Week 8: Case and agreement
LEL2D Morphosyntax
Case and agreement
This week we discussed case and agreement as well as their alignment. Agreement is the expression of features of a controller on a target. The controller is often an argument, for example a subject or an object, while the controller is often a predicate, for example a (finite) verb. We only really discussed cases in which an argument controls agreement on a verb.
Case is morphological marking on particular arguments in a clause. It is often only found on a subset of arguments (for example the objects of transitive clauses or the subjects of transitive clauses), rather than all arguments. Case systems are quite distinct across languages, ranging from languages with no morphological case at all to languages with more than 20 cases (for example the Uralic languages Hungarian and Finnish).
Both case and agreement can be differential, meaning that not all arguments of a particular type are marked in the same way. In the lectures, we saw examples of this from Urdu (Butt 2006), where only some direct objects are marked with the accusative clitic =ko and others remain morphologically unmarked.
Alignment
A crucial aspect of how case and agreement vary across languages is their alignment. In the lectures, we focused on nominative–accusative and ergative–absolutive alignment. These types refer to the grouping of particular arguments based on their morphological expression (case) or behaviour (agreement).
When subjects of intransitive (S) and transitive clauses (A) have the same morphological marking but objects of transitive clauses (P) have a different marker, we refer to their grouping as nominative–accusative alignment. The case of the transitive object P is called accusative, while the often unmarked case that S and A have in common is called nominative.
Conversely, if the intransitive subject (S) and the transitive object (P) have the same case-marker, and the transitive subject (A) has a different one, we call their alignment ergative–absolutive. The case of the transitive subject is called ergative, while the often unmarked case of S and P is called absolutive.
For agreement, there are two ways we discussed in which alignment can work. First, agreement can be with S and A arguments, but not P. This is nominative–accusative agreement. Second, agreement can track S and P, but not A, and show ergative–absolutive agreement. This often happens when the transitive subject A has ergative case. Third, in many Mayan languages, for example, agreement markers show ergative–absolutive alignment. One set of markers is used for S and P arguments, while another set of markers is used to indicate agreement with A.
Practical
In the practical, you explored data from grambank to explore the distribution of subject and object agreement cross-linguistically. You found that subject agreement is much more common than object agreement, that many languages have both, but that few languages have only object agreement but no subject agreement. Why could that be?