双语者的两种语言是同时激活的（Kroll, Dussias, Bice, & Perrotti, 2015; Thierry & Wu, 2007），因此，当双语者在语言间进行转换时，离不开双语控制系统的参与。研究者发现，双语控制会对一般性执行控制产生影响（Bialystok, 2017; Bialystok & Viswanathan, 2009; Costa et al., 2009; Coderre & van Heuven, 2014; Hilchey & Klein, 2011; Kovács & Mehler, 2009; Poarch & van Hell, 2012），并提出了双语者的抑制控制优势假设（inhibitory control advantage hypothesis, ICA）和执行加工优势假设（executive processing advantage hypothesis, EPA）。这两种假设都认为，双语控制提高了个体的一般性执行控制能力，但前者认为双语控制提高了一般性抑制能力，后者则认为提高了一般性监控能力（Coderre & van Heuven, 2014; Hilchey & Klein, 2011）。然而，前人的研究主要集中于语言产生中的双语转换对一般性执行控制的影响。我们知道，语言产生和语言理解密切相关，双语者在产生两种语言时离不开对两种语言的理解，但语言产生和语言理解本身的加工过程并不相同。鉴于此，本研究旨在探讨语言理解中的双语转换对一般性执行控制的影响。 具体来讲，我们选取非平衡的汉-英双语者为被试，通过图-词匹配任务设置了单语和双语理解环境，并比较不同语言理解环境下的一般性执行控制表现。本文共包括四个研究，7个实验。研究一和研究二采用行为实验的方法，分别从听觉理解和阅读理解两个方面，考察双语理解转换对一般性执行控制的影响。研究三和研究四利用ERP技术，进一步考察双语理解转换对一般性执行控制的影响，以及如何产生影响。 研究一包括三个实验。实验1采用跨任务适应范式，通过交叉呈现听觉理解试次和Flanker试次，探讨双语听觉理解转换对一般性执行控制的影响。采用图-词匹配任务设置了汉语（L1）、英语（L2）和汉-英双语三种理解条件，比较Flanker任务在三种条件下的表现。结果发现，相较于单语条件，双语理解条件下Flanker任务一致试次和不一致试次的反应时更短，说明双语理解转换提高了个体的一般性监控能力，这一结果支持了EPA假设。 实验2采用Block任务范式，进一步探讨双语听觉理解转换对一般性执行控制的影响。该任务范式是依次呈现听觉理解任务的block和Flanker任务的block。结果同样发现，双语理解条件下，Flanker任务一致试次和不一致试次的反应时更短，该结果进一步说明了双语理解转换提高了个体的一般性监控能力。 为了检验实验1的结果是否与双语条件下被试的警觉性提高有关，实验3采用跨任务适应范式，要求被试在单语和双语理解条件下分别完成非冲突的任务。结果显示，在不同语言理解条件下，非冲突任务的表现相当，说明了实验1的结果并非是个体在双语条件较高的警觉性造成的。 研究二包括两个实验，考察双语阅读理解转换对一般性执行控制的影响。实验4采用跨任务适应范式，交叉呈现视觉的图-词匹配试次和Flanker试次。结果发现，双语条件下Flanker任务一致试次和不一致试次的反应时更短，说明双语阅读理解转换提高了个体的一般性监控能力。实验5则利用非冲突任务，进一步考察实验4的结果是否与双语条件下被试的警觉性提高有关。结果发现，非冲突任务在三种阅读理解条件下的表现相当，说明了实验4的结果不是由个体在双语条件下的高警觉性造成的。 研究三包括一个实验。实验6采用高时间分辨率的电生理学技术，进一步探讨双语听觉理解转换对一般性执行控制的影响。结果显示，在Flanker任务中，双语理解条件下的反应时短于单语理解条件，P3波幅小于单语理解条件。而且，这种差异同时出现在Flanker任务的一致试次和不一致试次中。在图-词匹配任务中，双语理解条件所诱发的晚期正成分（LPC）的波幅大于单语理解条件。这些结果说明了双语理解转换提高了个体的一般性监控能力，并且这种促进作用可能与词汇水平的双语转换控制有关。 语言理解包含词汇的音/形加工，也包含更高级的语义信息加工。基于前面研究的结果，尚不清楚双语条件对一般性执行控制的影响是由最基本的音/形转换加工造成的，还是需要语义加工的参与。研究四（实验7）采用电生理学技术，进一步考察了双语条件对一般性执行控制影响的来源。为此，实验7要求被试学习两种新语言（德语和日语）的词汇，并在学习之前和之后两个阶段，分别对其一般性执行控制的表现进行测查。在学习之前，被试无法对德语和日语词汇进行语义加工，双语条件只包含语音层面的转换；学习之后，被试可以理解两种语言的词汇，双语条件同时包含了语音和语义层面的加工。结果显示，在学习之前，单语和双语条件下Flanker任务的表现相当；但在学习之后，双语理解条件下Flanker任务一致试次和不一致试次的N2波幅小于单语理解条件，P3波幅大于单语理解条件。实验7的结果说明，双语理解条件对一般性执行控制的影响来源于语义的加工，并非只要有语音的转换就能产生这种影响。 总之，本研究发现，双语理解转换影响个体的一般性执行控制，表现为提高了个体的一般性监控能力。双语理解转换在一般性执行控制中的作用具有语义加工的特异性，离不开语义信息的参与。依据双语控制领域的语言模式连续体假设（Grosjean, 2001, 2012），我们提出了双语执行控制效应的连续体假设，强调双语者的语言理解对一般性执行控制的影响是连续变化的。本研究的结果不仅揭示了双语理解转换对一般性执行控制的影响，也尝试为当前的理论争议提供新的解释。

Usually, the two languages of bilinguals are simultaneously active during language production and language comprehension. Due to the characteristic of parallel activation, bilinguals have to employ language control mechanism to coordinate two languages, and then have an effect on domain-general executive control. Some evidence revealed the positive effect of language control on domain-general executive control in bilinguals (Bialystok, 2017; Bialystok & Viswanathan, 2009; Costa et al., 2009; Coderre & van Heuven, 2014; Hilchey & Klein, 2011; Kovács & Mehler, 2009; Poarch & van Hell, 2012). And bilingual inhibitory control advantage hypothesis and bilingual executive processing advantage hypothesis have been proposed, respectively emphasizing the positive effect of bilingual language control on inhibitory control and monitoring (Coderre & van Heuven, 2014; Hilchey & Klein, 2011). Even though the effect of language production switching has been examined in bilinguals, it is not clear about the effect of language comprehension switching, which is accompanied with language production. Language comprehension is as important as production, but they involve different control processing in bilinguals. Therefore, it is necessary to explore whether and how language comprehension switching influence domain-general executive control in bilinguals. In order to explore the effect of bilingual comprehension switching on domain-general executive control, the current study employed the Chinese-English bilinguals, and compared the executive control performance in single- and dual-language comprehension contexts. There were four sections in the current study, totally including seven experiments. The first two sections included five behavioral experiments to examine the effect of listening comprehension switching (Experiment 1-3) and reading comprehension switching (Experiment 4-5) on domain-general executive control in bilinguals. The last two sections examined how language comprehension switching influence domain-general executive control in bilinguals by ERP experiments. There were three behavioral experiments in the first section, in order to explore the effect of listening comprehension switching on domain-general executive control in bilinguals. Experiment 1 used the cross-task adaptation paradigm, and created three language comprehension contexts, with L1 block, L2 block and a dual-language bilingual block. The results showed a positive effect of listening comprehension switching on the congruent and incongruent trials of Flanker task, suggesting that there was a positive effect of listening comprehension switching on monitoring in bilinguals. Experiment 2 employed the Block-task paradigm to further examine the effect of comprehension switching. In Block-task paradigm, language comprehension task block and Flanker task block were presented in sequence, resulting in the same task demand of Flanker task block among three language contexts. The results showed that the overall performance of Flanker task was enhanced in dual-language bilingual context in comparison to single-language contexts, in line with the results of Experiment 1. Experiment 3 examined the potential confounding effect of alertness factor by a control task without conflict control demand. The results of Experiment 3 showed the similar performance of control task between single- and dual-language contexts. Combined Experiment 3 with Experiment 1, the similar performance of control task excluded the potential effect of alertness. There were two behavioral experiments in second section, in order to explore the effect of reading comprehension switching on domain-general executive control in bilinguals. The cross-task adaptation paradigm was adopted in Experiment 4, and found that reading comprehension switching enhanced the overall performance of Flanker task, suggesting that reading comprehension switching in dual-language bilingual context facilitated domain-general monitoring control. Moreover, Experiment 5 excluded the confounding effect of alertness factor, demonstrating the facilitative effect of reading comprehension switching was independent of alertness factor. The third section continuously examined the effect of language comprehension switching on domain-general executive control by event-related brain potentials (ERPs) technique. Experiment 6 used cross-task adaptation task, and simultaneously recorded the language comprehension processing and Flanker task processing. The results of Experiment 6 showed that in Flanker task, both congruent and incongruent trial of Flanker task elicited smaller P3 amplitude in dual-language context, with shorter RT in dual-language context. Moreover, in picture-word matching task, bilingual comprehension elicited a larger LPC amplitude compared with single-language comprehension, suggesting that bilingual comprehension relied more on the language control in lexical level. Experiment 6 provided electrophysiological evidence for the effect of bilingual comprehension switching on domain-general executive control. Language comprehension contains phonological processing and semantic processing. However, it is not clear that the positive effect of language comprehension switching originated from the phonological processing or semantic processing in dual-language bilingual context. Therefore, in the fourth section, Experiment 7 asked a group of bilinguals learn the new words of German and Japanese, and examined the domain-general executive control before and after word learning phase. There was only phonological processing in the pre-test phase, without semantic processing because participants had no knowledge about German or Japanese. Whereas, in the post-test phase, participants could process and control new words, containing both phonological processing and semantic processing. The results showed, in the pre-test phase, the Flanker task performance was similar between single- and dual-language contexts; whereas, in the post-test phase, there was a positive effect of bilingual comprehension switching on the overall performance of Flanker task. The results in Experiment 7 suggested the positive effect on domain-general executive control was related with semantic processing in bilingual comprehension switching. In summary, the dissertation found that language comprehension switching played a role in domain-general executive control in bilinguals, with the positive effect of bilingual comprehension on domain-general monitoring control. Moreover, the positive effect originated from the semantic processing in bilingual comprehension switching. Thus, based on the language-mode continuum hypothesis (Grosjean, 2001, 2012), we proposed the executive control continuum hypothesis in bilinguals, emphasizing the continuous variation of bilingual language control effect on domain-general executive control. Together, the dissertation revealed the effect of bilingual comprehension switching on domain-general executive control.