Journal of Pidgin and Creole Languages - Volume 6, Issue 2, 1991

A conspicuous feature of Palenquero (P) morphosyntax is the unusual, though by no means unique, often clause-final or sentence-final placement of the predicate negator nu 'not', e.g., i [NU] kelé bae aya NU 'I don't want to go there'. Previous descriptions, often cited in the literature, present oversimplified (and, therefore, misleading) accounts of the data. This paper attempts to rectify the descriptions found in Bickerton & Escalante [B&E] (1970), Lewis (1970), Friedemann & Patino Rosselli [F&P] (1983), and Megenney (1986).Based on an extensive corpus of informal spoken data, it is demonstrated that Palenqueros (Ps) systematically use not one but two basic creole negation patterns — preverbal and postverbal nu — and that the selection of construction signals important discourse pragmatic information. The study goes beyond earlier accounts by providing evidence that the syntax of postverbal nu is not, as generally claimed, utterance-final. The frequent placement of this negator at the end of a clause or sentence depends rather on the absence of constructions which occur infrequently. Such facts, together with the argument that a preverbal negation strategy cannot be viewed as a simple case of interference from Spanish (S), show that, contrary to the opinion of previous investigators, P negation exhibits virtually no signs of decreolization.

A substantial pidgin text recorded in the western Solomons in 1893 is analyzed and placed in historical context. Showing complex syntactic pat-terns, the text is strikingly similar to twentieth century Solomons Pijin. It corresponds closely to an earlier text from the same area and to contem-poraneous texts from New Guinea, lending support to arguments for an early expansion of pidgin and for a monolectal origin for modern Melane-sian Pidgin dialects. The text, including hem i sequences and the "predi-cate marker" i, attests to the early development of an Oceanic pronominal pattern in the Solomons.

The use of the Tok Pisin particle i, traditionally known as the "predicate marker," is shown to be overwhelmingly subject to two constraints, i.e., that of "nondeictic" use (as controlled by subject arguments), and that of "resumptive" use, with the latter sometimes overriding the former. Here, nondeictic means any nonpronominal or pronominal nonfirst and nonsecond person NP. The hypothesis concerning the first constraint seems weakened by apparent counterevidence in the use of i as triggered by -pela pronouns, by the failure of prima facie nondeictic subjects in equa-tional constructions to trigger the use of i, and by the use of i with follow-up verbs in serial constructions and with downstairs verbs in causative constructions. However, all these apparent exceptions are shown either to serve primarily resumptive rules or to be in fact compatible with the non-deixis rule. A corollary reviews openness of i to reanalysis as a copula.