INTRODUCTION: Over the last few years, Hakka Creole (Taiwan) has gained considerable attention in linguistic circles, due to several truly strange features of the language. This article presents details of yet another such surprising discovery—one which may well revolutionize the way linguists deal with texts of substantial size in the future.

THEORETICAL PREREQUISITES: (Editor’s note: Mr. Slater’s article was too long. Therefore, we cut out the part about theoretical prerequisites entirely. If, as a result, you have trouble understanding the rest of the article, too bad. If you really want to know what he said here, contact Babel’s editorial offices.)

DISCOURSE GENDER IN HAKKA CREOLE: Once considered an essentially semantic tool (correlating with “real” gender in European languages), gender is now politely recognized for the purely grammatical functions it actually performs. Primarily, grammatical gender serves to classify words (usually nouns) into distinct categories, with the result that speakers are able to keep them mentally distinct, as well. Although not present in all languages, this feature is common enough to suggest that it is a universal tendency of human linguistic systems.

A systematic study of 526 naturally occurring Hakka Creole texts has revealed that an equivalent of grammatical gender operates at the higher level of discursive structure, serving essentially the same discriminatory function that its better-known cousin serves within the syntactic matrix.

The discourse gender distinction is decidedly bipartite in Hakka Creole, although this is by no means logically necessary; field analysts are encouraged to posit as many discourse genders as seem necessary to adequately characterize whatever variation they may encounter in their language. Once it has been conclusively demonstrated (as per this study) that discourse gender exists, analysts are free to make use of it in whatever form they find adequate.

It should further be noted that discourse gender goes a long way towards explicating many of the inter-propositional relationships which linguists have until now been unable to characterize with any regularity. Such higher-level relationships as have often been labelled, for example, “reason-result” and “protasis-apodosis” make far more sense within a framework that allows for discourse gender than ever before; what once required essentially ad hoc description may now be systematically classified with respect to organization at a substantially higher level of structural embedding.

Furthermore, and perhaps most exciting of all, it will no longer be necessary to refer to texts with such artificially constructed labels as “hortatory,” “narrative,” and “procedural.” Now, the analyst is enabled to group texts together in a way both etically and emically significant, having recognized the psychologically real distinctions actually imposed by the speakers of the language in question!

	Letters to the Editor
	Moving Greek Letters—Gianlorenzo Bernini
	Babel Vol I, No 1 Contents