Occasional commenter and full-time beerologist Martyn Cornell, aka Zythophile, has a long post that begins:
It’s a claim you will find repeated in dozens – possibly hundreds – of places: that the so-called “Hymn to Ninkasi”, a poem in the Sumerian language to the goddess of beer, at least 3,900 years old, known from three fragmentary clay tablets found in and around the ancient city of Nippur, which stood between the Euphrates and the Tigris, is “effectively a Sumerian recipe for brewing beer”, “the oldest beer recipe in history”, with a description of “the detailed brewing process” that “modern researchers have used to recreate Sumerian beer.” The Hymn to Ninkasi, according to one American publication, “served not only as spiritual homage but also as detailed brewing instructions for the beverage that came to be known as beer.”
Unfortunately, that is all total steaming nonsense.
I can’t do justice to it here, but I’ll quote a few excerpts:
Since we don’t actually know what Sumerian brewing and malting techniques were, our interpretations are pretty much guaranteed to be wrong, in part or whole, and we certainly can’t be certain of what we have got right. In particular, the various translations of the Hymn into modern languages vary considerably in what they claim the Hymn actually says, with differences that are extremely important in trying to work out from the Hymn how Sumerians brewed beer. So we cannot be certain that any given translator has got anything right at all. […]
Prince’s translation from not quite 110 years ago does not, in fact, mention beer or brewing at all. Miguel Civil, professor of Sumerology at the University of Chicago’s Oriental Institute, who made by far the best-known translation of the Hymn into English, in 1964, described Prince’s efforts as “less than satisfactory”, which is the polite academic way of saying “absolutely fakkin dreadful”. […]
Meanwhile, Civil’s translation seems to have become the one regarded, in the English-speaking world, as the definitive, accurate and only account there is of the Hymn in modern language. It is pretty much the only one ever mentioned in accounts of the Hymn today. However, there are at least two more translations, and crucially, they differ from Civil at vitally important points. One is by Walther Sallaberger, an Austrian-born Assyriologist, and professor of Assyriology at the University of Munich. His translation was published, in German, in 2012, in a collection of 20 essays in French, German and English in different subjects in the field of Assyrology, put together to honour Pascal Attinger, a Swiss-born peofessor at the University of Bern, and leading expert in the Sumerian language. Sallaberger calls Civil’s 1964 translation of the Hymn “fundamental for all subsequent work”. However, he says, “the interpretation of the Ninkasi-Hymne offered there contradicts in some respects all other, particularly documentary, evidence on beer brewing.” In other words, Sallaberger reckoned Civil had got his account of Sumerian brewing wrong. […]
Unfortunately, Civil’s translation, like Prince’s, contains lines that are complete nonsense, with one hugely important term, “bappir”, clearly a vital ingredient in the making of Sumerian beer, left untranslated because nobody, then or now, has any good idea exactly what bappir was. […]
There is, in fact, a third modern translation of the Hymn, into Russian, by Veronica Afanasyeva, a Sumerologist and poet born in 1933 who worked for the Hermitage Museum in St Petersburg. Again, Afanasyeva, like Civil and Sallaberger someone who had been studying the Sumerian language for decades, comes up with yet another translation differing in important places from the others. […] With the other important but disputed words in the Hymn, Afanasieva translates titab – “cooked mash” according to Civil. “spent grain cake” for Sallaberger – as сусло. This is where my complete lack of knowledge of Russian causes a big problem, because in Russian сусло can apparently mean both “wort” and “mash”, and I don’t know which one Afansieva meant. […] Finally, dida Afansieva translates as “noble beer”, very different from both Civil’s “sweetwort” and Sallaberger’s “dry beer”. […]
This considerable disagreement among experts as to how individual words, phrases and lines in the Hymn should be translated, even without the confusing and unclear description of the processes, means the Hymn to Ninkasi is, in fact, very little to no help in trying to work out how the Sumerians brewed beer, if we can’t be anywhere near sure what the Hymn actually says.
Indeed, the late Peter Damerow, of the Max Planck Institute for the History of Science in Germany, who died in 2011 aged 72, queried whether the Sumerians actually made beer at all. In a paper published in 2012, after his death, he wrote: “Given our limited knowledge about the Sumerian brewing processes, we cannot say for sure whether their end product even contained alcohol.” […]
But that was only a brief touch on the problems with the Hymn, and the need to finish off another couple of projects prevented me from pulling together a complete argument against the popular view of the Hymn until a month or two ago.
And then, just as I was about to start, along came a man considerably more qualified than me to write about the Hymn to Ninkasi, and the subject of beer and brewing in ancient Mesopotamia – Tate Paulette, archaeologist, and associate professor of history at North Carolina State University.
Paulette has written In the Land of Ninkasi, subtitled “A History of Beer in Ancient Mesopotamia”, just published by Oxford University Press (at least it’s published in the US: it appears to be not due out in the UK until November). It is, undoubtedly, the most important book published in the field of beer history for some time, because it looks at the whole 2,000-year story of brewing in the land between the Euphrates and the Tigris with the sceptical eyes of a scholar who is an expert in Middle East archaeology, knowledgeable enough about the ancient languages spoken in the region to be able to explain them to a lay person, and also knowledgeable enough about brewing to have participated in an attempt with professionals to recreate Sumerian beer (not the Anchor effort, but one with Great Lakes Brewing of Cleveland, Ohio).
Paulette has invented the word “šikarologist” to describe students of Mesopotamian beer, from the Akkadian word šikaru, meaning beer, the cuneiform for which looked like one of those pointy brewing jars laid on its side: 𒁉. I will definitely be stealing that. Strangely, a cognate of šikar in another Semitic language, Hebrew, the word shekhár, meaning “intoxicating drink”, is the root of the English word “cider”, via Greek, Latin and French. Shekhár is also the root of the Yiddish word for “drunk”: I remember my Jewish ex-father-in-law, who came from Vienna, staggering around at his 70th birthday party, surrounded by his extensive family in his large garden in rural Surrey, saying: “Ich bin so shicker!”
I was relieved to find that Paulette pretty much agrees with me that the Hymn to Ninkasi is not actually very helpful in working out how Sumerians really brewed.
There is much more about all the above topics at the link, and Cornell ends with his own “suggestions as to what might have been happening in a Sumerian brewhouse.” If any of this is intriguing, click on through, and I hereby fire up the Schwa-signal in the hopes that ə de vivre will take note and appear. (I reported on Miguel Civil’s death in 2019.) Thanks, Mike!
Based on this, is there any reason to believe that the majority of translations of Sumerian are accurate? I’ve always been a little skeptical (without knowing enough to have an informed opinion) of how Sumerologists had apparently gotten so good at translating a dead language isolate with so many different readings for each cuneiform sign.
It’s misleading to attribute the difficulty of interpretation here to the (fairly major) imperfections of the script. Assyriologists are very good at getting the right value of the many polysemous signs from context. (Japanese is just as bad in its own way, and nobody expresses serious doubt as to whether that can be read.)
The problem here is that we don’t know what the relevant technical vocabulary means.
Russian сусло is generally a filtered liquid (or perhaps, in applications other than beer brewing, a liquid to be fermented from which the coarse solids are removed in some other way, like by gravity?) but I am not sure the esteemed Sumerologist Veronika Afanasieva knew every detail of the beer-brewing technology. Or perhaps her intent was to avoid the Russian word for “mash” – затор – due to the simple fact that there are other, more common meanings of “затор” which would inevitably confuse a non-specialist. In contrast, сусло has only one meaning, and it was relatively familiar to the Soviet masses because it was sold in grocery stores as glass jars of concentrated kvass wort from which people would brew their own kvass.
In Afanasieva’s poem, сусло appears twice on two consecutive lines (Italicized below) – the liquid before being filtered through a bed of reeds, and after filtering. Therefore, both mash and wort are covered there AFAICT.
О Нинкаси, Энки-Нудиммуд – родитель твой,
Нинти, Абзу госпожа, – родительница.
Ты зерно проросшее большой лопатой сгребаешь,
С травами душистыми закваску в чане мешаешь.
Ты закваску ту в большой печи печешь.
Зерно для солода в кучи собрано, очищено.
То зерно для солода ты землей покрываешь, водой поливаешь,
Псам-сторожам стеречь его доверяешь.
Ты в котлы пивоварные солод с водой кладешь,
Волны падают, волны вздымаются.
На подстилки из тростника ты сусло кидаешь,
Ты спеченное сусло прохладою обдаешь.
Ох, и знатное пиво готовишь ты,
Мед, вино смешав, ты по капле льешь,
Пивоваришь воистину отменно ты,
Для цежения пива ты чаны берешь.
А цедильный чан так славно журчит,
На дне кадки огромной воистину ты сама.
Пивом цеженным в могучую кадку изливаешься ты –
Тигра и Евфрата наводнение.
PS: one particular source of Zythophile’s confusion might be explained by a simple cutting-and-pasting mistake: trying to make sense of Afanasieva’s lines 3 and 4, he pasted lines 5 and 6 into Google Translate instead…
Oops! Perhaps under the influence of his eponymous brew…
Surprisingly seemingly related: German vernacular “angeschickert”, meaning intoxicated. Presumably from Yiddish like a good number of words in that colloquial, often jocular register, like “Mischpoke”, “Tinnef”, “abzocken” (cf https://de.wikipedia.org/wiki/Liste_deutscher_W%C3%B6rter_aus_dem_Hebr%C3%A4ischen_und_Jiddischen).
I will from now on try to remember to write “angeschickert” as “ange𒁉t”.
This
Classification in Sumerian cuneiform and the implementation of iClassifier
Gebhard J Selz, Bo Zhang
Journal of Chinese Writing Systems
https://doi.org/10.1177/25138502231215875
is a stunning-looking account of the Sumerian classifier system (if you can get access). It reminds me those ants frozen in amber from the Jurassic.
a stunning-looking account of the Sumerian classifier system
Indeed. Abstract:
References in the classifier paper impressed me a whole lot, too. I had no clue that classifier systems exist in today’s spoken languages and have so many parallels with the first writing systems of antiquity!
Back to beer, for nitpicking sake, I also wanted to add that знатное пиво becomes “noble beer” only in Google Translate’s imagination. In reality it’s “remarkable beer” or just simply “good drink” (with the same stem as in verbs “to know” / “to recognize”)
Where are the ants frozen in amber ? Were they mentioned merely to lure people into discussing classifier systems ? The subject must be moribund if such wiles are needed to attract attention to it. Ants are interesting per se.
@ Stu
… answers to queries in the iClassifier may help to illuminate ‘the diverse mindsets of ancient Mesopotamia’.
I’m looking forward to the time soon when I can choose `mindset’ on my Sumerian fone app – i didn’t know such tools as iClassifier – [I don’t know what it does but the pictures are interesting] – exist. Sumerian is of course the perfect Petrie dish/ benchmark for this kind of classification and concordancing because it’s a complete (mod new archaeology) corpus.
The discussion of the [HIDE] classifier is almost lyrical, it subsumes lots of leather technology.
This is iClassifier, referred to above, a project dealing with the analysis of classifier graphemes in various scripts.
See also here, an article in the same issue as the Sumerian article.
Ants only evolved in the Cretaceous, and there’s very little known Jurassic amber. The ant in amber shown here is only Eocene in age – 50 million instead of 150 million years.
From your link: “In 1938, 12-year-old David Attenborough (brother of Richard who played John Hammond in Jurassic Park) was given a piece of amber containing prehistoric creatures from his adoptive sister;”.
I wonder how the sister managed to get her prehistoric creatures into the amber. Perhaps she merely supplied them, and someone else got them in.
She was adoptive. She could have been quite older.
See his memoir My Cretaceous Sister.
“This this book is dedicated to the Amber that came to me thanks to my Siblings, Pine Resin and Deep Time.”
Martyn hits on a lot of the problems of understanding a Sumerian text. The main one is that they assume a lot of cultural knowledge we just don’t have. Bappir, for example, is a big one. From administrative documents, we know it contains grain, is used in beer production, and can be stacked and stored for long periods.
The grammatical problem is that, given our current knowledge of Sumerian grammar, it’s usually the case that we can identify the participants in any given sentence, but who did what to whom is often up for debate. This is especially true when the grammar gets more complicated that “Dick sees Jane” (or “Lugalanda sees Baragnamtarak”) types of sentences. It’s also true that our knowledge of Sumerian grammar has come along way recently, so translations earlier than the 90s (yes, I still maintain that the 90s are “recently”) are often not to be trusted.
For example, the Stanza VI that he gives the wildly diverging translations for (where NOM/LOC means that the morpheme could either be a nominalizer, a locative case marker, or both):
munu₄ sahar ŋar-ra a sig₉-ga-zu
malt dirt place-NOM/LOC water put-NOM/LOC-2.SG.POSS
ur me-me nam gam-gam-ma-am₃
dog ?? lord? bend-REDUP-NOM/LOC-COP
Munu₄ unambiguously means ‘malt.’ Martyn makes a case that this is dry malt, but whether dry or green, we have to guess from context.
Saḫar usually means ‘dirt’ or ‘dust’ in a more material sense. It’s what Enki gets stuck under his fingernail and turns into a gala demon. ‘Ground,’ as a flat plane you would spread something like malt on is usually ‘ki.’ Additionally, if the malt was placed on the ground/dirt, saḫar should be in the locative case, saḫar-a.
Ŋar is ‘to put’ when you set something down onto a surface, not when you put something into something else (like putting malt into the ground). When you set bricks onto the ground, you ŋar them.
A means water. No one disagrees about this word.
Sig₉ is the usual verb for putting something into something else. When you put clay into brick-moulds, you sig₉ it.
Translators agree on a syntax where munu₄ is followed by two relative clauses, but differ over what role munu₄ plays in the relative clauses. Translators who put the malt on the ground have parsed this sentence as [munu(i) [(i NOM) saḫar ŋar-a] [(i) a sig-a]]-zu. “Your malt, which has been placed on the ground, which has had water put on/into it.”
Translators who put the malt under the earth parse the first relative phrase differently: munu(i) [saḫar (i LOC) ŋar-a]. Munu has been moved from a position in the locative case (which, as expected, leaves no trace of the case marking) rather than the nominative. This reading makes more grammatical sense. It explains why saḫar isn’t locative, and it explains why the text uses saḫar rather than ki. The substance saḫar is what you’d put or sprinkle over something, not the flat plane ki. Unfortunately, this reading doesn’t make much real world sense. It’s certainly not impossible that our understanding of Sumerian grammar is in error rather than our ability to imagine what this phrase means (it is the children who are wrong!), but this is part of the problem. Given our tenuous grasp of Sumerian syntax (and the possibility that the writing is just defective or contains an error), when do we let “real world plausibility” override what we provisionally understand as the rules of Sumerian grammar?
Briefly about the second line, ur is a dog. Me-me is perplexing. Me is a concept in Sumerian culture that means some kind of divine gift of an aspect of civilization. As far as I know, it’s never a verb, which it would have to be to modify ur. Nam alone usually means ‘prince/lord,’ but it can also be a verbal prefix, so you could also read nam-gam-gam-ma. Gam is to bend down, which translators understand as bending down in respect or bending down to protect. Without more research I can’t even hazard a guess what this means beyond what the translators have already said.
Thanks, I was hoping you could provide some details like that!
I still maintain that the 90s are “recently”
I fervently agree.
Would a possessive prefix really show up at the other end of the sentence from the possessed noun? Is it a clitic that attaches to whole clauses like ‘s in English?
Anyway, meme is what sheep make when they’re speaking Hattic. Dogs – not so much.
@DM
Yes, Sumerian contains very few (maybe none? That seems unlikely) true suffixes. It’s clitics all the way down. There’s also a construction where a possessive on a nominalized bare verb root marks the subject of a backgrounded action (something like “as you were sig9-ing), but in that case we’d expect “sig9-ga-zu-ne”.
“Recently” includes everything since 1453.
Sumerian contains very few (maybe none? That seems unlikely) true suffixes
There are plenty of languages in which virtually all morphology works via clitics rather than affixes, so it doesn’t seem implausible a priori, though it may well be implausible for Sumerian in particular (you would know …)
Partly it depends on where you draw the line between clitics and affixes.
Which always reminds me of
https://www.researchgate.net/publication/244421136_Cliticization_vs_Inflection_English_N'T
A certain kind of scholar of “Niger-Congo” makes out that the proclitic subject pronouns seen in a great many West African languages are actually subject prefixes, in line with their belief that proto-“Niger-Congo” was pretty much identical with proto-Bantu structurally. They are wrong.
My Kusaal grammar analyses two morphemes which are treated as suffixes in all previous accounts as actually being enclitics. One of them would actually be a suffix by Zwicky and Pullum’s criteria, but I am unmoved: I reckon the phonology makes my own analysis inescapable. The only alternative would be to set up a unique category for it of “suffix that behaves just like an enclitic object pronoun phonologically.”
I realize the nominalizing suffix I alluded to is a real suffix. It goes on the end of the verb and nothing else (although Sumerian is pretty strictly verb-final. I don’t know if you could test this without elicitation from a native speaker…)
Sumerian is odd in that it’s pretty robustly prefixing (even its limited derivational morphology is prefixing) SOV language, which is clearly evidence for a Sumerian–Dene-Yeniseian family, or at least an areal grouping!
Bantu is weird in having prefixing flexion (exclusively so in nouns, and mostly so in verbs) and almost entirely suffixing derivation (the only exception being tweaking noun-class membership for derivational purposes, which happens all over Volta-Congo.) The same is true of “Kwa” insofar as derivational morphology is still recognisable at all.
It seems pretty evident to me that this has come about historically via the turning of proclitic pronominals and particles into prefixes, but Bantuists are often remarkably unwilling to entertain this idea. The long shadow of nineteenth-century notions about the mystical superiority of inflexion to agglutination, I reckon.
There is at least one SOV Bantu language (Tunen.) Proto-Oti-Volta was probably SOV at one point, too, but Oti-Volta is thoroughly suffuxing throughout, except where noun-class prefixes have developed secondarily.