Is Sumerian related to Akkadian? Or maybe even its direct ancestor? I can find nothing on its genealogical classification, but the section about its grammer prevents me from thinking it is a Semitic language. But I can't be sure anyway. Perhaps someone can tell me?--Caesarion 12:41, 10 Feb 2005 (UTC)
from our article: Sumerian has been controversially identified as related to Tibeto-Burman (Jan Braun) and Ural-Altaic languages such as Hungarian (Miklos Erdy). (in other words, no, no relationship with another language is generally accepted, and it is certainly not Semitic). dab (ᛏ) 12:55, 10 Feb 2005 (UTC)
Did you know that About 125 Sumer words have Turkish origin ?
84.160.196.73 21:20, 13 Mar 2005 (UTC)
No evidence whatsoever. Just some pro-Turkish lunatic again whose pasttime involves trolling language articles and spreading his fanatical graffiti under an anonymous IP. Turkish is part of the Altaic family and Sumerian is described by linguists as an isolate language. There have been more "reasonable" attempts at connecting it to DravidianorElamo-Dravidian strangely enough, particularly under the controversial Nostratic grouping. However, no evidence as yet is sufficient to prove any relationship to a known language group. That's the answer we're going to have to accept until somebody serious comes along with scientific rigour to show us otherwise. --Glengordon01 13:13, 28 July 2006 (UTC)[reply]
You will always find a few apparent common words between two totally unrelated languages, by the laws of chance. Anthony Appleyard 09:13, 13 September 2005 (UTC)[reply]
Under "Classification", we are told that the ergative case, marked -e, is the subject of a transitive verb and that the absolutive shows no case ending. Yet in the "Grammar" section, the example shows "temple", the object of a transitive verb, with the ergative marker ".0", and the subject "Ur.Nammu" with no case at all, as if it were in the absolutive case. This seems to be a glaring contradiction. I made a provisional edit, but I think it needs to be looked at and clarified by someone actually knowledgable in the field.
It's a bit of a shame that the sample sentence here is not only unattested (as the article notes), but apparently also impossible as it stands. As far as I can tell, the ergative marker on the name Ur.Nammu always comes in the form of the sign ke4 (KID), which indicates that the name actually has a silent (or at least unwritten) k at the end that only becomes apparent when followed by a vowel. This would make sense for a name that is etymologically Ur.Nammu.ak, "Man of Nammu".
Strictly speaking, the sentence should really read something like Inanna, nin.ani.r, Ur.Nammuk.e e.0 mu.na.n.du, but then you'd face the inevitable questions about where the [k] came from. The whole problem could be finessed by adding lugal ("king") in apposition to Ur.Nammu and tacking the ergative marker onto that. The sentence would then be Inanna, nin.ani.r, Ur.Nammu lugal.e e.0 mu.na.n.du. MTCicero 06:21, 6 January 2006 (UTC)[reply]
"...second, the sparseness of linguistic data..." I've read somewhere that there are about 40,000 tablets written in Sumerian (which is a huge number.) The clay tablets proved to be best medium for preservation for thousands of years. I have a hard time to believe that the linguistic data is sparse. It is just that the interest is not there. I would speculate that once the IE linguists figured out that it is not an IE language (and that is why it was hard in the first place) they lost interest in Sumerian.
"...The Sumerian text corpus is huge in ancient terms; at least 100,000 documents have been excavated..." [[1]] AverageTurkishJoe
in ancient terms. that's nothing compared to later languages, considering that most documents just have a couple of words on them. dab (ᛏ) 08:49, 2 January 2006 (UTC)[reply]
I´ve just read the English version of the Sumerian language article (I wrote major parts of the German one.) And there are some problematic things.
So, I hope someone of the authors read (and answer) all this stuff. If nobody does riot against it, I will change all these bugs next week. Greetings, Thomas Goldammer, Feb 2, 22:54 CET.
I'll try to answer it. I'll start with the part about Suffixaufnahme, since I am personally responsible for it. I must admit that my only source regarding that phenomenon was Wikipedia itself and the links provided here. Obviously the very definitions vary, hence the controversy. But this source [2] claims that Sumerian does have Suffixaufnahme and apparently equates it with double case / case stacking or whatever you call the Sumerian principle. However, it is true that the other source cited in the wikipedia article [3] clearly differentiates between double case and Suffixaufnahme along the same lines as you do (while warning against confusing them in a way that suggests that this "confusion" does occur among linguists). As for the very entry in Wikipedia, it doesn't say anything about whether the head noun also receives the affix in question. Concerning suffixes/postpositions - well, they have been called suffixes very often in literature on the topic, and I don't find the distinction very meaningful to start with. Anyway, since the thing is obviously controversial at best, I'll remove it.
Concerning your other suggestions: Ur-Nammu used to be the standard version of the name, but I think you're right that ur-namma has gained acceptance. 2000 BC used to be the standard - very approximate - date, too; I don't know what people meant by 2300 BC, but I suppose it has something to do with the claims I've met that Sumerian was already dying out during The Akkadian Empire, as the Semitic element became more and more prevalent, and that its position in Ur III was more due to government policy. Whatever. ETCSL and others still use du3 instead of ru2, I really don't think ru2 is generally accepted as of yet. As for the "ventive" mu- and the other prefixes of the same type (i-, bi-, ba-, im-mi, im-ma- etc.), I suppose that Thomsen's (and maybe other people's) explanations should be mentioned, but I understand perfectly well why the contributor regarded them as neither definitive nor plausible. 90% of the usage of that "ventive" seems more or less unjustfied. My personal experience with Sumerian has made me an agnostic in these matters. Only ba- and mu- make sense - but rarely. I think you're right about ur-{d}namma-ke4 as opposed to ur {d}namma [-e]. The contributor assumes a genitive relation without a case marker, but in practice I find that ergative ur-{d}namma-ke4 occurs in ETCSL (although I found also the other possibility!). However, you're quite wrong as regards "du3"; I think it's obvious that the user who wrote the example in question wanted to provide the pronunciation and the actual morphological structure of the sentence, and not its transliteration! He wrote Inanna, nin.ani.r, Ur. Nammu.e e.0 mu.na.n.du His purpose had clearly never been to write: {d}inanna nin-a-ni-ir ur {d}nammu e2 mu-na-an-du3.
I'd say that portion of the article is written more like an essay than like an encyclopedia article. It shouldn't look like that, but I don't feel like re-writing it.
Thank you for your suggestions and explanations. Yes, perhaps interlinerization (or whatever it's called in English) would be enough, but such a radical change would be a lot of work, especially if no valuable information is to be lost, and I prefer the article to contain too much rather than too little information. If the essay were to be trimmed, I somehow feel that the trimmer would be obliged to provide a proper and detailed grammar overview in its place, and - as I said -I really don't feel like it right now. Anyway, I think it's decent as it is now - perhaps too large already.
Oh, yes - concerning the conjugation prefixes: I just noticed that also the 2005 overview by a certain prof. dr. Klausen that your German language page has a link to states only:
Konjugationspräfixe. Es gibt zwei Arten dieser Präfixe, die sich nicht wesentlich in der Funktion, aber sicher in der Herkunft unterscheiden. Jede finite Verbform benötigt einen dieser sechs Marker. Ihre Herkunft, Bedeutung und Verwendung ist bis heute nicht gänzlich geklärt, obwohl darüber sehr viele Arbeiten in den letzten Jahren geschrieben wurden. In der Übersetzung werden sie meist nur als ein Hinweis auf die 'finite Verbform' ohne weitere semantische Relevanz angesehen (das hielten schon die Akkader in den Bilinguen so).
It also calls -da etc "Kasussuffixe". So - well, it's all very messy.
Greetings!
There are recent works where the "conjugation prefixes" are described in a very detailed way. (My Sumerian teacher herself wrote something about them, but I don´t know the exact reference, I have to ask her.) Just take Edzard´s and Attinger´s grammars. But, of course, all these prefixes seem to have inobvious meanings, unfortunately.
Of course there are works where the prefixes are described in a detailed way. The question is how much of what they describe is actually certain and convincing enough to be written in an encyclopedia. I'm not sure that one should write much more about this than Kausen did. By the way, could you please consider the DR topic below? --85.187.203.123 20:14, 6 February 2006 (UTC)[reply]
Oh, you've already done it, thanks! I reckoned that's what the guy meant, but I wasn't sure..
The Sumerian phonemic inventory contains a cluster DR? May I ask for the sources for this statement? It occurs in the middle of the word, but so do other clusters, don't they?
The overview in ETCSL says "an early stage of the language is thought to have had a further consonant whose identity remains uncertain. In the period of the corpus this consonant appears to have merged in some contexts with d and in others with r, or sometimes simply to have been dropped." "The period of the corpus" starts in Gudea's times, if not earlier. So - shouldn't you add something like "at an early stage of the language" (how early, actually)? Or maybe your grammars tell you something else, since you prefer to use ŕ yourself? --85.187.203.123 20:35, 6 February 2006 (UTC)[reply]
Okay. Since the newest complete edition of a Sumerian grammar I currently have at home is from the 1950s -!!!- I'll have to leave that to your conscience. :) Although the guys in ETCSL are Sumerologists if there ever were any. Never mind. I only wish that we don't get a one-sided story in what is after all a very debated field. p.s. I'll think about the account. !grtngs! --85.187.203.123 21:28, 6 February 2006 (UTC)[reply]
Thank you, Garzo, for making the phonemic inventory clearer (and for de-vandalizing this discussion page <grrr!>). But are you sure that s/z/ś/š all were fricatives? (I reconstructed z and š to be affricates ([ts] and [tʃ] respectively), according to M.P.Streck´s 2005 reconstruction of the Akkadian sibilants.) Many greetings, --Thomas Goldammer 11:13, 10 February 2006 (UTC)[reply]
I just changed two small things. Now it´s perfect (at least for me, but, of course, open for further suggestions). OK, for s and ś, I have not yet an idea what was the distinctive feature (I guess I prefer +ant, as you wrote in the article, thus alveolar vs. postalv.), but they collapsed to one single phoneme not later than Ur-III.
Here comes a passage of my reconstruction paper (it is an old version without mentioning of ś (I was not convinced of its existence at the time I wrote it); you may delete it after reading (or before :-) ) if you want. I will finish a renewed version of it in some weeks):
"Since affricates, rather than fricatives, are very common in Akkadian, we have the problem to distinguish between these two groups of sounds in the Sumerian language. Probably, all four Akkadian sibilants were affricates (Streck 2005). But, as we know, there are only three sibilant signs in Sumerian. The lateral affricate was written by Š-signs, the other affricates by Z-signs. The S-signs were used for sibilantic fricatives, emerging from the affricates in certain contexts. Therefore, we have to assume, that S was a real fricative rather than an affricate in Sumerian. With the same argumentation, we have to predict an affricate pronunciation for Z. All Akkadian affricates were at an anterior coronal place of articulation, which is by far the most common place of articulation for affricates in the language world. With these facts, the Sumerian sounds S and Z were also at this place. But if we look at the Akkadian loans, we can see, that Sumerian S freely changes to Š in Akkadian. This can only be explained by the assumption of a non-apical articulation for S in Sumerian, because Akkadian deaffricated affricates are to transcribe as [s z], but both of them are apical and anterior, and there would no explanation for a change to non-apical Š, which was deaffricated to [ɬ]. Since the lateral fricative is a really uncommon sound in the world´s languages, it would be very unprobable, if Sumerian had it as its only coronal fricative. What remains is either a dental or an alveolar non-apical, and thus laminal, fricative. The dental fricative ([θ]) can not be excluded, because Akkadian had this sound in a former stage of the language, see Ehret (1995) for Proto-Semitic. The sound developed to Š later. So the variation of Sumerian S to Š in Akkadian could be a reflex of this sound change. The alveolar non-apical fricative [s̻], a rarely attested sound that occurs in Basque for instance (Hualde and Ortiz de Urbina (2003)), is also possible, because the noice of this sound is much closer to the lateral fricative than to the noice of [s] or [θ], which would also explain the variation to Š. I prefer the latter variant, because it is also very rarely, if ever, attested, that in a language with one single coronal fricative, this one is dental. Now we could adopt this reconstruction for Z, too. But than we would expect a variation to Š in Akkadian also for Sumerian Z, which does not occur. Thus, Z was really an apical alveolar affricate. Whether S lost its laminal articulation in favour to an apical sound [s], remains as unclear as the question, whether or not the sound became voiced in later Sumerian. Since laminal sounds are highly marked and voiceless obstruents are highly unmarked, the former change is much more probable than the latter one."
That´s enough of boring stuff about sibilants. For ḫ, I reconstruct two phonemes (but unsure if really distinguished), namely a velar and an uvular fric. But: Looking at all the data we have, I have to reconstruct a [h] (glottal fric.), too, which disappeared around Ur-III. (That´s why I do never use h for ḫ.). Edzard in his 2003 grammar also suggests the existence of [h] in earlier Sumerian. (I don´t want to bore you again with another section of my paper here, but shortly: There are some loans in western Semitic lgs. as haikal, Hideqqel etc. which force an explanation as above.)
Sorry, I always write too much... Many greetings, --Thomas Goldammer 13:44, 10 February 2006 (UTC)[reply]
Yes that´s another argument for ś = [ʃ]. In general, for me, voicing seems to be non-phonemic in Sumerian (until Ur-III, at least), because Sumerian B, D, G were changed to P, T, K in older Akkadian loans. (apkallu for Sum. ABGAL, and the like.) And, still after Ur-III, there was a neutralization in syllable-final position probably to the plain voiceless obstruents. (VC-signs don´t distinguish voiced/voiceless.) But that´s maybe another discussion.
Many greetings, --Thomas Goldammer 16:14, 10 February 2006 (UTC)[reply]
Speaking of voicing: Prior to phonemic voicing, wasn't there assumed to have been an aspirated/non-aspirated contrast in its place? (a vague memory from a vague article)If yes, why not write it? --85.187.203.123 23:12, 10 February 2006 (UTC)[reply]
You´re right. Most Sumerologists assume that. (So I do.) OK, I will add it. But now its nearly 1 AM and I don´t feel like doing it now. :-) Greetings, --Thomas Goldammer 23:47, 10 February 2006 (UTC)[reply]
And, concerning the affricates - I am a layman, and I can only ask questions, take no offence: do MOST Assyriologists nowadays accept the affricate reconstruction for Akkadian and Sumerian? Was it popular before Streck 2005 - 'cause it obviously hasn't even had the time to become accepted since then? You know, people have been writing things like ""sh" as in ship" and all that for ages. Do even your latest Sumerian grammars talk about affricates? "z" as /ts/ does at least sound familiar to me, but "tsh"? Of course, you know that only what is generally accepted should be stated in articles, plus controversies if there are any. Not to mention original research. Greetings, --85.187.203.123 12:37, 11 February 2006 (UTC)[reply]
Thank you for the very detailed and interesting discourse, it does sound very logical and convincing and I will certainly take time to go deeper into it. However, I must remind you of something more basic: Wikipedia is about conventional knowledge, it shouldn't include "Original research" (http://en.wikipedia.org/wiki/Original_research). Now, you are citing your own reconstruction (I suppose it's published in a reputable publication, so it's fortunately not "original research" in the strict sense of the term). Anyway, you shouldn't include its results in an encyclopedia article, or at least not until a significant portion of the scientific community has accepted them. Or rather, you may include them, but not as the only truth. If both the Jagersma and Zolyomi believe(d) š to have been a fricative, then you should write that this opinion exists or is predominant, and the extent to which you mention your own point of view should depend on the extent to which it finds support in conventional Sumerology. The formulation "In der Semitistik setzt sich die Erkenntnis durch" indicates 1. that the author himself is a supporter of the affricate theory; 2. that he acknowledges that many others still aren't. I understand that Streck's arguments have convinced you, since you based your reconstruction on them; but have they convinced the vast majority of his peers already? If not, then they should be mentioned only as ONE existing idea, not as the ONLY one. And here, we're not even talking about the validity of Streck's Akkadian theory, but about your own Sumerian conclusions from it. Don't get me wrong, they certainly should be mentioned (although I guess a rigorous interpretation of "original research" rules would prohibit even that, but then, almost nothing meaningful could be written). But not as THE truth.
I'll try to find out more about this in a week or so, when I have time. But judging from your own words, the article shouldn't look the way it looks now. Again, I want to stress that I do not question your research, quite on the contrary, I am deeply impressed by it. Best regards, --85.187.203.123 00:24, 12 February 2006 (UTC)[reply]
Thank you. That'll do for a start. Did/do Jagersma and Zolyomi believe that š was a voiceless postalveolar fricative? --85.187.203.123 07:26, 12 February 2006 (UTC)[reply]
Thanks, then we can write that, too! :) Now it looks alright to me. Do write or correct if you find my wording inexact in some way!
The two genders have not only separate pronouns, but also separate PNG (person-number-gender) verbal prefixes and different marker options (thus, the plural marker -ene, the ergative -e and the dative marker -ra are restricted to the animate gender, while the directive -e and the locative -a are restricted to the inanimate gender). See [4] for facts and current terminology.
You can say that the verbal PNG prefixes ("infixes") are in fact pronominal, but that is not the same as being "pronouns", since "pronouns" is a word class and not a morpheme class. Thus, the conjugation of verbs by means of suffixes (amo-amas-amat) does not involve pronouns, even though it does signal person and number. In Semitic languages and some Slavic ones, verbs have a gender distinction, too. The Sumerian verbal prefixes (-n-, -b-, not to mention -a-, -e- etc.) can't occur independently, they are part of the verb complex, hence no pronouns.
Even aside from the prefixes, the noun paradigms are different, and in one case the same suffix (-e) has completely different meanings depending on gender. --85.187.203.123 10:17, 24 February 2006 (UTC)[reply]
Does anybody know if there's any way to re-orient the photo so we don't have to crank our heads sideways to try and read the sumerian? 65.79.30.55 18:44, 23 May 2006 (UTC)Eric Smith[reply]
I seem to remember that the pronunciation of Sumerian signs was deduced from how they were used in Akkadian. If this is true, I am curious how any sounds could be assigned to Sumerian that were not also in Akkadian. For instance, if Sumerian had an ö sound, Akkadian would not have been able to render it accurately - this may tie in with the remark in the article that Sumerian appears to have had lots of homophones. Maybe this is a dumb question - my exposure to Sumerian was about 45 years ago (I studied for a semester under K. Oberhuber)... Jpaulm 18:06, 10 July 2006 (UTC)[reply]
"It jas also been argued that Sumerian was a constructed language, intended from the very start to be used for mainly religious, scientific or literary purposes, to be written or spoken by a select or initiated few, such as priests. This, it is argued, may be the origin of the traditional Babylonian and Assyrian usage of it referred to above." First of all, "has" not "jas". Second of all, where are people getting this nonsense? Third of all, has everyone with a modicum of sense finally left Wikipedia in disgust? I'm taking it out. You can battle with me here on Talk:Sumerian language about it. If Sumerian was really a constructed language, it's funny how natural it looks combined with the fact that it came in different dialects just like a natural language (Eme-gir and Eme-sal). --Glengordon01 13:17, 28 July 2006 (UTC)[reply]
Speaking of crazy... what happened to Eme-sal?? Do the editors of Sumerian language even know what they're talking about? Odd. I see it in another language (Emesal (swedish)) but can't find it in English! So does that mean that English-speakers are just badly educated? --Glengordon01 13:25, 28 July 2006 (UTC)[reply]
--Glengordon01 10:17, 16 September 2006 (UTC)[reply]
Hello, good evening. I've read in sumerian textes the number 3 as if it was a letter (du3, lah4 or e2). What does it sound like? And then, what's the correct pronunciation of『eme-ĝir』,『eme-ŋir』or "eme-gir"?KekoDActyluS 19:36, 31 December 2006 (UTC)[reply]