Measured Language: Quantitative Studies of Acquisition, Assessment, and Variation

Introduction

xiii ■Introduction JEFF CONNOR-LINTON AND LUKE WANDER AMOROSO Georgetown University ■ THE GEORGETOWN UNIVERSITY ROUND Tables on Languages and Linguistics (GURT) and their attendant published volumes of selected papers have historically offered seminal contributions to the field of linguistics; papers from GURT volumes are frequently referenced in journal articles and are often required reading for graduate linguistics courses. The continued relevance of GURT publications stems from their coverage over the years of the many different areas of inquiry that constitute linguistics . With topics ranging from language teaching and education to discourse analysis, technology, and language-specific research, GURT conferences draw linguists from around the world. The 2012 Round Table focused on the ways in which various aspects of language can be quantified, and how measurement informs and advances our understanding of language. The metaphors and operationalizations of quantification serve as an important lingua franca for much linguistic research, allowing methods and constructs to be translated from one area of linguistic investigation to another. A primary goal of the conference and this volume was to provide a forum for exploring relations among constructs developed from seemingly disparate theoretical and methodological perspectives. While many previous GURT volumes have focused on a single method of linguistic analysis or considered a particular theoretical problem from many different perspectives, we took a different approach. We called for studies that employed quantitative methods of measuring language acquisition, assessment, and variation in the belief that quantification for one particular purpose would prove useful to researchers investigating other linguistic questions. We believe that researchers in these areas of linguistics have much to learn from each other, both conceptually and methodologically, so conference contributors were asked to share the relevance of their perspectives and findings to other areas of linguistic inquiry. For this volume we have selected papers that illustrate forms of measurement and quantitative analysis that are current in diverse areas of linguistic research, from language assessment to language change, from generative linguistics to psycholinguistic experimentalism, and from longitudinal studies to classroom research. The range and clarity of the research collected here ensures that even linguists who would not traditionally use quantitative methods will find this volume useful. Quantification is central to many of the core principles of the modern scientific method—from falsifiability of hypotheses to arguments for the reality of underlying constructs, reliable measurement of those constructs, and the use of inferential xiv Jeff Connor-Linton and Luke Wander Amoroso statistics to estimate the risks of generalizing results of a study. As the philosopher of science Karl Popper pointed out, we cannot conclusively affirm a hypothesis—we can only conclusively negate it. Therefore, the validity of a claim (knowledge) is related to the probability that that claim can be falsified. Measurement and quantification help us to operationalize constructs and relations that comprise hypotheses. The more specifically and concretely a hypothesis is stated, the greater the possibility that it can be negated, and the more clearly its validity can be tested in the “trial by fire” of replicated research. The importance of replicable research—and therefore replicable measurement—to linguistics cannot be overstated; it is what separates a coherent field of study from a loose collection of individual case studies. Quantification also enhances the scientific process dialectically through the rigor of explicitly operationalizing a construct, developing a reliable way to measure it, and validating that measurement. One of the more obvious values of quantification is that it facilitates classification and comparison of phenomena. In his plenary paper, “The Ubiquitous Oral versus Literate Dimension: A Survey of Multidimensional Studies,” Douglas Biber first reviews his original multidimensional analysis of register variation in English speech and writing, in which he compared the frequencies of use of a variety of linguistic features across genres. His analysis of functionally motivated co-occurrence relations among linguistic features transmuted the qualitative folk notion of genres into measurable function-motivated text-types. The foundation of co-occurrence as an indicator of a shared function then allows him to compare how languages and their cultures build and organize their distinctive repertoires of text-types. Where Biber uses factor analysis to identify underlying co-occurrence patterns, Emily Gasser uses Bayesian modeling in her paper titled “Subgrouping in Nusa Tenggara: The Case of Bima-Sumba” to expose underlying lexical similarities and differences—and therefore historical relations—between languages in the Central Malayo-Polynesian family. Gasser provides an example of an approach which, in addition to historical applications, has been useful in modeling morphology, child language acquisition, and sentence...