Ecological Approach to Vocabulary Density Estimation

Abstract

This study scrutinizes to what degree an extension of the Capture-Recapture (CR) approach, extrapolated from ecology, gives us further insight into second language (L2) productive lexical density. Previous CR studies that have adopted this method to estimate L2 productive vocabulary size have typically taken only two written samples, produced within a few days of each other, from the participants. Results have invariably shown that this approach does not measure absolute vocabulary size. Another issue that compromises the CR scores is the kind of task used to elicit samples of productive vocabulary. In ecology there is an assumption that all animals of the target population have an equal chance of being captured. In a lexical context this assumption would translate as that all vocabulary items have an equal chance of being produced. This assumption presents a serious problem to overcome because different lexical items have different probabilities of being elicited. This paper addresses both issues. On the one hand, three samples from 26 Thai students of Spanish as L2 have been taken within a period of two weeks (referred to as “an extension of the CR method”). On the other hand, a context-free task based on the Roman alphabet was used to elicit productive vocabulary. The results of this approach were compared to those of the CR method (using only the first two samples). The overall estimates are higher when using three samples. However, once again, the scores do not give absolute vocabulary size. A discussion of the method’s assumptions is provided together with an interpretation of the results.