Sumerian/Grammar/Lesson Seven - Ergativity in Sumerian

Subjects and Objects
In English, we're used to talking about the subject and object of a sentence. In fact, almost every language in the Indo-European language family uses this mechanism in its syntax. Consider the sentence:

The student studied the Sumerian tablet.

The subject, in English anyway, comes just before the verb, and the object after. So here we have the student as our subject, and the Sumerian tablet as our object.

Only slightly more complicated is the distinction between direct and indirect objects. Consider the sentence:

The student gave the curator a Sumerian tablet.

Here, as before, our subject the student lies in front of the verb. But now, we have two noun phrases after the verb: the curator and a Sumerian tablet. In English, we call a Sumerian tablet the direct object (here, the object actually being given) and we call the curator the indirect object.

Hopefully you're familiar with these terms. If you're not, you can brush up with any decent English grammar or linguistic introductory text that focuses on English.

Transitive vs Intransitive
But there are other types of sentences too, ones that have no object at all. These types of sentences are called "intransitive", and sentences with objects are called "transitive". Consider the intransitive sentence:

The student napped.

We don't need anything after napped, as we did before with gave. These types of sentences are equally common in English. Again, though, we call the noun phrase before the verb the subject.

Agents and Patients
Ergative languages, in contrast, have another way to mark these roles in a sentence. They use "agents" and "patients" instead of subjects and objects. Consider, for instance, the transitive sentence:

The curator woke up the student.

In ergative terms, the curator is the agent of this sentence - the one doing the action. Also, the student is called the patient - the one to whom the action is being done.

So what happens in intransitive sentences? Let's take a look at the sentence:

The student awoke.

Here, the student is the one to whom the awakening is happening - and so we would assign the syntactic role of patient to this noun phrase.

So we can see how the "subject" in English becomes the "agent" of a transitive sentence, but the "patient" of an intransitive sentence! It's all in your perspective, I guess.

Split Ergativity
As it turns out, quite a few languages use ergativity. However, it's extremely uncommon for a language to be completely ergative. Sumerian, like most languages with ergative features, is actually "split-ergative", meaning that sometimes Sumerian uses ergative cases, and sometimes it uses the familiar "subject/object" cases.

We'll get into more detail in later lessons about when Sumerian uses the subject/object case system, but for now, we're just going to focus on the ergative aspects of the language.

A transitive example
As it turns out, we've already seen the case markers for the ergative cases, both in lesson three and in lesson six. Let's look at our example sentence from lesson six again, from "Enki's journey to Nibru":

Eridug.a e.Ø gu.a bi.n.du

In this sentence, the temple (e) is the thing being built, our patient. We see that the .Ø marker is used for the absolutive case. Remember, since this is a transitive sentence, the object of action is called the patient, and the case used is called the absolutive case.

Note, however, that the agent in this sentence is implicit - since we don't know who exactly the agent is, we just translate the subject into English as "he". In this sentence, we have no noun phrases marked in the ergative case. Below, we'll take a look at a sentence that has an explicit marker.

An intransitive example
Let's also look at an intransitive sentence, from "Enlil and Ninlil" (ETCSL number 1.2.1.91):

Enlil i.ĝen

We begin, as always, by analyzing our phrases. First up is the name Enlil, who was a deity in Sumerian lore. This phrase is unmarked, leading us to guess that it is actually in the absolutive case.

Our final phrase is the verb chain. The root of the chain is ĝen, meaning to go or to travel. i. at the front is just a filler vowel that Sumerian uses if a verb has no other markings, but we want to put it in the "finite aspect" (which you can think of as past tense). So we translate this verb simply as

Enlil went.

Couldn't be simpler! Now, Enlil being the subject of an intransitive verb, we can confirm our suspicions that it was indeed in the absolutive case, and hence unmarked.

An example with the ergative case marker
So far we haven't run into the ergative case marker, only the absolutive. That's because our sentences that had ergative subjects have all lacked explicit subjects to this point. Let's look at the following example, from "The Lament for Urim" (ETCSL reference 2.2.2.172):

Enlil.e ud.e gu ba.n.de

By now you know that the first thing to do is to analyze our phrases.

First up is Enlil.e. As you probably guessed, the .e is our elusive ergative marker! It's attached to the name Enlil, which we've seen before, and is the name of an important Sumerian deity. So we find that this phrase marks Enlil as the agent of the sentence.

Next up is ud.e. The base noun here, ud, means storm. And just like in the first phrase, we have a .e case marker. Now, it's unlikely that we have another phrase in the ergative case - but it turns out that the terminative case also uses .e as a marker! Ambiguity can arise when analyzing cases, so usually you just make a best guess and see from context if your guesses hold up. For now, let's guess that we translate this phrase as to the storm.

An interesting feature of this sentence is that the verb here is actually a two-word mash-up! This is really common in Sumerian - many different concepts can be described by combining several smaller concepts. In this case, gu is a noun meaning voice, and de is our root verb meaning to pour. So our verb means something like to pour the voice. If you are starting to think like a Sumerian, you're probably thinking that this could translate best as to speak. In fact, that's how Sumerologists read this verb in most texts! Occasionally, the translation might work better in english as to call or to tell, but you get the idea.

Now let's look at the verb chain proper. As we just noted, the root verb is gu ... de meaning to speak. We see our familiar friend .n. just before the root, and we remember seeing that before as the (third-person singular inanimate) cross-reference for the agent of a transitive verb. And just before that, we see ba.. This one is new, and means something like "the interests of the subject are immediately affected". While you might find it odd that Sumerian has a verbal prefix expressing such a concept, to Sumerians, it was completely natural. However, in translation, we usually ignore this in the English. So we can translate this verb roughly as spoke, but with this extra meaning on top.

Putting it all together, we translate this sentence as:

Enlil spoke to the storm.

I love Sumerian literature. It really uses anthropomorphism in wonderful ways! But more importantly, we see here that we can be pretty sure that we guessed right in our assignment of the ergative and terminative case markers.

Verbal Chain Cross-References
As we've already noticed, Sumerian verb chains have markers cross-referencing ergative and absolutive elements of a sentence, when present. This is no different that a verb in any of the other cases - if we had a noun phrase in the ablative case, we'd expect to see the ablative cross-reference marker in the verb chain.

So far, though, the only markers we've seen are the .n. prefixed just before the verbal root, signifying that the patient was third-person singular inanimate, and the .Ø after the verbal root, signifying that the agent was third-person singular animate.

Obviously, there are markers for all the different persons and pluralities, but we'll get to those in later lessons.

Quick Quiz
Okay, now that you've got the basics of ergativity, consider the following sentence, from "Dumuzid's dream" (ETCSL reference 1.4.3.18):

igi.ani cu bi.n.kij

First, let's get some vocabulary.

Quiz Vocabulary

 * igi = eye
 * cu = hand
 * kij = to seek

Quiz Analysis
We start by analyzing our first noun phrase, igi.ani. We know that igi means eye. Remember lesson two, possessives? Look back at the table and see if you can figure out what .ani means, and hence what the whole noun phrase means.

Next, we see that we have a noun, cu, and a verb kij. Sumerian, as we have already learned, has a lot of multiword verbs, and this is no exception! Here, literal translation gives us to seek with the hand. A little creativity, and we realize that in Sumerian, this is an idiom that means to rub. Pretty neat, eh? So really, we can add the following multiword verb to our vocabulary:


 * cu ... kij = to rub

Now, let's tackle our verb chain. As usual, we see a .n. just before the verbal root. Do you remember what this means? If not, look back at the example sentence in lesson six to refresh your memory.

Also, we start the verbal chain with the bi. particle. We saw that already in this lesson - see if you can remember what it means.

Quiz Answers
For convenience, I've put the meaning of each phrase in the hovertext in the following sentence. Once you have figured out what your phrases mean, you can check here.

igi.ani cu bi.n.kij

How did you do? Excellent, I'm sure! At this point, I'm sure you can supply the meaning of the entire sentence given the individual phrases. Now you're really reading Sumerian!

Previous (Lesson Six - A Sumerian Sentence) : Up (Sumerian Grammar - Main Page) : Next (Lesson Eight - The Case System)