higher-level noun phrase. Various types of relationship to the verb are possible, a basic
distinction being between phrases that are essential to the verb’s syntax (its
complements) and those which are less so (its adjuncts). In Sumerian, two case markers
occur only with adjuncts: the similative (/gin/ with a basic meaning such as “like”) and
the adverbiative (/esˇ/ “in the manner of ”). Other cases, termed relational, mark what
can be a complement or an adjunct depending on the verb with which they occur, an
example being the dative (/ra/ “for,” reducing to /r/ after a vowel), again restricted to
phrases with a human head (ama-zu-ur 2 , /amazur/, ama+ /zu/ + /r/, “mother” + “your”
- DAT, “for your mother”). The other relational cases are the comitative (/d(a)/
“with”), ablative (/t(a)/ “from”), allative (/sˇ(e)/ “to”), directive (/e/ “toward”), and
locative (/’a/ “in,” the glottal stop indicating here, as elsewhere, that typically it does
not contract with a preceding vowel).
A few core cases mark complements only: the absolutive (a zero marking) specifies
a phrase with a noun at its head as the subject of an intransitive verb or the direct
object of a transitive verb, while a noun-initial phrase that is the subject of a transitive
verb is marked by the ergative (/e/, which appears to contract with any preceding
vowel, the resulting long vowel sometimes being indicated by writings like ama-a,
/ama ̄/, “mother”). However, if a subject phrase has instead a personal pronoun as its
head the analysis differs. These phrases are zero-marked: as this system marks the
intransitive and transitive subject in the same way it can be regarded as following
nominative–accusative principles.
The expression of syntactic relationships within a noun phrase is dominated in
Sumerian by the genitive case marker (/ak/ “of,” in a very basic representation of its
morphology). Such possessor noun phrases are typically embedded within a higher-
level phrase that itself ends in a case marker indicating its syntactic relation to the
verb:dumu lugal-la-ra(“child” [“king” + GEN] + DAT, “for the child of the king”).
Occasionally, however, they are displaced to the left, into a syntactically isolated but
more emphatic position, their original slot being marked by a possessive determiner:
lugal-la dumu-ni-ir([“king” + GEN] “child” + “his” + DAT).
VERBAL MORPHOLOGY
One feature of Sumerian verbs is their use of the other type of bound morpheme –
affixes (prefixes coming before a base, suffixes following it and circumfixes surrounding
it), their finite forms being much more morphologically complex than their non-finite
ones. This morphology serves partly to make distinctions of tense (locating a situation
in time) and/or aspect (expressing a situation’s temporal quality). Aspect labels are
adopted here, to a degree simplifying the complexity of these grammatical categories.
The principal distinction made is between completive and incompletive aspects,
that is, between verbal forms specifying an event as finished or as ongoing. Just as
stative verbs are excluded from continuous aspect in English (I am knowingbeing
unacceptable), so too stative verbs are excluded from incompletive aspect in Sumerian.
The following discussion focuses on affixation; in addition, the language’s few irregular
verbs have a different base in incompletive aspect.–– The Sumerian language ––