On So-called Triplication in Colloquial Singapore English and Thao: A Response to Blust

Abstract

A morphological pattern originally exemplified by forms from Thao, an aboriginal language of Taiwan, finds further exemplification in Colloquial Singapore English, and resultant theoretical clarification.

1. Introduction.

In his richly documented analysis of aspects of the semiaffixal¹ morphology of Thao, Blust (2001) identifies the following morphological processes: (1) full reduplication, (2) CA- reduplication, and (3) rightward reduplication. These processes are grouped together under the traditional taxonomic label "reduplication," and various combinations of them are referred to by him as "serial reduplication," "triplication," and "quadruplication." B holds the view that whereas "serial reduplication" is not a unified morphological process, "triplication" is, and he proceeds to offer some insightful observations on the domain and interpretive limitations of the latter. The main point of B's paper therefore is to argue for the existence of "triplication" and to suggest some of its possibly universal properties.

In this response, we argue that (1) although B deserves full credit for recognizing the purely morphological nature of the processes listed above, his analysis seems to suffer from a confusion between morphological processes and the patterns that result from the operation of these processes. The revisions made necessary by the removal of the confusion inadvertently introduced and nurtured by B, we shall argue, in fact constitute improvements. In order to avoid problems of the sort B creates for himself, we shall use terms like "triplication" to describe specific patterns of morphological marking that result from an appropriate application of the morphological processes identified above. The derivation of these patterns is, in other words, an entirely different kettle of fish (we shall return to the matter later in this response).²

(2) His very interesting empirical findings regarding the nature of triplication are, unfortunately, not given any theoretical grounding; they merit serious attention and, we believe, when pursued to their logical end, lead to some important conclusions regarding the relationship between iconicity and reduplication as a pattern of morphological marking, derived, we remind the reader, from an application of any one of the processes identified above.

(1) and (2) are related in that, as we shall see, B claims that triplication only occurs with verbs. For this purpose, his claim that "serial reduplication," by which he means derivations requiring two "unrelated" (333) morphological operations, does not count as triplication is important because the sort of derivational history it entails is apparently not needed for verbs, but only for numerals. In our response, although we will argue against B's attempt to separate serial reduplication from triplication, we will attempt to preserve his insight that there is indeed a special relationship of iconicity involving triplication (= a pattern, we repeat, obtained by more than one application of a reduplication process or by application of two distinct reduplication operations/processes) and verbs.

In section 2, we provide a brief review of B's work. This is followed in section 3 by a presentation of data from Colloquial Singapore English (CSE),³ an analysis of which we believe contains implications for B. Section 4 contains a number of speculations on the nature of the kind of iconicity involved.

2. Blust 2001.

According to B (drawing on Chang 1998), there are three major reduplication processes in Thao: full reduplication, CA- reduplication, and rightward reduplication.⁴Table 1 includes some of the examples he gives of each of these processes. While the term "full reduplication" is used, consonant codas are actually invariably omitted so that the reduplication is in fact slightly less than full. More relevant for the discussion here, however, is the fact that full reduplication may operate on either verbs (to signal "repetitive or continuative aspect") or adjectives (to signal intensity) (Blust 2001:325). In the case of CA- reduplication, the syntactic category of the base is also variable; it includes nouns in addition to verbs. Rightward reduplication is described as having a "generally similar semantic content" to full reduplication (326). Presumably this is because rightward reduplication also tends to indicate repetition or continuation with verbs; it also seems to...