In lieu of an abstract, here is a brief excerpt of the content:

  • Irregularity and Pronominal Markedness:Where Favoritism Sets In
  • Mark Donohue

The principles associated with the ordering of two agreement markers on one verb are discussed in detail for Asmat. An argument is presented that some languages have agreement ordering systems that are simply irregular and cannot be modeled, short of being totally stipulative.

1. Pronominal Systems.

Much work has been published on the different types of pronominal systems to be found in different languages, and indeed in different subsystems of languages. This paper is concerned with the relative ordering of pronominal agreement suffixes with respect to one another when occurring bound to a verb in Asmat, a language of southwestern New Guinea (Drabbe 1953, 1963; Voorhoeve 1965, 1980). In Asmat, the majority of the verbal paradigm regularly shows a V-o-s ordering, but the position of the 1SG suffixes is difficult to model.

2. Ordering Agreement.

There have not been many attempts to explain the different orders of subject and object agreement markers on verbs. In his survey of characteristics of the Papuan languages of New Guinea, Foley (1986:105) classifies the relative orderings of affixes for subject and object (at that time, the existence of Papuan languages with ergative-absolutive agreement systems was unknown), showing that all possible combinations of positions for subject marker, object marker, and verb stem are found. In this environment, we can find four out of the six agreement ordering possibilities (of those that involve agreement with two arguments; agreement with one argument is not discussed) that have both affixes on the same side of the verb: V-o-a, o-a-V, V-a-o, and a-o-V. These are all robustly attested in the languages of New Guinea (except for o-a-V, which is only attested in part of some paradigms, never as the sole option in an agreement set).

3. Disorganization.

In addition to the systems described above, which are all amenable to modeling with differing degrees of complexity, we can also [End Page 409] recognize languages in which there simply is not a set of nonstipulative principles governing the organization of the morphemes. There is a degree of irregularity in an otherwise ordered system.

3.1 Asmat.

Asmat (Asmat-Kamoro family, southwestern West Papua) is a language with verbs containing suffixes for both subject and object. We examine materials from a range of dialects of Asmat, and analyze the resulting paradigms of subject and object suffixing to show that, while the majority of the paradigm is regular, there are some unexplainable but not quite suppletive discrepancies in the forms.

The verb por- 'see' has the forms shown in table 1 for different combinations of subject and object in the ultimate past paradigm, which is typical of the other paradigms but lacks complicating tense suffixes (Voorhoeve 1965:107). (Some cells have been left blank, because they represent combinations that can only be expressed by means of reflexive morphology.)1 It is clear that the root does not have morphophonemic interactions with the suffixes, and also that there are no suffixes other than those needed to mark the person and number of the subject and object. Furthermore, these suffixes do not significantly interfere with either other, and do not result in portmanteau forms. The only variation we see are some vowel-harmony phenomena that are easily factored out. The one exception to this is the cell for 3PL SUBJ ® 2PL OBJ, in which only the object form appears, not the expected *pornés.

Basic segmentation of the suffixes produces the groups shown in table 2, where each suffix is subscripted "s" if it indicates the subject (transitive or intran-sitive) and "o" if it indicates the object of the verb. We note that there is not a simple one-to-one assignment of form to person/number category-marking function in Asmat. The O-marking suffix -n, for instance, appears in the table to mark a 1SG, 2SG, AND 2PL object. In contrast to this, the 1PL object is marked with -aw, which does not serve any additional functions.

In table 3, we can see that the usual order of suffixes and verb...


Additional Information

Print ISSN
pp. 409-420
Launched on MUSE
Open Access
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.