View in EDS HTML Full Text PDF Full Text

When Language Switching Is Cost-Free: The Effect of Preparation Time

Saved in:
Bibliographic Details
Title: When Language Switching Is Cost-Free: The Effect of Preparation Time
Language: English
Authors: Mosca, Michela (ORCID 0000-0002-1209-1304), Manawamma, Chaya, de Bot, Kees (ORCID 0000-0001-7442-2693)
Source: Cognitive Science. Feb 2022 46(2).
Availability: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
Peer Reviewed: Y
Page Count: 24
Publication Date: 2022
Document Type: Journal Articles
Reports - Research
Descriptors: Language Usage, Indo European Languages, English (Second Language), Bilingualism, Code Switching (Language), Cognitive Processes, Time
DOI: 10.1111/cogs.13105
ISSN: 1551-6709
Abstract: Previous research has shown that language switching is costly, and that these costs are likely to persist even when speakers are given ample time to prepare. The aim of this study was to determine whether there are cognitive limitations to speakers' ability to prepare for a switch, or whether a new language can be prepared in advance and any cost to switch language eliminated. To explore this, language switching costs were measured in a group of Dutch-English (L1-L2) bilinguals who named pictures in their two languages while the preparation time was manipulated. The participants were given either no time to prepare (cue to stimulus interval, CSI: 0 ms), or some time to prepare, for the target language (CSI: 250, 500, and 800 ms). The results revealed that when speakers had no time to prepare, language switching was costly. However, when preparation time was provided, switching costs disappeared. This suggests that there might be no cognitive limitations to the ability to prepare for a language switch, and that, provided enough preparation time, the effort to switch language could be eliminated. This finding might also explain why normal code-switched conversations seem effortless, as speakers typically have ample time to prepare for the language switch.
Abstractor: As Provided
Notes: https://osf.io/bnfs8
Entry Date: 2022
Accession Number: EJ1330021
Database: ERIC
Full text is not displayed to guests.
FullText Links:
  – Type: pdflink
    Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwHX0P4rQzp4bpN9FobrBex0AAAA4TCB3gYJKoZIhvcNAQcGoIHQMIHNAgEAMIHHBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDMqWYhuEfmRNF3je-QIBEICBmc_g0L4Olu8Hxamig25zwFgnqykd8-oCNfKmzY67euVTXCHv4kfX77WNQsOOX-Fc3AFbOpD1lt_p85KbzH_qo1asI1X_iqiQfrdevhL_6aS9CY_rWXIOfwNejoTyW09WO1ke2AgSr8kfq-cJKUvabRfroVPOl2hpDgweriA6IMmK6PNMVlbG6itWeF_sWUyNm7R7B_zL20muMQ==
Text:
  Availability: 1
  Value: <anid>AN0155381511;cgn01feb.22;2022Feb24.03:42;v2.2.500</anid> <title id="AN0155381511-1">When Language Switching is Cost‐Free: The Effect of Preparation Time </title> <p>Previous research has shown that language switching is costly, and that these costs are likely to persist even when speakers are given ample time to prepare. The aim of this study was to determine whether there are cognitive limitations to speakers' ability to prepare for a switch, or whether a new language can be prepared in advance and any cost to switch language eliminated. To explore this, language switching costs were measured in a group of Dutch‐English (L1‐L2) bilinguals who named pictures in their two languages while the preparation time was manipulated. The participants were given either no time to prepare (cue to stimulus interval, CSI: 0 ms), or some time to prepare, for the target language (CSI: 250, 500, and 800 ms). The results revealed that when speakers had no time to prepare, language switching was costly. However, when preparation time was provided, switching costs disappeared. This suggests that there might be no cognitive limitations to the ability to prepare for a language switch, and that, provided enough preparation time, the effort to switch language could be eliminated. This finding might also explain why normal code‐switched conversations seem effortless, as speakers typically have ample time to prepare for the language switch.</p> <p>Keywords: Bilingualism; Language switching; Preparation time; Language reconfiguration; Inhibition</p> <hd id="AN0155381511-2">Introduction</hd> <p>Many studies in the last two decades have shown that although bilinguals are remarkably good at switching between their languages, the process is far from effortless. In fact, language switching is not a one‐step process but is supported by a series of complex cognitive processes referred to collectively as "language control," and includes shifting attention to the relevant language, maintaining the new language, and ignoring the other one.</p> <p>Initially, it had been suggested that language control might be supported by a mental device (Penfield & Roberts, 1959) that allowed bilinguals to turn a language on and off (Macnamara, Krauthammer, & Bolgar, 1968). But subsequent studies showed that when bilinguals process one language, the other is also activated, discouraging the idea that languages can be completely turned off (e.g., Colomé, 2001; Poulisse & Bongaert, 1994). Neuroimaging studies then revealed that language switching might be supported by a network of cortical and subcortical brain areas, which are closely related to more general executive functions (e.g., Abutalebi & Green, 2007, 2008; Ma, Li, & Guo, 2016; for a recent review, see Calabria, Costa, Green, & Abutalebi, 2018). Of the executive functions, inhibition is seen as a key mechanism that allows humans to control their behavior (e.g., Aron, 2007; Norman & Shallice, 1986) as well as language switching (e.g., Green, 1986, 1998; Kroll, Bobb, Misra, & Guo, 2008).</p> <p>According to the inhibitory account of language control, when one language is used, the other is suppressed (e.g., Green, 1998, but see La Heij, 2005; Philipp, Gade, & Koch, 2007; Verhoef, Roelofs, & Chwilla, 2009 for alternative accounts). The amount of inhibition applied to a language depends on the dominance of the language, with more inhibition required to suppress the stronger language (e.g., the native language or "L1") than the weaker one (e.g., a second acquired language or "L2"). In the case of language switching, the suppression previously applied to a language needs to be overcome for the language to be successfully activated and later selected. The inhibitory account can be viewed in conjunction with the assumption that the processes at play in the preceding task persist for some time before decaying (e.g., Allport, Styles, & Hsieh, 1994; Wylie & Allport, 2000). Interestingly, recent studies have shown that the interfering effects of a preceding language task can be long lasting, persisting even when the previous language is no longer part of the linguistic task (aka "language after‐effect," e.g., Branzi, Martin, Abutalebi, & Costa, 2014; Wodniecka, Szewczyk, Kałamała, Mandera, & Durlik, 2020).</p> <p>Successful language switching, however, requires more than just the decay of the interference coming from a preceding language task. To change from one language to another, a sort of "mental gear" is needed that actively prepares the system for a new task (e.g., Mayr & Kliegl, 2000, 2003; Meiran, 1996; Rogers & Monsell, 1995; Rubinstein, Meyer, & Evans, 2001). The reconfiguration of the new language includes processes, such as shifting the attention to the new language, retrieving the rules of that language (such as grammatical and phonological rules), and inhibiting the nonintended language. Both interference decay and language preparation need some time to become effective, leading to the so‐called "switching costs."</p> <hd id="AN0155381511-3">The cost to switch language</hd> <p>Language switching costs are measured using a language switching paradigm. In most cases, bilinguals are prompted to name pictures, or digits, in the language indicated by a language cue (e.g., Costa & Santesteban, 2004; Costa, Santesteban, & Ivanova, 2006; Declerck, Kleinman, & Gollan, 2020; Meuter & Allport, 1999; Schwieter & Sunderman, 2008). It is well established that speakers are faster and more accurate when the stimulus is named in the same language as the preceding trial ("repetition trial," e.g., L1‐L1) compared to when a different language is used relative to the previous trial ("switch trial" e.g., L2‐L1). Switching costs are usually measured for all the languages involved in the switching task (e.g., in both the L1 and L2, in the case of a bilingual task). For each language separately, switching costs are calculated as the difference in performance between repetition and switch trials. Note that in this context, repetition and switch trials differ with respect to the language of the preceding trial, while the current language remains the same (e.g., switching costs in the L1 are calculated as L1‐L1 vs. L2‐L1). However, an alternative method is also in use (e.g., Bultena, Dijkstra, & van Hell, 2015a; 2015b), where repetition and switch trials share the language of the preceding trial, but not the language of the current trial (e.g., switching costs in the L1 are calculated as L2‐L2 vs. L2‐L1).</p> <p>While switching cost measurements are a widely used methodology to investigate language control, and while they have been used in this study to measure the cost to switch language, they represent just one way to measure the effort to change language. By looking at speakers' general performance on both repetition and switch trials, some studies have shown that bilinguals might exhibit a "paradoxical" or "reversed language dominance effect," namely, an overall better performance in the weaker than in the stronger language (e.g., Christoffels, Firk, & Schiller, 2007, Christoffels, Ganushchak, & La Heij, 2016; Costa & Santesteban, 2004; Kleinman & Gollan, 2018; Mosca & de Bot, 2017). The "reversed language dominance effect" has often been explained as the tendency, during a mixed language task, of bilinguals to improve the performance in their weaker language by making the stronger language less accessible, namely, by applying some inhibition to it. In some cases, however, the amount of inhibition applied is more than needed, leading to a general performance cost in the stronger language (Declerck et al., 2020).</p> <p>Another indicator of effortful language switching is the so‐called "n‐2 language repetition cost" (e.g., Babcock & Vallesi, 2015; Declerck & Philipp, 2018; Declerck, Thoma, Koch, & Philipp, 2015; Guo, Liu, Chen, & Li, 2013; Guo, Ma, & Liu, 2013). This type of cost is measured by comparing performance in an ABA versus a CBA language sequence (where "A," "B," and "C" represent three different languages). Performance in language A is usually slower in the ABA relative to the CBA sequence, and this has been interpreted as evidence that inhibition persists in a freshly abandoned language, affecting its reactivation (such as in an ABA sequence). Further evidence of persisting inhibition, and so of effortful language switching, comes from studies investigating how exposure to one language can negatively impact the ability to use the other one, even though the previously used language is no longer part of the task (the "language after‐effect," e.g., Branzi et al., 2014; Kreiner & Degani, 2015; Wodniecka et al., 2020).</p> <hd id="AN0155381511-4">Preparing for a switch</hd> <p>While much research has investigated the processes at play during language switching (e.g., inhibition of the nonrelevant language), much less is known about the mechanisms that allow the system to prepare for such a switch. Some insights on how people might prepare for a switch come from the task switching literature.</p> <p>A typical way to investigate switch preparation is by manipulating the interval between the display of the cue––indicating the language of the task to perform, and the presentation of the stimulus––the object on which the language will be used. This is known as the Cue to Stimulus Interval(CSI). The idea is that longer intervals provide more time for task preparation, thus reducing the switching costs. According to the reconfiguration approach (e.g., Meiran, 1996; Monsell, 2003; Monsell & Mizon, 2006; Rogers & Monsell, 1995), to switch tasks requires a kind of mental rearrangement to take place. Task reconfiguration includes processes like shifting attention to a new task, inhibiting the previous task, retrieving new goals (what to do), as well as new rule actions (how to do it), and enabling a new response set. According to this approach, when people are given some time to prepare, these reconfiguration processes can be encoded beforehand, facilitating the switch from one task to another. Providing participants with time to prepare can also reduce the interference coming from the previous task (e.g., Allport et al., 1994; Wylie & Allport, 2000) which, in turn, also facilitates the reconfiguration of the new one.</p> <p>Yet, people's ability to prepare for a switch trial (i.e., a trial where the task to be performed is different than in the trial before) seems to be limited. Originally, it was assumed that preparing for a switch might be only partially possible so that switching costs could be reduced but never fully eliminated (Rogers & Monsell, 1995). It was later proposed that it might be possible to completely prepare for a switch and so eliminate switching costs, but not always. According to de Jong (2000), people can fully prepare a switch but sometimes this mechanism fails. It is not entirely clear what prevents people from consistently preparing for a switch trial despite abundant preparation time, however, to de Jong (2000), the reason might be attributed to a weak link between cue (e.g., blue screen) and action (e.g., use English when the screen color is blue). This weak link can be observed when participants cannot see a benefit in preparing for a switch in advance, or when their fatigue levels are high. This observation would link switch preparation to participants' motivation and alertness levels. However, switch preparation might also fail because of persisting competition from the previously relevant task (e.g., de Jong, Berendsen, & Cools, 1999). To Nieuwenhuis and Monsell (2002), the problem lies in a more general limitation to retrieve the relevant information to prepare the switch, and this is clearly visible even in young, highly motivated, and nonfatigued participants. The failure to prepare a switch was demonstrated by de Jong (2000) through a computational model showing the distribution of reaction times in a group of participants switching between tasks. The model revealed that when people had little time to prepare (CSI: 150 ms), switch trials were responded to slower than repetition trials, however, with longer preparation time (CSI: 1200 ms), the switch trial distribution became more mixed, with only a portion of trials still showing that participants responded slower during switch trials than on repetition trials. Crucially, in another subset of trials, switch trials were responded to as fast as repetition trials. De Jong suggested that when advanced preparation is correctly triggered, no difference is detected between switch and repetition trials. However, failure to prepare causes slower responses on switch compared to repetition trials. As a result, switching costs will be eliminated on some trials but persist on others.</p> <hd id="AN0155381511-5">Language preparation</hd> <p>With regard to language control, past studies have consistently shown that language switching costs can be reduced if the speakers are informed about the language to prepare (e.g., Costa & Santesteban, 2004; Declerck, Koch, & Philipp, 2015; Fink & Goldrick, 2015; Guo et al., 2013; Khateb, Shamshoum, & Prior, 2017; Ma et al., 2016; Mosca & Clahsen 2016, but see Philipp et al., 2007). However, similar to what was observed in task switching, these studies have revealed that people's ability to prepare for a language switch might in fact be limited.</p> <p>Language switching costs are likely to persist even when speakers are given ample time to prepare (e.g., CSI up to 800 ms in Costa & Santesteban, 2004; Declerck, Ivanova, Grainger, & Duñabeitia, 2020; CSI up to 1000 ms in Philipp et al., 2007; CSI up to 1175 ms in Graham & Lavric, 2021; CSI up to 1500 ms in Fink & Goldrick, 2015; Verhoef et al., 2009), but also when the language to use next is predictable (i.e., when the same language sequence is used throughout a task such as in Declerck et al., 2015; Declerck, Philipp, & Koch, 2013; Festman, Rodriguez‐Fornells, & Münte, 2010; Jackson, Swainson, Cunnington, & Jackson, 2001; 2004) or when the speakers voluntarily switch language (e.g., de Bruin, Samuel, & Duñabeitia, 2018; Gollan & Ferreira, 2009; Gollan, Kleinman, & Wierenga, 2014; Gross & Kaushanskaya, 2015). Yet, a recent finding seems to challenge the hypothesis that speakers are unable to fully prepare for a language switch (Mosca & Clahsen, 2016).</p> <p>While investigating the effect of preparation time on language switching in a group of bilinguals performing a cued picture naming task, Mosca and Clahsen (2016) reported significant L1 and L2 switching costs when speakers had no time to prepare (CSI: 0 ms), but no switching costs, in either language, when the same experiment (same participants and same pictures) was conducted with a CSI of 800 ms. To our knowledge, this is the only case of switching cost elimination in a cued paradigm during language production, which makes the finding unique but also in need of further investigation. The result could be explained with the idea that the relatively long interval between trials (>3000 ms) used in the study had minimized the interference coming from the preceding language task, giving space for a language switch to be fully reconfigured when speakers were given time to prepare. If this assumption is correct, persisting switching costs might simply reflect the fact that language preparation is not completed yet, rather than a limitation of the system to prepare all switch trials. More precisely, it would imply that all trials (and not just a portion as previously assumed) can be successfully prepared as long as the interfering effects of a previously relevant language task have decayed.</p> <p>Though the theoretical and methodological issues brought up by Mosca and Clahsen (2016) are potentially relevant, their study relied on a basic experimental design. In their study, in fact, speakers were either provided with some preparation time (800 ms) or none (0 ms). A more fine‐grained manipulation of preparation time should better define the preparatory processes underlying language control.</p> <p>The goal of the present study is to systematically investigate the effect of preparation time on language switching while minimizing the interference coming from the previously relevant language task. To do this, language switching costs were measured in the L1 and L2 of a group of bilinguals performing a picture‐naming task. The interference from the previous language task was neutralized by using intervals of 3200 ms between trials. The ability to prepare for a language switch was assessed by gradually increasing the interval between the language cue and the stimulus to be named: CSI: 0, 250, 500, and 800 ms. The more extreme CSIs of 0 and 800 ms mirror the ones used in Mosca and Clahsen (2016). The halfway intervals of 250 and 500 ms have been introduced to investigate the effects of preparation time on language control at a more granular level.</p> <hd id="AN0155381511-6">Method</hd> <p></p> <hd id="AN0155381511-7">Participants</hd> <p>Thirty participants (age range 19–27 years; six males) took part in the experiment. All participants were Dutch native speakers, born and brought up in the Netherlands, and either undergraduate or postgraduate students living in Groningen at the time of testing. An adaptation of the paper‐based Language History Questionnaire LHQ 2.0 (Li, Zhang, Tsai, & Puls, 2014) was administered to assess their linguistic background. All participants reported having acquired Dutch from birth and learning English after early childhood (mean AoA: 10.2 years, see also Table 1). When asked how frequently they engaged in activities, such as listening to the radio, watching TV, reading for work, and reading on the web, for each of the languages, English and Dutch, 20 participants reported that they did most of these activities in English rather than Dutch. The remaining participants reported engaging in such activities more in Dutch than in English. As to language switching, 18 participants reported mixing words or sentences from the two languages in their speech, while 12 participants answered that they did not usually engage in language switching behaviors. A 7‐point self‐rating scale (1 = very poor and 7 = native‐like) was incorporated into the questionnaire to assess participant's current levels of proficiency in reading, writing, speaking, and listening for the two languages. The self‐rating test revealed that participants considered themselves native speakers of Dutch (mean = 7) and advanced learners of English (mean = 5.6); see Table 1 for details.</p> <p>1 TableMean self‐rating scores (standard deviation in brackets) of language proficiency in Dutch (L1) and English (L2) based on a 7‐point scale (1 = very poor, 2 = poor, 3 = fair, 4 = functional, 5 = good, 6 = very good, and 7 = native‐like)</p> <p> <ephtml> <table><thead><tr><th /><th>Reading</th><th>Writing</th><th>Speaking</th><th>Listening</th><th>Total</th></tr></thead><tbody><tr><td>Dutch (L1)</td><td>7</td><td>7</td><td>7</td><td>7</td><td>7</td></tr><tr><td>English (L2)</td><td>6</td><td>5.3</td><td>5.3</td><td>6</td><td>5.6</td></tr><tr><td>(0.7)</td><td>(0.8)</td><td>(0.7)</td><td>(0.8)</td><td>(0.8)</td></tr><tr><td>English AoA</td><td>10.2</td><td>10.4</td><td>9.9</td><td>–</td><td>10.2</td></tr><tr><td>(1.6)</td><td>(1.4)</td><td>(1.9)</td><td>–</td><td>(1.7)</td></tr></tbody></table> </ephtml> </p> <p>1 Note: At the bottom, mean age of acquisition (standard deviation in brackets) of English in reading, writing, and speaking.</p> <p>Apart from the LHQ questionnaire, a lexical decision task in English, known as LexTale, was administered in order to obtain a more objective measure of their English proficiency. LexTale is considered a good predictor of English vocabulary as well as of general English proficiency (Lemhöfer & Broersma, 2012), requiring participants to decide if the sequence of letters appearing on the screen is an English word or not. The test consists of 60 trials and can lead to a maximum score of 100%. The average score of the participants in the present experiment was 81.8% (sd: 12.19), which, according to the authors, corresponds to a C1 level of the Common European Framework of Reference. Based on the responses from the LHQ questionnaire and the LexTale results, participants of the current study are deemed to be moderate to high proficient Dutch‐English (L1‐L2) bilinguals.</p> <hd id="AN0155381511-8">Materials</hd> <p>Twenty pictures of simple, common objects were used as experimental stimuli, and another 10 pictures were used as practice items. All 30 pictures were taken from the colorized Snodgrass and Vanderwart pictures set (Rossion & Pourtois, 2004; Snodgrass & Vanderwart, 1980). Only pictures with a single response (i.e., no pictures with two or more names, or pictures that yield uncertain responses) were selected. Using the values of the CELEX web database (<ulink href="http://celex.mpi.nl/">http://celex.mpi.nl/</ulink>), words in the two languages were matched for length (number of syllables, all <emph>p</emph>>.05) and frequency (lemma and word form, all <emph>p</emph>>.05). Only visually simple pictures were used, namely, pictures with a visual complexity score ≤ 3 (this is based on the rating scale, where 1 = very simple and 5 = very complex, provided by Rossion & Portouis, 2004). Stimuli were presented at the center of a white, 15.6‐inch computer screen with a resolution of 1366 × 768 pixels using the E‐Prime 2.0 software (Psychology Software Tools, Pittsburgh, PA; Psychology Software Tools Inc., 2012). See Appendix A for the list of the stimuli.</p> <hd id="AN0155381511-9">Procedure</hd> <p>Participants were tested individually in a soundproof laboratory room. Before starting the picture naming task, they were asked to fill out the LHQ questionnaire and complete the LexTale test. They were then instructed to name the pictures displayed on a computer screen in the language that had been indicated by the language cue, with a green background signaling that the image should be named in Dutch, and a blue background signaling the use of English. The color of the language cue was counterbalanced across participants. Participants were asked to be as fast and accurate as possible. The experiment consisted of two types of trials: repetition and switch trials. Repetition trials were trials in which the language of response was the same as the preceding trial (e.g., L1 preceded by L1). Switch trials were trials in which the language of response was different from the preceding trial (e.g., L1 preceded by L2). The experiment consisted of 320 trials organized in pseudo‐randomized chunks. Each chunk included 75% repetition and 25% switch trials. For instance, an "L1‐L1‐L1‐L2" chunk implied that the first three trials had to be named in the L1, while the fourth one had to be responded to in the L2. Only the last two trials of the chunk were included in the analysis (X‐X‐L1‐L2), while the first two elements were used as fillers. The reason for this lies in the fact that we aimed at testing repetition trials coming from "pure" repetition trials (note difference between <emph>L1</emph>‐L1‐L1‐L2 and <emph>L2</emph>‐L1‐L1‐L2) as well as at avoiding any uncontrolled n‐2 repetition effect (e.g., L1‐L2‐L1‐L2, for a review on this effect, see Koch, Gade, Schuch, & Philipp, 2010). There was an equal number of L1‐L1‐L1‐L2 and L2‐L2‐L2‐L1 chunks. Preparation time was manipulated by using different intervals between the language cue and the stimulus. There were four experimental blocks, one for each CSI: 0, 250, 500, or 800 ms. Each block included th e same number of chunk types. The order of the language presented first, as well as the order of the CSI block presentation, was counterbalanced across participants. Within each CSI block, the order of picture presentation was also pseudo‐randomized. This was done to prevent either the same picture or the same picture sequence from appearing twice in a row. The experiment started with a short practice session and included self‐paced pauses between the experimental blocks. The order in which the stimuli were presented was unpredictable.</p> <p>All trials commenced with a blank screen lasting 200 ms, followed by a fixation cross ("+"). When preparation time was given, the fixation cross was displayed together with a colored background (indicating the language to use) for 250, 500, or 800 ms. After this, a picture appeared for 1500 ms and was followed by a 1500 ms blank screen. When no preparation time was given, the fixation cross appeared on a neutral background (white) for 200 ms and was followed by the picture to be named. The picture was displayed together with the language cue for 1500 ms and was followed by a blank screen for 1500 ms. Hence, the interval between trials was kept constant (stimulus to cue interval or SCI = 3200 ms). Participants' responses were recorded through an external microphone connected to the computer. Vocal responses were checked and measured manually using the software Praat version 5.4.08 (Boersma & Weenink, 2015).</p> <hd id="AN0155381511-10">Data analysis</hd> <p>The dependent variables were accuracy rates and naming latencies, the latter being calculated as the interval between the display of the stimulus and the onset of speech. A small portion of data (2.5%) had to be removed because of technical issues. Both correct and incorrect responses were used in the analysis of the accuracy rates, while only correct responses were included in the analysis of the naming latencies. A response was considered incorrect in the case of a microphone miss‐triggering (e.g., hesitation, self‐repair, etc.) and when the wrong word and or language was selected. Missing responses were also classified as incorrect. Based on these criteria, 5% and 3% of the L1 and L2 data, respectively, were excluded from subsequent analyses. Most of the incorrect responses were due to missing responses (L1: 55%; L2: 55%), followed by the selection of the wrong language (L1: 43%; L2: 40%), and finally, incorrect answers in the target language (L1: 2%; L2: 5%).[<reflink idref="bib1" id="ref1">1</reflink>] Selection of the wrong language was more likely to occur on switch trials (85%) than on repetition trials (15%). This trend was similar for the L1 and L2; see Fig. A1. We also observed that the occurrence of an incorrect response increased the naming latency of the next trial, even when the upcoming response was correct. This seemed especially the case for switch rather than for repetition trials; see Fig. A2. Correct responses following incorrect trials were, therefore, excluded from further analysis.</p> <p>Visual inspection of the correct naming responses revealed that the data were positively skewed. To approach normality, a reciprocal square root transformation was performed as suggested by the Box–Cox function (λ = –0.26) of the MASS package in R (Venables & Ripley, 2002). All the analyses were carried out in GNU‐R (R Development Core Team, 2018) using the lme4 package (Bates, Maechler, Bolker, & Walker, 2014). Naming latencies were fit into linear mixed‐effect models, while the accuracy data were binary coded (correct = 1; incorrect = 0) and fit into generalized mixed‐effect models with a logistic link function (Bates, 2010). The factors Language (L1 vs. L2), Condition (Repetition vs. Switch), and Preparation Time (0, 250, 500, and 800 ms) were included in all models as fixed effects. The R package lmerTest (Kuznetsova, Brockhoff, & Christensen, 2017) was used to obtain the <emph>p</emph>‐values of the fixed effects returned by the lmer function. The models also included crossed random factors for participants and items, which allowed us to capture differences across the speakers and stimuli selected for the experiment (Baayen, Davidson, & Bates, 2008).</p> <p>To select the random structure, we first built a maximally fit model following Barr and colleagues' (2013) recommendation. When the random structure is maximally fit, the model takes into account not only that subjects/items might generally differ from each other (i.e., through random intercepts for subjects and items), but also, that each subject/item can be differently affected by the main effects and interactions under scrutiny (i.e., through random slopes for subjects and items). Unfortunately, often the amount of information in the data (i.e., the number of observations) is not enough to unambiguously estimate all these parameters. In such a situation, the random structure of the model might need to be simplified (Bates, Kliegl, Vasishth, & Baayen, 2018). This was also the approach we took for selecting our best‐fit model: In the case of model nonconvergence or singularities, we reduced the complexity of the random structure based on the amount of variance explained by the data (as suggested by Bates et al., 2018). Following this procedure, the best‐fit lmer model included Language as a by‐subject random slope, and Language and Condition as a by‐item random slope. The best‐fit glmer model included by‐subject and by‐item random intercepts but no random slopes.</p> <p>Main effects and interactions were calculated using sum contrasts when we expected the two levels of a factor to be theoretically equal (i.e., the factor Language), treatment contrasts when we expected the two levels to be theoretically different (i.e., the factor Condition), and repeated contrasts coding when we expected multiple levels of a factor to be theoretically different (i.e., the factor Preparation Time). Significant interactions in the models were explored using the emmeans package in R (continuation of the lsmeans package; Lenth, 2016), which estimates marginal means. When looking for simple effects, this method considers all the other factors included in the best‐fit model of reference and adjusts for multiple comparisons (default = tukey method). Finally, to evaluate the strength of an effect, we measured its Bayes factor with the R package bayestestR (Makowski, Ben‐Shachar, & Lüdecke, 2019).</p> <hd id="AN0155381511-11">Results</hd> <p></p> <hd id="AN0155381511-12">Naming latencies</hd> <p>Mean naming latencies and switching costs are shown in Figs. 1 and 2, respectively. The output of the best‐fit model can be found in Table 2. For an in‐depth overview of the naming latencies and switching costs, see Table A1.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01feb22/cogs13105-fig-0001.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs13105-fig-0001.jpg" title="1 Mean naming latencies of correct responses. Note. Mean naming latencies of correct responses in the L1 (Dutch) and L2 (English) as a function of Condition (Repetition vs. Switch trials) and Preparation time (0, 250, 500, and 800 ms). Error bars represent standard errors." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01feb22/cogs13105-fig-0002.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs13105-fig-0002.jpg" title="2 Mean switching costs of correct responses. Note. Mean switching costs of correct responses as a function of Preparation time (0, 250, 500, and 800 ms). Switching costs are measured as the difference in naming latencies between Switch and Repetition trials. Error bars represent standard deviations." /> </p> <p></p> <p>2 TableEstimated coefficients, standard errors (SE), and t‐values from the best‐fit linear mixed‐effects models for the naming latencies</p> <p> <ephtml> <table><thead><tr><th>Naming latencies</th></tr><tr><th /><th>Fixed effects</th><th>Random effects</th></tr><tr><th /><th>Estimate</th><th>SE</th><th><italic>t</italic>‐value</th><th>by subject SD</th><th>by items SD</th></tr></thead><tbody><tr><td /><td /><td /><td /><td /><td /></tr><tr><td>Intercept</td><td>–35.12</td><td>0.52</td><td>–67.16<sup>*</sup></td><td>1.64</td><td>1.69</td></tr><tr><td>Language</td><td>0.45</td><td>0.17</td><td>2.67<sup>*</sup></td><td>0.46</td><td>0.37</td></tr><tr><td>Condition</td><td>0.08</td><td>0.35</td><td>0.23</td><td /><td>1.12</td></tr><tr><td>Prep. Time 0 versus 250 ms</td><td>–0.79</td><td>0.20</td><td>–3.85<sup>*</sup></td><td /><td /></tr><tr><td>Prep. Time 250 versus 500 ms</td><td>0.02</td><td>0.20</td><td>0.13</td><td /><td /></tr><tr><td>Prep. Time 500 versus 800 ms</td><td>–0.13</td><td>0.20</td><td>–.67</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition</td><td>–0.17</td><td>0.16</td><td>–1.06</td><td /><td /></tr><tr><td>Language<sup>*</sup>Prep. Time (0 vs. 250 ms)</td><td>–0.21</td><td>0.20</td><td>–1.04</td><td /><td /></tr><tr><td>Language<sup>*</sup>Prep. Time (250 vs. 500 ms)</td><td>0.02</td><td>0.20</td><td>0.10</td><td /><td /></tr><tr><td>Language<sup>*</sup>Prep. Time (500 vs. 800 ms)</td><td>–0.15</td><td>0.20</td><td>–0.72</td><td /><td /></tr><tr><td>Condition<sup>*</sup>Prep. Time (0 vs. 250 ms)</td><td>–1.20</td><td>0.28</td><td>–4.16<sup>*</sup></td><td /><td /></tr><tr><td>Condition<sup>*</sup>Prep. Time (250 vs. 500 ms)</td><td>–0.08</td><td>0.28</td><td>–0.29</td><td /><td /></tr><tr><td>Condition<sup>*</sup>Prep. Time (500 vs. 800 ms)</td><td>0.38</td><td>0.28</td><td>1.32</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition<sup>*</sup>Prep. Time (0 vs. 250 ms)</td><td>0.07</td><td>0.29</td><td>0.24</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition<sup>*</sup>Prep. Time (250 vs. 500 ms)</td><td>0.05</td><td>0.29</td><td>0.17</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition<sup>*</sup>Prep. Time (500 vs. 800 ms)</td><td>0.37</td><td>0.28</td><td>1.29</td><td /><td /></tr></tbody></table> </ephtml> </p> <ulist> <item>2 Asterisks (<sups>*</sups>) indicate <emph>p</emph><.05.</item> <item>3 Formula: RT ∼ Language<sups>*</sups> Condition<sups>*</sups> Prep. Time (Language |subject) + (Language+Condition |item).</item> </ulist> <p>As expected, the analysis of the naming latencies showed that speakers responded significantly faster when they were given some time to prepare compared to when no preparation time was provided (<emph>t</emph> = –3.85, <emph>p</emph><.05 for CSI: 0 vs. 250 ms). However, response times were not affected by the amount of preparation time (CSI: 250 vs. 500 ms, and CSI: 500 vs. 800 ms, <emph>p</emph>>.05). There was a significant main effect of Language, with responses in the L2 being faster than in the L1 (<emph>t</emph> = 2.67, <emph>p</emph><.05). The difference between repetition and switch trials was not significant (<emph>p</emph>>.05), indicating that overall language switching was not costly. Yet, the significant interaction between Condition and Preparation Time showed that the difference between repetition and switch trials was larger when speakers had no time to prepare compared to when some preparation time was given (<emph>t</emph> = –4.16, <emph>p</emph><.001 for CSI: 0 ms vs. CSI: 250 ms). A deeper exploration of the interaction revealed that language switching was costly when speakers had no time to prepare (<emph>t</emph> = 2.28, <emph>p</emph><.05, df = 18, for Repetition vs. Switch trials in CSI: 0 ms), but not when preparation time was provided (<emph>t</emph><2, <emph>p</emph>>.05, df = 19.1, 18.9, and 19.4, for Repetition vs. Switch trials in CSI: 250, 500, and 800 ms, respectively). All the other interactions were not statistically significant.</p> <p>To provide more statistical evidence to the fact that switching language was costly when speakers could not prepare, but became cost‐free when some preparation time was provided, we ran an additional Bayesian hypothesis test. The analysis of the inverse Bayes factor showed that when the speakers had no time to prepare, the alternative hypothesis (i.e., Repetition ≠ Switch trials) was very strongly preferred over the null hypothesis (BF<subs>10</subs> = 329; Raftery, 1995). However, when the speakers had time to prepare, the null hypothesis was preferred (BF<subs>10</subs><1 for all the other CSIs). Altogether, the Bayes factor analysis supports the finding that repetition and switch trials are statistically different when speakers have no time to prepare, but not when preparation time is provided.</p> <hd id="AN0155381511-15">Accuracy rates</hd> <p>Mean accuracy rates and the best‐fit model for the error data can be found in Fig. 3 and Table 3, respectively. For an in‐depth overview of the accuracy rates, see Table A1.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01feb22/cogs13105-fig-0003.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs13105-fig-0003.jpg" title="3 Mean accuracy rates. Note. Mean accuracy rates in the L1 (Dutch) and L2 (English) as a function of Condition (Repetition vs. Switch trials) and Preparation time (0, 250, 500, and 800 ms). Error bars represent standard errors." /> </p> <p></p> <p>3 TableEstimated coefficients, standard errors (SE), and z‐values from the best‐fit linear mixed‐effects models for the accuracy rates</p> <p> <ephtml> <table><thead><tr><th>Accuracy rates</th></tr><tr><th /><th>Fixed effects</th><th>Random effects</th></tr><tr><th /><th>Estimate</th><th>SE</th><th><italic>z</italic>‐value</th><th>by subject SD</th><th>by items SD</th></tr></thead><tbody><tr><td>Intercept</td><td>3.15</td><td>0.20</td><td>15.69<sup>*</sup></td><td>0.87</td><td>0.18</td></tr><tr><td>Language</td><td>–0.32</td><td>0.10</td><td>–3.22<sup>*</sup></td><td /><td /></tr><tr><td>Condition</td><td>–0.57</td><td>0.13</td><td>–4.23<sup>*</sup></td><td /><td /></tr><tr><td>Prep. Time 0 versus 250 ms</td><td>0.45</td><td>0.28</td><td>1.57</td><td /><td /></tr><tr><td>Prep. Time 250 versus 500 ms</td><td>–0.50</td><td>0.28</td><td>–1.79</td><td /><td /></tr><tr><td>Prep. Time 500 versus 800 ms</td><td>0.26</td><td>0.26</td><td>1.02</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition</td><td>0.12</td><td>0.13</td><td>0.98</td><td /><td /></tr><tr><td>Language<sup>*</sup>Prep. Time (0 vs. 250 ms)</td><td>0.03</td><td>0.28</td><td>0.11</td><td /><td /></tr><tr><td>Language<sup>*</sup>Prep. Time (250 vs. 500 ms)</td><td>0.02</td><td>0.28</td><td>0.10</td><td /><td /></tr><tr><td>Language<sup>*</sup>Prep. Time (500 vs. 800 ms)</td><td>0.01</td><td>0.26</td><td>0.06</td><td /><td /></tr><tr><td>Condition<sup>*</sup>Prep. Time (0 vs. 250 ms)</td><td>–0.39</td><td>0.36</td><td>–1.10</td><td /><td /></tr><tr><td>Condition<sup>*</sup>Prep. Time (250 vs. 500 ms)</td><td>0.31</td><td>0.35</td><td>0.88</td><td /><td /></tr><tr><td>Condition<sup>*</sup>Prep. Time (500 vs. 800 ms)</td><td>0.09</td><td>0.33</td><td>0.28</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition<sup>*</sup>Prep. Time (0 vs. 250 ms)</td><td>–0.25</td><td>0.36</td><td>–0.69</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition<sup>*</sup>Prep. Time (250 vs. 500 ms)</td><td>0.14</td><td>0.35</td><td>0.40</td><td /><td /></tr><tr><td>Language<sup>*</sup>Condition<sup>*</sup>Prep. Time (500 vs. 800 ms)</td><td>–0.01</td><td>0.33</td><td>–0.04</td><td /><td /></tr></tbody></table> </ephtml> </p> <ulist> <item>4 Asterisks (<sups>*</sups>) indicate <emph>p</emph><.05.</item> <item>5 Formula: Accuracy ∼ Language<sups>*</sups> Condition<sups>*</sups> Prep. Time (1 |subject) + (1 |item).</item> </ulist> <p>The analysis of the errors revealed that responses were more accurate on repetition than switch trials (<emph>z</emph> = –4.23, <emph>p</emph><.001). Also, speakers made fewer errors in the weaker L2 than in the stronger L1 (<emph>z</emph> = –3.22, <emph>p</emph><.01). This paradoxical pattern resembles the reversed language dominance effect found in the RT data, where it was observed that there were faster responses in the weaker rather than in the stronger language. The main effect of Preparation time was not significant (<emph>p</emph>>.05), meaning that responses were named equally well irrespective of whether or not speakers had been given time to prepare. All the other interactions were not statistically significant (<emph>p</emph>>.05).</p> <hd id="AN0155381511-17">Discussion</hd> <p>Past research had suggested that there might be some limitations to people's ability to prepare for a switch, even when abundant time to prepare is provided (de Jong, 2000; Rogers & Monsell, 1995). In language control research, this assumption has long been backed by empirical evidence showing that persisting switching costs occurred no matter how much time speakers were given to prepare for a language switch (e.g., Costa & Santesteban, 2004; Fink & Goldrick, 2015; Graham & Lavric, 2021; Philipp et al., 2007, Verhoef et al., 2009).</p> <p>The goal of this study was to systematically investigate the effect of preparation time on language switching. To do this, language switching costs were measured in a group of bilinguals prompted to name pictures in their L1 or L2, based on a color cue. The ability to prepare for a language switch was investigated by giving the speakers some time to prepare for the upcoming trial (i.e., the language cue appeared before the stimulus) or none (i.e., language cue and stimulus appeared simultaneously). Importantly, the interval between trials was kept constant and relatively long (3200 ms), so to avoid the effects of the previous language task interfering with the switch preparation. The choice of interval was based on the assumption that the interference coming from a previous task can persist for some time before decaying (Allport et al., 1994; Wylie & Allport, 2000).</p> <p>The results showed that when no preparation time was given, language switching was costly. However, when participants were given some time to prepare, language switching costs disappeared. Language switching became cost‐free at 800 and 500 ms, when preparation time was relatively long, but also when preparation time was as short as 250 ms. While switching costs elimination has been previously reported in a study where speakers had a relatively long time to prepare (CSI: 800 ms; Mosca & Clahsen, 2016), the abolition of language switching costs with shorter preparation times is a novel finding.</p> <p>Altogether, the results of the study indicate that there might be no limitations to our ability to prepare for a language switch. They also suggest that 250 ms of preparation time might be enough for switch trials to be successfully prepared for, and so for switching costs to be eliminated. The results, therefore, undermine the hypothesis that humans are structurally unable to prepare for a switch (Rogers & Monsell, 1995), or that the ability to prepare a switch is limited to a subset of trials only (de Jong, 2000). It must be noted, however, that these findings refer to speakers switching between languages in a relatively easy task, that is, in single object naming. The question of whether a cost‐free language switch is also possible during more complex tasks, such as in action naming (i.e., verb naming task), or during sentence production, remains unanswered. It also remains unclear whether the ability to prepare for a language switch is restricted to language production, or whether similar mechanisms are at play during language recognition.</p> <p>The results of this study contest the results found in earlier experiments, when similar or even longer preparation times were provided (e.g., Costa & Santesteban, 2004; Fink & Goldrick, 2015; Graham & Lavric, 2021, Guo et al., 2013; Khateb et al., 2017; Ma et al., 2016; Verhoef et al., 2009). The main difference between this and previous studies lies in the relatively long interval (3200 ms) between consecutive trials used in the present study. We assumed that when the interval between trials is long enough, the interference coming from the previous language task has time to completely dissipate and would not affect the preparation of the new language (cf. Allport et al., 1994; Wylie & Allport, 2000). In such a context, the language control system needs less than 250 ms to prepare a language switch. It remains to be determined, however, at which intertrial interval under 3200 ms, the interference from the previous trial can no longer dissipate, and at which preparation time under 250 ms, language switching starts to be costly. It would be pertinent for future studies to determine whether these timing constraints are related to language dominance, to speakers' executive control abilities, or to both.</p> <p>Regarding the accuracy rates, the analyses revealed that switching language elicited more errors than language repetition. The effect was comparable across the four preparation intervals, suggesting that, sometimes, regardless of preparation time, the language control system reacted to a language cue incorrectly. Specifically, errors on repetition trials might indicate that the system occasionally triggered a language switch even if this was not needed. Errors on switch trials might indicate that, sometimes, a language switch was not even initiated despite a language cue prompting the participant to do so. This seems to contradict the assumption that speakers can prepare for the upcoming language if they are given the time to do so, but we propose that this is not necessarily the case. As with other decision tasks, a language switching paradigm might elicit responses that are externally triggered (i.e., by the language cue) and others that are internally driven (i.e., through predictive mechanisms); cf. Kleinsorge and Scheil (2016). Some of the errors, therefore, might be associated with "incorrect predictions" rather than "incorrect reactions" to the language cue. In this experiment, repetition trials occurred more often than switch trials (i.e., each language chunk was composed of 75% repetition and only 25% switch trials). The higher error rates for switch compared to repetition trials might, therefore, be attributed to the erroneous prediction of a language repetition trial.</p> <p>As to switching costs, in this study they are supposed to reflect the effort to prepare a new language, that is, shifting attention to this language, retrieving its rules (e.g., phonological rules), and inhibiting the other language.[<reflink idref="bib2" id="ref2">2</reflink>] The process of switch preparation seems to be comparable for the two languages, in that we did not find any difference in the switching cost reduction between the L1 and L2 (the three‐way interaction of Language, Condition, and Preparation Time was not significant). Moreover, the data indicate that when bilinguals had no time to prepare, switching costs in the two languages were equivalent (nonsignificant Language by Condition interaction). These findings support the hypothesis that the cost to prepare a language does not depend on its dominance (Costa & Santesteban, 2004). However, that also raises a question. If persisting inhibitory effects depend on language dominance (so that inhibition coming from a stronger language takes longer to decay compared to that coming from a weaker language; Green, 1998), why is the cost to prepare a new language (entailing inhibition of the other language) dominance independent? If stronger languages are more suppressed than weaker languages, we would expect longer preparation times for the weaker language (when the stronger language needs to be inhibited) than for the stronger one (when the weaker language must be inhibited). A possible answer is that, during language preparation, the effort to suppress either language is equivalent when those languages do not differ much in their proficiency level (as in the case of L2 highly proficient speakers). This explanation fits with the hypothesis that the time needed to recover from the inhibition that has been previously applied might be the same when speakers' have a comparable proficiency level in their two languages (cf. Meuter & Allport, 1999). However, further studies are needed to determine to what extent the effort to prepare a language is influenced by the relative dominance in L1 and L2. This could be done, for instance, by employing bilinguals with varying L2 proficiency levels.</p> <p>The data also revealed that performance in the weaker language was faster and more accurate than in the stronger one (aka "reversed language dominance effect"). This phenomenon has often been attributed to the likelihood that participants unconsciously employ a strategy to help the weaker language in a language switching context (e.g., Christoffels et al., 2007, Christoffels et al., 2016; Costa & Santesteban, 2004), which is achieved either by facilitating the weaker language and/or by hampering the stronger language (e.g., Christoffels et al., 2007; Meuter, 2005). An alternative view, however, ascribes this language reversal dominance to the cumulative effect of global inhibition and repetition priming (Kleinman & Gollan, 2018). Specifically, it has been observed that performance slows down every time unrelated pictures in the other language are named. This "global inhibition" is especially true for the stronger language. Naming the same picture in the other language also reduces performance, increasing the effect of global inhibition. In contrast, performances become faster every time a picture is named in the same language. This "repetition priming" applies especially to the weaker language. Hence, the reversed dominance effect might stem from the greater L1 global inhibition together with the larger L2 repetition priming. Importantly, global inhibition/facilitation appears to rely on a sustained effect accumulating throughout the task while switching costs seem to be supported by a transient mechanism, which is elicited by a sudden change (the language switch) that dissipates very quickly (cf. Kleinman & Gollan, 2018). The results of this study support the view that the two mechanisms are separate and different in nature. For example, while the preparation costs disappeared when bilinguals were given enough time to prepare, the reversed dominance effect persisted throughout the four preparation time intervals (nonsignificant interaction of Language by Preparation Time). Moreover, whereas the former did not seem to be affected by language dominance, the latter was determined by language dominance, as suggested by the slower responses for the stronger than for the weaker language. This is, to our knowledge, the first study reporting such a dissociation, where switching costs vanish but a language paradox effect persists.</p> <p>A way to explain the observed dissociation between disappearing switching costs and reversed dominance effect is to assume that the longer intertrial interval allows not only the dissipation of the persisting interference but also fosters the system to reach a state of readiness to respond in either language. Once this state is reached, a minimal preparation interval might be enough to select the correct language schema and so switch without apparent cost. The dissociation with the reversed dominance effect might arise because, while adjusting the activation level of the two competing languages, the stronger language must be gated from entering the speech plan (e.g., by enhancing its activation threshold). This type of control process, however, is not always uniform. In some cases, bilinguals might overshoot while trying to make the two languages equally accessible, unintentionally making the stronger language much less accessible than the weaker one (Declerck et al., 2020). As a result, instead of having equally ready language schemas, the weaker language becomes more available than the stronger one, leading to a reversed dominance effect.</p> <hd id="AN0155381511-18">Conclusion</hd> <p>The aim of this study was to assess whether bilinguals can overcome language switching costs when given some time to prepare, or whether there is a limitation to people's ability to do so. The study suggests that the switching process is correctly triggered when enough preparation time is provided, with switch costs disappearing when the language cue appears at least 250 ms in advance. This effect was the same for the stronger L1 and the weaker L2 language when tested in a group of bilinguals where the L2 was highly proficient, suggesting that language preparation is not closely related to language dominance. In contrast, the data revealed persisting slower responses in the stronger relative to the weaker language, independent of preparation time.</p> <p>The results of this study challenge some established theoretical views and traditional, experimental practices. The study proposes that, in contrast with what was previously assumed, speakers might be able to fully prepare for a switch and so overcome switching costs, at least during simple object naming. On the practical side, the study has stressed the importance of parameters like timing by showing how apparently minor changes in the experimental design can have significant effects on the subject's behavior.</p> <p>The data might also clarify why often spontaneous code‐switched language production does not seem to be costly. In a way, speakers might "see a switch coming" and then have ample time to prepare for it. It must be noted, however, that though language switching in real‐life communication might be perceived as fluent, disruptions might still occur at deeper levels (e.g., subtle shifts in intonation and speech rate before a language switch). Also, the ability to prepare for a language switch might be related to the speaker's switching habits or their general executive function abilities. Future studies are needed to help determine whether switching costs can be eliminated in real‐life communications, whether this depends on the speaker's linguistic or nonlinguistic switching habits, and if this is affected by the number of languages involved in the interactional context.</p> <hd id="AN0155381511-19">Acknowledgments</hd> <p>We are grateful to David Green, Alba Casado, and two anonymous reviewers for their helpful comments on previous versions of this paper. We would also like to thank Rick Cooper as handling editor for the constructive comments and suggestions.</p> <hd id="AN0155381511-20">Funding</hd> <p>This research was partially funded by the European Union under the framework "Erasmus+: Erasmus Mundus Master Course (EMMC)," partnership agreement 2010‐0111 in the joint Master Program "European Master's in Clinical Linguistics" (EMCL).</p> <hd id="AN0155381511-21">Disclosure</hd> <p>The authors declare they have no conflicts to disclose.</p> <hd id="AN0155381511-22">Open Research Badges</hd> <p>This article has earned Open Data. Data are available at https://osf.io/bnfs8/.</p> <p>Appendix</p> <p>Materials (L1 Dutch, L2 English)</p> <p>L1: <emph>aardbei, ananas, boom, borstel, hond, kerk, kers, lepel, mand, mes, mier, paard, paraplu, riem, rok, sleutel, stoel, stropdas, vlieger, vlinder</emph>.</p> <p>L2: <emph>strawberry, pineapple, tree, brush, dog, church, cherry, spoon, basket, knife, ant, horse, umbrella, belt, skirt, key, chair, tie, kite, butterfly</emph>.</p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01feb22/cogs13105-fig-0004.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs13105-fig-0004.jpg" title="A1 Incorrect responses as a function of Language and Condition. Note. Total amount of incorrect responses based on Condition (Repetition vs. Switch trials) and Language (L1 Dutch vs. L2 English). Incorrect responses are classified depending on the type of error: Incorrect responses in the intended language ("Same Language" error), responses in the nonintended language ("Other Language" error), and missing responses ("Timeouts")." /> </p> <p></p> <p> <img src="https://imageserver.ebscohost.com/img/embimages/rdk/CGN/01feb22/cogs13105-fig-0005.jpg?ephost1=dGJyMNHX8kSepq84v%2bvlOLCmsE6epq5Srqa4SK6WxWXS" alt="cogs13105-fig-0005.jpg" title="A2 Correct responses as a function of Preceding Trial and Condition. Note. Mean naming latencies of correct responses as a function of Condition (Repetition vs. Switch trials) and Preceding Trial (Correct vs. Incorrect). Error bars represent standard errors." /> </p> <p></p> <p>Table Mean naming latencies of correct responses and accuracy rates (standard deviation in brackets) in the L1 (Dutch) and L2 (English) as a function of Condition (Repetition vs. Switch trials) and Preparation Time (0, 250, 500, and 800 ms)</p> <p> <ephtml> <table><thead><tr><th /><th /><th>L1</th><th>L2</th><th>Mean</th></tr></thead><tbody><tr><td>Preparation time 0 ms</td><td /><td /><td /><td /></tr><tr><td>Repetition</td><td>Naming latencies</td><td>895 ms</td><td>847 ms</td><td>870 ms</td></tr><tr><td>(200)</td><td>(178)</td><td>(190)</td></tr><tr><td /><td>Accuracy rates</td><td>91%</td><td>96%</td><td>93%</td></tr><tr><td>(29)</td><td>(21)</td><td>(25)</td></tr><tr><td>Switch</td><td>Naming latencies</td><td>936 ms</td><td>880 ms</td><td>908 ms</td></tr><tr><td>(185)</td><td>(178)</td><td>(183)</td></tr><tr><td>Accuracy rates</td><td>90%</td><td>91%</td><td>91%</td></tr><tr><td>(31)</td><td>(28)</td><td>(29)</td></tr><tr><td>Switching costs</td><td /><td>41 ms</td><td>33 ms</td><td>38 ms</td></tr><tr><td>Preparation time 250 ms</td><td /><td /><td /><td /></tr><tr><td>Repetition</td><td>Naming latencies</td><td>853 ms</td><td>812 ms</td><td>833 ms</td></tr><tr><td>(214)</td><td>(187)</td><td>(202)</td></tr><tr><td /><td>Accuracy rates</td><td>94%</td><td>97%</td><td>96%</td></tr><tr><td>(24)</td><td>(17)</td><td>(21)</td></tr><tr><td>Switch</td><td>Naming latencies</td><td>849 ms</td><td>794 ms</td><td>821 ms</td></tr><tr><td>(207)</td><td>(176)</td><td>(194)</td></tr><tr><td /><td>Accuracy rates</td><td>88%</td><td>93%</td><td>91%</td></tr><tr><td>(32)</td><td>(25)</td><td>(29)</td></tr><tr><td>Switching costs</td><td /><td>–4 ms</td><td>–18 ms</td><td>–12 ms</td></tr><tr><td>Preparation time 500 ms</td><td /><td /><td /><td /></tr><tr><td>Repetition</td><td>Naming latencies</td><td>851 ms</td><td>808 ms</td><td>829 ms</td></tr><tr><td>(207)</td><td>(155)</td><td>(184)</td></tr><tr><td /><td>Accuracy rates</td><td>91%</td><td>95%</td><td>93%</td></tr><tr><td>(29)</td><td>(22)</td><td>(26)</td></tr><tr><td>Switch</td><td>Naming latencies</td><td>841 ms</td><td>793 ms</td><td>817 ms</td></tr><tr><td>(200)</td><td>(173)</td><td>(189)</td></tr><tr><td /><td>Accuracy rates</td><td>87%</td><td>90%</td><td>89%</td></tr><tr><td>(33)</td><td>(30)</td><td>(31)</td></tr><tr><td>Switching costs</td><td /><td>–10 ms</td><td>–15 ms</td><td>–12 ms</td></tr><tr><td>Preparation time 800 ms</td><td /><td /><td /><td /></tr><tr><td>Repetition</td><td>Naming latencies</td><td>846 ms</td><td>810 ms</td><td>829 ms</td></tr><tr><td>(210)</td><td>(189)</td><td>(201)</td></tr><tr><td /><td>Accuracy rates</td><td>93%</td><td>96%</td><td>94%</td></tr><tr><td>(26)</td><td>(20)</td><td>(23)</td></tr><tr><td>Switch</td><td>Naming latencies</td><td>857 ms</td><td>780 ms</td><td>819 ms</td></tr><tr><td>(213)</td><td>(174)</td><td>(199)</td></tr><tr><td /><td>Accuracy rates</td><td>91%</td><td>93%</td><td>92%</td></tr><tr><td>(29)</td><td>(26)</td><td>(27)</td></tr><tr><td>Switching costs</td><td /><td>11 ms</td><td>–30 ms</td><td>–10 ms</td></tr></tbody></table> </ephtml> </p> <ref id="AN0155381511-25"> <title> Footnotes </title> <blist> <bibl id="bib1" idref="ref1" type="bt">1</bibl> <bibtext> A microphone miss‐triggering (e.g., hesitation, self‐repair, etc.) was classified as a missing response, incorrect response in the target language, or incorrect language selection based on the type of answer that followed the microphone activation.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref2" type="bt">2</bibl> <bibtext> The reader should note that, in this study, a language switch always overlapped with a cue switch. This means that the observed switching costs are likely to reflect not only the effort to prepare a new language, but also cue‐switching costs. Based on a recent study, however, the cue‐related costs should be minimal compared to the language‐related ones (Heikoop, Declerck, Los, & Koch, [40]). This implies that the switching costs in this study should mostly reflect language control processes.</bibtext> </blist> </ref> <ref id="AN0155381511-26"> <title> References </title> <blist> <bibtext> Abutalebi, J., & Green, D. W. (2007). Bilingual language production: The neurocognition of language representation and control. Journal of Neurolinguistics, 20, 242 – 275. https://doi.org/10.1016/j.jneuroling.2006.10.003</bibtext> </blist> <blist> <bibtext> Abutalebi, J., & Green, D. W. (2008). Control mechanisms in bilingual language production: Neural evidence from language switching studies. Language and Cognitive Processes, 23, 557 – 582. <ulink href="http://doi.org/10.1080/01690960801920602">http://doi.org/10.1080/01690960801920602</ulink></bibtext> </blist> <blist> <bibl id="bib3" type="bt">3</bibl> <bibtext> Allport, A., Styles, E. A., & Hsieh, S. (1994). Shifting intentional set: Exploring the dynamic control of tasks. In C. Umilta & M. Moscovitch (Eds.), Conscious and nonconscious information processing: Attention and performance XV (pp. 421 – 452). Cambridge, MA : MIT Press.</bibtext> </blist> <blist> <bibl id="bib4" type="bt">4</bibl> <bibtext> Aron, A. R. (2007). The neural basis of inhibition in cognitive control. Neuroscientist, 13, 214 – 228. https://doi.org/10.1177/1073858407299288</bibtext> </blist> <blist> <bibl id="bib5" type="bt">5</bibl> <bibtext> Baayen, R. H., Davidson, D. J., & Bates, D. M. (2008). Mixed‐effects modeling with crossed random effects for subjects and items. Journal of Memory and Language, 59, 390 – 412. https://doi.org/10.1016/j.jml.2007.12.005</bibtext> </blist> <blist> <bibl id="bib6" type="bt">6</bibl> <bibtext> Babcock, L., & Vallesi, A. (2015). Language control is not a one‐size‐fits‐all languages process: Evidence from simultaneous interpretation students and the n‐2 repetition cost. Frontiers in Psychology, 6, 1622. https://doi.org/10.3389/fpsyg.2015.01622</bibtext> </blist> <blist> <bibl id="bib7" type="bt">7</bibl> <bibtext> Barr, D. J., Levy, R., Scheepers, C., & Tily, H. J. (2013). Random effects structure for confirmatory hypothesis testing: Keep it maximal. Journal of Memory and Language, 68, 255 – 278. https://doi.org/10.1016/j.jml.2012.11.001</bibtext> </blist> <blist> <bibl id="bib8" type="bt">8</bibl> <bibtext> Bates, D. M. (2010). lme4: Mixed‐effects modelling with R. Retrieved from http://lme4.r‐forge.r‐project.org/book</bibtext> </blist> <blist> <bibl id="bib9" type="bt">9</bibl> <bibtext> Bates, D., Kliegl, R., Vasishth, S., & Baayen, H. (2018). Parsimonious mixed models. arXiv preprint arXiv:1506.04967.</bibtext> </blist> <blist> <bibtext> Bates, D. M., Maechler, M., Bolker, B., & Walker, S. (2014). lme4: Linear mixed‐effects models using S4 classes. [ https://CRAN.R‐project.org/package=lme4 ]. R package version 1.1–7.</bibtext> </blist> <blist> <bibtext> Boersma, P., & Weenink, D. (2015). Praat: Doing Phonetics by Computer [Computer Program]. Version 5.4.08. Retrieved from <ulink href="http://www.praat.org/">http://www.praat.org/</ulink></bibtext> </blist> <blist> <bibtext> Branzi, F. M., Martin, C. D., Abutalebi, J., & Costa, A. (2014). The after‐effects of bilingual language production. Neuropsychologia, 52, 102 – 116. https://doi.org/10.1016/j.neuropsychologia.2013.09.022</bibtext> </blist> <blist> <bibtext> Bultena, S., Dijkstra, T., & van Hell, J. (2015a). Language switch costs in sentence comprehension depend on language dominance: Evidence from self‐paced reading. Bilingualism: Language and Cognition, 18, 453 – 469. https://doi.org/10.1017/S1366728914000145</bibtext> </blist> <blist> <bibtext> Bultena, S., Dijkstra, T., & van Hell, J. G. (2015b). Switch cost modulations in bilingual sentence processing: Evidence from shadowing. Language, Cognition and Neuroscience, 30, 586 – 605. https://doi.org/10.1080/23273798.2014.964268</bibtext> </blist> <blist> <bibtext> Calabria, M., Costa, A., Green, D., & Abutalebi, J. (2018). Neural basis of bilingual language control. Annals of the New York Academy of Sciences, 1426, 221 – 235. https://doi.org/10.1111/nyas.13879</bibtext> </blist> <blist> <bibtext> Christoffels, I., Firk, C., & Schiller, N. O. (2007). Bilingual language control: An event related brain potential study. Brain Research, 1147, 192 – 208. https://doi.org/10.1016/j.brainres.2007.01.137</bibtext> </blist> <blist> <bibtext> Christoffels, I. K., Ganushchak, L., & La Heij, W. (2016). When L1 suffers: Sustained, global slowing and the reversed language effect in mixed language context. In J. W. Schwieter (Ed.), Cognitive control and consequences of multilingualism (pp. 171 – 192). Ontario : John Benjamins.</bibtext> </blist> <blist> <bibtext> Colomé, A. (2001). Lexical activation in bilinguals' speech production: Language‐specific or language‐independent? Journal of Memory and Language, 45, 721 – 736. https://doi.org/10.1006/jmla.2001.2793</bibtext> </blist> <blist> <bibtext> Costa, A., & Santesteban, M. (2004). Lexical access in bilingual speech production: Evidence from language switching in highly proficient bilinguals and L2 learners. Journal of Memory and Language, 50, 491 – 511. https://doi.org/10.1016/j.jml.2004.02.002</bibtext> </blist> <blist> <bibtext> Costa, A., Santesteban, M., & Ivanova, I. (2006). How do highly proficient bilinguals control their lexicalization process? Inhibitory and language‐specific selection mechanisms are both functional. Journal of Experimental Psychology: Learning, Memory, and Cognition, 32, 1057 – 1074. https://doi.org/10.1037/0278‐7393.32.5.1057</bibtext> </blist> <blist> <bibtext> de Bruin, A., Samuel, A. G., & Duñabeitia, J. A. (2018). Voluntary language switching: When and why do bilinguals switch between their languages? Journal of Memory and Language, 103, 28 – 43. https://doi.org/10.1016/j.jml.2018.07.005</bibtext> </blist> <blist> <bibtext> de Jong, R., Berendsen, E., & Cools, R. (1999). Goal neglect and inhibitory limitations: Dissociable causes of interference effects in conflict situations. Acta Psychologica, 101, 379 – 394. https://doi.org/10.1016/S0001‐6918(99)00012‐8</bibtext> </blist> <blist> <bibtext> Declerck, M., Ivanova, I., Grainger, J., & Duñabeitia, J. A. (2020). Are similar control processes implemented during single and dual language production? Evidence from switching between speech registers and languages. Bilingualism: Language and Cognition, 23, 694 – 701. https://doi.org/10.1017/S1366728919000695</bibtext> </blist> <blist> <bibtext> Declerck, M., Kleinman, D., & Gollan, T. H. (2020). Which bilinguals reverse language dominance and why? Cognition, 204, 104384. https://doi.org/10.1016/j.cognition.2020.104384</bibtext> </blist> <blist> <bibtext> Declerck, M., Koch, I., & Philipp, A. M. (2015). The minimum requirements of language control: Evidence from sequential predictability effects in language switching. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41, 377 – 394. https://doi.org/10.1037/xlm0000021</bibtext> </blist> <blist> <bibtext> Declerck, M., & Philipp, A. M. (2018). Is inhibition implemented during bilingual production and comprehension? N‐2 language repetition costs unchained. Language, Cognition and Neuroscience, 33, 608 – 617. https://doi.org/10.1080/23273798.2017.1398828</bibtext> </blist> <blist> <bibtext> Declerck, M., Philipp, A. M., & Koch, I. (2013). Bilingual control: Sequential memory in language switching. Journal of Experimental Psychology: Learning, Memory and Cognition, 39, 1793 – 1806. <ulink href="http://doi.org/10.1037/a0033094">http://doi.org/10.1037/a0033094</ulink></bibtext> </blist> <blist> <bibtext> Declerck, M., Thoma, A. M., Koch, I., & Philipp, A. M. (2015). Highly proficient bilinguals implement inhibition–Evidence from n‐2 language repetition costs. Journal of Experimental Psychology: Learning, Memory, and Cognition, 41, 1911 – 1916. https://doi.org/10.1037/xlm0000138</bibtext> </blist> <blist> <bibtext> De Jong, R. (2000). An intention‐activation account of residual switch costs. In S. Monsell & J. Driver (Eds.), Control of cognitive processes (pp. 357 – 376). Cambridge, MA : MIT Press.</bibtext> </blist> <blist> <bibtext> Festman, J., Rodriguez‐Fornells, A., & Münte, T. F. (2010). Individual differences in control of language interference in late bilinguals are mainly related to general executive abilities. Behavioral and Brain Functions, 6, 5. https://doi.org/10.1186/1744‐9081‐6‐5</bibtext> </blist> <blist> <bibtext> Fink, A., & Goldrick, M. (2015). Pervasive benefits of preparation in language switching. Psychonomic Bulletin and Review, 22, 808 – 814. https://doi.org/10.3758/s13423‐014‐0739‐6</bibtext> </blist> <blist> <bibtext> Graham, B., & Lavric, A. (2021). Preparing to switch languages versus preparing to switch tasks: Which is more effective? Journal of Experimental Psychology: General, 150, 1956 – 1973. https://doi.org/10.1037/xge0001027</bibtext> </blist> <blist> <bibtext> Green, D. W. (1986). Control, activation and resource. Brain and Language, 27, 210 – 223. <ulink href="http://doi.org/10.1016/0093‐934X(86)90016‐7">http://doi.org/10.1016/0093‐934X(86)90016‐7</ulink></bibtext> </blist> <blist> <bibtext> Green, D. W. (1998). Mental control of the bilingual lexico‐semantic system. Bilingualism: Language and Cognition, 1, 67 – 81. https://doi.org/10.1017/S1366728998000133</bibtext> </blist> <blist> <bibtext> Gollan, T. H., & Ferreira, V. S. (2009). Should I stay or should I switch? A cost–benefit analysis of voluntary language switching in young and aging bilinguals. Journal of Experimental Psychology: Learning, Memory, and Cognition, 35, 640 – 665. https://doi.org/10.1037/a0014981</bibtext> </blist> <blist> <bibtext> Gollan, T. H., Kleinman, D., & Wierenga, C. E. (2014). What's easier: Doing what you want, or being told what to do? Cued versus voluntary language and task switching. Journal of Experimental Psychology: General, 143, 2167 – 2195. https://doi.org/10.1037/a0038006</bibtext> </blist> <blist> <bibtext> Gross, M., & Kaushanskaya, M. (2015). Voluntary language switching in English‐Spanish bilingual children. Journal of Cognitive Psychology (Hove, England), 27, 992 – 1013. <ulink href="http://doi.org/10.1080/20445911.2015.1074242">http://doi.org/10.1080/20445911.2015.1074242</ulink></bibtext> </blist> <blist> <bibtext> Guo, T., Liu, F., Chen, B., & Li, S. (2013). Inhibition of non‐target languages in multilingual word production: Evidence from Uighur‐Chinese‐English trilinguals. Acta Psychologica, 143, 277 – 283. https://doi.org/10.1016/j.actpsy.2013.04.002</bibtext> </blist> <blist> <bibtext> Guo, T., Ma, F., & Liu, F. (2013). An ERP study of inhibition of non‐target languages in trilingual word production. Brain and Language, 127, 12 – 20. https://doi.org/10.1016/j.bandl.2013.07.009</bibtext> </blist> <blist> <bibtext> Heikoop, K. W., Declerck, M., Los, S. A., & Koch, I. (2016). Dissociating language‐switch costs from cue‐switch costs in bilingual language switching. Bilingualism: Language and Cognition, 19, 921 – 927. https://doi.org/10.1017/S1366728916000456</bibtext> </blist> <blist> <bibtext> Jackson, G. M., Swainson, R., Cunnington, R., & Jackson, S. R. (2001). ERP correlates of executive control during repeated language switching. Bilingualism: Language and Cognition, 4, 169 – 178. https://doi.org/10.1017/S1366728901000268</bibtext> </blist> <blist> <bibtext> Jackson, G. M., Swainson, R., Mullin, A., Cunnington, R., & Jackson, S. R. (2004). ERP correlates of a receptive language‐switching task. Quarterly Journal of Experimental Psychology, 2, 223 – 240. https://doi.org/10.1080/02724980343000198</bibtext> </blist> <blist> <bibtext> Khateb, A., Shamshoum, R., & Prior, A. (2017). Modulation of language switching by cue timing: Implications for models of bilingual language control. Journal of Experimental Psychology: Learning, Memory, and Cognition, 43, 1239 – 1253. https://doi.org/10.1037/xlm0000382</bibtext> </blist> <blist> <bibtext> Kleinman, D., & Gollan, T. H. (2018). Inhibition accumulates over time at multiple processing levels in bilingual language control. Cognition, 173, 115 – 132. https://doi.org/10.1016/j.cognition.2018.01.009</bibtext> </blist> <blist> <bibtext> Kleinsorge, T., & Scheil, J. (2016). Guessing versus choosing an upcoming task. Frontiers in Psychology, 7, Article 396. https://doi.org/10.3389/fpsyg.2016.00396</bibtext> </blist> <blist> <bibtext> Koch, I., Gade, M., Schuch, S., & Philipp, A. M. (2010). The role of inhibition in task switching – A review. Psychonomic Bulletin and Review, 17, 1 – 14. https://doi.org/10.3758/PBR.17.1.1</bibtext> </blist> <blist> <bibtext> Kreiner, H., & Degani, T. (2015). Tip‐of‐the‐tongue in a second language: The effects of brief first‐language exposure and long‐term use. Cognition, 137, 106 – 114. https://doi.org/10.1016/j.cognition.2014.12.011</bibtext> </blist> <blist> <bibtext> Kroll, J. F., Bobb, S. C., Misra, M. M., & Guo, T. (2008). Language selection in bilingual speech: Evidence for inhibitory processes. Acta Psychologica, 128, 416 – 430. <ulink href="http://doi.org/10.1016/j.actpsy.2008.02.001">http://doi.org/10.1016/j.actpsy.2008.02.001</ulink></bibtext> </blist> <blist> <bibtext> Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. B. (2017). lmerTest package: Tests in linear mixed effects models. Journal of Statistical Software, 82, 1 – 26. https://doi.org/10.18637/jss.v082.i13</bibtext> </blist> <blist> <bibtext> La Heij, W. (2005). Selection processes in monolingual and bilingual lexical access. In J. F. Kroll & A. M. B. de Groot (Eds.), Handbook of bilingualism: Psycholinguistic approaches (pp. 289 – 307). Oxford : Oxford University Press.</bibtext> </blist> <blist> <bibtext> Lemhöfer, K., & Broersma, M. (2012). Introducing LexTALE: A quick and valid lexical test for advanced learners of English. Behavior Research Methods, 44, 325 – 343. https://doi.org/10.3758/s13428‐011‐0146‐0</bibtext> </blist> <blist> <bibtext> Lenth, R. V. (2016). Least‐squares means: The R package lsmeans. Journal of Statistical Software, 69, 1 – 33. https://doi.org/10.18637/jss.v069.i01</bibtext> </blist> <blist> <bibtext> Li, P., Zhang, F., Tsai, E., & Puls, B. (2014). Language history questionnaire (LHQ 2.0): A new dynamic web‐based research tool. Bilingualism: Language and Cognition, 17, 673 – 680. https://doi.org/10.1017/S1366728913000606</bibtext> </blist> <blist> <bibtext> Ma, F., Li, S., & Guo, T. (2016). Reactive and proactive control in bilingual word production: An investigation of influential factors. Journal of Memory and Language, 86, 35 – 59. <ulink href="http://doi.org/10.1016/j.jml.2015.08.004">http://doi.org/10.1016/j.jml.2015.08.004</ulink></bibtext> </blist> <blist> <bibtext> Macnamara, J., Krauthammer, M., & Bolgar, M. (1968). Language switching in bilinguals as a function of stimulus and response uncertainty. Journal of Experimental Psychology, 78, 208 – 215. https://doi.org/10.1037/h0026390</bibtext> </blist> <blist> <bibtext> Makowski, D., Ben‐Shachar, M., & Lüdecke, D. (2019). bayestestR: Describing effects and their uncertainty, existence and significance within the Bayesian framework. Journal of Open Source Software, 40, 1541. 10.21105/joss.01541</bibtext> </blist> <blist> <bibtext> Mayr, U., & Kliegl, R. (2000). Task‐set switching and long‐term memory retrieval. Journal of Experimental Psychology: Learning, Memory, and Cognition, 26, 1124 – 1140. https://doi.org/10.1037/0278‐7393.26.5.1124</bibtext> </blist> <blist> <bibtext> Mayr, U., & Kliegl, R. (2003). Differential effects of cue changes and task changes on task‐set selection costs. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29, 362 – 372. https://doi.org/10.1037/0278‐7393.29.3.362</bibtext> </blist> <blist> <bibtext> Meiran, N. (1996). Reconfiguration of processing mode prior to task performance. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22, 1423 – 1442. https://doi.org/10.1037/0278‐7393.22.6.1423</bibtext> </blist> <blist> <bibtext> Meuter, R. (2005). Language selection in bilinguals: Mechanisms and processes. In J. Kroll & A. De Groot (Eds.), Handbook of bilingualism: Psycholinguistic approaches (pp. 349 – 370). New York, NY : Oxford University Press.</bibtext> </blist> <blist> <bibtext> Meuter, R., & Allport, A. (1999). Bilingual language switching in naming: Asymmetrical costs of language selection. Journal of Memory and Language, 40, 25 – 40. https://doi.org/10.1006/jmla.1998.2602</bibtext> </blist> <blist> <bibtext> Monsell, S. (2003). Task switching. Trends in Cognitive Sciences, 7, 134 – 140. https://doi.org/10.1016/S1364‐6613(03)00028‐7</bibtext> </blist> <blist> <bibtext> Monsell, S., & Mizon, G. A. (2006). Can the task‐cueing paradigm measure "endogenous" task‐set reconfiguration? Journal of Experimental Psychology: Human Perception and Performance, 32, 493 – 516. https://doi.org/10.1037/0096‐1523.32.3.493</bibtext> </blist> <blist> <bibtext> Mosca, M., & de Bot, K. (2017). Bilingual language switching: Production vs. recognition. Frontiers in Psychology, 8, 934. https://doi.org/10.3389/fpsyg.2017.00934</bibtext> </blist> <blist> <bibtext> Mosca, M., & Clahsen, H. (2016). Examining language switching in bilinguals: The role of task demand. Bilingualism: Language and Cognition, 19, 415 – 424. https://doi.org/10.1017/S1366728915000693</bibtext> </blist> <blist> <bibtext> Nieuwenhuis, S., & Monsell, S. (2002). Residual costs in task switching: Testing the failure‐to‐engage hypothesis. Psychonomic Bulletin & Review, 9, 86 – 92. https://doi.org/10.3758/BF03196259</bibtext> </blist> <blist> <bibtext> Norman, D. A., & Shallice, T. (1986). Attention to action: Willed and automatic control of behaviour. In R. J. Davidson, G. E. Schwartz, & D. Shapiro (Eds.), Consciousness & self‐regulation, 4 (pp. 1 – 18). New York : Plenum Press.</bibtext> </blist> <blist> <bibtext> Penfield, W., & Roberts, L. (1959). Speech and brain mechanisms. Princeton, NJ : Princeton University Press.</bibtext> </blist> <blist> <bibtext> Philipp, A. M., Gade, M., & Koch, I. (2007). Inhibitory processes in language switching? Evidence from switching language‐defined response sets. European Journal of Cognitive Psychology, 19, 395 – 416. https://doi.org/10.1080/09541440600758812</bibtext> </blist> <blist> <bibtext> Poulisse, N., & Bongaerts, T. (1994). First language use in second language production. Applied Linguistics, 15, 36 – 57. https://doi.org/10.1093/applin/15.1.36</bibtext> </blist> <blist> <bibtext> Psychology Software Tools Inc. (2012). Psychology Software Tools Inc. [E.‐Prime 2.0]. Retrieved from <ulink href="http://www.pstnet.com">http://www.pstnet.com</ulink></bibtext> </blist> <blist> <bibtext> R Development Core Team. (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing.</bibtext> </blist> <blist> <bibtext> Raftery, A. E. (1995). Bayesian model selection in social re‐search. In P. V. Marsden (Ed.), Sociological methodology (pp. 111 – 196). Cambridge, MA : Blackwell.</bibtext> </blist> <blist> <bibtext> Rogers, R. D., & Monsell, S. (1995). Costs of a predictable switch between simple cognitive tasks. Journal of Experimental Psychology: General, 124, 207 – 231. doi: 037/0096‐3445.124.2.207</bibtext> </blist> <blist> <bibtext> Rossion, B., & Pourtois, G. (2004). Revisiting Snodgrass and Vanderwart's object set: The role of surface detail in basic‐level object recognition. Perception, 33, 217 – 236. https://doi.org/10.1068/p5117</bibtext> </blist> <blist> <bibtext> Rubinstein, J., Meyer, D., & Evans, J. (2001). Executive control of cognitive processes in task switching. Journal of Experimental Psychology: Human Perception and Performance, 27, 763 – 797. https://doi.org/10.1037/0096‐1523.27.4.763</bibtext> </blist> <blist> <bibtext> Schwieter, J., & Sunderman, G. (2008). Language switching in bilingual speech production: In search of the language‐specific selection mechanism. Mental Lexicon, 3, 214 – 238. https://doi.org/10.1075/ml.3.2.06sch</bibtext> </blist> <blist> <bibtext> Snodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174 – 215. https://doi.org/10.1037/0278‐7393.6.2.174</bibtext> </blist> <blist> <bibtext> Venables, W. N., & Ripley, B. D. (2002). Modern applied statistics with S (4th edition). New York : Springer.</bibtext> </blist> <blist> <bibtext> Verhoef, K., Roelofs, A., & Chwilla, D. (2009). Role of inhibition in language switching: Evidence from event related brain potentials in overt picture naming. Cognition, 110, 84 – 99. https://doi.org/10.1016/j.cognition.2008.10.013</bibtext> </blist> <blist> <bibtext> Wodniecka, Z., Szewczyk, J., Kałamała, P., Mandera, P., & Durlik, J. (2020). When a second language hits a native language: What ERPs (do and do not) tell us about language retrieval difficulty in bilingual language production. Neuropsychologia, 141. https://doi.org/10.1016/j.neuropsychologia.2020.107390</bibtext> </blist> <blist> <bibtext> Wylie, G., & Allport, A. (2000). Task switching and the measurement of "switch costs." Psychological Research, 63, 212 – 233. <ulink href="http://doi.org/10.1007/s004269900003">http://doi.org/10.1007/s004269900003</ulink></bibtext> </blist> </ref> <aug> <p>By Michela Mosca; Chaya Manawamma and Kees de Bot</p> <p>Reported by Author; Author; Author</p> </aug>
Header DbId: eric
DbLabel: ERIC
An: EJ1330021
AccessLevel: 3
PubType: Academic Journal
PubTypeId: academicJournal
PreciseRelevancyScore: 0
IllustrationInfo
Items – Name: Title
  Label: Title
  Group: Ti
  Data: When Language Switching Is Cost-Free: The Effect of Preparation Time
– Name: Language
  Label: Language
  Group: Lang
  Data: English
– Name: Author
  Label: Authors
  Group: Au
  Data: <searchLink fieldCode="AR" term="%22Mosca%2C+Michela%22">Mosca, Michela</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0002-1209-1304">0000-0002-1209-1304</externalLink>)<br /><searchLink fieldCode="AR" term="%22Manawamma%2C+Chaya%22">Manawamma, Chaya</searchLink><br /><searchLink fieldCode="AR" term="%22de+Bot%2C+Kees%22">de Bot, Kees</searchLink> (ORCID <externalLink term="https://orcid.org/0000-0001-7442-2693">0000-0001-7442-2693</externalLink>)
– Name: TitleSource
  Label: Source
  Group: Src
  Data: <searchLink fieldCode="SO" term="%22Cognitive+Science%22"><i>Cognitive Science</i></searchLink>. Feb 2022 46(2).
– Name: Avail
  Label: Availability
  Group: Avail
  Data: Wiley. Available from: John Wiley & Sons, Inc. 111 River Street, Hoboken, NJ 07030. Tel: 800-835-6770; e-mail: cs-journals@wiley.com; Web site: https://www.wiley.com/en-us
– Name: PeerReviewed
  Label: Peer Reviewed
  Group: SrcInfo
  Data: Y
– Name: Pages
  Label: Page Count
  Group: Src
  Data: 24
– Name: DatePubCY
  Label: Publication Date
  Group: Date
  Data: 2022
– Name: TypeDocument
  Label: Document Type
  Group: TypDoc
  Data: Journal Articles<br />Reports - Research
– Name: Subject
  Label: Descriptors
  Group: Su
  Data: <searchLink fieldCode="DE" term="%22Language+Usage%22">Language Usage</searchLink><br /><searchLink fieldCode="DE" term="%22Indo+European+Languages%22">Indo European Languages</searchLink><br /><searchLink fieldCode="DE" term="%22English+%28Second+Language%29%22">English (Second Language)</searchLink><br /><searchLink fieldCode="DE" term="%22Bilingualism%22">Bilingualism</searchLink><br /><searchLink fieldCode="DE" term="%22Code+Switching+%28Language%29%22">Code Switching (Language)</searchLink><br /><searchLink fieldCode="DE" term="%22Cognitive+Processes%22">Cognitive Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Time%22">Time</searchLink>
– Name: DOI
  Label: DOI
  Group: ID
  Data: 10.1111/cogs.13105
– Name: ISSN
  Label: ISSN
  Group: ISSN
  Data: 1551-6709
– Name: Abstract
  Label: Abstract
  Group: Ab
  Data: Previous research has shown that language switching is costly, and that these costs are likely to persist even when speakers are given ample time to prepare. The aim of this study was to determine whether there are cognitive limitations to speakers' ability to prepare for a switch, or whether a new language can be prepared in advance and any cost to switch language eliminated. To explore this, language switching costs were measured in a group of Dutch-English (L1-L2) bilinguals who named pictures in their two languages while the preparation time was manipulated. The participants were given either no time to prepare (cue to stimulus interval, CSI: 0 ms), or some time to prepare, for the target language (CSI: 250, 500, and 800 ms). The results revealed that when speakers had no time to prepare, language switching was costly. However, when preparation time was provided, switching costs disappeared. This suggests that there might be no cognitive limitations to the ability to prepare for a language switch, and that, provided enough preparation time, the effort to switch language could be eliminated. This finding might also explain why normal code-switched conversations seem effortless, as speakers typically have ample time to prepare for the language switch.
– Name: AbstractInfo
  Label: Abstractor
  Group: Ab
  Data: As Provided
– Name: Note
  Label: Notes
  Group: Note
  Data: https://osf.io/bnfs8
– Name: DateEntry
  Label: Entry Date
  Group: Date
  Data: 2022
– Name: AN
  Label: Accession Number
  Group: ID
  Data: EJ1330021
PLink https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1330021
RecordInfo BibRecord:
  BibEntity:
    Identifiers:
      – Type: doi
        Value: 10.1111/cogs.13105
    Languages:
      – Text: English
    PhysicalDescription:
      Pagination:
        PageCount: 24
    Subjects:
      – SubjectFull: Language Usage
        Type: general
      – SubjectFull: Indo European Languages
        Type: general
      – SubjectFull: English (Second Language)
        Type: general
      – SubjectFull: Bilingualism
        Type: general
      – SubjectFull: Code Switching (Language)
        Type: general
      – SubjectFull: Cognitive Processes
        Type: general
      – SubjectFull: Time
        Type: general
    Titles:
      – TitleFull: When Language Switching Is Cost-Free: The Effect of Preparation Time
        Type: main
  BibRelationships:
    HasContributorRelationships:
      – PersonEntity:
          Name:
            NameFull: Mosca, Michela
      – PersonEntity:
          Name:
            NameFull: Manawamma, Chaya
      – PersonEntity:
          Name:
            NameFull: de Bot, Kees
    IsPartOfRelationships:
      – BibEntity:
          Dates:
            – D: 01
              M: 02
              Type: published
              Y: 2022
          Identifiers:
            – Type: issn-electronic
              Value: 1551-6709
          Numbering:
            – Type: volume
              Value: 46
            – Type: issue
              Value: 2
          Titles:
            – TitleFull: Cognitive Science
              Type: main
ResultId 1