Iver Larsen iver_larsen at
Sat Oct 27 16:59:41 EDT 2001

> Take a look at John 6:45:
> Looks to me like QEOU is functioning as the semantic agent here. So once
> again, what are you saying about the genitive case and semantic roles?

It is more a general linguistic semantic notion that the genitive in pretty
much any language serves to link two nominals. This also applies to
languages that do not have anything like the case system of Greek.
When I talk about semantic roles I am thinking primarily of finite verbs. An
infinitive is a type of nominal and a participle is also a type of nominal.
Even an adjective can function as a nominal, what we call the substantive
In your example above I would take DIKAKTOI as a substantive adjective, a
nominal "taught-ones" or "people that are taught" When combined with QEOU we
get "taught-ones of God".
God is an agent in the deep semantics but that is not what the genitive
indicates. The genitive only links the two words together and how they are
linked is a matter of analysis of the underlying verb. If the sentence had
been "God will teach everybody" we would have a finite verb with the agent
as subject. If we make it passive, we could have "all will be taught by

Now in Greek there are certain verbs that require a genitive form as object.
That would be a different matter. In such cases the genitive may correspond
to a semantic role like patient. But those are special cases for Greek. In
general linguistics, the genitive is simply an associative marker that links
two nominals.

Iver Larsen

More information about the B-Greek mailing list