The new big picture for linguistics: Complex systems

Abstract In the history of linguistics there have been crucial moments when those of us interested in language have essentially changed the way we study our subject. We stand now at such a moment. In this presentation I will review the history of linguistics in order to highlight some past important changes in the field, and then turn to where we stand now. Some things that we thought we knew have turned out not to be true, like the systematic, logical nature of languages. Other things that we had not suspected, like a universal underlying emergent pattern for all the features of a language, are now evident. This emergent pattern is fractal, that is, we can observe the same distributional pattern in frequency profiles for linguistic variants at every level of scale in our analysis. We also have hints that time, as the persistence of a preference for particular variants of features, is a much more important part of our language than we had previously believed. We need to explore the new realities of language as we now understand them, chief among them the idea that patterned variation, not logical system, is the central factor in human speech. In order to account for what we now understand, we need to get used to new methods of study and presentation, and place new emphasis on different communities and groups of speakers. Because the underlying pattern of language is fractal, we need to examine the habits of every group of speakers at every location for themselves, as opposed to our previous emphasis on overall grammars. We need to make our studies much more local, as opposed to global. We do still want to make grammars and to understand language in global terms, but such generalizations need to follow from what we can now see as the pattern of language as it is actually used.