Sumerian Verb Conjugator (SVC) (under construction)

In SVC, this verb pattern is coded as {V+a+mu+mu_oo}. In order to reproduce the verb form with SVC, one has to select the options "suppress -b- imperf." and "1st/2nd prefix: early".

Note that if {i+mu_oo+BI+e} were allowed, it would have to appear as i3-ĝa2-ĝa2, whereas {i+mu_vent+BI+e}, "he placeth it here", appears as i3-im-ĝa2-ĝa2.

to top

Ventive imperatives with prefix/suffix -um, -am3 {mu} and -um-ze2-en, -am3-ze2-en {mu+zen}

Jagersma, p. 504, writes: "In an imperative form that lacks a further affix after the ventive, the ventive prefix has the form /um/, probably through metathesis." OB literature has the following example:

However, the vowel a can also be used:

Indeed, OBGT BII in the version of the Chicago Oriental Institute provides these forms:
➔ ĝa2-nu-um {BP+mu}- al-kam - "(thou) come!"
➔ ĝa2-nu-um-ze2-en {BP+mu+zen} - al-ka-nim - "(ye) come!"

... whereas the version from the Ur excavations provides the following forms:
➔ ĝen-am3 - al-kam - "(thou) come!"
➔ ĝen-am3-ze2-en - al-ka-nim - "(ye) come!"

SVC shows both versions, the one with -um by default, and the other with -am if the option "-a-CV in imperative" is activated.

As expected, the suffix {um} can also loose the syllable-final -m- and then appears as -u3, e.g. in this passage:

And it is very likely that imperatives ending with -a can have an invisible ventive /m/, as well.

Moreover, when the ventive appears as (m-)ma- {mu+ba}, then it also seems that the preceding vowel need not be /a/, e.g. -am3-ma- or .a-ma, but can also be /u/:

Compare this with:

For more information on ventive imperatives, see also Foxvog, "Introduction to Sumerian Grammar" (2016), p. 112ff.

to top

IO and OO (Indirect and Oblique Object) in the 2nd person plural

With the IO prefix of the 2nd person plural, SVC provides three different forms. Jagersma in his "Introduction to Sumerian Grammar" (2017), p. 43, and (2018), p. 44f., focused on the older language, only mentions the prefix -e-ne-, but without any examples. However, in DGS (2010), p. 399 and p. 407, he states that the prefix of the 2nd person singular, namely -ra-, was also used for the 2nd plural, and that the plural form could have a facultative suffix -en-ze2-en. Since there are no attestations from his corpus, he can only provide examples from the OB period, namely the following from Inanna's Descent:

Attinger (1993) states on p. 210 and 231 that the "theoretical" form /ene/ is only found in grammatical texts (OBGT BII 191-246), written -e(n)-ne-. He considers the forms with -ra- to be Old Babylonian "néologismes(?)". Thomsen p. 221 shares this view.

SVC shows all these forms.

Note that the form -ra-...-en-ze2-en has the disadvantage that the 2nd plural suffix suppresses the direct object or absolutive marker in perfective forms and the subject or ergative marker in imperfective tense. On the other hand, if the suffix -en-ze2-en is omitted, then the number of the indirect object is lost, but the ergative or absolutive suffix can appear. The speaker may have given preference to the one or the other solution depending on the situation.

to top

With the OO prefix of the 2nd person plural, just as with the IO prefix, three different forms are possible. Jagersma in his "Introduction to Sumerian Grammar" (2017), p. 45, as well as (2018) p. 48, states that the prefix is -e-ne-, thus the same as with the IO. In DGS, p. 428, he provides an example, which he takes from Attinger (1993, p. 237):

While the same prefix is also attested in grammatical texts (Attinger (1993), p. 210), OB literature attests an alternative construction analogous to the IO prefix -ra-. The prefix of the 2nd person singular -ri- is taken and the suffix of the 2nd plural optionally added to the the verb stem. According to Attinger, p. 237, this may be a "forme néologique(?)".

Example:

Again, SVC provides all these forms.

The phenomenon is thus exactly analogous to the one described with the indirect object. The form -ri-...-en-ze2-en has the disadvantage that the 2nd plural suffix suppresses the direct object or absolutive marker in perfective forms and the subject or ergative marker in imperfective tense. On the other hand, if the suffix -en-ze2-en is omitted, then the number of the oblique object is lost, but the ergative or absolutive suffix can appear. The speaker may have given preference to the one or the other solution depending on the situation.

to top

Why is there an option "add -biOO-"?

SVC shows two options for oblique object prefixes: "OO" and "add -biOO-". At first glance this seems to be unnecessary because -bi- is included in the option "OO". The reason is that, while IO and OO prefixes normally cannot be combined, the OO prefix -bi- ("in(to) contact with") can appear together with IO prefixes, however only in verb forms where it stands immediately before the verb stem in the form -b-.

Jagersma (p. 435) provides the following example from the Ur III period:

A {bi} that is due to the option -biOO- is never rendered as a causative in translations in SVC. If you need a causative, then please use the option "OO".

to top

Co-existence of OO prefixes with other prefixes

According to Jagersma (who refers to the earlier language), oblique object prefixes cannot be combined with any of the following prefixes: IO, -da-, -ta-, -ši-, -ni-, -bi/e-, -bi-/OO. In addition, the middle marker ba- cannot coexist with OO -bi-. (p. 442).

SVC therefore does not allow any such combinations, with the following exceptions, which are allowed by the system of slots (Jagersma (2018), p. 84):
- IO prefixes can be combined with a prefix -b- = {bi_oo} immediately before the verb stem. See preceding and next paragraph.
- All OO prefixes except the 1st person {mu_oo} and the 3rd person non-human {bi} can be combined with the IO prefixes 1st person {ma} and the 3rd person non-human {ba}. See next paragraph.
- OO can be combined with -ni-, but then appears in the form of an IO. See paragraph "OO+ni+ appears as IO+ni+".

to top

OO combined with IO?

Old Babylonian Grammatical Texts (BIII §8) provide the following verb forms which, at first glance, could be interpreted as combinations of IO {nna} with OO {nni} (cf. Huber (2018), p. 28):

OBGT BIII §8:
➔ kas4 ⸢du11-ga⸣-na-ni-ib2 {BP+a+nna+nni+b; šu-ul-si2-im-šum},
 "make him run for him."
➔ kas4 ⸢ga-na⸣-ni-ib2-du11 {ga+nna+nni+b+BP; [lušalsimšum]},
 "let me make him run for him."
➔ kas4 ⸢ḫe2-na⸣-ni-ib2-du11 {ḫa-i+nna+nni+b+BP; [lišalsimšum]},
 "they certainly made him run for him".

Huber (2018, p. 18) believes that in these examples -ni- marks the "subordinate subject", thus the causee or the person who is caused to "run for him" (lit. "to do kas4 for him"). The Akkadian version of it, šulsimšum etc., seems to support this interpretation.

However, Jagersma does not allow this combination (Jagersma 2010, p. 442). In his opinion, such forms are "due to imperfect language learning by non-Sumerian scribes". They may result from the fact that Akkadian Š-stems tend to be rendered by the prefix {ni}. (Jagersma, communication of 9 May 2019; cf. Attinger 1993, pp. 59f. and 583f.) It should be noted, however, that the verb forms of OBGT BIII §8 can also be reproduced using SVC by selecting the options 'OO' + '-ni-', because this combination of prefixes transforms the OO into an IO (OO+ni→IO+ni = nna+ni). (See paragraph "OO+ni+ appears as IO+ni+" further below.) With this interpretation the causee would be marked by /na/, not by /ni/ in these examples. The correct interpretation of OBGT BIII §8 would then be:

➔ kas4 ⸢du11-ga⸣-na-ni-ib2 {BP+a+nni→nna+ni+b; šu-ul-si2-im-šum},
 "make sb. run for him."
➔ kas4 ⸢ga-na⸣-ni-ib2-du11 {ga+nni→nna+ni+b+BP; [lušalsimšum]},
 "let me make sb. run for him."
➔ kas4 ⸢ḫe2-na⸣-ni-ib2-du11 {ḫa-i+nni→nna+ni+b+BP; [lišalsimšum]},
 "they certainly made him run for him".

Old Babylonian literature also has the following instances where at first glance {nna} and {nni} seem to appear in the same verb forms:

However, Attinger's new edition of the text in Babel und Bibel, p. 25, has only the variants mu-un-na-ni-ĝal2 and mu-na-ni-in-ĝar. Jagersma thinks it is actually {nna+ni}. (private communication of 9 May 2019). Attinger's literal translation is (p. 39 foot note 147):
"il y a placé pour elle" "un bon festin".
This solution is also supported by the following parallel where a causative meaning can be ruled out:

Nevertheless, Jagersma does allow IO+OO if the two prefixes occupy different slots. One such example (from the ED IIIb period) is found in his grammar of 2010 on p. 409 (ex. 44):

The /e/ of the second person may become invisible, but nevertheless be present in mu-ni-.-:

In the following example, the same phenomenon also occurs after the IO prefix {nna}:

And in the following examples, the same apparently happens after the prefix {ḫa}:

The following example seems to demonstrate that the same kind of phenomenon appears not only with the 2nd, but also with the 1st person:

In the following instance, where the agent is a 3rd person, the /e/ is difficult. A locative {e} seems to be ruled out as well, because it is incompatible with locative {ni}:

In the following case, which is exceptional, the 2nd person /e/ seems to appear before the OO prefix {nni}:

to top

Old Babylonian mu-e-re-, ba-e-re-, etc.

A similar "metathesis" of the final 2nd person prefix {e} can also take place with the ablative prefix {ta}. (cf. Attinger (1993), p. 219f., remarques 2, footnote 500; Foxvog, Ventive, p. 95, note 88). In this case, {mu+ta+...+e+B} becomes mu-e-re-...-B and {ba+ta+...+e+B} becomes ba-e-re-...-V. Unlike the forms with -e-ni-, the forms with -e-re- do not often replace the final person prefix by an {n} or {b}.

Note that the form ba-e-re- has the advantage that it cannot be analysed as {bara+e+}. It is clearly {ba+ta/ra+e+}.

Examples:

SVC shows the above forms only if the option "1st/2nd prefix: late" is selected, because they appear only in Old Babylonian literature.

There are a few cases from royal poetry of the first dynasty of Isin, where mu-e-te- and me-te- is used instead of mu-e-re-. All of these examples are verb forms of the compound verb a2 nun ĝal2:

The following example might belong in the same category, but mu-e-ta- appears as mu-u3-ta-:

In the following examples, the e seems to stand for locative {e} rather than for the 2nd person prefix:

And in the following example, e-re seems to stand for IO {ra} plus locative {e}:

to top

Old Babylonian mu-e-ši-, ba-e-ši-, etc.

Attinger also mentions that there are cases of "(-)mu-e-ši-, -me-ši-= {mu + ši + e(r)} (rare; cf. Jacobsen, AS 16 93 n. 15 et ZA 78 199 sq.)" (Attinger (1995), p. 219, remarques 2). Examples for {+ši+e+B} becoming {+e+ši+n+B} are rare and hard to search because {+e+ši} mostly stands for "to thee" or also "to me". However, in ETCSL there are the following likely examples:

In the following examples, the /e/ before {ši} seems to be a locative prefix {e}:

A related phenomenon might be the following, where there seems to be a metathesis of the /e/ of the verb root e-re7 with a preceding prefix {ši}:

SVC does not produce the above forms.

to top

Old Babylonian mu-e-de3-, ba-e-de3-, etc.

Analogous forms for the comitative prefix {da}, namely forms of the type -mu-e-de3(da)- {mu+b+da+e+} or -ba-e-de3(da)- {ba+da+e+} are difficult to find because in most cases, -e-de3- means either "with thee" or "thou canst". However, the following cases are beyond doubt:

In the following example, the "metathesis" seems to occur with the 1st rather than the second person:

The "metathesis" seems to also take place when the comitative prefix refers to a person, however, the vowel /e/ is not lost before the verb stem:

In the following example, the same phenomenon occurs after the IO prefix {nna}:

There is also the following example, where the /e/ seems to stand for a locative {e}:

In the last examples, the outer personal prefix {n} before {da} is apparently supplanted by the locative {e}. This does not seem to happen in the following cases:

And again, the /e/ of the verb root e-re7 can appear before the prefix {da}:

to top

Old Babylonian im-mu-e-, am3-mu-e, etc.

And there is another phenomenon in OB language that appears with the 2nd person final prefix, namely that {i+mu+ba+e+} does not become im-ma(-e), but im-mu-e:

In the following example, it seems that /m-mu-e/ is contracted to /m-mi/, thus {mu+ba+e+} can apparently appear as im-mi-:

In other cases, the /e/ after /m-mu/ is omitted, as in the following example:

And in the following examples, it seems that the /e/ after /m-mu/ is replaced by an /n/. Note that {+mu+ba+n+} would usually appear as .m-ma-an-, not as .m-mu-un-, thus a 2nd person is likely.

A related phenomenon might be the appearance of ma-e-... as mu-e, as in the following examples:

Interestingly, the expected form ma-e- is not found at all in ETCSL, with only one exception, where, however, the /e/ seems to be a locative prefix:

to top

Groups of persons treated as non-human class

On p. 103, Jagersma states:
"... a group of persons is usually treated as non-human and not as human. ... Normally, a group of persons is non-human, but if the persons are important as separate individuals, they are treated as human..."
and:
"If the subject of the verb refers to two individuals, each of them mentioned by name, it is nearly always expressed with the human person-affixes {n} and {eš}, but sometimes with the non-human prefix {b}. E.g.:

As an analogon in imperfective tense, the following example can be quoted (p. 298):

Jagersma in DGS, p. 298, gives this as an example for "mismatches in number". However, in a communication of 15 Feb. 2020 he agrees that it can also be considered a case where a group of persons is handled like a non-human singular.

In the following example from the OB period, any plural marker is missing in the second verb form:

If it were a perfective form and "they" were the subject, as is suggested by ETCSL's translation, then it is interesting to note that there is a human inner person prefix {n}, although according to Jagersma's examples above {b} should be used. Thus, we would have expected either ne mu-un-su-su-ub-be2-eš or ne i3-im-su-su-ub. A reference to the oblique object "him" is missing in any case. We would expect ne mu-ni-in-su-su-ub-be2-eš or ne mu-ni-(ib2-)-su-su-ub. On the other hand, if the subject is "he", thus he kissed them, then the verb form looks more correct, but a reference to the oblique object "them" is also missing in the verb form. We would expect bi2-in-su-su-ub or im-mi-in-su-su-ub. Moreover, it could also be a passive: "he was kissed (by them)". The /n/ before the verb stem would be the oblique object prefix {nni}. And finally, it could perhaps be an imperfective, with the expected suffix {ene} missing.

Here, the perfective passive solution looks cleanest, and the plural subject is just suppressed by using a passive.

Examples for a group of persons as a direct object or intransitive subject are:

And in imperfective (from DGS p. 341):

The phenomenon also appears with prefixes of the OO and IO (from Jagersma, DGS p. 432, ex. 59 and p. 412, ex. 58):

And it also appears with the possessive suffix {bi} ({be2}) (from Jagersma, DGS p. 214, ex. 43):

to top

Plural objects and subjects in perfective

Ergative subject 3rd plural

Expressions like "they beat me" (perfective) have the problem that the plural ergative "they" requires a suffix -eš, whereas the absolutive "me" requires a suffix -en, but both cannot be combined. A form like *in-tud-eš-en is not allowed. Edzard (2003, p. 88) is not sure how this case was treated in Sumerian.

However Foxvog (2016, p. 64) writes:
"The minimal verbal sentence ba-an-zi-ge-eš could thus actually represent either 'He/she (-n-) caused them (-š) to rise' or 'They (-n-√-(e)š) caused me/you/him/her/it/them to rise'!"

Such equivocation concerning the direct object can be reduced if the plurality of the ergative subject is expressed through the non-human prefix {b} (cf. Jagersma, p. 339). Then no additional suffix is required to express the number of the subject, and the suffix slot can be used to indicate the direct object. E.g. (by SVC)
ba-ab-zig3 either means "it (-b-) caused him/her/it to rise" or "they (-b-) caused him/her/it to rise".
This has the advantage that the direct object can be made clearer:
ba-ab-zig3-ge-en: "It/they caused me/thee to rise"
ba-ab-zig3-ge-en-de3-en: "It/they caused us to rise"
ba-ab-zig3-ge-en-ze2-en: "It/they caused you to rise"
ba-ab-zig3-ge-eš: "It/they caused them to rise"

SVC does not show these alternative forms. It is very easy for the user to keep in mind that the subject row "it" provides alternative forms for the subject row "they".

A combination of both methods to mark a plural subject is given in the following examples, where the non-human subject prefix {b} is combined with the plural suffix {eš}:

Ergative subject 1st and 2nd plural

Jagersma (2017, p. 38) provides the following examples:

Thomsen (p. 143 and 149) and Edzard (1976b p. 160) also mention the following instance:

From this, it seems logical to build the following forms:
*ba-zig3-ge-en-de3-en {i+ba+?+BP+enden}, "we caused thee/him/her/it/you/them to rise".
*ba-e-zig3-ge-en-ze2-en {i+ba+e+BP+enzen}, "ye caused me/him/her/it/us/them to rise".

Direct object or subject in plural

From the above quotation from Foxvog, it follows that in perfective forms the plural suffixes {eš}, {enden}, and {enzen} either indicate the subject or the direct object. E.g.

ba-an-zi-ge-eš - 'He/she caused them {eš} to rise'
ba-an-zi-ge-eš - 'They {eš} caused me/you/him/her/it/them to rise'.

Even if the prefix is {b} (instead of {n}), we cannot be sure whether the suffix {eš} marks the subject or the direct object.

Plural verb root with singular prefixes and suffixes

There are instances where the plural verb root is used because there is a plural direct object or intransitive subject, but the plural markers (-n-)V-eš are missing. It seems, however, that this usage is mainly found with the non-human class, like animals or dead objects. Examples are found with the verb gub/sug2, with the verb tum2/laḫ4, and with the verb lug/se12:

However, the plural stem does not seem to be mandatory with a plural of animals:

The phenomenon of a plural verb stem, but without a plural suffix, also appears with the human class:

Thus, in the verb stem the intransitive subject is treated as a plurality, whereas in the suffix it can be treated like a "group" and a non-human singular or numberless. Other examples could be these:

In these examples, however, we cannot be sure whether the suffix {eš} has just disappeared due to defective spelling. Jagersma (DGS, p. 316, ex.18) analyses the verb form e-da-se as "'i-n-da-se12-Ø", but in another place (p. 197f., ex. 300a,b) he analyses it as "'i-n-da-se12-eš". According to a private communication (6 February 2020) he is not sure about the correct analysis.

Note, however, that the plural verb stem is not mandatory if a group of persons is treated as a non-human singular. E.g.:

In a nutshell, it appears that the form i3-sug2 {i+BPpl} or i3-im-sug2 {i+mu+BPpl} is a possible expression for "they stood", where "they" can refer to a group of either objects, animals, or human beings. SVC therefore assumes that with verbs that do not have a separate plural stem, too, a verb form can be pluralic even when there is no plural mark. E.g. i3-ĝar can be equivalent with i3-ĝar-re-eš. Thus, in the verb form i3-gar, the stem ĝar could be interpreted either as a singular or as a plural stem. SVC therefore displays e.g. the following table fields:

Plural verb stems

All prefixes and suffixes that unequivocally indicate a plural belong to the human class. Prefixes and suffixes of the non-human class can never be plural. If they refer to a group or collective of persons, they are nevertheless treated like a non-human "singular".

The situation is different with singular and plural verb stems. The plural verb stem does not necessarily refer to a plurality of human beings. It could also occur with a plurality of animals or inanimate objects. Thus the distinction between the categories human and non-human, of which the former does allow plurals, the latter does not, does not apply to verb stems.

to top

Plural objects and subjects in imperfective

Plural inner personal prefixes

The 3rd person plural imperfective has the prefix {ne} or {nne}, but it is well attested only in the early language and seems to disappear after the Old Akkadian period. There is no unambiguous testimony for the double n. Nevertheless it is likely that it had the same form as the outer personal prefix {nne}. (Jagersma pp. 339f.) Therefore we always render it as -(V)n-ne-.

The outer prefix nne must be distinguished from the inner person prefix nne. Both Jagersma (p. 340) and Zólyomi (2016, p. 84) quote the following case, where IO -ra- appears combined with direct object -ne- (= nne). I quote it with a bit more context:

Thus, forms like *ḫu-mu-ne-ne-šum2-mu = *ḫu-mu-un-ne-en-ne-šum2-mu {ḫa+mu+nne_io+nne+BI+e}, "he must give them to them", seem to be possible and will actually be shown by SVC.

The following example from the OB period at first glance seems to belong in the same category:

However, Edzard (ZA 81 (1991), pp. 216-217) thinks that -ne- here is an unusual spelling for -ni-. Thus the verb would have to be analysed as {ḫa+mu+ra+ni+BBP+en}:
"Fine flour of the gods let me bring to thee into the mountain lands."
Indeed, this gives better sense because all of the extant words form one complete sentence.

It thus seems that after the Old Akkadian period, preference was given to the non-human personal prefix {b} to mark the 3rd person plural human. (Jagersma p. 340f.) If the verb has a separate plural stem, the form can nevertheless be unambiguous because the direct object prefix {b} will be combined with the plural verb stem.

The following Old Babylonian example can be given for the 1st person plural. The expected prefix {me} can also be written /mu-e/ (Cf. Attinger (1993), p. 227):

In line 240, a metathesis of the direct object prefix {me} with the comitative prefix {da} might have taken place.

The imperfective inner prefix of the 2nd person plural is not attested, but it is assumed that it was {ene}, just like the outer prefixes used for IO and OO and with {da} and {ši}.

Plural suffixes

The personal suffixes which mark the ergative or intransitive subject in imperfective are unambiguous as long as their spelling is not defective. E.g. {eš} can appear as -e, which could be interpreted as a 3rd singular, and {enden} can appear as -e-de3, which could be understood as a 3rd singular form with {ed}.

It should be noted, however, that a 3rd singular form can also be used for a collective or a group of persons (even only two persons). Thus:
- intransitive forms with a human plural subject can have a final -e (ed) instead of -e-de3(-eš) {ed+eš};
- transitive forms with a human plural subject can have a final -e {e} instead of -e-ne {ene}.

It also happens that both forms, the one with suffix {e} and the one with suffix {ene} appear beside each other, e.g. in the following text:

In the following example, we have a plural intransitive subject expressed by a plural verb stem but without a plural suffix:

SVC therefore shows the following alternative forms:

For the usage of the plural stem, please read the previous chapter.

to top

to top

Treatment of Persons as Non-Persons

The Sumerian language distinguishes persons from non-persons in many ways, e.g. by using the personal prefix {n} for persons and {b} for non-persons. Nevertheless, there are cases where {b} is applied to persons. For example, {b} can be used to mark a plurality of persons. However, there are cases where {b} is also used for a single person.

Examples:

Attinger writes in his note 437 on verse 310:
"Aux ll. 310 // 333 // 343 (avec /tum/) et 327 // 337 // 343 (avec zi2-iĝ3), la divinité destinée d'abord à être emmenée dans le monde infernal, mais qu'Innana ne livre pas aux démons, est traitée dans la quasi-totalité des duplicats comme un non-personnel et vue donc comme déjà morte; sur l'association des morts avec le nonpersonnel, cf. provisoirement E. Flückiger-Hawker, OBO 166 (1999) 176; P. Attinger, ZA 105 (2015) 237. A la l. 357, Dumuzi est considéré comme un non-personnel dans y, mais comme un personnel dans U."

A comparable instance could be:

Very interesting is the following case:

In the following example Dumuzid is treated partly as human and partly as non-human:

We would have expected either mu-un-da-an-šub or im-da-ab-šub.

With other instances, Attinger's explanation obviously does not seem to hold:

In the following examples, a young woman is called a lu2-tur and treated as non-human:

Is it rather something similar as exists in German neuter words like "Kind", "Mädchen", "Fräulein", "Schätzchen", "Herzchen"? In the latter, the suffixes "-chen" and "-lein" are diminutives, with the connotation of endearment. Note, however, that lu2-tur is not always treated as non-human. No certain conclusion can be given here, just the space be opened for a greater variety of possible explanations.

And in the following example, Inanna treats the "man of my heart" as a non-human being:

In the following example, the mother seems to be treated as non-human:

And in the following examples, Enlil is treated as non-human:

Here, the non-human prefix {b} clearly refers to Enlil, who cannot be considered "as good as dead".

to top

Treatment of Non-Persons as Persons

Sometimes, non-persons are treated as persons, too. This happens when non-persons (e.g. animals or inanimate beings) behave or speak like human beings, e.g. in myths:

Note, however, that in the above two examples, the fly (nim) has the directive suffix {e}, which is used for non-human dative.

Here, the skip-rope game is treated like a human being with a human prefix {n}, perhaps because one has to imagine that Inanna is running the game. CDLI, P346100 has kug-ga, which would result in the translation:
  "I want to make him (mount Ebiḫ) run in Inanna's skip-rope game."
Then mount Ebiḫ is treated as a human being.

The phenomenon also appears in the literary genre of "debate poems" or in proverbs where things or living beings are humanised:

In the following instance, the human prefixes may be used because the water is doing something that is typically done by human beings:

In the following example the wind, although not really behaving like a human being is treated as human, whereas Euphrates is not:

to top

The plural suffix {en} (in cohortative)

Jagersma discusses a likely old plural suffix {en} on pp. 322f., referring to publications by Krecher (1965: 29-30) and Cavigneaux (1987: 47-48). SVC supports this suffix only as a marker of the 1st person plural in cohortatives. Jagersma provides two examples, the isolated verb forms ga-na-ab-be2-en and ga-i-re7-en. We quote the whole sentences in which they appear and provide our own morphological gloss:

The former example is transitive and uses a plural stem when the subject is in plural, the latter is intransitive with a plural stem.

In the following examples, the verb does not require a plural stem when the subject is in plural. Nevertheless, it makes good sense to assume a plural subject:

An Emesal example is found in Thomsen, p. 200:

If {en} marks a plural, it should actually be translated
  "We will set up a lament"
However, Thomsen believes that this verb form is incorrect. Edzard (p. 115) mentions our first example, ga-na-ab-be2-en, but thinks it is an irregular form equivalent to ga-na-ab-dug4 in Gudea. He therefore translates it in singular: "let me say ... to her".

to top

Nominalisation suffix {a} after imperative

Hard to imagine, but it does occur:

Programme option: "reflexive forms"

In Sumerian, reflexive verb forms are constructed using a "reflexive pronoun", more precisely using the noun ni2, "fear, aura, self", to which a personal possessive suffix is appended. In the third person human, ni2-te+Suffix is used. The verb itself then does not contain the direct object prefix that corresponds to the subject, but to "it", i.e. the "self". It is a bit reminiscent of English: "myself" = "my self", however, it has a wider range of uses. For details, see Jagersma, p. 234ff. and Thomsen 78f. Here, we only discuss its use with reflexive verb forms.

The reflexive pronoun can be in the absolutive case if it is a direct object or also in directive case, if it is an oblique object, or also in locative or terminative. Examples:

The reflexive pronoun of the 1st and 2nd person plural is not known with certainty. I assume that they are just ni2 with the regular possessive suffixes, i.e. ni2-me and ni2-zu-ne.

The possessive suffix can be missing, e.g. in the following examples:

In default mode, SVC does not display any reflexive verb forms. This means that table fields for verb forms like "I know myself", "I get to know myself" etc. are empty. However, if the option "reflexive forms" is activated, we find it rendered as ni2-ĝu10 i3-zu and ni2-ĝu10 ib2-zu-zu-un.

In addition, the fields for "I Xed us" and "we Xed me", "thou Xedst you (plural!)" and "ye Xed thee" are also empty in default mode. Even in English such constructions are unusual, although, in principle, it seems possible to say, e.g., "I praise us and thou praisest you (plural!)". Interestingly, in the 2nd person the person addressed is the singular, not the plural. Similarly, constructions such as "we adorn me" and "ye adorn thee" could perhaps be thought of. Therefore, such forms can be studied if the option "reflexive forms" is activated.

With the following examples, existing translations render ni2 as "terror". Although this might be correct, it must be kept in mind that, strictly speaking, all of them are ambiguous because they could also be reflexive in meaning:

It seems that reflexive IO and -da- are not expressed using the "reflexive pronoun" ni2. For example, the abilitative comitative must be considered a reflexive form, but it is formed without ni2:

According to Jagersma, p. 236, "The reflexive pronoun is not used as an indirect object, because an indirect reflexive is expressed with the prefix {ba} ...". And p. 490: "One middle use of {ba} is that it expresses the indirect reflexive. It then indicates that the verb has an indirect object which refers to the same person or thing as the subject."

However, in the following examples, the indirect object prefixes seem to have reflexive meaning, i.e. the dative prefix refers to the subject and thus is actually reflexive:

However, in the following examples, ni2 in terminative is used:

... and in the following examples ni2 with the directive case as an OO:

In the following examples, ni2(-te) appears in locative:

If the reflexive pronoun is used in ergative, the person may be treated as a non-human being as is also the case in all other usages of ni2. E.g. there is the following example; unfortunately, however, no inner personal prefix is shown:

SVC is also able to build a verb form with two reflexive ni2, one expressing a direct object and the other an oblique object. E.g. it can produce verb forms like *ni2-te-ni ni2-te-ni bi2-in-uš2, "He got himself to kill himself", although such double reflexives are apparently not attested.

Note that the reflexive verb form always uses the singular verb base. E.g. we have ni-te-ne-ne in-gub-be2-eš, "they placed themselves", lit. "they placed their self", with the singular verb base gub. By contrast, English "they placed them" is rendered as in-sug2-ge-eš with the plural verb base sug2.