Introduction

According to a UNESCO report (Bjeljac-Babic 2000), there are some 6,000 spoken languages in the world. Anyone who has worked in the field in Africa, Italy, South Asia, or many other parts of the world may rightly feel that this is an underestimation. Whatever the actual number, the variations we can witness are enormous, and this diversity is something to celebrate. ¹ A part of our work as linguists is to delve deeply into the workings of a particular language in order to try to come to an understanding of its internal logic. At the same time, as linguists working in the generative tradition, we want to know what these languages have in common, for the basic tenet that guides our work is that each language exemplifies a general structure employed by all human languages (Chomsky 1975, 77). The combination of looking deeply into a particular language and at the same time keeping the universality of human language in mind naturally leads to a way to do our work that is both interesting and illuminating. When studying a language, we tend to focus on some element that is unusual about that language, and we take it up as a challenge to try to figure out how the chosen element works. We see this pattern repeated over and over in linguistics research, with studies of such phenomena as the unusual transitive expletive construction in Icelandic, the mysterious auxiliary selection in Romance, and the extraordinary agreement system in Bantu. It is quite a feat simply to elucidate how, for example, the Bantu agreement system works, and one can devote a lifetime to such a task without coming even close to a full understanding of the system at work. These unusual features of a particular language are inherently interesting because they represent the diversity of human language, how languages can vary, and vary greatly. At the same time, as we learn the inner workings of these unusual facets of a particular language, we begin to explore how they tap the resource of the general linguistic structure, thereby illuminating what the underlying universal system holds. In this way, the unique properties of a particular language do not comprise what are essential only to that language, which would isolate the language from the rest, but rather, they demonstrate the possibilities of the general system at work, thereby uniting the language under study with the untold others that are spoken around the world.