Defining Ergativity

Ergativity is an umbrella term used in modern typological linguistics to cover a variety of grammatical phenomena. Basically, ergativity is a grammatical pattern in which nouns are inflected or marked for case. Fabricius observed this phenomenon (which he named as nominativus transitivus) in Greenlandic Eskimo in 1801, but in fact Adolf Dirr used the term ‘ergative’ (based upon the Greek word "ergon", plural "erga", ‘work, task’) for the first time in 1912.

The essential thing about ergativity is that this pattern challenges a universal characterization of “subject”. Up to now, Asian, Australian and American (Eskimo, Mayan and various Ge-Amazonian) languages have been claimed to have some ergative properties. It is interesting to note that ergativity is virtually nonexistent in Europe (with Basque of the western Pyrenees as the single exception) and Africa. Needless to say, ergative languages are still being discovered.

Quite simply, ergative case occurs on nominal case marking on the nouns and/or on the verbal agreement system. To see how this works, let us consider first that every natural language has three types of clause: a minor type and two major types of verbal phrases. These three are:

1. EQUATIVE (also equational), which involves two noun phrases (NPs). For example, “My daughter is a teacher” (some languages require copula, i.e. copular verb, like English is; others do not) in which the subject and complement refer to the same person;

2. INTRANSITIVE, which involves a verb and one core NP (as in “The baby cries”) – that is, intransitive verbs are followed by no obligatory element; and

3. TRANSITIVE, which involves a verb and two or more core NPs.

As a result, every language should always find the means to distinguish between the three basic grammatical categories, namely: intransitive subject (Si), transitive subject (St), and direct object (O). More precisely, the languages of the world use systems of case marking to show the function of the noun or noun phrase in a clause. We can say that the chief function of case marking systems is to disambiguate contexts, which is to permit the hearer to identify the semantic role of the denotation of an NP.

Plainly, there are five types of logically possible systems for assigning case to Si, St, and O, to know: the nominative-accusative (the pattern known from the major European languages), in which all subjects are assigned a single marker (nominative) distinctly from O (accusative); the absolutive-ergative, in which Si and O have the same morphological marker (absolutive, usually null), while a different marker is used for St (ergative); the tripartite (also three-way) with three distinct markers (this system is relatively rare); the neutral, in which the same marker (possibly null) is used for all three categories; and the horizontal (also double-oblique), in which St and O are identically marked in opposition to Si (so far unattested).

In short, according to case marking displayed by subjects and objects languages can be grouped into five case types. Naturally, every language has a lot of possibilities in order of expressing the nominal case marking. Those comprise: case inflections, particles and adpositions, word order, cross-referencing and verbal agreement, etc.

There are two basic types of ergative languages: the classic type and the active-stative type. Let us explain:

· THE CLASSICAL definition says that an ergative language treats the subjects of intransitive verbs (Si) in the same way as the objects of transitive verbs (O) – that is, Si/O behave alike since absolutive zero marks both --, and differently from a transitive subject (St), which receives an ergative special marker. For example, in West Greenlandic Eskimo the ergative marker is the suffix ‘-p’, whereas in Basque is ‘-k’.

· THE ACTIVE-STATIVE type marks some Si (so-called non-actives or ‘unaccusatives’) with O, whereas another class of Si (so-called actives/agentives or ‘unergatives’) has the same marking as St. In other words, there are two subgroups of intransitive verbs: ‘unaccusative’ (e.g. ‘sleep’) and ‘unergative’ (e.g. ‘dance’). Note that these semantics terms are currently used for Basque, which is an ergative active language.

Now let us consider that since ergative marking may occur also in tripartite languages, as noted above, it follows that there is no absolutive bond in tripartite (O is marked by accusative) and in active (Si is marked by ergative) languages. Hence, we should recognize that the absolutive grouping (Si/O) is not a necessary feature of ergative languages.

Since R. M. W. Dixon’s classic study on the Australian language Dyirbal (1972) it is known that there are at least two classes of ergative languages: a small class of syntactically ergative languages (also inter-clausal) and a large class of morphologically ergative languages (also intra-clausal). The former show an absolutive pivot, i.e. a grammatical subject grouping (Si/O) on clausal coordination, while the latter have nominal case marking and/or verbal agreement or cross-referencing of persons on the verbs. For the most part, morphological ergative languages may have both systems at the time, i.e. display ergative pattern for case marking and accusative for agreement.

Finally, it is need to be stressed that ergative languages are never pure. This is because many languages show mixed nominative-accusative and absolutive-ergative features. This means that two ways of organizing case marking and agreement occur in ergative languages. More precisely, ergative languages have a split ergative pattern in which some but not all transitive clauses are ergative constructions. This phenomenon is called split ergativity, which it is to say that an agent is marked for ergative case or left unmarked depending on its position on the animacy hierarchy, or on the tense/aspect/mood of the clause, or on the grammatical status of the clause, whether it is main or subordinate, etc.

Marco Bomfoco copyright 2006 All Rights Reserved.