Sentence structure: how words fit together to make meaning

So far we have mostly been looking at individual words, or in groups of two or three. But we also need to look at how they interact to form different sorts of sentences, and to pay closer attention to the function words that hold sentences together.

When you have read this page, you might like to test your understanding by trying Cuneiform exercise 4.

Word order

In the course of the pages on Nouns and adjectives, Possession and existence, and Verbs and adverbs, we have already come across several rules about which words belong together in a typical Akkadian sentence:

Naturally, these rules combine to govern the structure of more complex sentences. In this example the superscript numbers allow you to compare the word order in the Akkadian sentence and its English translation:

kalbum1 ṣehrum2 ša3 awīlim4 damqim5 wardam6 ina7 bīt8 bēlišu9 danniš10 iššuk11
"The good5 man4's3 small2 dog1 strongly10 bit11 the slave6 in7 his master's9 house8."

You can see, then, just how different Akkadian and English word order are. Fortunately, the case endings on the nouns also give clear signals about who is doing what to whom.

Function words

So far we have mostly been discussing nouns, adjectives, verbs and adverbs—all types of words whose meaning in Akkadian derives partly from their morphology or shape, and partly from their root consonants. Collectively, we call these content words. But like many other languages, Akkadian also has lots of function words, which are the grammatical glue that hold the sentence together. In that last sentence, "but", "like", "also", "of", "which", "the" and "that" are all function words. Without them it would be much more difficult to understand the sentence: "Other languages Akkadian has lots function words are grammatical glue hold sentence together." Just as there are different sorts of content words, we can usefully distinguish several types of function words in Akkadian. In fact, we have mentioned some of them already:

Articles (e.g., "the" and "an"):
Akkadian, like many languages, does not use articles.
Conjuctions (e.g., u "and" and ū "or"):
Conjunctions create relationships between two words of the same type, without affecting their morphology (e.g., noun case): bītum ū ēkallum, "house or palace".
Negatives (ul and "not"):
As shown above, ul usually precedes verbs (except after šumma "if") and statives, while precedes adjectives, and verbs in conditional and subordinate clauses. Akkadian's use of negatives with questions, wishes, and commands gets more complicated, but that is beyond our scope.
Prepositions (e.g., ina "in, from" and ana, "to, for"):
In Akkadian, prepositions often locate or move a noun in time or space: ina bītim, "in the house"; ana dārītim, "for eternity". Nouns following prepositions are always in the genitive case.
Pronouns (e.g., anāku "I", šuāti "him"):
Pronouns stand in for nouns to save repetition. In Akkadian independent pronouns are only used for emphasis. All verbs automatically mark their subjects with prefixes, while their objects are often indicated with suffixes.

We will now discuss some other types of function words, but (as usual) this is not an exhaustive treatment. Good introductory Akkadian textbooks will tell you much more.

Sentences with šumma "if"

All of king Hammurabi's laws begin by imagining a problematic situation and then propose a legal solution. Here, for instance, is the start of Law 195:

šumma mārum abašu imtahaṣ
"If a son has hit his father"

The sentence begins with šumma "if" and the verb is in the perfect tense. It could equally well be in the preterite (with imhaṣ, "he hits"). We call this "if" fragment a conditional clause, because it puts forward a hypothetical situation (a condition) and is not a complete sentence in itself although it contains a verb.

Another technical name for a conditional clause is a protasis. It is almost always paired with an apodosis, the second half of the sentence which describes the consequences of the protasis. The apodosis of Law 195 runs:

rittašu inakkisū
"they shall cut off his hand."

Unlike the protasis, the apodosis is always a sentence in itself. Also unlike the protasis, the apodosis is in the present-future or durative tense. And if the sentence had negative verbs in it, the protasis would use while the apodosis would use ul:

šumma mārum abašu imtahaṣ rittašu ul inakkisū
"If a son has not hit his father: they shall not cut off his hand."

Sequential actions with -ma

The protases of many laws detail sequences of actions or events, each of which is the consequence of the preceding one. In this case, the verbs are all in the preterite except the final outcome or key event, which is in the perfect tense. Further, the verbs are all connected by the particle -ma, "and then, so that". Here for instance, is the protasis of Law 22, slightly modified:

šumma awīlum hubtam ihbutma iṣtabtūšu
"If a man commits a robbery and (as a result) they have caught him"

The key event here is catching the robber, for he cannot be punished if he is not caught. Compare this with the conjunction u "and" that connects events that are not logically related: awīlum hubtam ihbut u kalbam imhaṣ, "the man committed a robbery and (incidentally) hit the dog." The man could have hit the dog whether or not he committed a robbery; but (we hope!) he would not have been caught if he were innocent of the crime.

Be careful not to confuse verbal -ma, "and then", with -ma on nouns, "it is". When you are reading a long Akkadian text in cuneiform, it is a good first strategy to look for -ma on the ends of verbs. This will help you divide it up into manageable clauses and sentences.

Subordinate clauses with relative pronoun ša

Quite often we want to add extra information in a sentence, which is relevant but not grammatically essential to it. If that information is expressed with a verb, we call it a subordinate clause. In the first sentence of this paragraph, for instance, everything after the comma is subordinate. Subordinate clauses often begin with a relative pronoun: words like "which", "that", "who", "whom" in English. Akkadian just has one relative pronoun, ša, which is the equivalent of all of these English words. That makes subordinate clauses easy to identify, but it can be tricky to work out exactly how to translate ša. We will look at some examples below.

First, though, you might be worrying about possible confusion with ša "of". (It's already bad enough to discover that there are two sorts of -ma!) But just like the two -ma's, one ša ("of") belongs with nouns, while the other ša ("which", etc.) belongs with verbs. And to make subordinate clauses easier to identify, verbs in these clauses take a subjunctive suffix -u if they do not have a long vowel suffix already: kalbum ša imhaṣu, "the dog that he hit". (In cuneiform writing, however, it is often difficult to distinguish between long and short vowels, so without more context we would be equally justified in reading 𒊭 𒅎 𒄩 𒍪 ša im-ha-ṣu2 as ša imhaṣū, "that they hit".)

In order to know how to translate ša correctly, we need to pay close attention to the cases of the nouns in the subordinate clause, and any pronoun suffixes on the verb. Here are some examples:

Akkadian also has subordinate clauses after subordinating conjuctions (like "when", "until", "because"), and with construct nouns, but we do not have space to go into them here. As usual, you should consult a good Akkadian grammar book for more information.

Infinitives as nouns

Finally, we need to come back to two ideas we have met before: the prepositions ina and ana, and the infinitive of the verb, such as parāsum. So far we haven't said anything about the grammatical function of infinitives within sentences. In fact, in many contexts they actually behave like nouns. You may have already noticed that all Akkadian infinitives have the nominative case ending -um. They can also exist in the other cases, in the construct, and with possessive suffixes. We have been translating infinitives as parāsum, "to cut, to decide", but sometimes it is more appropriate to say "cutting, deciding" instead.

Intentions after ana
After ana, infinitives denote intentions: ana parāsim, "(in order) to decide", ana ālam kašādim, "(in order) to conquer the city". In both of these examples, the infinitive is in the genitive case because it follows the preposition ana. But in the second example, ālam "city" is in the accusative case because it is the object of kašādum "to conquer". An alternative expression of this idea, ana kašād ālim, "for the conquering of the city", gives a slightly different nuance. We can use possessive suffixes in this context too: ana parāsišu, "for his deciding", or "in order for him to decide"; ana ālam kašādiya, "for my conquering of the city", or "in order for me to conquer the city". Which translation you choose depends on what sounds best in the context.
Concurrent events after ina
After ina, on the other hand, infinitives denote an action taking place at the same time as the main verb: ina parāsim, "on deciding", ina parāsiša, "when she decided", ina ālam kašādišu, "as he conquered the city" or "during his conquering of the city".

Content last modified on 02 Apr 2024.

Eleanor Robson

Eleanor Robson, 'Sentence structure: how words fit together to make meaning', Knowledge and Power, Higher Education Academy, 2024 [http://oracc.museum.upenn.edu/saao/knpp/CuneiformRevealed/Akkadianlanguage/Sentencestructure/]

 
Back to top ^^
 
© Higher Education Academy, 2007-11. Since 2015, SAAo is based at the Ludwig-Maximilians-Universität München, Historisches Seminar (LMU Munich, History Department) - Alexander von Humboldt Chair for Ancient History of the Near and Middle East. Content released under a CC BY-SA 3.0 [http://creativecommons.org/licenses/by-sa/3.0/] license, 2007-20.
Oracc uses cookies only to collect Google Analytics data. Read more here [http://oracc.museum.upenn.edu/doc/about/cookies/index.html]; see the stats here [http://www.seethestats.com/site/oracc.museum.upenn.edu]; opt out here.
http://oracc.museum.upenn.edu/saao/knpp/CuneiformRevealed/Akkadianlanguage/Sentencestructure/