Bilingual speakers show a response time (RT) cost when switching between languages. These costs could reflect the organization of language in a shared bilingual lexicon (Grainger, Midgley, & Holcomb, 2010) or a domain general cognitive processing cost (Green & Abutalebi, 2013). To test these accounts, we analysed RT distributions of bilingual (Spanish-English) performance on generalized lexical decision (GLD) tasks using Ratcliff (1978) diffusion model. Experiment 1 revealed that language switches decrease the rate of evidence accumulation (drift rate) and slow the cognitive processes that occur prior to decision-making (non-decision time). Experiment 2 showed that the anticipation of language switches did not change these effects. The results suggest that language switch costs originate from a combination of at least two loci: lexical access and a task-specific decision process.

中文翻译：

---

扩散双语词典：语言转换成本的基于任务和词汇的组成部分

双语者在语言之间切换时会显示响应时间 (RT) 成本。这些成本可以反映共享双语词典中的语言组织（Grainger、Midgley 和 Holcomb，2010 年）或域一般认知处理成本（Green 和 Abutalebi，2013 年）。为了测试这些帐户，我们使用 Ratcliff (1978) 扩散模型分析了在广义词汇决策 (GLD) 任务上双语（西班牙语-英语）性能的 RT 分布。实验 1 表明，语言转换会降低证据积累率（漂移率）并减慢决策之前（非决策时间）发生的认知过程。实验 2 表明，对语言转换的预期并没有改变这些影响。结果表明，语言转换成本源于至少两个基因座的组合：