| This study attempts to measure by objective means whether there exist near-native speakers (NNSs) of English who, having begun their acquisition of English after the critical period (set at 12 years old for this experiment) have attained a competence level similar to that of native speakers (NSs). All subjects (20 anglophones, 20 francophones, and 20 sinophones) (1) narrated a 7.5 minute silent film, (2) answered questions and (3) completed sentences about the film. This recorded data was transcribed and the tense/aspect of each verb used by each subject was counted. From these verb counts, average percent anglophone verb usage for each tense/aspect was calculated, and all subjects were measured against these native speaker averages.;In addition, the subjects (4) gave scalar grammaticality judgments on 85 sentences, half of which contained ungrammatical or problematic tense/aspect usage, and (5) filled in a forced-choice cloze test of 10 blanked-out verbs. Once again, native-speaker averages served as a yardstick in comparing the grammaticality judgments and the verb choices of all 60 subjects.;When the data was compared at the group level, clear group differences showed up in all 5 experiments. As a general rule, the francophones' linguistics and metalinguistic productions were more similar to those of the anglophones. Two different discriminant analyses were performed--one based on linguistic production data and the other on metalinguistic judgment data. The former classified 81.67% of the subjects in their correct linguistic groups, and the latter was 91.67% accurate in classifying the subjects according to linguistic group.;Combining the data in various ways, it was possible to rank the subjects from 1 (most anglophone-like) to 60 (least anglophone-like) by four different ranking methods: two based on linguistic performance, and two others based on metalinguistic judgments. Calculating Kendall's coefficient of concordance for these four different ranking methods showed a highly significant correlation between them (W =.6065, df = 59, p =.001), establishing the coherence between the subjects' linguistic performance and their metalinguistic judgments.;Averaging the four different rankings into an overall ranking discovered four NNS subjects indistinguishable from anglophones based on their linguistic and metalinguistic performance.;Therefore, this study provides evidence that some NNSs who acquire English after the critical period do appear to attain a native-like competence, as far as verb tense and aspect are concerned. In addition, a clearly significant correlation was demonstrated between linguistic performance data and metalinguistic judgment data. |
