Number and Events: Verbal Reduplication in Niuean

Abstract

In this paper we examine Niuean verbal reduplication, and, building on Sperlich (1997a,b, 2001), we show that there are several morphophonological types of reduplication (monomoraic prefixation, bimoraic suffixation, bimoraic prefixation), and there are several different meanings for reduplicated verbs (singularization, intensity, iterativity) that all lie within the semantic domain of pluractionality. We show that, in general, each different phonological type of reduplication corresponds to a particular meaning. It is difficult to predict, however, what a reduplicated verb will mean from its form alone, in part because most words in the language are bimoraic, and there is no difference in form between bimoraic suffixation and prefixation for such words. Once we add in the lexical aspectual semantics of the base verbs, however, we find that it is to a large extent predictable what the meaning will be for a given reduplicated verb, depending on its aspectual class. As well as examining verbal reduplication in general, we focus on agreement or accord reduplication in Niuean, and argue that it is, in fact, not an independent syntactically governed type of reduplication, but instead it is a subtype of the reduplication associated with iterative semantics. The accord effect comes into play if an affecting or a stative verb is iterated, because if such a verb is iterated, the relevant argument must be plural for semantic reasons.

1. Introduction.¹

This paper addresses the forms and functions of verbal reduplication in Niuean, an Oceanic language of the Tongic subgroup, spoken mainly in Niue and New Zealand. As in other Oceanic languages, the reduplication system in Niuean is a rich and interesting one (Ioane 1999; Krupa 1982; McEwen 1970; Seiter 1980; Sperlich 1997a,b, 2001), in which verbs, nouns and modifiers can be reduplicated to yield a variety of derived meanings. But this paper will concentrate on verbal reduplication, and will take as a starting point the phenomenon referred to as agreement by Seiter (1980) and as syntactically governed reduplication by Sperlich (1997) and Ioane (1999). Our exploration of the types of verbs that undergo this agreement process will reveal that it is not an agreement phenomenon, nor a syntactic phenomenon distinct from the other semantically based functions of reduplication, but that it instead falls in quite naturally with the other uses of reduplication, once we consider the lexical aspectual class of the reduplicating verb within the semantics of pluractionality (Lasersohn 1995; Newman 1980, 1990). As well as bringing this class of reduplication in line with others, our paper will examine several other aspects of the morphophonology and semantics of Niuean verbal reduplication.

2. Reduplicative Accord.

Seiter (1980) notes that some verbs in Niuean reduplicate to agree with a plural absolutive subject of an intransitive verb, as in (1). With a few verbs, plural agreement is indicated by suppletion rather than by reduplication (2). Certain transitive verbs indicate agreement with an absolutive object (3). Other verbs do not exhibit agreement at all (4).²

(1)

1. a. Ne   hoko    mai  a     Sione.
   PAST  arrive     there  ABS  Sione
   'Sione arrived/came there.'

2. b. Ne   hohoko mai  a     laua.
   PAST  arrive.PL  there  ABS   they
   'They arrived/came there.'

(2)

1. a. To   fano    a     au ...
   FUT   go       ABS  I
   'I will go ...'

2. b. To    ō      a    tautolu ...
   FUT   go.PL    ABS  we (INCL)
   'We will go ...'

(3)

1. a. Kua  hala    e     ia   e     lā   akau.
   PERF  cut       ERG   he    ABS  branch tree
   'He cut the branch.'

2. b. Kua  hahala  e    ia    e    tau   lā    akau.
   PERF  cut.PL    ERG  he    ABS   PL    branch tree
   'He cut the branches.'

(4)

1. a. Kua  tomo   e     vaka.
   PERF  sank      ABS   canoe
   'The canoe sank.'

2. b. Kua  tomo   e    tau   vaka.
   PERF  sank      ABS   PL    canoe
   'The canoes sank.'

Seiter refers to this reduplication as agreement, but it has a number of unusual properties in comparison with standard verb-argument agreement. For example, not all verbs agree with a plural argument, and agreement is just for number, and not for person. In addition, agreement is not triggered by a consistent surface grammatical function. This last point is illustrated in (5), from...