Detail #295: Agreement with the Subject in Case Marking

This post uses the distinction between argument and adjunct without really ever providing any specific defition. I Should probably sometime actually 

Imagine a language where the case markers of direct arguments of the verb show some kind of agreement with the gender and number of the subject. We come up with a set of object markers that - maybe with some syncretism - thus marks mostly the agreement features of the subject, but to some extent also marks some agreement features of the object.

For the other cases, for arguments of the verb the congruence marker only codes the gender-number of the subject, and there's a separate morpheme marking the number of the argument. This number marker marks the gender of the noun itself, not that of the subject.

For adjuncts of the finite verb and for both the arguments and adjuncts of infinitives, the noun takes its own gender's case marker.

To illustrate such a system, I guess a three-gender system is suitable. Let's assume a very IE one, i.e. with three genders, one masculine, one feminine and one neuter.

In the system I came up with , the masculine and feminine objects take their own gender's accusative marker as object marker when the subject is neuter. When the subject is of the same gender as the object, they take their own gender's default morpheme - the one used for adjuncts and infinitive objects.

However, when a masculine subject and feminine object or vice versa occur, a special set of morphemes are used. Slightly more complicated in the singular than the plural, however, with more syncretism in the plural.

Finally, for the other cases as arguments, a masculine singular subject triggers masculine singular case markers, a feminine plural subject triggers feminine plural case markers, etc. If the argument and the subject are of the same gender-number marking, the argument's own number-morphology is lost.