Research on the cognitive consequences of bilingualism typically proceeds by labeling participants as “monolingual” or “bilingual” and comparing performance on some measures across these groups. It is well-known that this approach has led to inconsistent results. However, the approach assumes that there are clear criteria to designate individuals as monolingual or bilingual, and more fundamentally, to determine whether a communication system counts as a unique language. Both of these assumptions may not be correct. The problem is particularly acute when participants (...) are asked to classify themselves or simply report how many languages they speak. Participants' responses to these questions are shaped by their personal perceptions of the criteria for making these judgments. This study investigated the perceptions underlying judgments of bilingualism by asking 528 participants to judge the extent to which a description of a fictional linguistic system constitutes a unique language and the extent to which a description of a fictional individual's linguistic competence qualifies that person as bilingual. The results show a range of responses for both concepts, indicating substantial ambiguity for these terms. Moreover, participants were asked to self-classify as monolingual or bilingual, and these decisions were not related to more objective information regarding the degree of bilingual experience obtained from a detailed questionnaire. These results are consistent with the notion that bilingualism is not categorical and that specific language experiences are important in determining the criteria for being bilingual. The results impact interpretations of research investigating group differences on the cognitive effects of bilingualism. (shrink)

Trust and trustworthiness are central concepts in contemporary discussions about the ethics of and qualities associated with artificial intelligence (AI) and the relationships between people, organisations and AI. In this article we develop an interdisciplinary approach, using socio-technical software engineering and design anthropological approaches, to investigate how trust and trustworthiness concepts are articulated and performed by AI software practitioners. We examine how trust and trustworthiness are defined in relation to AI across these disciplines, and investigate how AI, trust and trustworthiness (...) are conceptualised and experienced through an ethnographic study of the work practices of nine practitioners in the software industry. We present key implications of our findings for the generation of trust and trustworthiness and for the training and education of future software practitioners. (shrink)

Attention has recently been proposed as the mechanism underlying the cognitive effects associated with bilingualism. However, similar to bilingualism, the term attention is complex, dynamic, and can vary from one activity to another. Throughout our daily lives, we use different types of attention that differ in complexity: sustained attention, selective attention, alternating attention, divided attention, and disengagement of attention. The present paper is a focused review summarizing the results from studies that explore the link between bilingualism and attention. For each (...) level of attention, a brief overview of relevant theoretical models will be discussed along with a spotlight on paradigms and tasks used to measure these forms of attention. The findings illustrate that different types and levels of attention are modified by the variety of bilingual experiences. Future studies wishing to examine the effects of bilingualism on attention are encouraged to embrace the complexity and diversity of both constructs rather than making global claims about bilingualism and attention. (shrink)

Researchers have recently begun to question the specificity and reliability of conflict adaptation effects, also known as sequential congruency effects, a highly cited effect in cognitive psychology. Some have even used the lack of reliability across tasks to argue against models of cognitive control that have dominated the field for decades. The present study tested the possibility that domain-general processes across tasks might appear on more sensitive mouse-tracking metrics rather than overall reaction times. The relationship between SCE effects on the (...) Stroop and Flanker tasks were examined for the first time using a mouse-tracking paradigm. Three main findings emerged: Robust SCEs were observed for both the Stroop and Flanker tasks at the group level, Within-task split-half reliabilities for the SCE across dependent variables were weak at best and non-existent in many cases, and SCEs for the Flanker and Stroop tasks did not correlate with each other for overall reaction times, but did show significant correlations between tasks on more dynamic measures that captured processes before response execution. These findings contribute to the literature by highlighting how mouse-tracking may be a fruitful avenue by which future studies can examine the specificity and reliability of conflict adaptation and tease apart different theoretical models producing the effects. (shrink)