​CHAPTER 21. THE SUMERIAN LANGUAGE

The previous chapter ends at Ereshkigal's threshold. What follows is formula work. But formula work in a dead language requires more than phonetic memorization. A sound without the form is noise; a formula without grammatical bones cannot be modified or built from scratch.

Sumerian is a language isolate. It has no known relatives, no ancestor tongue that modern scholarship has identified. It was spoken in southern Mesopotamia from at least the fourth millennium BCE and died as a living spoken language sometime around 2000 BCE, surviving for another two thousand years as a sacred and scholarly language in temple schools and ritual contexts, the way Latin survived in medieval Europe long after it ceased to be anyone's mother tongue. The tablets that preserve it were written in cuneiform — wedge-shaped impressions pressed into wet clay.

The principles of later chapters require the grammar as their foundation.

​Agglutination

The single most important fact about the language for the purposes of formula construction is that Sumerian is agglutinative. In an agglutinative language, words are built by attaching morphemes (meaningful units) to a root in a chain. Each morpheme carries one piece of information. The root carries the core meaning; the attached morphemes add direction, case, tense, number, and verbal mode. They remain distinct and identifiable, stacked in a fixed order.

Sumerian words can be very long chains of short units, and each unit in the chain is legible independently. Hé-en-da-gub-bu breaks into hé- (precative prefix: "may it be so") + en- (verbal prefix) + da- (comitative: "with") + gub (root: "stand") + -bu (suffix). Every element plays a specific role.

Sumerian is therefore the right language for the practice. Its agglutinative structure makes it transparent: every element of a formula can be traced back to its role, and the formula’s geometry is visible in its very construction. The word-count that determines which Weaving the formula carries is legible because the language's units are discrete and countable.

​SOV Structure

Sumerian follows Subject-Object-Verb word order. The verb comes last. In any Sumerian sentence or formula, the force accumulates through the preceding elements and is delivered by the verb at the end. The subject establishes who or what acts, the object — what receives the action. The intervening elements (place-words, directional operators) build the context in which the verb will land.

The principle of formula construction is as follows: build from domain to vector to potency, end with the verbal seal. Gub, zi, ba, du, tu, íl — these roots cannot go anywhere else in a grammatically coherent Sumerian utterance. They are always the final element. A formula that places the verbal seal anywhere elsewhere than the end is grammatically impossible.

​Nominal Chains and Postpositions

The grammatical marker that in English comes before a noun ("in the steppe," "from above") comes after it in Sumerian. This produces nominal chains where the root comes first and its relational markers accumulate after it:

edin = steppe

edin-na = in the steppe (-na: locative)

edin-ta = from the steppe (-ta: ablative)

edin-šè = toward the steppe (-šè: directive)

The same root takes different postpositions depending on the direction of force. A formula that moves KI.TA force toward KUR uses kur-šè; a formula that establishes presence inside KUR uses kur-ra.

The operators AN.TA and KI.TA are nominal chains of the same type. AN.TA = AN (sky) + TA (ablative postposition: "from") — literally "from the sky," the descending pressure. KI.TA = KI (earth/below) + TA — "from below," the uprising current. They are ordinary Sumerian grammar describing spatial relations, which the ritual tradition recognized as cosmological operators because the spatial relations they describe are the same relations that govern the membrane.

​Verbal Prefixes and the Precative

The Sumerian verb carries a complex chain of prefixes before the root. For formula construction, two prefix types matter most.

The precative prefix hé- (sometimes spelled hé-en- or he-) transforms a statement into a command with divine authorization: "may it be so," "let it stand," "let it rise." Hé.gub = "may it stand, let it establish presence." Hé.en.da.gub.bu = "may it stand here with me". When the magician uses hé-, they are invoking the same mode that the gods use when they decree fate.

The verbal prefix mu- marks first-person active agency: "I do / I make happen." Mu-da-sikil = "I cleanse with / I purify alongside." Formulas that combine both (hé- for the authorization and mu- for the agency) establish the relationship between the force being invoked and the subject directing it.

​Nominal Classes: Human and Non-Human

All nouns are divided into two classes: human (animate, rational) and non-human (everything else). Gods, humans, and certain divine animals belong to the human class. Demons, ordinary animals, objects, and abstract concepts belong to the non-human class.

Divine names with the d- determinative are always human class. Udug, a-lá, gedim, maškim, the demonic forces, are non-human class. The verbal seals in this book's formulas are selected for compatibility with the class of the primary force being invoked. When a formula addresses both Ninlil and a demonic force in the same utterance, the verbal form aligns with the higher-class entity.

​The Bilingual Tradition

From the Old Babylonian period onward, Sumerian ritual texts were written in parallel with Akkadian translations in interlinear format: one line of Sumerian, one line of Akkadian below it. The Udug Hul series and the Maqlû series are both bilingual texts. The tradition preserved both versions as equally valid — the Sumerian as the original, the Akkadian as the translation that carried the same force.

Akkadian is a Semitic language, different from Sumerian: it uses prefixes and suffixes differently, it has a VSO tendency rather than SOV, and its roots follow the Semitic triconsonantal pattern. But in the bilingual ritual tradition the two languages are treated as equivalent vessels for the same content. The Sumerian carries the precision of the agglutinative chain; the Akkadian bears the semantic density of the Semitic root. Together they weave a bilingual formula in which the two layers reinforce one another.

The formulas in the following chapters draw from both traditions in the same way the ancient bilingual texts did, using Akkadian elements where the Semitic vocabulary has a specific semantic weight that the Sumerian equivalent does not. Kaššāptu, kitû, mūši — these Akkadian terms enter the formula-palette as the bilingual layer that the ritual tradition utilizes.

​