Good news for those of us who are indiscriminate fans of ancient languages; I quote the Altorientalisches Institut – Universität Leipzig’s Facebook post:
We are delighted to announce the launch of a new long-term dictionary project!
The Leipzig Akkadian Dictionary (LAD) at the Saxon Academy of Sciences and Humanities in Leipzig will start in January 2025. The 17-year project aims to create a new, up-to-date digital online dictionary of Akkadian.
The existing major Akkadian dictionaries, the Chicago Assyrian Dictionary and von Soden’s Akkadisches Handwörterbuch, are outdated. Numerous cuneiform texts have been published since their completion, containing new words and facilitating more detailed and precise descriptions of known words.
The LAD will collect the vocabulary of Akkadian in its entirety. It is a reference dictionary that not only translates the words into English, German, French, and Arabic, but also documents their contexts, uses, and etymologies. The existing print dictionaries will be digitized and integrated into LAD. Links will lead to glossaries and indices of other online projects. The digital publication is based on a database structure and allows the vocabulary to be analyzed one corpus at a time rather than alphabetically. The first intermediate objective is to analyze the vocabulary of Akkadian literary texts (including royal inscriptions).
The project is headed by Michael P. Streck at the Institute of Ancient Near Eastern Studies at Leipzig University.
The full press release (in German) is here.
The digital publication is based on a database structure and allows the vocabulary to be analyzed one corpus at a time rather than alphabetically.
That’s a very peculiar sentence. It seems to say “water is wet”, and also to make a bidirectional downward comparison.
The digital publication is based on a database structure …
Every book is a structure, and its content is the data. So a book is a form of structured data. Any kind of structured data is a “database”.
The point of having “structure” is to help a reader find things. Typical structure elements of a book are chapter divisions, page numbers, page headers, table of contents and one or more indexes.
A digital “database” has all these elements and more, just under different names. For example, each “corpus” might represent a chapter. So you can focus on one corpus like you would focus on one chapter.
Nothing new there. “Databases” and books are both organized structures. Water and milk are both wet.
… allows the vocabulary to be analyzed one corpus at a time rather than alphabetically
Is there something old-fashioned about working “alphabetically” ? It seems to be implied that before this new “database”, people had to search on average through half the pages of a printed dictionary to find a given word, because the dictionary words were not printed in alphabetical order – and there were no indexes, page headers or page numbers.
Apart from that, it sounds very exciting for Akkadian enthusiasts. Implementing it is bound to cost a passel of money.
Agreed on both counts! I hope the money doesn’t run out in the coming age of austerity…
I’m no expert, but I would imagine that being able to search by corpus is a big plus in the case of a language with two thousand years of recorded history and two major geographical dialects even before you get into its extensive use as a L2 international lingua franca. You may very well be specifically looking for what some Canaanite scribe in the latter second millennium meant by some word, for example, rather than some Babylonian in the time of Hammurabi.
It somewhat throws me that modern Akkadian dictionaries/vocabularies are arranged in alphabetical order, instead of by root consonants. If you want to look up a verb, you need to vocalise it right before you can find it, which always seems weird to someone like me who is really only used to looking up Hebrew, Aramaic and Arabic words in dictionaries of Semitic languages. Presumably this is no big deal for your actual Assyriologist, though.
So that’s what they were deprecating !
modern Akkadian dictionaries/vocabularies are arranged in alphabetical order, instead of by root consonants.
What, the consonant groups are not arranged in alphabetical order ?
Sure. But the vowels between them are, too. As the vowels in Semitic verbs aren’t actually part of the verb stem itself, but inserted in varying patterns depending on aspect and other things, this has a bit of the same discombobulating effect as you get from Bantu language dictionaries ordered alphabetically, where (for example) you’ll find Swahili watu under M, and vitabu under K.
Someone here was complaining that in a Welsh dictionary you find (for example) fam under M and fuwch under B, while I was stoutly maintaining that this is not a big problem in practice if you actually know how the Welsh initial mutations work. Presumably Assyriologists similarly don’t have any great trouble being able to trace (say) bullutum under balātum (though it does get worse than this …)
The digital publication is based on a database structure and allows the vocabulary to be analyzed one corpus at a time rather than alphabetically.
Every book is a structure, and its content is the data. So a book is a form of structured data. Any kind of structured data is a “database”.
I’ll bet they meant this definition from the OED:
“1962–
A structured set of data held in computer storage and typically accessed or manipulated by means of specialized software. Also in extended use. Cf. base n.1 I.ii.12c.
“A database in its simplest form typically consists of a single file containing multiple records (record n.1 A.I.9), each of which has data for a particular set of fields (field n.1 III.19).”
And I wonder whether they meant “one corpus at a time in addition to alphabetically” (for some value of “alphabetically”).
And I wonder whether they meant “one corpus at a time in addition to alphabetically”
I was presuming that they did. The advantage of a digitalised system is that you can have it both ways.
Come to think of it, one problem with organising an Akkadian dictionary by consonants alone, in the proper Semitic manner, is that the Assyrians and Babylonians had carelessly lost quite a few consonants. (I blame the Sumerians.)
I’d also add that, because Akkadian lost its exotic fricatives, even if you believe that consonant roots are the true and best atomic units of meaning in all Semitic languages, there are consonant roots that are only distinguished by their vowels. For example, a proto-Semitic *ʕa becomes /e/ in Akkadian, while *ʔa becomes /a/. So what were the roots ʕCC and ʔCC would, in Akkadian be eCC vs aCC. Alphabetical order lets you show that arābum and erēbum are two distinct words more easily. I guess you could order your dictionary by proto-Semitic roots, but that sounds actively cruel to the reader.
That said, The Etymological Dictionary of Akkadian orders its entries by root, which is handy for etymology, which I think is illustrative. A dictionary for presenting etymology chooses roots, while a dictionary aimed at, among other things, more general use and corpus linguistics chooses alphabetical order.
I for one welcome a dictionary that gestures in the direction of machine readability. Maybe one day we’ll be able to send queries directly to the LAD API.
Someone should start a collection of all the ancient near east open data lexicographical projects of use for the LAD: the Mesopotamian API Directory for the Leipzig Akkadian Dictionary, MADLAD.
I take “The digital publication is based on a database structure” to mean something a little more specific. The true inner form is a database of tables, rows, and columns, which are then assembled typographically into dictionary entries. In other words, it’s a content management system where the elements are not entire articles but rather the head words, stems, derivations, lemmas, definitions, usage notes, etc.
David Eddyshaw: aside from Akkadian and its writing system having lost some Semitic consonants, the other issue is that much of the Akkadian vocabulary is not Semitic and does not have a trilateral root. So organizing dictionaries around those roots was tried and rejected (there are handy lists of verbs organized that way)
Classical Arabic words for material culture have more loan words from the Greek than I had expected eg. qlm “reed, pen” from Greek kalamos not Akkadian qanû
Most by way of Aramaic, I suppose?
Y: Aram. qlm “pen” and trs “to shied” = Arabic turs “shield” (from Greek thureos “long shield used by European barbarians”) do seem to be attested as far back as Qumran, but basically in the Roman world Greek was the language of talking about how to make specialized things, just like in the early 20th century German was the language of chemistry and chemical engineering. So I don’t know whether borrowings directly from Greek or through Aramaic were more common.
I’m not an Aramaic philologist unfortunately and the Comprehensive Aramaic Lexicon is hard to interpret as to dates.
17 years? I wonder if I’ll be still alive (and sound of mind) by then. And such estimates are usually optimistic, especially in Germany; probably it will take at least twice the time. Especially at a time, when apparently many German universities want to shut down departments dedicated to so-called Orchideenfächer, because there is no job market for graduates anyway, and thus the money invested is simply wasted (this is from an article in a recent edition of Die Zeit.
Incidentally, the chief editor of the dictionary Michael P. Streck published the article “Großes Fach Altorientalistik: Der Umfang des Keilschriftlichen Textkorpus.” (and followups) which might be of interest to Hattics: Mitteilungen der Deutschen Orient-Gesellschaft zu Berlin 142 (2010) https://www.academia.edu/38124922/Gro%C3%9Fes_Fach_Altorientalistik_Der_Umfang_des_keilschriftlichen_Textkorpus_in_M_Hilgert_ed_Altorientalistik_im_21_Jahrhundert_Selbstverst%C3%A4ndnis_Herausforderungen_Ziele_Mitteilungen_der_Deutschen_Orient_Gesellschaft_142_2010_ersch_2011_35_58