Cuneiform 102 · Chapter 02

Warm-up · 6 quick retrievals from earlier chapters
  1. 𒁉 [bi] — what does it mean?

    its · BI · its/that; [bi] (a vessel in origin)

  2. rope — draw it, then check

    𒂠 [še] · EŠ2 · rope; the terminative -še3

  3. 𒈨 — read it

    [me]rites · ME · rites, divine powers; to be

  4. 𒌨 [ur] — what does it mean?

    dog · UR · dog; in names, "hound of" a god or king

  5. lord — draw it, then check

    𒂗 [en] · EN · lord; high priest

  6. 𒈗 — read it

    [lugal]king · LUGAL · king

The case that hides

Three signs bought you the sentence’s skeleton — verb last, tags on the nouns, a doer that can be marked. This chapter’s grammar bite is smaller and sneakier: the tag that means “belonging to,” which is everywhere in real texts and which likes to vanish from the spelling the moment you stop watching it.

Three new signs

SignNameKeyReadsMeansNotes
𒊏RAtoward[ra] [ra]; "to, for" (a person) — the dative tag origin opaque — a grammar workhorse from its first appearance
𒂵GAmilk[ga] milk; [ga] the sign lists tie it to the milk jug
𒀝AKdo[ak] to do, to make origin unclear; ak is the case's sound

Five times each, aloud as you write: ra, ga, ak.

The genitive — the case that hides

English marks possession by changing the possessor: the kingthe king’s. Sumerian does something else entirely. It leaves the possessor alone and tags it, and — this is the part worth slowing down for — it puts the possessed thing first:

X Y-ak = “Y’s X” (literally “X of-Y”)

House, king-of: 𒂍 𒈗 e2 lugal(-ak), “the king’s house.” The possessed noun (e2, “house”) leads; the possessor (lugal, “king”) follows, carrying the genitive tag. Now count the signs: two — and the transliteration shows a tag the clay does not. That mismatch is not a typo; it is this chapter’s whole lesson, coming right up.

Now the honest part. Say that phrase aloud and you will notice the /k/ barely wants to exist: e2 lugal-ak tends to surface, and get written, as something closer to e2 lugal-a, or even just e2 lugal, with no genitive tag visible at all. The rule of thumb from the standard grammars is that the /k/ of -ak mostly surfaces only when another vowel follows it and gives it something to lean on; standing at the end of a phrase, alone, it routinely hides in the spelling. You will meet e2 lugal-ak far more often looking like e2 lugal-a than looking like its full form. Learn the pattern (possessed first, possessor-plus-tag second) and expect the tag itself to be shy.

Contrast that with a case you can now write in full, no hiding at all — the dative, -ra, marking the person something is done for:

𒈗𒊏 lugal-ra, “for the king.”

The dative says its whole name every time. The genitive, notoriously, does not. But when the shy /k/ does step into the open — when a vowel follows and gives it footing — the syllable it forms is exactly what your new sign 𒀝 writes. Keep that pairing: the case is /ak/; the sign is 𒀝; the clay shows it only when the sound has somewhere to stand. (𒀝 also works a plainer day job as the verb “to do, to make” — and it has one more surprise waiting for you at the end of chapter 05.)

The genitive, schematically: possessed first, possessor-plus-tag after possessed X "the house" possessor Y + (ak) "the king('s)" 𒂍 𒈗 e₂ lugal-(ak) — "the king's house"
Schematic, per this course's habit — the parenthesized (ak) marks the tag's honest instability: present in the grammar, frequently absent from the clay.

Reading: milk names from Fara

The same Early Dynastic IIIa name lists that gave 101 and chapter 00 their readings include a small run of children’s names built squarely on this chapter’s new sign. Two neighboring entries:

𒂵𒅥𒅥ga-gu7-gu7"milk, eat, eat" — approximate; a nursery-ish name
𒂵𒅘𒅘ga-nag-nag"milk, drink, drink" — approximate; a nursery-ish name
SF 061, ED IIIa (ca. 2600–2500 BCE), Šuruppak (mod. Fara). CDLI P010652 via Nabu, urn:nabu:cdli:p010652 · license: attribution. Name translations are approximate by nature.

Look closely at the doubled signs and you will recognize old friends: 𒅥 and 𒅘 are 101’s sign-workshop compounds, built on the mouth — “eat” and “drink” doubled here the way Sumerian doubles a verb for repeated or ongoing action. The mouth eats and drinks again, twice over, and somebody’s baby got a name out of it. Hold the milk loosely, though: 𒂵 has a second, grammatical job waiting in chapter 09, and when you own it, these same two names will open a second — and better — reading. The clay keeps; the course will come back.

Reading: a name with a hidden genitive

One more Fara name, and it puts this chapter’s grammar bite to work without a single spelled-out -ak:

TSŠ 0181, ED IIIa (ca. 2600–2500 BCE), Šuruppak. CDLI P010772 via Nabu, urn:nabu:cdli:p010772 · license: attribution. Name translations are approximate by nature.

e2-gal, “big house” — the palace — is 101 chapter 07’s word, and here it sits as the possessed thing a man belongs to: not ur e2-gal-ak spelled in full, but ur e2-gal — “Ur,” of the “palace,” tag hidden exactly the way this chapter warned you it would be. Ur, the classic 101 sign for “dog” doing service as “man/servant” in names, plus palace, plus a genitive you have to supply yourself. That is the whole game from here on: knowing a tag is there even when the clay declines to write it.

Drill before you go on

  1. Write 𒊏, 𒂵, 𒀝 from memory; say ra, ga, ak.
  2. Compose in transliteration, then in signs: lugal-ra (“for the king,” dative, fully spelled), and e2 lugal-ak (“the king’s house,” genitive — then say it again the way the clay usually would, with the /k/ dropped).
  3. In the two milk-names above, name which sign is doubled for repetition and which two 101 compounds are hiding inside them.

Next: say it twice — doubling, and the abstract nam-.

Read it cold

This chapter's last reading, once more — script only. Read it aloud before you unfold.

Check yourself
TSŠ 0181, ED IIIa (ca. 2600–2500 BCE), Šuruppak. CDLI P010772 via Nabu, urn:nabu:cdli:p010772 · license: attribution. Name translations are approximate by nature.