How beliefs about the presence of machine translation impact multilingual collaborations

Abstract

Traditional communication tools tend to make their presence known, e.g., "when my collaborators and I are using IM to discuss our work, how could we not realize the actual presence of IM?" In the case of machine translation (MT) mediated collaborations, however, the absence or presence of MT is not obvious. English sentences with poor grammar can result from both a partner's lack of fluency and errors in the MT process. We hypothesize that partners' attributions about the source of the errors affects their collaboration experience. To test this hypothesis, we conducted a laboratory experiment in which monolingual native English speaking participants collaborated with bilingual native-Mandarin speakers on a map navigation task. Participants were randomly assigned into a 2 (beliefs about MT: absence vs. presence) by 2 (actual mediation of MT: absence vs. presence) experiment design. Beliefs about presence of MT significantly impacted the collaboration experience, opening new opportunities for both research and design around MT-mediated collaborations.

Effect of Machine Translation in Interlingual Conversation

“…Effective MT-mediated communication requires solutions for hurdles posed by the technology [8,25,26]. Yamashita et al found that MT-mediated chat imposes extra difficulties of building common ground between multiple parties due to mistranslation and peoples' lack of awareness thereof [25,26].…”

“…Yamashita et al found that MT-mediated chat imposes extra difficulties of building common ground between multiple parties due to mistranslation and peoples' lack of awareness thereof [25,26]. This not only hinders effective mutual understanding, but also reduces perceived quality of communication [8].…”

Examining Agency, Expertise, and Roles of AI Systems in AI-Mediated Communication

“…One line of future work in AI-MC could investigate attributions of responsibility in the form of blame or credit. Research by Hohenstein and Jung [56] and Gao et al [45] suggest that in instances where people are using AI-generated text suggestions or machine translations in interpersonal communication, and there is some sort of communication problem, people are more likely to attribute agency to the AI system. As a result, people tend to blame the AI system for the problem to at least some degree, lessening the blame that is put on the communication partner.…”

Two is Better Than One

“…Showing the fact that there are alternatives doesn't just provide a resource for repair; it also foregrounds the fact that translation is happening in the first place and that it is imperfect. Functionally, this has benefits such as encouraging people to attribute problems to the technology rather than to collaborators, in turn improving social outcomes [11]. Philosophically, supporting reflection [30] on the way systems impact our communication is also beneficial, paralleling discussions about how search and recommendation algorithms affect our experience of information [20] and how systems that track our behavior make both correct and incorrect inferences about us [23].…”

“…Because previous work has shown that beliefs about the presence or absence of MT affect communication and collaboration [11]., more research is needed to separate the effects of having two translations on message understanding vs. attributions about the causes of misunderstandings.…”