable with respect to vowel height around the middle of the 3rd millennium BCE, see Lesson 2, section 2.2. above.
In verbal forms containing a morpheme in any of S3–11 the occurrence of the finite-marker prefix is constrained by a rule which is clearly phonological by nature: it occurs only in verbal forms that would start with a consonant cluster without the finite-marker prefix. In ex. (198) below, for example, the verbal form would start with a geminated consonant, /mm/, without the finite-marker. The verbal form in ex. (199) has no finite-marker, and its starts with a single consonant.
(198) Gudea Cyl. A 18:8 (Lagash, 22nd c.) (ETCSL 2.1.7)
e₂-e im-ma-ŋen
e=e S2i-S4m-S5b-S7a-S12ŋen-S14ø house=DAT.NH FIN-VEN-3.SG.NH-DAT-go-3.SG.S
“He went to the temple.”
(199) Gudea Cyl. A 2:26 (Lagash, 22nd c.) (ETCSL 2.1.7) kug dŋa₂-tum₃-dug₃-ra mu-na-ŋen
kug ŋatumdug=ra S4mu-S6nn-S7a-S12ŋen-S14ø holy DN=DAT.H VEN-3.SG.H-DAT-go-3.SG.S
“He (= Gudea) went to holy Gatumdug.”
In ex. (200) below the finite-marker is present, without it the verbal form would again start with a geminated consonant, /nn/. And the verbal form starting with a single consonant in ex. (201) has no finite marker either.
(200) MVN 3 36 obv. 3:2–7 (Isin, ca. 25–24th c.) (P222204)
1/2 kug ma-na, maḫir, dumu amar-šubaₓ(MUŠ₃.KUR.ZA)-ra,
1/2 kug mana=ø maḫir dumu amaršuba=ra
1/2 silver unit=ABS PN₁ child PN₂=L3.H
a-pa-e₃ azlag₂, sam₂-še₃, i₃-ni-gu₇
apa’e azlag=e sam=še S2i-S6nn-S10i-S11n-S12gu-S14ø PN₃ fuller=ERG price=TERM FIN-3.SG.H-L3-3.SG.H.A-eat-3.SG.P
“Apa-e, the fuller, made Mahir, the son of Amar-šuba, receive (lit. ‘eat’) half a mana of silver as the price.”
(201) Gudea Cyl. A 4:4 (Lagash, 22nd c.) (ETCSL 2.1.7) kar niŋin₆ki-na-ke₄ ma₂ bi₂-us₂
kar niŋin=ak=e ma=ø S5b-S10i-S11n-S12us-S14ø
quay GN=GEN=L3.NH boat=ABS 3.SG.NH-L3-3.SG.H.A-be.next-3.SG.P
“He (= Gudea) moored the boat at the quay of Nigin.”
In verbal forms in which the finite-marker prefix is the only prefix before the verbal base in S12 the rule just described does not hold. In ex. (202) below the presence of the finite-marker has no phonological reason, it is apparently only there to indicate that the verbal form functions as a predicate but not as a modifier.
(202) Gudea Cyl. A 12:12 (Lagash, 22nd c.) (ETCSL 2.1.7) gu₃-de₂-a i₃-zig
The phonological nature of the rule that predicts the presence of the finite-marker suggests that the system we know is the outcome of a linguistic change that resulted in the disappearance of the finite marker in many morphophonological environments. In verbal forms containing a morpheme in any of S3–11, the finite-marker is preserved only in environments where its disappearance would result in a consonant cluster at the beginning of a word, which is, however, not allowed by the rules of Sumerian phonotactics.
Our texts consequently only preserved the remnants of an old system in which the prefix /i/- and /a(l)/- had a complementary distribution and probably occurred in all finite verbal forms. In this system, the finite-marker /i/- was used in preterite verbal forms expressing an action. Verbal forms expressing a state, see exx. (203) and (204), a state resulting from a previous action, see ex. (205), or present-future verbal forms expressing an action, see exx. (206), (207) below, were prefixed with /a(l)/-.
(203) CUSAS 26, 45 rev. 1:3 (unknown, 23rd c.) (P427623) inim-be₂ al-til
inim=be=ø S2al-S12til-S14ø word=DEM=ABS FIN-complete-3.SG.S
“This case is finished.”
(204) E-ana-tum 1 obv. 4:20–22 (RIME 1.9.3.1) (Lagash, 25th c. BCE) (P222399) e₂-an-na, dinana, eb-gal-ka-ka a-tum₂
eana inanak ebgal=ak=ak=ʾa S2a-S5b-S10(i>)ø-S12tum-S14ø TN₁ DN TN₂=GEN=GEN=L2.NH FIN-3.SG.NH-L2-worthy-3.SG.S
“He is worthy of the E-ana of Inana of the Ebgal”
(205) En-ana-tum I 18 2:3–7 (RIME 1.9.4.18) (Lagash, 25th c.) (Q001086) šu-ni-al-dugud, kindagal, nam-nu-banda3 e2-šag4-ga,
P1šunialdugud kindagal namnubanda ešag=ak=ø
P1PN chief.barber overseership inner.room=GEN=ABS
an-na-taḫ-ḫa, KIB mu-dim2-dim2
S2a-S6nn-S7a-S12taḫ-S14ø-S15ʾa=e KIB=ø S4mu-S11n-S12dim~dim-S14ø
FIN-3SG.H-DAT-add-3.SG.S-SUB=ERG nail=ABS VEN-3.SG.H.A-create~PL-3.SG.P
“Shuni-aldugud, the chief barber, who has been assigned to the office of the personal quarters’ overseer, fashioned numerous inscribed clay nails.”
(206) Rimush 18 20–22 (RIME 2.1.2.18) (Agade, 23rd c.) (Q000842)
lu₂, im-sar-ra-e, ab-ha-lam-me-a
lu imsara=e=ø S2a-S11b-S12halam-S14e-S15ʾa=ø person inscription=DEM=ABS FIN-3.SG.NH.P-obliterate-3.SG.A-SUB=ABS
“Whoever obliterates this inscription, ….”
The prefixes of Slots 2–6
(207) En-metena 7 15–16 (RIME 1.9.5.7) (Lagash, 25th c.) (P222539) gur₄-gur₄ kug luḫ-ḫa i₃ itud!-da
gurgur kug luh-ʾa=ak i itud=ak=ø
vessel silver clean-PT=GEN oil month=GEN=ABS
dnin-ŋir₂-su₂-ke₄ ab-ta-gu₇-a,
ninŋirsuk=e S2a-S5b-S9ta-S12gu-S14e-S15ʾa=P5ø DN=ERG FIN-3.SG.NH-ABL-eat-3.SG.A-SUB=P5ABS mu-na-dim₂
S4mu-S6nn-S7a-S11n-S12dim-S14ø
VEN-3.SG.H-DAT-3.SG.H.A-create-3.SG.P
“(For the god Ningirsu, En-metena) fashioned a gurgurvessel of purified silver, from which Ningirsu consumes the monthly oil (offering).”
This reconstructed original system almost completely disintegrated in the Sumerian used in the South (preserved mainly in texts from Lagash) by the end of the 3rd millennium BCE. Here the the prefix /i/-replaced the prefix /a(l)/-in almost all verbal forms, except for the imperative, see ex. (208), and ex. (392) in Lesson 14, section 14.3, exx. (481) and (482) in Lesson 16, section 16.2 below.
(208) NG 208 obv. 1:17 (Lagash, 21st c.) (P135750) kišib-ŋu₁₀ zi-ra-ab
kišib=ŋu=øS1zir-S2a-S11b
seal=1.SG.POSS=ABS break-FIN-3.SG.NH.P
“Destroy my sealed tablet!”
In the North (preserved mainly in texts from Nippur and Adab) the original distribution and functions of the two finite-marker prefixes remained more or less intact; and the verbal forms with the prefix /a(l)/- acquired even an additional function compared to texts from the South: the finite-marker prefix /a(l)/-came to be used as a marker of passive verbal forms, see ex. (209) below, corresponding to verbal forms with the middle prefix /ba/-in Sumerian in the South, see section 11.3 below.
(209) OSP 2, 58 3–4 (Nippur, 23rd c.) (P216212)
den-lil₂-le-ma-ba-ra, an-na-la₂
enlilemaba=ra S2a-S6nn-S7a-S12la-S14ø PN=DAT.H FIN-3.SG.H-DAT-weigh-3.SG.S
“(16 shekels of silver, (which is) the value of the corvée obligation,) were paid to Enlile-maba.”
LESSON11
148
Exx. (210) and (211) below show that passive forms expressed with the middle marker /ba/-in Ur, may be expressed with the finite-marker /a(l)/-in Nippur.
(210) MS 4096 rev. 6 (Ur, 21st c.) (P253193)
igi-⸢be₂⸣-še₃ mu lugal ba-pad₃
igi=be=še mu lugal=ak=ø S5ba-S12pad-S14ø face=3.SG.NH.POSS=TERM name king=GEN=ABS MID-call-3.SG.S
“The promissory oath was sworn before them.”
(211) MVN 3, 330 obv. 9 (Nippur, 21st c.) (P113890)
mu lugal-be₂ al-pad₃
mu lugal=ak=be=ø al-pad-ø
name king=GEN=3.SG.NH.POSS=ABS FIN-call-3.SG.S
“The promissory oath concerning it was sworn.”
The old function of the /a(l)/-finite-marker to mark present-future forms was apparently also known to the Nippurian compilers of the so called Old Babylonian Grammatical texts, which date to the first part of the 2nd millennium BCE. In OBGT VII (MSL 4 88–89), for example, which is an Akkadian-Sumerian paradigm of the verb ŋen = alākum, “to go”, the present-future Akkadian verbal forms are translated with present-future Sumerian verbal forms containing the finite-marker /a(l)/-:
(212) OBGT VII 31–33 (= A 24189) (Nippur) 31. al-du i-⸢il⸣-lak
32. al-du-un a-al-lak 33. al-du-un tal₂-lak