Neuroscientific evidence

In two imaging studies, aspiring interpreters were scanned at the beginning and end of a 15-month-long SI program, including 10 weekly hours of formal SI instruction and additional practice sessions. In both cases, the control group was composed of UMs matched for L2 proficiency. One of those studies (Hervais-Adelman, Moser-Mercer, Murray & Golestani, Reference Hervais-Adelman, Moser-Mercer, Murray and Golestani2017) found that, by the second assessment, only SI trainees had developed increased cortical thickness in temporal, parietal, and dorsal premotor regions supporting phonetic, lexico-semantic, and executive functions. The other one (Hervais-Adelman, Moser-Mercer & Golestani, Reference Hervais-Adelman, Moser-Mercer and Golestani2015), based on a subset of participants from the abovementioned study, analyzed fMRI recordings during a shadowing and an interpreting task. Relative to UMs, the trainees exhibited distributed patterns of functional changes between scans, with SI performance in the latter group being characterized by reduced recruitment of the right caudate nucleus, a region implicated in bilingual control (Luk, Green, Abutalebi & Grady, Reference Luk, Green, Abutalebi and Grady2011) and word translation (García, Reference García2013). More recently, structural and functional brain changes were reported after only nine months of training in SI (as opposed to translation), mainly in temporo-parietal, frontostriatal, and fronto-cerebellar circuits which have been previously implicated in diverse functions, including cognitive control (Van de Putte, De Baene, García-Pentón, Woumans, Dijkgraaf & Duyck Reference Van de Putte, De Baene, García-Pentón, Woumans, Dijkgraaf and Duyck2018).

Interestingly, neuroanatomical and neurofunctional differences in many of these regions have also been reported in PSIs relative to UMs (Elmer, Hänggi & Jäncke, Reference Elmer, Hänggi and Jäncke2014; Elmer, Meyer & Jäncke, Reference Elmer, Meyer and Jäncke2010; Elmer & Kühnis, Reference Elmer and Kühnis2016) and consecutive interpreters and translators (Becker, Schubert, Strobach, Gallinat & Kühn, Reference Becker, Schubert, Strobach, Gallinat and Kühn2016), even in proportion to the subjects’ amount of training. Note, however, that some such findings are mixed and others have been criticized on methodological and interpretive grounds (Hervais-Adelman, Moser-Mercer & Golestani, Reference Hervais-Adelman, Moser-Mercer and Golestani2018; Koshkin & Ossadtchi, Reference Koshkin and Ossadtchi2017) – details can be found in a separate review within this special issue.

Evidence on verbal domains

Complementary evidence has revealed behavioral enhancements in specific verbal domains. Using a 15-minute-long SI task, Tzou, Eslami, Chen, and Vaid (Reference Tzou, Eslami, Chen and Vaid2012) recorded the output of the participants, had their renditions judged in terms of two translation quality measures, and observed that second-year trainees outperformed first-year trainees and that the latter outperformed UMs. In another study (Hervais-Adelman et al., Reference Hervais-Adelman, Moser-Mercer and Golestani2015), aspiring interpreters performed shadowing and interpreting tasks at the beginning and end of a 15-month-long SI program, with UMs being tested at the same time-points. Comparisons of performance between testing sessions showed that only the SIs had significantly improved their interpreting outcomes, whereas neither group exhibited changes in the shadowing condition. Taken together, these studies show that SI training is linked to a greater capacity for achieving linguistic correspondence between source and target texts.

More notably, additional evidence shows that early advantages can be tracked in specific sub-domains encompassed by SI. In an auditory shadowing experiment, Darò (Reference Darò1989) found that first-year ISs made significantly fewer errors than UMs. While these results should be taken with reserve due to the extremely low sample size of both groups (n ≤ 5), a more robust study found that a sample of interpreting students (combined with a few professionals) proved more precise than UMs and translators in respeaking – a form of shadowing whereby live discourse from audiovisual materials is immediately repeated into speech recognition software (Szarkowska, Krejtz, Dutka & Pilipczuk, Reference Szarkowska, Krejtz, Dutka and Pilipczuk2018). Together, these studies suggest higher abilities in interpreting trainees for handling simultaneous comprehension and production, both inter- and intra-lingually.

For their part, Bajo, Padilla & Padilla (Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000) assessed first-year ISs and UMs via semantic categorization and sentence comprehension tasks. While no between-group differences were reported in a first evaluation, the trainees were the only ones exhibiting either nearly-significant effects (trends) or significant improvements across all tasks upon retesting one year later. Still, these results should be taken with caution as they were obtained with small sample sizes (n ≤ 10) and without a systematic assessment of potential confounds, such as L2 proficiency. Actually, not all studies have found differences between ISs and proficient bilinguals in linguistic tasks, as shown by measures of word reading speed (Chincotta & Underwood, Reference Chincotta and Underwood1998) and verbal fluency (Van de Putte et al., Reference Van de Putte, De Baene, García-Pentón, Woumans, Dijkgraaf and Duyck2018).

Briefly, SI training seems to confer verbal processing advantages in some but not all domains examined so far. Suggestively, the same seems to be true of executive functions, as discussed next.

Evidence on executive domains

Linguistic processes in the bilingual mind constantly interact with (domain-general) executive mechanisms, including STM, WM, cognitive coordination, inhibitory control, mental-set shifting, self-monitoring, and attentional skills (Miyake & Friedman, Reference Miyake and Friedman2012). Notably, performance in relevant tasks seems to increase in proportion to the subjects’ bilingual abilities (Linck et al., Reference Linck, Osthus, Koeth and Bunting2014). Given that many executive domains are markedly taxed during SI training, at least some of them might be honed in ISs.

Beginner ISs have shown advantages over UMs in free-recall tasks, but only under articulatory suppression (Köpke & Nespoulous, Reference Köpke and Nespoulous2006). Moreover, results from WM tasks requiring concurrent sentential processing systematically indicate better performance in both languages for ISs than UMs at the end (Tzou et al., Reference Tzou, Eslami, Chen and Vaid2012) and even at early stages (Köpke & Nespoulous, Reference Köpke and Nespoulous2006) of their training. This WM advantage was corroborated in longitudinal studies testing ISs before and after five (Antonova Ünlü & Sağın Şimşek, Reference Antonova Ünlü and Sağın Şimşek2018) and four (Chmiel, Reference Chmiel2016) semesters of practice. However, only the first of these investigations was able to show the specificity of this effect – by testing UMs at the same time-points and finding no comparable results. Interestingly, this robust pattern does not emerge when intervening cognitive tasks involve non-verbal materials (Babcock, Capizzi, Arbula & Vallesi, Reference Babcock, Capizzi, Arbula and Vallesi2017), highlighting its specificity for language-mediated operations. Neither does it seem to hold for sign-language ISs (Macnamara & Conway, Reference Macnamara and Conway2014), arguably because of particularities of bimodal interpreting, such as sensorimotor differences between input-output interfaces for oral and signed languages. Nevertheless, these subjects do seem to exhibit training-related gains in WM coordination and transformation functions after four semesters, probably due to enhancements in managing joint comprehension and production of two languages (Macnamara & Conway, Reference Macnamara and Conway2014).

STM assessments (via simple span tasks) have also yielded mixed results, likely reflecting discrepancies in the stimuli used and particularities of the samples tested in each study. While Tzou et al. (Reference Tzou, Eslami, Chen and Vaid2012) reported better digit span for second-year ISs than UMs, the latter group was undifferentiated from beginner (Köpke & Nespoulous, Reference Köpke and Nespoulous2006; Tzou et al., Reference Tzou, Eslami, Chen and Vaid2012) and advanced (Antonova Ünlü & Sağın Şimşek, Reference Antonova Ünlü and Sağın Şimşek2018; Chincotta & Underwoord, Reference Chincotta and Underwood1998) ISs in various STM measures. Moreover, longitudinal evidence indicates that interpreting training can trigger improvements in STM for letters but not for digits or spatial locations (Babcock et al., Reference Babcock, Capizzi, Arbula and Vallesi2017; Antonova Ünlü & Sağın Şimşek, Reference Antonova Ünlü and Sağın Şimşek2018). Therefore, interpreting training does not seem to induce systematic STM advantages in the course of pre-professional preparation. Nevertheless, the great heterogeneity in the tasks and procedures used precludes any firm conclusions and calls for replication studies including all reported experimental approaches.

Inconclusive results have also been documented for different aspects of cognitive control. On the one hand, mental-set shifting skills seem to be enhanced in first-year ISs relative to UMs, and even more so in second-year trainees (Dong & Xie, Reference Dong and Xie2014). On the other, assessments of attentional (Babcock et al., Reference Babcock, Capizzi, Arbula and Vallesi2017), inhibitory (Dong & Xie, Reference Dong and Xie2014; Köpke & Nespoulous, Reference Köpke and Nespoulous2006; Van de Putte et al., Reference Van de Putte, De Baene, García-Pentón, Woumans, Dijkgraaf and Duyck2018), and switching (Babcock et al., Reference Babcock, Capizzi, Arbula and Vallesi2017; Van de Putte et al., Reference Van de Putte, De Baene, García-Pentón, Woumans, Dijkgraaf and Duyck2018) skills do not point to any coherent behavioral advantage for ISs after one or even two years of training. As will be argued below, the absence of effects in these domains could reflect their marginal relevance for successful performance of SI (see “Discussion”).

Roughly speaking, then, approximately 12 months of SI training seems enough to induce significant neurocognitive changes. Beyond neuroimaging outcomes, the most robust findings reveal improvements in SI proper as well as recall (especially under articulatory suppression), set-shifting and, above all, WM skills. However, assessments of STM, attention, inhibition, and switching have yielded mixed or consistently null results. Although the evidence for some of these patterns is limited, inconsistent, or based on unevenly reliable paradigms, it would thus seem only specific functions, directly taxed in the exercise of SI, benefit from sustained practice of this activity. Importantly, neuroanatomical, hemodynamic, and behavioral findings from pre/post-training experiments show that these changes are directly triggered by SI practice, thus ruling out their prior existence in prospective interpreters. Remarkably, at least some of those changes prove independent of other experience-related factors, such as L2 proficiency. An integrative treatment of these patterns will be offered in the “Discussion” section, after we consider additional empirical constraints from research on PSIs, as detailed below.

Evidence from professional interpreters

A second set of studies illuminates the issue by focusing on PSIs. This empirical corpus includes several experiments on verbal and executive functions, which we address separately in the following sections.

Evidence on verbal domains

Interpreting expertise has been linked to enhanced auditory perception. Elmer, Klein, Kuhnis, Liem, Meyer and Jäncke (Reference Elmer, Klein, Kuhnis, Liem, Meyer and Jäncke2014) tested PSIs, professional musicians, and non-expert controls in a sound categorization task involving music-to-noise, speech-to-noise, and speech-to-music continua. PSIs were the only ones exhibiting a bias towards linguistic stimuli in the speech-to-music condition, and they also showed enhanced skills for categorizing musical items. Though very tentatively due to the study's modest sample sizes and reduced stimulus set, this pattern might reflect the development of enhanced, multifaceted echoic memory representations, arguably because successful SI hinges on efficient extraction and recognition of diverse auditory signals.

As regards linguistic processes proper, some reports have compared PSIs and UMs via sentence- and discourse-level tasks. Preliminary evidence from a self-paced reading paradigm (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000) suggested better sentence comprehension in the former population. A similar finding was reported by Yudes, Macizo, Morales and Bajo (Reference Yudes, Macizo, Morales and Bajo2013), who found that PSIs outperformed UMs with similar L2 proficiency in answering open-ended questions after a text-reading task. Compatibly, in a small-scale study with reduced sample sizes, Dillinger (Reference Dillinger1990) compared PSIs and UMs in interpreting and recalling measures, and found that the former exhibited greater correspondence between the propositions of the source and target texts. Together, these studies suggest that experts in SI possess an elevated capacity to understand unfolding texts – another competence that is markedly taxed in the profession. Still, a better control of L2-related confounds would be necessary to assess the specificity of this pattern (see “Discussion” section).

Other studies have focused on error detection skills. In a dichotic listening test (Fabbro, Gran & Gran, Reference Fabbro, Gran and Gran1991), PSIs recognized more semantic inaccuracies than ISs in translated sentence pairs, but the opposite was true for syntactic errors. Similarly, using a text-based task, Yudes et al. (Reference Yudes, Macizo, Morales and Bajo2013) found that PSIs outperformed UMs in detecting semantic (but not lexical or syntactic) mistakes. Plausibly, the need to properly comprehend input could involve greater sensitivity to semantic features across unfolding pieces of discourse.

Beyond such results, most evidence comes from lexical paradigms. A pivotal finding is that PSIs know more words than bilingual university students, but they present no differences relative to L2 teachers with similar amounts of professional experience (Christoffels, de Groot & Kroll, Reference Christoffels, de Groot and Kroll2006). The same seems true of phonological and semantic fluency tasks. These have revealed better vocabulary-search skills in PSIs relative to UMs (Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño & García, Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018) but not as compared to L2 teachers (Stavrakaki et al., Reference Stavrakaki, Megari, Kosmidis, Apostolidou and Takou2012). Therefore, interpreting expertise may be related to greater word knowledge and fluency, but such an effect can also emerge through other sustained practices taxing bilingual verbal skills. Importantly, however, since both groups in Santilli et al. (Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018) were matched in key L2-related variables (competence, age of acquisition, years of study, weekly exposure), that effect seems directly – though non-exclusively– related to interpreting expertise, arguably due to the continual need to retrieve specific words under pressing time constraints (Chernov, Reference Chernov2004).

Moreover, two studies indicate that PSIs also present advantages in word translation. First, Christoffels et al. (Reference Christoffels, de Groot and Kroll2006) reported that PSIs were faster than UMs in both L2-L1 and L1-L2 translation, but their performance was similar to that of L2 teachers. While the authors attributed the latter pattern to high L2 proficiency in both professional groups, more recent evidence suggests otherwise. Indeed, Santilli et al. (Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018) found that PSIs had shorter response times than UMs in both translation directions despite being matched in L2 competence, age of acquisition, years of study, and weekly exposure. Also, PSIs proved more accurate and faster than UMs with similar language proficiency at detecting whether serially presented words belonged to their L1 or L2 (Aparicio, Heidlmayr & Isel, Reference Aparicio, Heidlmayr and Isel2017). Thus, sustained engagement in interlingual reformulation by PSIs may optimize cross-linguistic processing and language identification skills, even beyond the impact of L2 proficiency.

However, other operations within bilingual memory do not seem boosted in interpreters. Indeed, two separate experiments reported non-significant differences between PSIs and UMs or L2 teachers in picture naming (Christoffels et al., Reference Christoffels, de Groot and Kroll2006; Santilli et al., Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018). Neither are PSIs characterized by consistent enhancements in other conceptually-mediated word-level processes. Semantic congruency judgment tasks involving pairs of (real) words have yielded mixed results, with one study showing interpreter advantages relative to L2-matched UMs (Elmer & Kühnis, Reference Elmer and Kühnis2016) and another one yielding non-significant behavioral differences and distinct neurophysiological responses in only one out of four conditions (Elmer et al., Reference Elmer, Meyer and Jäncke2010). Null effects of interpreting experience were also observed for semantic categorization of real words (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000). However, the latter study showed that PSIs did outperform UMs in categorizing non-typical exemplars (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000). Though undermined by its relatively low sample size and small number of trials, this result might suggest that the effects of SI experience on conceptual mechanisms may be confined to unfamiliar lexical fields.

Interestingly, a similar pattern seems true of lower-order language skills. Widely experienced PSIs and UMs of comparable L2 proficiency have similar outcomes in over-practiced or shallow-processing tasks, such as repeatedly counting from one to ten (Signorelli, Haarmann & Obler, Reference Signorelli, Haarmann and Obler2012), performing lexical decisions on real words (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000), reading such items in L1 or L2 (Santilli et al., Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018), or repeating them, even under distractive conditions (Hiltunen et al., Reference Hiltunen, Paakkonen, Vik and Krause2016). Conversely, PSIs do outperform UMs when asked to recognize (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000) or repeat (Signorelli et al., Reference Signorelli, Haarmann and Obler2012) non-words, and they also surpass L2 teachers when asked to recall such stimuli (Stavrakaki et al., Reference Stavrakaki, Megari, Kosmidis, Apostolidou and Takou2012). Accordingly, PSIs’ enhancements in form-level processing may be evident only for unfamiliar, non-rotely-learned items – which are arguably highly recurrent in professional SI settings.

In sum, only some bilingual verbal skills are behaviorally enhanced in PSIs. Relative to UMs, these experts would appear to possess superior skills for auditory perception, discourse comprehension, and semantic-error detection. Moreover, although their wide vocabulary may not exceed that of other language professionals, they seem characterized by greater efficiency for time-constrained lexical search, word translation, and processing of unfamiliar items. Nevertheless, they possess no clear advantages in conceptually-mediated word-level processes. Finally, their expertise seems to have no influence on lexical- and syntactic-error detection or shallow-processing tasks. Therefore, linguistic enhancements in this population seem confined to those mechanisms directly taxed in professional settings (see “Discussion” below).

Evidence on executive domains

Additional results have been reported for executive domains. Numerous studies assessing STMFootnote ³ and WM have consistently revealed advantages for PSIs over UMs. Although digit-span tasks have yielded significant (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000), marginal (Stavrakaki et al., Reference Stavrakaki, Megari, Kosmidis, Apostolidou and Takou2012), and null (Köpke & Nespoulous, Reference Köpke and Nespoulous2006; Santilli et al., Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018) effects, STM advantages have been found for letters (Babcock & Vallesi, Reference Babcock and Vallesi2017), spatial locations in visual matrices (Babcock & Vallesi, Reference Babcock and Vallesi2017), non-words (Stavrakaki et al., Reference Stavrakaki, Megari, Kosmidis, Apostolidou and Takou2012), and words in L1 and L2 (Christoffels et al., Reference Christoffels, de Groot and Kroll2006). Compatibly, PSIs outperformed translators in a letter memory task (Henrard & Van Daele, Reference Henrard and Van Daele2017). Considering that STM effects prove much less systematic in ISs (see above), this might suggest that gains in these domains tend to consolidate once expert standards are attained (see “Discussion”).

Superior memory performance has also been shown in span tasks involving concurrent cognitive operations, such as reading ever-larger groups of phrases (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000; Christoffels et al., Reference Christoffels, de Groot and Kroll2006), uttering written sentences (Signorelli et al., Reference Signorelli, Haarmann and Obler2012; Yudes, Macizo & Bajo, Reference Yudes, Macizo and Bajo2011; Yudes et al., Reference Yudes, Macizo, Morales and Bajo2013), producing spontaneous statements (Christoffels et al., Reference Christoffels, de Groot and Kroll2006), judging sentence sensibility (Chmiel, Reference Chmiel2016), and performing mathematical operations (Babcock & Vallesi, Reference Babcock and Vallesi2017). Therefore, PSIs seem characterized by increased ability to transiently store information while performing additional tasks, a critical requisite for successful SI. However, results from a listening span task showed no differences between PSIs and L2 teachers (Stavrakaki et al., Reference Stavrakaki, Megari, Kosmidis, Apostolidou and Takou2012), suggesting that other forms of expert bilingual training may also involve WM gains. Therefore, in line with evidence of enhanced WM in consecutive interpreters (Dong & Liu, Reference Dong and Liu2016), it would seem that one and the same executive domain may significantly benefit from different professional activities taxing dual-language mechanisms.

Evidence from free- and cued-recall paradigms reinforces the view that mnesic advantages in PSIs are related to the particular demands they encounter in professional settings. When participants must simply recall as many words as possible from a given list, no differences emerge between PSIs and UMs (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000; Köpke & Nespoulous, Reference Köpke and Nespoulous2006; Signorelli et al., Reference Signorelli, Haarmann and Obler2012; Stavrakaki et al., Reference Stavrakaki, Megari, Kosmidis, Apostolidou and Takou2012). The same is true in comparisons with L2 teachers or consecutive interpreters, both for word lists (Hiltunen et al., Reference Hiltunen, Paakkonen, Vik and Krause2016) and prose passages (Hiltunen & Vik, Reference Hiltunen and Vik2017). Conversely, when the task involves articulatory suppression (by repeating a predefined syllable during encoding), PSIs exhibit significantly better outcomes than UMs (Bajo et al., Reference Bajo, Padilla, Padilla, Chesterman, Gallardo and Gambier2000; Köpke & Nespoulous, Reference Köpke and Nespoulous2006). In fact, they even outperform advanced ISs in complex (though not in standard) suppression conditions (Yudes, Macizo & Bajo, Reference Yudes, Macizo and Bajo2012). Therefore, recall efficiency in this population only seems significantly greater when subvocal rehearsal is impeded – i.e., in scenarios that replicate the processing conditions of actual SI. In fact, whereas interpreting performance correlates with WM under articulatory suppression, no such association emerges when subvocal rehearsal is allowed (Injoque-Ricle et al., Reference Injoque-Ricle, Barreyro, Formoso and Jaichenco2015).

Likewise, PSIs seem better than UMs and other language professionals (translators and consecutive interpreters) at handling specific cognitive control functions. They were observed to perform faster on a dual task requiring joint identification of tones and shapes (Becker et al., Reference Becker, Schubert, Strobach, Gallinat and Kühn2016; Strobach, Becker, Schubert & Kühn, Reference Strobach, Becker, Schubert and Kühn2015) and in a dual (visual-auditory) version of the n-back-task – where participants had to identify any item that matched stimuli presented one, two, or three trials earlier (Morales, Padilla, Gomez-Ariza & Bajo, Reference Morales, Padilla, Gomez-Ariza and Bajo2015). Therefore, expertise in SI would seem related to better coordinating functions, a critical component of successful SI (Chernov, Reference Chernov, Lambert and Moser-Mercer1994, Reference Chernov2004; García, Reference García2014; Gerver, Reference Gerver and Briskin1976). In particular, this skill would be needed to manage joint input and output demands while refreshing critical information in memory segment after segment.

However, research on other executive functions has produced inconclusive findings. Comparisons of overall processing speed between PSIs and other language professionals have yielded both significant and null results (Christoffels et al., Reference Christoffels, de Groot and Kroll2006; Henrard & Van Daele, Reference Henrard and Van Daele2017). Also, whereas the PSIs assessed by Yudes et al. (Reference Yudes, Macizo and Bajo2011) proved more efficient and precise than UMs (matched for L2 proficiency) at the Wisconsin Card-Sorting Test (a classical measure of mental-set shifting skills), no between-group differences were observed by Santilli et al. (Reference Santilli, Gonzalez, Mikulan, Martorell, Muñoz, Sedeño and García2018) on the same task. While this discrepancy might be partly related to the analysis of only one outcome measure (number of correctly inferred categories) in the latter study, further evidence attests to the unsystematic impact of SI expertise on cognitive control. Indeed, when both groups are asked to switch between two tasks (color and shape discrimination), PSIs exhibit smaller mixing costs but not smaller switching costs (Babcock & Vallesi, Reference Babcock and Vallesi2017), even when compared with translators and consecutive interpreters (Becker et al., Reference Becker, Schubert, Strobach, Gallinat and Kühn2016). Interestingly, the former measure is negatively associated with grey matter density in the left frontal pole, a region implicated in cognitive control (Becker et al., Reference Becker, Schubert, Strobach, Gallinat and Kühn2016). Neither did PSIs differ from translators in a plus-minus task assessing flexibility (Henrard & Van Daele, Reference Henrard and Van Daele2017). This pattern is consistent with the demands of professional SI, where intrusions must be minimized during a single dual-language activity, without a need to alternate between tasks involving different modalities.

Moreover, it seems that PSIs possess no advantages in tasks assessing aspects of attention and inhibition. Research on alerting and orienting functions shows that attentional domains are not particularly honed in these subjects, either in terms of efficacy (accuracy) or efficiency (reaction times) (Babcock & Vallesi, Reference Babcock and Vallesi2017; Morales et al., Reference Morales, Padilla, Gomez-Ariza and Bajo2015). Similarly, no interpreter advantages have emerged in conflict-resolution experiments that assessed the capacity to suppress the effects of lexical (Babcock & Vallesi, Reference Babcock and Vallesi2017; Köpke & Nespoulous, Reference Köpke and Nespoulous2006) and locational (Yudes et al., Reference Yudes, Macizo and Bajo2011) information on perceptually-driven processes – but see Henrard and Van Daele (Reference Henrard and Van Daele2017) for different results. In fact, evidence from a Stroop task shows that PSIs respond faster but significantly less accurately than UMs (Aparicio et al., Reference Aparicio, Heidlmayr and Isel2017). Though based on varyingly robust tasks, these null and negative results might reflect the fact that successful practice of SI does not require active suppression of interference among linguistic, spatial, and perceptual modalities.

Overall, as is the case with verbal domains, only some executive mechanisms seem enhanced in PSIs. Although the evidence is not fully consistent, these experts appear to possess elevated STM and WM skills, higher recall capacity under articulatory suppression, and greater abilities to perform simultaneous tasks and prevent mixing between alternating processes. However, results concerning task-switching, attentional, and inhibitory functions indicate null effects of interpreting experience. Once again, among all executive domains, only those distinctively taxed during SI appear to be boosted in interpreters (see “Discussion” below).