A new term is beginning in Austria, and that seems as good a time as any to talk about some of the reasons why I don’t have much time for blogging. Since September I have been learning Sumerian. I could speak about the controversies about Sumerian grammar, or the difficulties of the script, but in this post I will try to talk about the intersection of both.
I have tried to write this post in plain English while following the format which is traditional in historical linguistics (such as writing * before an expression which is ungrammatical, misspelled, or hypothetical). I have also converted the special characters used in transcribing Sumerian such as š and ğ into their Latin equivalents (/sh/ as in “ship” and /ng/ as in “running”). I do not know if I have succeeded, but I hope that this post is at least understandable to people without a background in historical linguistics.
Sumerian cuneiform evolved out of a logographic accounting system which originally only had signs for common nouns and numbers. As time passed characters acquired a further phoenetic connotation, initially from the sound of Sumerian words which translated the various connotations of a sign. Because so many Sumerian texts are damaged and published as drawings, students have to become familiar with different forms of the script, whereas students of Greek or Latin can chose to read nothing but modern printed texts with a standard font. There are about 600 common Sumerian characters, so in this respect at least Sumerian is easier than Chinese.
Sumerian is an agglutinative language, one which builds words by sticking together short elements which stay more-or-less the same regardless of which word they are attached to. As always, that “more or less” is the rub. Sumerian cuneiform is partially logographic, and even when signs are used phoenetically, they probably only corresponded roughly to actual pronunciation. Our understanding of Sumerian is heavily influenced by lists of phoenetic values compiled by speakers of Semitic languages, and they may not have pronounced Sumerian exactly as native speakers did. Many signs seem to have the same pronunciation, which are indicated with forms like du5 (“doo five”) and è (“ee three”) when Sumerian is transcribed in Latin letters. This lets the reader identify the original character, because some words which were pronounced the same were written with different characters. In addition, some syllables seem to lose their first or last letter in some contexts.
Gudea patesi lagash.ke4 e2.gal mu.na.du3
Gudea, prince of Lagash (patesi lagash.ke4), has built him (mu.na.du) the temple (e2.gal).
Sumerian marks possession with an /ak/ after the word, and marks the subject of a transitive sentence with the suffix /e/. Since the clause “Gudea, prince of Lagash” is the subject, the /e/ very logically goes at the end of the clause, right after the possessive marker /ak/. For some reason which made sense to them, Sumerian scribes wrote /*ak.e/ as /ke4/, a sign which is almost identical to the sign for “house” /e2/. (Mercifully, later forms of the script make it easier to distinguish between these two signs).
The sign /ra/ is easy to recognize and often appears near the end of a line. It can mark a person who receives something (the dative case), …
For (ra) the goddess (dingir) Ningirsu
or link a word ending in /r/ with a suffix beginning with /a/ …
His/her (ani) god/goddess (dingir)
I already mentioned that the last consonant in a word tends to disappear. Suffixes are not immune. That gives charming forms such as e2gishgigir.ra. One might guess that this is “to (/ra/) the house (e2) of the gishgigir” but in fact it is “the house of (/a(k)/) the gishgigir,” the house of the chariot.
These suffixes are often absent when we would expect them. The dedicatee of an inscription is often named without a /ra/. One possible explanation is that scribes left these suffixes out when they expected the reader to understand them; another is that speakers left them out to save time and effort and the scribes spelled words as they were spoken. Sumerian grammar is simple, but it is difficult to nail down the rules, especially as the language changed over its long documented history.
Last, some string of characters have meanings which do not follow from the phoenetic or logographic meanings of individual characters. Shir.bur.laki is the city of Lagash. The noun pa.te.si seems to be the same as ensi, the ruler of a city. As fate would have it, most students learn to read Sumerian from the inscriptions of Gudea, patesi of Lagash, so they get very familiar with these strings.
Edit: A number of people seem to arrive here looking for advice on learning Sumerian. That is hard, and harder without a teacher and a group of fellow-students to commiserate with. I can recommend Konrad Volk’s A Sumerian Chrestomathy, the Electronic Text Corpus of Sumerian Literature, and the Electronic Pennsylvania Sumerian Dictionary as resources for beginners. Beginners will also want a grammar, but I do not know any which are set up for teaching beginners with the most common aspects earlier and drills to help them learn; there is a list here, to which advanced students can add A.H. Jagersmaa’s A Descriptive Grammar of Sumerian (Doctoral Thesis, Leiden, 2010). I am most familiar with Marie-Louise Thomsen, The Sumerian Language and Otto Edzard Dietz, Sumerian Grammar and the Liepzig-Münchner Sumerischer Zettelkasten.
Volk’s texts and list of common signs, the EPSD, and memorizing the verb prefixes and suffixes and the case markers and the pronouns from a grammar will get someone started learning Sumerian.