Chapter 2
THE GRAMMAR OF FIRST THOUGHT
Sumerian grammar reveals a different way of organizing reality.
2.1 Not Like Us
Every language makes choices.
When you describe an event, you have to decide: Who did what to whom? How do you mark the doer? How do you mark the thing done to? These seem like trivial questions—surely there’s only one way to do it?
There isn’t. And Sumerian’s way is profoundly different from English.
Consider two sentences:
The king arrived. The king built the temple.
In English, “the king” works the same way in both sentences. It’s the subject. It comes first. It determines whether the verb is singular or plural. Whether the king is arriving (doing something by himself) or building (doing something to something else), the grammar treats him identically.
This feels natural. It isn’t.
Sumerian belongs to a different grammatical family—not genetically, but typologically. It uses what linguists call ergative-absolutive alignment. In this system:
- The subject of “the king arrived” is treated one way.
- The agent of “the king built the temple” is treated differently.
- But the object of “the king built the temple” is treated the same way as the subject of “the king arrived.”
In Sumerian:
lugal-Ø i-gen — “The king arrived.” lugal-e é mu-dù — “The king built the temple.”
Notice that “-e” on “lugal” in the second sentence. That’s the ergative case marker. It appears only when the noun is the agent of an action done to something else. In the first sentence—arriving, an action without an object—the king takes no marker at all (that’s what the “Ø” indicates: zero marking, or “absolutive” case).
Now here’s the strange part. Look at “é” (temple) in the second sentence. It also has no marker. The thing being built and the person arriving are grammatically parallel—both unmarked—while the person building takes special marking.
Why would a language work this way?
One theory: ergative languages focus on the effect of actions rather than the actor. The king arriving and the temple being built are both states that result from events. The king-as-builder is the external force that caused one of those states. Ergative grammar highlights causation rather than agency.
Another theory: it’s simply a different way of categorizing—not better or worse, just different. Languages evolve solutions to the problem of encoding meaning, and there’s more than one workable solution.
What matters for us is simpler: the Sumerians organized sentences differently than we do. When a Sumerian scribe described an action, his grammar forced him to make distinctions we don’t make and ignore distinctions we find fundamental. His language was carving reality at different joints.
We cannot know if this affected how Sumerians thought—the relationship between language and thought is fiercely debated. But we can know that it affected how they spoke, which affected what their texts preserve, which affects how we reconstruct their world.
The first written language wasn’t just unfamiliar vocabulary. It was an unfamiliar way of structuring meaning itself.
2.2 Building Blocks
English builds sentences by arranging words in order. “The dog bit the man” means something different from “The man bit the dog.” Word order carries the weight.
Sumerian works differently. It’s agglutinative—from Latin agglutinare, “to glue together.” Meaning is constructed by attaching morphemes (meaningful units) to a root, one after another, like beads on a string.
Consider the Sumerian word:
nam-lugal-la-ka-ni
This isn’t five words. It’s one word with five parts:
| Morpheme | Meaning |
|---|---|
| nam- | abstract noun prefix (“the state of being…”) |
| lugal | king |
| -la | (phonetic glue, eases pronunciation) |
| -ka | genitive (“of”) + locative (“in”) |
| -ni | “his/her” |
Combined meaning: “in his kingship” or “during his reign.”
English needs four words for what Sumerian does with one. And this isn’t an extreme example—Sumerian verbs are far more complex.
The agglutinative system has consequences. Each piece has a stable meaning. Unlike English, where “ran” and “run” involve a vowel change that just has to be memorized, Sumerian morphemes are consistent. Once you know what nam- does, you can recognize it anywhere. Once you know -ak marks possession, you can parse any genitive.
This makes Sumerian, in some ways, easier to learn than English. The system is regular. The parts snap together predictably.
But it also means that single words can become enormously long. The scribe writing nam-lugal-la-ka-ni had to hold the whole structure in mind—root plus five suffixes, each in its proper place. Reading required the same: unpacking the chain, identifying each morpheme, assembling the meaning.
Imagine thinking in such a language. Every noun arrives with its relationships attached—whose it is, where it is, what role it plays in the sentence. You don’t say “the house of the king”; you say “king-house-of.” The possession is built into the word.
Some linguists argue that agglutinative languages promote a certain kind of precision. Every grammatical relationship is explicitly marked. Nothing is left to context or word order. The meaning is in the word, not floating between words.
Others argue this makes no difference to thought—that all languages can express the same ideas, just with different tools. The debate continues.
What we can say for certain: Sumerian scribes operated with a building-block system that English speakers find unfamiliar. When they composed a sentence, they were assembling morphemes as much as arranging words. The architecture of their language was fundamentally different from ours.
2.3 The Verbal Engine
If Sumerian nouns are complex, Sumerian verbs are extraordinary.
A single verb can contain what English would express in an entire clause. Subject, object, indirect object, location, direction, aspect, mood—all packed into one word, morpheme by morpheme.
Consider:
mu-na-ni-in-dù
This is one word. Let’s unpack it:
| Position | Morpheme | Function |
|---|---|---|
| 1 | mu- | “ventive” — motion toward speaker or reference point |
| 2 | -na- | dative — “for him/her” |
| 3 | -ni- | locative — “in it” |
| 4 | -in- | ergative — third person agent (“he/she” as doer) |
| 5 | dù | verb root — “to build” |
Combined meaning: “He built it in it for him.”
One word. Five morphemes. A complete sentence with subject, verb, direct object (implied), indirect object, and locative phrase.
The Sumerian verb operates on a template—a series of slots, each with its designated function. Scholars diagram it roughly like this:
[Modal] - [Conjugation] - [Dimensional prefixes] - [STEM] - [Agreement suffixes]Different verb forms fill different slots. The system is highly regular but extremely dense. A scribe had to master which prefixes went where, in what order, with what meanings.
This is not how Indo-European verbs work. In Latin or Greek, verbs inflect—they change form to mark person, number, tense. Amo, amas, amat—I love, you love, he loves. The changes are at the end, modifying a stem.
Sumerian verbs have prefixes and suffixes, and the prefixes encode far more information. Direction of action, beneficiary of action, location of action—all marked before the root. It’s as if English had separate verb forms for “build-toward-for-someone-inside” versus “build-away-from-against-someone-outside.”
Why does this matter?
First, it shows that human languages can be organized in radically different ways. The Sumerian verbal system isn’t a primitive version of something better. It’s a complete, sophisticated solution to the problem of encoding actions—just not our solution.
Second, it affected what was easy to say. Every language makes some things simple and other things cumbersome. In Sumerian, specifying the direction and beneficiary of an action was easy—you just filled the slots. In English, those same specifications require extra words, extra clauses, extra effort.
Third, it challenges translation. When we render Sumerian texts into English, we lose the compactness. A Sumerian poem where each line is a single dense verb becomes an English paragraph. The rhythm changes. The weight changes. Something is always lost.
We read Sumerian in translation. We should remember what translation leaves behind.
2.4 What This Reveals
Does grammar shape thought?
This is one of the oldest questions in linguistics. The strong version—that language determines what its speakers can think—is mostly rejected today. People can think about things their language lacks words for. Concepts travel between languages despite grammatical differences.
But a weaker version persists: language influences habitual thought. If your language forces you to mark certain distinctions every time you speak, you become practiced at noticing those distinctions. If your language ignores certain distinctions, you can still perceive them, but you’re not constantly reminded.
What distinctions did Sumerian enforce?
Agency and causation. The ergative system required scribes to distinguish, in every transitive sentence, between the agent (marked with -e) and the patient (unmarked). English speakers can ignore this distinction—“the king” is “the king” whether he’s acting or being acted upon. Sumerian speakers couldn’t ignore it. It was built into every sentence.
Spatial relationships. The verbal prefixes encoded direction and location with precision. Motion toward, motion away, action inside, action outside—these weren’t afterthoughts but fundamental parts of the verb. A Sumerian speaker describing an event was always, grammatically, situating it in space.
Relationships between participants. Who benefited? Who was affected? The dative and other dimensional prefixes made these relationships explicit. You couldn’t just say “he built”; you had to say for whom, in what, toward what. The grammar demanded specificity about social connections.
None of this means Sumerians were incapable of abstract thought. Their literature, mathematics, and theology prove otherwise. But it suggests their default mode of description was concrete, spatial, relational. The grammar pulled toward specificity.
Compare this to Classical Chinese, which has minimal morphology—no case markers, no verb conjugation, minimal marking of any kind. Meaning depends almost entirely on word order and context. Scholars have suggested that Classical Chinese pulls toward abstraction, toward meanings that hover between possible interpretations.
Whether these grammatical tendencies shaped philosophy, literature, or worldview is debatable. What we can say is that Sumerian grammar encoded a particular set of priorities: causation, space, social relationship. These priorities are visible in every sentence the Sumerians wrote.
When we read Sumerian texts, we’re not just reading content. We’re reading the traces of a grammatical system that organized reality differently than ours. The words are translated, but the structure behind the words was alien.
The Sumerians didn’t just speak a different language. They navigated meaning with a different map.