SumerianGrammar-3.htm

Nouns

"Singular and Plural"

The Sumerian noun is not inflected to show a plural, and is interpreted as singular or plural according to the context. However, to indicate an entire class of objects (Thomsen 1984:59), it is reduplicated:

kur-kur, 'all foreign countries';
contrasting with kur, 'one or more foreign countries' {Thomsen 1984:59}

In an attempt to force Sumerian into more familiar patterns, a number of associated phenomena have been, at one time or another, considered by Sumerologists as plurals:


-e.ne (**-ên-ê)


In order to understand what -e.ne represents, we need to be aware that a demonstrative pronoun of the form ne(-e) has been identified, supposedly with the meaning 'this'. It is also found in the form ne-en, which can only be ne + the nominalizer -**n(a). I believe nê represents, instead, the simplest form of the demonstrative pronoun without a specific deictic reference: simply 'one' rather than 'this one, that one,' or 'that one yonder'; and, as such, is a combination of PL NA, 'one' + ¿E, 'like', yielding Sumerian **nê, 'one'.

For Sumerian, we must also reconstruct a demonstrative **anê, composed of Proto-Language ?A, 'here' + the above-mentioned nê, meaning 'this one'. In Old Sumerian and archaicizing Late Sumerian, **anê¹⁷ + ê (from PL ?A-¿E, 'nearby'), i.e. **an-ê, written a.ne, is the "ergative" subject pronoun of the 3rd person singular, (loosely) 'he/she'. In later Sumerian, the place of **an-ê is taken by **ên-ê, which is analyzed as ?A-¿E, 'nearby' + **nê, 'one', meaning 'that (nearby) one (**ênê)' + ê, 'nearby', producing 'that (nearby) one ("ergative")'. A class designation, which serves as the 3rd person plural pronoun (really: 'all of that (nearby)'), is formed by simple reduplication: ê + nênê + -ê (ergative), yielding **ê-nê-nê-ê, written e.ne.ne, 'all those nearby', (loosely) 'they (ergative)'. But, an even fuller form, -e.ne.ne.ne is also rarely attested. The former form served for the possessive pronouns in the singular: anê + -î (-¿E), which yielded **-anî, written -a.ni, 'his/her'. With the plural formant, -(nê)nê, it produced -a.nî₅.nê (read by Sumerologists as -a.ne.ne), 'their' (a + nê + î + nê(nê)).

The so-called plural ending -e.ne is written identically to the "ergative" 3rd personal singular person pronoun (in plene spellings, -e.e.ne, strongly suggesting the length of, at least, the first vowel), sometimes with resumption of the final consonant of the noun (Ce-ne) but, after a vowel, simply as -ne (we can surmise that after -a or -ê, the result was ê; after -u, ü; after -i, probably ^yê). Written identically, it is nearly certain that it must be related morphologically. I believe it is simply a reduced form of the 3rd person plural pronoun ên(ê)nê (written -ê.nê) less the "ergative" formant -**ê, and thus has the basal meaning of 'those'.

Since we have the reading of ana and ana₃ for two different signs meaning 'one' as a cardinal number, it suggests that, in Sumerian, an early specialization of forms meaning 'one' based on **na was made for cardinal numbers while forms based on **nê were restricted to demonstrative applications. On the other hand, a compound based on ?E-¿E-NA, 'that one (yonder)', probably is the basis for Sumerian in₃, another reading of Jaritz Sign #1, also read ana, and meaning '(that) one'; exactly this formulation underlies IE *ei-n-, 'one', listed under *e-, 'that, he'; and Old Indian e:-na, 'he', shows that in this formulation it could mean 'he' as well as 'one'.



-hi.a


It is now generally recognized that -hi.a is not a plural suffix (Thomsen 1984:62-63) but rather a subordinate clause meaning 'which are mixed', composed of the verb hi, 'mix' + -a, the nominalizer ('this') of subordinate clauses.



(-)me.eš


In early Sumerian, this combination is frequently written meš₂, composed of the verb me, 'to be' + -eš = 'they are . . .'. The ending -eš is considered to be the indicator of the intransitive 3rd personal plural form of the verb so the full form, -me-eš would simply mean 'they are . . .'. An interesting sentence utilizing the full form is:

lugal-ra lu₂ mu-ši-re₇-eš-a lu₂ hi-hi-a-me-eš

The men who verily came up (-ra) towards (-ši-) (=against) the king were men, they are (those) who are always (or 'all') mixed (up{?}).


In late Sumerian, -me.eš was employed simply as a graphic sign of the plural to facilitate translation into Akkadian and eventually, other languages.



Nominal Case Formants and Associated Verbal Chain Adverbials

We have discussed above the case formants called by Thomsen (1984:88) ergative, absolutive, genitive, dative, locative, locative-terminative.

We will now complete the series by discussing the comitative, terminative, ablative-instrumental, and equative.

Comitative, -da, 'with'

The comitative, signifying 'with', is written with -da or -da₅ in Old Sumerian. Sumerologists have correctly looked to the meaning of the simple formant — Sumerian is, after all, an agglutinative language — and identified da as 'side', noting that the archaic sign (Jaritz Sign #629) actually depicts a 'hand with arm'. In my opinion, the basal meaning of this morpheme is that of PL T[?]A, 'hand'; and rather than comitative, the most literal meaning of the formant is 'at the hand of', i.e. 'at the side of': a lative.

This formant is very frequently apparently "resumed" in the verbal chain as -da-. This resumption should be regarded as basically redundant in the same way that we can say 'with' or 'along with' in English.

^(DIG2[~]IR)Nin.sikil.a-da a₂ mu-da-a(n)g³[~]

Verily, the arm (a₂) was bent along with ^(GOD)Ninsikila
(she was instructed by 'pointing out' things with the arm).


Sumerian -da also occurs in three important idiomatic expressions:


1. -bî-da, 'and (with its)': ^(ID2⁾Idigna ^(ID2⁾Buranun-bî-da, the ^(RIVER)Tigris and the ^(RIVER)Euphrates

2. -da nu-(better nÛ-)-mê-a, 'without (never being with)': ^(DIG2[~]IR)En.lil₂-da nu-mê-a, without Enlil

3. as a non-redundant part of the verbal chain, -da- has been supposed by Gragg (Thomsen 1984:226) to convey the idea of 'able to', which is termed "abilitative". The examples given might be interpreted in that way although, in the example given immediately below, the distinction between 'may' and 'can' ("I can divide the estate"; better: 'I may verily divide the estate.'[?]) seems not to be a strong consideration for the translator.

Personally, I do not favor this translation. It seems much more likely to me that this construction, -da- + Verb is an expression equivalent to our prefixing a verb with 'co-' in the first example given, with the 'co-' referring to the recipients of the apportionment. Since ba has the basic meaning of 'split apart', perhaps the -da- indicates apportioning in the sense of multiple 'splittings apart'.


ê₂ mu-da-ba-e-e[n]

The estate is verily something to be (co-)apportioned.



In another example given, -da- seems to refer to a missing communication:


kin.gi₄.a ( . . .) šu nu-mu-un-da-an-gi₄-gi₄

The emissary (messenger) is never verily coming back with (-da-)
(it) (-(u)n-) (in) hand (=he does not convey the message.).



4. There are also apparent examples of -da-, 'with', being used incorrectly in the verbal chain where, because of the meaning, we might expect -ta-, 'from', the "ablative-instrumental", discussed briefly below.


5. There are also two supposedly related forms of this formant:


a. As a part of the verbal phrase, -de₃- (better -dî₇-; also -di-, and -de₄- in Old Babylonian literary texts) is found. Sumerologists believe that this formant is due to vowel harmony, i.e. the adjacency of a front vowel, which subsequently disappears after modifying -da- but there are attestations which strongly suggest an interpretation of 'to the side', which we would analyze as having resulted from PL T[?]A, '(at) hand' + HHE-¿E, 'go-ing (to)':


u₃-na-de₃-(better -dî₇-)dah

Now (-na-) (there) is again (u₃-) adding (-dah) to the side (-dî₇-) (of it).



b. This formant, -de₃, also occurs after the Verb + Pronominal Possessive in the Old Babylonian period, as in:


ka₂ e₂.gal-la-še₃ g²[~]in-a-(n)g³[~]u₁₀-de₃ silim-ma lugal-
(n)g³[~]a₂-ke₄ en₃ li-bi₂-(better pi₅-)-in-tar

While I approached the gate of the palace, a call (en₃) regarding the health condition (silim-ma, 'what was healthy') of my king was not directed around (-bi₂-[better -pi₅-]).



I believe it possible and probable on semantic grounds that this -de₃ may represent rather -nê (which is written with the same sign: Jaritz Sign #339), adverbial 'inside, while', derived from PL NA, 'interior' + ¿E, 'interior-like=within'; but it must be admitted that this cannot be conclusively demonstrated at this time.



Terminative, -šê₃ / -Vš, 'toward'

The "terminative", signifying 'toward(s)' (in temporal phrases: 'until'), is written with -še₃ in Old Sumerian, and later appears also as -Vš. Sumerian šê₃ is written with Jaritz Sign #893, which has a multiplicity of readings and meanings, one of which is šê₁₀, 'excrement', a concept that is connected logically with 'separation' in a number of languages. The formant, it should be added, is considered by some Sumerologists to be actually "-eše". I believe this formant (and this noun) is ultimately referable to PL S[H]E, 'separate', specifically as S[H]E-¿E, 'separating (for)'.

This formant is often apparently "resumed" in the verbal chain as -šê₃- in Old Sumerian (and later as -ši-) but it is interesting to note that with verbs like g²[~]in, 'go', which imply directional movement, it often is not "resumed". This suggests that, when it is present, it is not so much resuming the element of Noun + -šê₃ but rather modifying the verbal meaning directly.

uru-šê₃ (n)g³a₂-ê ga-g²[~]in

I want to go toward the city.


It is also interesting to notice that -še₃- / -ši- in the verbal chain frequently does not resume Noun + -še₃:

tur du(n)g³₄-ga-zu mah du(n)g³₄-ga-am₃ šu ba-a-ši-ib₂-ti(better ti(n)g³[~]_x)

A small thing spoken by you: it was something important spoken (that) here had been pressed towards slightly (reached for) with the hand (=accepted).


And, in combination with postpositions:

lugal-ra dumu Adab^(KI)(-a) min-am₃ mu-(un-)ši-re₇^(RE)-eš

It was the two sons of Adab^(PLACE) that were those who verily charged^(COME) against ('toward him', -(u)n-ši- + 'for', -ra = 'against') the king. (This might be rather: . . . charged toward him for the king.)


Ablative-Instrumental, -ta (better -tû_x / -tu_x[?]), 'from, by'

The sign with which -ta is written, Jaritz Sign #240, has preserved the Akkadian caption "ta-tû"; and, although *tu_x is not a recognized reading of this sign, I believe the Akkadian caption suggests it strongly.

What Jaritz Sign #240 clearly depicts is a 'drill being applied to a cavity'. This is an activity which I have discovered is associated with PL T[?]SO, 'swing (around)', and 'by swinging around, remove or be removed', which, according to the correspondences we have postulated, would appear in Sumerian as tu.

Because of IE *dheu-, 'run, run (intransitive), disappear, be in strongly whirling motion', that we are dealing we a derivative of PL T[?]SO-FA rather than simply T[?]SO, which would imply Sumerian -tû.

Accordingly, even though I believe the proper vocalization of this formant is -tû in the meaning 'from', I am aware that in those cases in which it forms a closed syllable in a sequence like -ta-an- in the verbal chain, the resumptive vowel is obviously -a, indicating that it must also have been read as ta in some context but this sign is rather unusual in that no underlying nominal idea has yet been identified to provide the basis for its adverbial and adpositional employments. In this context, the Greek postposition -then (cf. IE -*1. dhen, 'run, run (intransitive), flow'), 'from', is of interest.

I suspect that in the meaning 'by', -tû_x (for -ta) may also be an attempt to render a second -tu_x, an unrecorded alternate reading of ta₂, which pictures a 'hand and arm', and has the Akkadian caption "da-dû" (Jaritz Sign #629).


Equative, -gin₇, 'like'

Sumerian gin₇ is written with Jaritz Sign #785, which pictures a maul, which which various materials were shaped and formed. Its basal meaning is 'make'; and I believe gin₇ is but a transfer of significance into the area of human reproduction, and means 'engender', which corresponds to IE *1. g^en- (PL K[?]E-NA, 'start to penetrate'), 'engender'. The implication is that something that is like something else is a 'chip off the old block' in a figurative sense.


Zabalam^(KI)-ê u₈ sila₄ gur₅-a-gin₇ sig₄ mu-da-gi₄-gi₄

Zabalam^(PLACE) verily was screaming therewith like a ewe separated (from) a lamb.

1. The first person plural possessive is generally recognized to be -me (Jaritz Sign #889, which also reads men₃ and mi₃). In the cases of the second and third persons plural possessives, -zü(.nê).nê and -a.nî₅.nê, these forms seem to clearly be built on the singular possessive forms (-zü and -a.nî) by addition of the element plural element -(ê.)nê(.nê), 'those nearby'. On this pattern, the first person plural should presumably also end in -(ê.)nê(.nê), and would be -**mê.nê(.nê). Conversely, a first person plural of the form -**mê.nê(.nê), from which we can easily understand a stress-accentual reduction to -**mên₂, an alternate reading of the sign with which -me is written, implies a first person singular possessive form, on the strength of the pattern of -a.nî, like -**mê, a reduction from -**ma + -î (PL -¿E, '-like')

Now just this postulated form, ma, is found in the Emesal dialect in combination as ma.a-ra, 'for me'. Now while it is true that the m in the Emesal dialect has examples that suggest strongly that the g[~] in the main dialect (Emegir), there are also examples where EG m = ES m (EG alim = ES elim, 'deer') so there is no necessity to apply this relationship.

And, in fact, the previously universally accepted reading for the first person singular possessive, -mu, could be the result of a -**mê that has been re-formed to -mu or -**mü on the analogy of -**Sü.

In spite of the attestation of ES ma.a-ra, the normal form of the first person singular ergative pronoun is me.e paired with ze for Emesal. We could hypothesize an original **ma-e that has become me.e under the influence of the stress-accent on the final -e. If we were to read the currently accepted EG g[~]a₂.e as ma₃.e, we would have the postulated form.

The accepted reading of the second person singular ergative is za.e but in addition to ES ze, we have another variant in some Old Babylonian texts of ze₂.e (Jaritz Sign #276), which also has the reading Se/i (S = s with lower dot, "emphatic" s). Sumerian -zu, which is Jaritz Sign #6, also has the readings Su₂ and su₂ while za(.e), Jaritz Sign #957, also has the reading Sa.

Although we may never know certainly what the original forms were for these two pronouns, I would like to suggest an interpretation that attends to all of the data:


a. In the companion essays comparing Sumerian and Indo-European, I have compiled a PL / IE / Sumerian Correspondences between the two languages, in which I demonstrate that Sumerian m = IE m, and Sumerian s and S = IE s from PL S([H])O.

b. PL S[H]O, 'clansman', shows up in IE as -s, the secondary ending of the second person singular active verbal inflection; this same format would correspond to Sumerian Su or su. Since both of these are alternate readings of the Sumerian second person singular possessive -zu, and because the third person singular possessive is -a.nî, the base of which is immediately recognizable as an IE demonstrative, *2. an-, 'there, on the other side' (PL ?A-NA, 'here-one'), I believe the emendation of -zu to -s/Su is hypothetically justified. On the basis of Sumerian -a.nî, 'his', compared with a.nê, 'he (ergative)', I believe it is additionally justified to hypothetically assume that Sumerian formed a possessive with PL -¿E, '-like', just as IE forms a possessive stem in -i. I have already indicated that I believe that Sumerian u can represent /ü/, the product of the combination of u + î, the Sumerian reflex of PL -¿E.

c. Therefore, I suggest that the base form for the second person singular "pronoun" in Sumerian is s/Su, and that its possessive, written -s/Su₂, represents -**s/Sü₂.

d. PL M[H]A, 'occupier, resident', shows up in IE as -m, the secondary ending of the first person singular active verbal inflection; this same format would correspond to Sumerian ma. Since mu is an alternate reading of the Sumerian second person singular possessive, currently read as -g[~]u₁₀, and because the third person singular possessive is -a.nî, the base of which is immediately recognizable as an IE demonstrative, *2. an-, 'there, on the other side' (PL ?A-NA, 'here-one'), I believe the emendation of -mu to -mü is hypothetically justified. On the basis of Sumerian -a.nî, 'his', compared with a.nê, 'he (ergative)', I believe it is justified to hypothetically assume that Sumerian formed a possessive adjective with PL -¿E, '-like', just as IE forms a possessive stem in -i. I believe the earlier Sumerian first person singular possessive was -**mê (**ma+-î), which, read -me, is currently believed to be the possessive of the first person plural. By false analogy with -**s/Sü, it was changed to -**mü, which was written with Jaritz Sign #102, normally read -mu, which, however, Sumerologists currently read as -g[~]u₁₀. It should be noticed that -mu has no subscript, indicating that it is the commonest reading for this sign. The subscript ₁₀ for g[~]u indicates that it is a very rare reading for the same sign, and far less likely to be the correct reading for this common meaning..

e. If we believe the base of the first person singular is Sumerian ma, we have an immediate correlation with Jaritz Sign #458, which, though currently read g[~]a₂, also reads ma₃; and, Akkadian glosses on the reduplicated ma₃-ma₃, indicate it means 'to be, rest', meanings that correlate with our postulated 'occupier, resident'. The Emesal form, me.e, first person singular ergative, especially in view of second person singular ergative ze (also reads s/Se₂), therefore is easily understandable as an assimilation of the -a- of **ma to the following -ê, marker of the ergative.

f. The currently read za.e for the second person singular ergative then is also easily understood as an adaptation of an earlier **s/Su.e by false analogy to **ma₃-e as s/Sa.e (or in an even greater assimilation to s/Sê or s/Sê₂) in exactly the same process that transformed earlier -**mê into -mü (written -mu) by false analogy with -s/Sü₂, currently read -zu.

g. Plural forms for the first and second persons ergative do not exist in early texts; and we can assume, on the demonstration of their nominal bases, that they, like other nouns, had no regular plural ending. We have already discussed above the basis of formation for the third person plural ergative pronoun, ê.nê.nê. The pattern of the plural possessive pronouns seems to clearly be attaching -(nê.)nê to the singular possessive form: third person plural, -a.ni₅+-nê; second person plural, -s/Sü+-(nê.)nê. On this basis, the expected first person plural possessive would be -**mê+-(nê.)nê. Now the sign used to write -me (Jaritz Sign #889), also reads men₂.When the earlier singular possessive (-**mê) was in use, the plural possessive probably was -**mênê, perhaps shortened to -**mên; however, we -**mê was made into -mü by false analogy with -s/Sü, the even earlier form with no indication of the plural, -mê, would have been distinguishable from the more recent -mü, and a plural marker (-(nê.)nê) would have been unnecessary. My best guess is that Jaritz Sign #889 may have indicated either reading, and that both forms were in use for the first person plural possessive at various time periods.

-î, to designate a marû (progressive/imperfective) form of the active transitive verb;

-ê.nê, which indicated a plural ergative subject of a marû (progressive/imperfective) form of the active transitive verb;

and -eš₂ / -eš, which indicated a plural absolute subject (object) of a hamTu active transitive verb or of the passive transitive verb; or the plural absolute subjects of an intransitive verb of motion.

continue to

return to Sumerian Grammar (Part One)

return to Sumerian Grammar (Part Two)

go to first 35+ root cognates (1-35) ?

PL MORPHOLOGICAL ELEMENTS IN SUMERIAN

press here to see

to investigate these phonological correspondences in detail, see the

NOTATIONAL CONVENTIONS

Combinatory Modifications

for modifications of the vowels and consonants in combination, see the

Table of Modifications

PROTO-LANGUAGE MONOSYLLABLES

In order for readers to judge the semantic plausibility of the analysis of Proto-Language (PL) compounds suggested here, I am including access to a table of Proto-Language monosyllables and the meanings I have provisionally assigned.

Most assignments can be exhaustively supported by data from actually attested forms but a few animates are very doubtful; and this list does not represent the "final" solution of these questions, which will only be approached when other scholars assist in refining it.

Patrick C. Ryan

Summer 1998

17. A simpler form (without -î) of basal anê may perhaps be seen in a.na, 'what? ('this one?') — if not â.na from PL ?A-?A-NA, 'here?-this one' nor simply from PL ?A-NA, 'here?-one', which is perhaps the most likely analysis)'.

18. But cf. -a.nê₂ (as if from a.na + -î), reconstructed by Falkenstein (1978, I: 233).

19. Reconstructed as -bê₂ by Falkenstein (1978, I: 233).

20. "2. Das auslautende -n der Personenzeichen der 1. und 2. ps. sg. und pl. schwindet in der Sprache Gudeas ausnahmslos, falls es nicht durch ein vokalisch anlautendes Suffix — dieses kann nur das Substantivierungssuffix -a und die 3. ps. sg. der enklitischen Kopula -àm (s. § 45) sein — geschützt wird . . . (¹) Das nicht geschützte auslautende -n der Personenzeichen der 1. und 2. ps. sg. erscheint erstmals in Texten aus der Zeit der III. Dynastie von Ur . . ."

21. A view of some older Sumerologists.

22. In view of the etymology of šig(u), I believe that ge₄, which is a reading of Jaritz Sign #605, is a scribal error for Jaritz Sign #131, of which #605 is the shaded (gunu) variant. #131, in addition to reading gi, has an Akkadian reading of gu_x, which is unacknowledged by Sumerologists; both signs mean 'turn, twist', properly the meaning to be associated with #605, which is the meaning we attribute to PL K[?]O. K[?]O would have the Sumerian reflex gu, the same reflex as we would expect from K[H]O, the final component of šig(u); accordingly, we believe the correct reading here is gu_x, standing for **gü, the phonetic result of gu + î.