Adjectives describe an object. For example, this website has a "green background".
In Sumerian, the adjective goes behind the noun it describes, like in French:
hemeda = purple ĝiri = dagger
ĝiri hemeda = purple dagger (lit. "dagger purple")
If you've checked the epsd or epsd2, you'll notice that there are very few "pure" adjectives. No worries. We can make them from verbs easily. Simple add:
"-a" to the very end of an unmodified verb root, and slap it immediately behind the noun you want to describe. Some call it the nominalising particle. Beware, it can also transform a verb into its passive variant, but still act as a verb. So, context matters. However, verbs very rarely are "naked" with only a nominalising particle on it, so if you see a lonely root with an "-a" tacked on, it's most likely an adjective. If such a root is being treated like a noun (I'll explain later), it's 100% an adjective.
What's all this about the verb you ask? I'll save it for the verb part of this website to avoid confusion. Just remember that if you want to turn a verb into an adjective (so, for example, "to be strong" into "strong"), put at -a on it!
Let's go through some examples:
kug = to be holy, šum = to give, di = case
di kug.a = di kug-ga = the holy case
di šum.a = di šum-ma = the given case
Now, say one interpreted "šum.a" to be "to give" in the passive sense. It would mean "to be given" as a verb. So, "I am given, you are given, he is given, etc...". How does this relate to using verbs to describe nouns? Well look at our example! "di šum.a" can mean "the given case" but, verbs are very rarely "naked" like this without any conjugation, so it's unlikely to be mistaken for "the case, to be given".
You may be wondering, "hey! 'kug' is already passive without the nominalising particle!". You're right! Many Sumerian verbs are automatically passive, such as "kug = to be holy". In these cases, a verb root with a "-a" on it is almost certainly being used to describe a noun. It makes it easy. It is only with verbs that do not have "be" in them, such as "šum = to give" that there is some need for context when you use the nominalising particle.
So, what if we wanted to add a case onto a noun with an adjective? For example, say "the king with the purple dagger" or "the purple dagger with the king"? Well, in Sumerian, if the case ending effects a noun with an adjective, the cases attach to the adjective.
For example:
da = with
ĝiri hemeda.da = with the purple dagger
lugal ĝiri hemeda.da = the king with the purple dagger
ĝiri hemeda lugal.da = the purple dagger with the king
Now, the same applies to demonstrative and possesive pronouns that are enclitic. So, if you wanted to say "the king with my purple dagger", you'd do:
lugal ĝiri hemeda.ĝu10.da = the king with my purple dagger
All in all, just imagine that the adjective steals all the elements that attach to the noun it is a part of and attaches them to itself