Frequency effects in language, vol. 1: Frequency effects in language learning and processing. Ed. by Stefan Th. Gries and Dagmar Divjak. (Trends in linguistics: studies and monographs 244.1.) Berlin: De Gruyter Mouton, 2012. Pp. v, 243. ISBN 9783110274059. $140 (Hb).
Reviewed by Natalia Levshina, Université catholique de Louvain
The fact that usage frequency has an effect on linguistic form and function has been known since George Zipf’s work. However, it is only recently that frequency has gained its central place in usage-based theory and methodology. Although there seems to be a general concensus that frequency ‘matters‘, the exact nature of its influence on language use, acquisition, and change is still not fully understood. The high-quality articles solicited for this book as well as for its companion book offer an important contribution to our understanding of what frequency effects are and why they exist.
The studies cover a wide range of topics: phonology and morphology, multiword units and syntactic variation, children’s language, second language acquisition, and language change. The data are impressively diverse and include corpus-based and experimental evidence. The latter ranges from classical acceptability judgments and questionnaires to less familiar techniques, such as continuous shadowing and sentence copying. Although most studies report convergent evidence from different sources, this methodological pluralism also allows the researchers to overcome the limitations of using only one method.
However diverse, the articles in the book are connected by one leitmotif: our application of frequency measures and intepretation of their effects should be very nuanced. This is illustrated by a quote from Albert Einstein in Nick C. Ellis’ article: ‘Everything that can be counted does not necessarily count; everything that counts cannot be necessarily be counted’ (7). As Ellis argues in his overview of the state of the art, corpus frequencies on their own are uninformative and should be integrated in multifactorial dynamic models of language learning, development, and use. The behaviorist-like understanding of frequency effects as mere repetition should be replaced with a contextually informed one. This contextualization can be interpreted, for instance, as the cumulative exposure of a word to specific phonological environments. On the other hand, frequency effects of multiword expressions can be contextually enhanced by the presence of additional lexical clues.
Speakers are also sensitive to relative frequency measures (e.g. the difference between the frequency of a base verb and that of a prefixed form), as shown by Eugenia Antić in her experiments with morphological productivity of prefixed verbs. Vsevolod Kapatsinski’s study provides experimental support for the well-known thesis of Joan Bybee that product-oriented generalizations of stimulus frequencies are more important that the source-oriented ones. Nuancing is also needed at the level of interpretation of linguistic facts; frequency should not be regarded as a magic all-explaining factor. Gunther De Vogelaer’s study of standardization and resemantization in the gender system of Dutch dialects shows that corpus frequency can predict processes related to Labovian diffusion (standardization), but not transmission (resemantization), where the age of acquisition is a better approximation of entrenchment.
The book can be recommended to anyone interested in empirical methods and theoretical developments in contemporary usage-based linguistics.