a) In any transitive verb construction, the completion of the idea of the verbal action, which produces a permanent affect, is the ultimate goal of the action. In Sumerian as in other related languages (PIE and Egyptian), this is most easily marked by verbal complete reduplication: gul, ‘destroy', i.e. ‘performs an act leading to total destruction of'; 'gul-gul, ‘accomplishes the act of total destruction of'. This type of reduplication is recognized by Sumerologists as ‘emphatic reduplication' and by PIEists as well; and it is equivalent to nominal complete reduplication in Sumerian: kur-kur, ‘all the foreign countries', from kur, ‘foreign country'.
-
a)) Complete reduplication has been accepted in Sumerian as a method to mark ‘repeated acts without the implication of achieving the ultimate goal of the verbal action'; in other words, it is considered a category of marû reduplication.
-
a))) Thomsen (1984: 304), in my opinion, incorrectly lists gul-gul as a marû form of gul, ‘destroy', citing the phrase: ". . . e3 gul-gul-lu-de3 . . .", . . . (that its) house(s) shall be destroyed . . .". We analyze this differently, as an emphatic reduplication: gul-gul-, ‘utterly destroy'; (-l-)-u-, frequentative; -d-, future; -e(3), ‘for' (see below)(2).
-
a)) We find partial reduplication of the basic verb as the marû form listed by Thomsen for kur9 (hamTu); ku4-ku4 (marû); ‘enter'; both are written with the same sign; i.e. they are both recorded readings of the same sign (Jaritz #99, derived from one of four archaic signs rhat depicts a ‘sprout coming out of a hole' or ‘being tall in a hole'). In order to fully understand all the ramifications of the situation, we must look at both the hamTu and marû manifestations of this verb.
-
a))) An example is the (hamTu) phrase rendered by Thomsen (1984:142) as:
'I entered'
In the hamTu third person singular, Thomsen has:
'he entered'
Why should the final -r of the verb, which is resumed by -re- in the first two persons of the singular, disappear in the third person? In other words, why is this not read kur9 for the third person singular? Does Sumerian not have enough homonyms as it is? Some Sumerologists have theorized that many final consonants are subject to deletion if not followed by a vowel as a part of an inflection added to the verbal root. Even Thomsen seems skeptical of this when she writes: "As a rule, we may perhaps presume (emphasis added) that the final consonant of a verbal stem is dropped in the marû reduplication. . .". And this is a cautious restatement of a principle that other Sumerologists have taken farther with, for example, the third person singular form cited above.
I think there is a better explanation (that allows the integrity of the verbal root to be maintained) than that of this final consonant being elided if no voweled inflection follows.
I propose that the third person singular hamTu form should be read:
'he entered'
First, let us remember that the reading ku4 is justified on theoretical grounds alone. If we operate under different premises, and propose, as we have, a reading of kur9, the signs would be exactly the same.
I propose as a counter-theory that the reason the final consonant is resumed in the first and second singular forms (as well as throughout the plural) is not to indicate retention of the final consonant of the root which would otherwise be elided but rather >u>to indicate a shift of the stress-accent to the second syllable:
'I entered'
One of the reasons that has been advanced by the proponents of the ‘disappearing final consonant' theory is that, in this verb, the marû form is a reduplication: written kur9-kur9. In marû forms with inflection, however, the final consonant is not resumed, hence, must have been elided:
'I enter'
The better explanation, in my opinion, is that Sumerian followed a pattern of reduplication known from PIE, namely that the first element of the reduplication receives the stress-accent: ‘kur9-kur9; and, as in the case of normal (non-emphatic) PIE reduplication, the final consonant of the first reduplicated root is rarely retained but, in verbs, generally elided: this would yield ‘ku4-kur9. That this was true of Sumerian as well is suggested strongly by the transcription babbar, ‘(being) white', for what we think we know is the result of a partial reduplication(rather than assimilation) : *barbar, presumably a reduplication of *bar11, ‘*white', another reading for the same sign (Jaritz #684). Now, it is always possible that this is an emphatic, complete reduplication, and that the missing final -r and presumed gemination for the b is a result of assimilation rather than elision; if this is true, the form suggests a stress-accent displaced to the second syllable.
The proposed, new reading of the first reduplicated sign kur9 has the advantage of explaining how the same sign came to be read ku4 as well as kur9; it would be the actual reading of the first element in a partial reduplication, with the sign retained to provide semantic continuity. Interestingly, as mentioned above, intensive reduplication is usually a complete reduplication as has been proposed for Sumerian, and is also known for PIE; and can be theorized for Egyptian (ex. sk.j, ‘destroying [‘is causing to perish']'; sksk, ‘[probably ‘utterly'] destroy'; sk.j, ‘is perishing [‘is being wiped off']', is probably a durative passive derivation from sk (3), ‘ wipes off [‘to dry']'). For anyone who believes that this development of meaning is strained, consider the usage of ‘liquidation' and ‘eradication' in modern English - being ‘liquified' or ‘erased' is quite a fatally serious matter.
If this analysis is correct, the reason the final consonant is not resumed in the marû reduplicated form is that the stress-accent, which, uninflected is on the first syllable, only shifts to the second syllable when combined with a voweled inflection (and not to the third syllable); consequently, final consonant resumption for the third syllable is inappropriate:
'I enter'
While in the third person singular, the reduplication, by itself, indicates a stress-accent on the first syllable so that no other stress-accent marker is necessary:
'he enters'
In the oldest recorded Sumerian, the third person singular of the marû form for non-reduplicating verbs shows -e with resumption of the final consonant of transitive CVC verbs: -sar-re, ‘he writes'. The indication we have from later Sumerian is that object elements, if present, precede the verbal root in a transitive employment of the verb (i(-)b2-sar-re. Accordingly, at least in this third person singular form, we can assume that, in the absence of the reduplication to indicate marû, i.e. this element (-e) indicates it. If my analysis of the vowels in Sumerian is correct, e can only occur as a reduction of the diphthong ai, a rather common phenomenon in the languages of the world. We further assume that, because of its derivation from a diphthong, that e is long, i.e. *ê, and can only be long. In originally CVCV nouns and verbs which bear the stress-accent on the first syllable, any trace of the final short vowel has disappeared. Further supporting this view is the fact that the final consonant is resumed, which we, above, assigned to an indication of a shift of the stress-accent to a non-initial syllable: thus sar-re represents *sar-‘rê; the long vowel has drawn the stress-accent from the first syllable. The PL monosyllable which seems the likeliest candidate for contributing an *a to the diphthong is HA, which we believe signified a large indefinite animate plural. If our assignment is correct, this would result in /a:/, the same formant we see in PIE collectives in -a:. This Sumerian *â could only be written a since Sumerian possessed no recognizable mechanism for distinguishing long vowels from short vowels (there is always the possibility that long vowels could be indicated by sign selection but this is difficult in extremis to demonstrate). It was, therefore liable to be confused with ?a or later â (both written a), a suffix indicating the locative from PL ?A, i.e. ‘top' becomes ‘on'. While the locative maintained its form, to differentiate it from the large indefinite animate plural designating the marû inflection, the marû inflection was modified by adding i, resulting in the *ai mentioned above which became *ê. Because its meaning was roughly ‘many(-like)', it was admirably suited to be employed as the plural personal inflections of any verb.
On the other hand, we find -e as a suffix in the first two persons of the singular of the hamTu and marû forms of intransitive verbs, where any indication of the subject is placed after the verb while the third person has no suffix. There is no justification for a plural element in the singular so this -e must have a different origin. We can see better what it represents if we know that in the oldest Sumerian, ‘he' as an ergative subject was a-ne; it became e-ne in later Sumerian; and the rare use of a demonstrative -e is attested that means ‘this'. What seems to me likeliest to conclude is that in a-ne, the first syllable represent PL ?A, here interpreted as ‘here' so that a-n is ‘this one' (the second element is PL NA, ‘one'; and that in later Sumerian, this element, a was combined with i, ‘-like', to produce *ê, written e, i.e. ‘here(-like)', corresponding to the demonstrative e. In the first two persons of the singular, then, the suffix -e designates the first and second persons as ‘here', i.e. participants in the conversation (without distinguishing farther between them), opposing the third person, which, also bears -e if the verb is not reduplicated to show the plurality of the verbal idea in the third person singular transitive marû.
Thus, only a later misunderstanding of the logic of the system could result in the addition of -e to a reduplicated third person singular in any of the verbal categories since the plurality of the verbal idea was already contained in the reduplication; and the third person singular was not a part of the speech situation; and, being singular, indicating the plural nature of the subject was not appropriate; The -e would be mandatory in any third person singular marû that was not reduplicated; and unnecessary in any third person singular hamTu form. With the phonological identity of these two elements (speech situation and verbal plurality), it is hardly to be wondered that later scribes were inconsistent. In the first two persons plural of unreduplicated marû verbs, both the verbal plurality e and the speech situation e would have been theoretically called for but if both were employed, that employment is invisible since they apparently coalesced as e. This is probably the reason why, after Old Sumerian, unreduplicated marû verbs added -n to the first and second singular as well as plural in an attempt to mark the speech situation (conversation participants) in addition to the- e marking verbal plurality.
Two final elements may be mentioned for the sake of completeness. In the third person plural of transitive marû verbs, in the final position of the inflection which theoretically called for a marker of speech participation, a marker to indicate plurality for the ergative subject was added. The nominal plural -e-ne (‘many-animate subject-ergative') termination, which was abbreviated to -ne was borrowed and employed so that the combined inflection became -e (verbal plurality) + -ne = -e-ne (ergative subject plurality) for unreduplicated verbs. For those with reduplication, only -ne was necessary.
For transitive hamTu verbs, in the third person plural, a prefix indicated the ergative subject (i(-)n-zig-). In the next to final position of the inflection, which theoretically called for a marker of plurality of the subject, that was indicated by -e-. In the final position which called for a marker of object plurality, the inanimate nominal plural suffix seen combined with me, ‘to be', in -me-eš was employed. This -eš is related to aš, ‘one, unique', (probably ‘those unique ones'), mentioned above. The idea seems to be that individual inanimate objects are being contemplated. This e and eš combine into -eš.
For intransitives, both hamTu and marû, in the inflection for the third person plural, we expect, a marker for subject plurality in addition to a marker of verbal plurality, either in the form of reduplication or e for the marû forms. The marker we find in this position is surprising: -eš. This marker, eš, as mentioned above, is essentially inanimate, and intrinsically unsuited to convey the plurality of an animate subject for an intransitive verb. The only possible reason I can think of for its use here is to differentiate between intransitive verbs inflections and the inflection of the third person plural of the transitive marû verb. In any case, we have -eš , which represents either e + eš = -eš or simply eš. The -eš of the oldest inscriptions is -eš2.
In later Sumerian, additional pronominal elements which were not present in oldest Sumerian apparently, were employed to differentiate the first (de3) and second persons (ze2) of the plural but though they are inferred for older Sumerian, they are not attested, and beyond the scope of our discussion. Later yet, -en was added to these elements, presumably representing an abbreviated form of -e-ne, ‘nominal animate plural.
-
a)) An example is gub/p (*gûp); with a singular subject, the hamTu form means ‘stands up/stood up/will stand up', as mentioned above; the marû form means ‘is standing/was standing/will be standing', or gub/'p+u (written gubu with one sign, Jaritz #410; also recorded: guba). It is important to note that this inflection can only be seen through recorded variants of the sign since final short vowels are not indicated in the system, and when the inflected verb is followed by additional elements: e.g. gu'b/pu, which yields gu'b/pun, written gu'b/p(u)-b/pu-un. To be explicit, bu has the recorded variant pu.
Current Sumerological theory explains the u of b/pu and the following -un as a consequence of the root originally containing a final u, to which the familiar -en is assimilated. This is certainly possible. But what militates against it is the existence of the reading gub/pa. The simplest explanation, I believe, is that this verb had the basic form *gûp, and that the readings with u and a represents stem formants in the form we should expect from PL FA/FHA (frequentative) and ?A/HA (stative). Since the great majority of verb forms in Sumerian are third person singular, we could tell only from the context if a stative *gû'pa were to be read, something along the lines of ‘the temple stands', with a condition implied. But the stative is also very suitable for what Thomsen calls "subordinate construction". The verbal root + a is used in constructions like the following:
inim dug4-ga, ‘the spoken word'
This, I believe, represents the stative since Sumerian has no discernable trace of a past or passive participle. Even when an ergative subject is included, the verb form is prefixed neither by i-, which is a particle meaning ‘then' (PIE e-; Egyptian j-), or a subject element (-n-) for a transitive verb). This is because no transitivity is associated with a stative except parenthetically. The frequentative describes an action that is repeated until a contemplation completion; the stative represents the state or condition achieved without reference to the preceding action,
END
Leaving reduplication and other less common marû devices aside, it is clear that Sumerian through sar, ‘has written' (momentary), and sar-‘re, ‘is writing' (durative), is following the pattern we have seen in PIE, Akkadian, Egyptian, and Arabic; and therefore, it is extremely likely that momentary/punctual and durative verbal ideas were principally distinguished in Nostratic by the stress-accent.
Since Sumerian separated from the evolving Proto-Language before it had reached the
stage of "nominative type" (Klimov), it offers support for the idea that variable stress-accent
(expiratory) was a feature of not only Nostratic (Afrasian and Indo-European) but also of the
Proto-Language.
BIBLIOGRAPHY
Edel, Elmar. 1955/64. Altägyptische Grammatik. Rome:
Pontificum Institutum Biblicum
Ehret, Christopher. 1995. Reconstructing Proto-Afroasiatic
(Proto-Afrasian): Vowels, Tone,
Consonants, and Vocabulary. University of California Publications in Linguistics: Vol.
126.
Berkeley and Los Angeles. University of California Press
Jaritz, Kurt. 1967. Schriftarchäologie der altmesopotamischen
Kultur. Graz: Akademische Druck-
u. Verlagsanstalt
Lehmann, Winfred P. 1974. Proto-Indo-European Syntax.
Austin, Texas and London: University of
Texas Press
Loprieno, Antonio. 1995. Ancient Egyptian: A Linguistic
Introduction. Cambridge: Cambridge
University Press
Thomsen, Marie-Louise. 1984. The Sumerian Language: An
Introduction to Its History and
Grammatical Structure. Copenhagen: Akademisk Forlag
Vergote, J. 1971. Egyptian (pp. 40-67) in Afroasiatic:
A Survey. Edited by Carleton T. Hodge. The
Hague/Paris: Mouton
HTTP://WWW.GEOCITIES.COM/proto-language/ProtoLanguage-MomentaryDurative.htm
PROTO-LANGUAGE@msn.com