About 1800 years ago, a tombstone for Regina was set near the Hadrian’s Wall in England. Regina was a slave who was freed and married Barates, a man from Palmyra. Barates erected a monument with a Latin inscription dedicated to her: “To the spirits of the departed and to Regina, his freedwoman and wife. Barates of Palmyra set this up. She was a Catuvellaunian by tribe, aged 30”. Interestingly, after this relatively neutral description, Barates added a lament in his native language, Aramaic: “Regina, the freedwoman of Barates, alas!”. Barates changed languages probably because his emotional experience was best captured in his native tongue. Here, we address whether language context affects the development of a basic emotion, fear.
Emotions and feelings are at the core of human nature. We all experience joy, sadness, sorrow, contempt, and fear, and we often use language to communicate these feelings. Interestingly, the vehicle used to express these emotions seems to affect how we experience them. Highly emotional words or expressions do not seem to prompt the same emotional reaction in the native and foreign languages, as Barates’ story reveals1,2. Why this happens is still unknown, but it is probably related to how native and foreign languages are typically acquired. While native languages are acquired in emotionally rich family contexts, foreign languages are often learned in more emotionally neutral academic contexts3,4.
This differential emotional reactivity has widespread effects in various cognitive domains, as revealed by the so-called foreign language effect on decision-making and language processing. For example, loss and risk aversion are reduced in foreign language contexts5,6. Also, people tend to opt for utilitarian options more often in a foreign language when facing moral dilemmas7, and foreign language contexts affect automatic stages of emotional processing, modulating responses to self-related stimuli8. Finally, processing highly arousing negative sentences elicit a reduced psychophysical response when sentences are presented in a foreign language9,10. Three non-mutually exclusive factors have been suggested to be behind the foreign language effect: a reduction in emotional reactivity, a reduction in cognitive fluency, and an increase in psychological distance (see Costa et al.11 and Hayakawa et al.12, for reviews).We do not know, however, how these factors interact to give rise to this effect. In any case, the notion that foreign language contexts increase emotional distance and/or reduce emotional reactivity is receiving strong empirical support. This opens doors to the potential use of foreign languages in scenarios where it would be desirable a less (or more) emotional involvement, like conflict resolution, moral judgment, healthy choices, or even psychotherapy. Here we focus on whether fear acquisition can be modulated by the language context in which people are set.
The role of native vs. foreign language use in psychotherapy has a long history13. As already noted by Freud and colleagues, bilingual patients sometimes preferred using their second language when talking about anxiety related topics, a phenomenon called “the detachment effect”14,15. Similarly, Costa and Dewaele16 and Dewaele and Costa17described how the use of a foreign language in psychotherapy allowed patients to talk about highly emotional material while feeling somewhat protected by this linguistic detachment. Thus, the use of a foreign language can promote certain level of psychological distance that can help people cope with emotional events. However, most of this evidence is somewhat anecdotic and comes mostly from clinical observations, lacking clear experimental evidence documenting this phenomenon. This is the first empirical study on the impact of using a foreign language for one important psychopathological mechanism, fear conditioning, which grounds one of the most effective psychotherapeutic techniques, exposure therapy. We asked a group of participants to complete a fear conditioning experiment in their native language, and another group completed it in their foreign language. This way, we explored whether a simple change in the linguistic context affects fear acquisition. If this were the case, not only would we better understand the mechanisms behind fear conditioning, but also foresee a new line of research and treatment where language context can help exposure therapy.
Fear conditioning is one the most important paradigms in psychology with clear clinical implications. The paradigm consists of presenting a neutral stimulus, the conditioned stimulus (CS), repeatedly in the presence of an aversive unconditioned stimulus (US), such as an electric shock. This pairing elicits a conditioned response of fear in the absence of the US, only presenting the CS. Fear acquisition and extinction were first conceptualized like low-level processes not involving cognitive processes, but this idea of mutual independence was rapidly discarded18. Fear can be acquired not only by direct learning (repeated pairing of CS-US), but also by observational learning (observing another person’s fear response to a neutral stimulus), and more importantly for our purposes, by verbal instruction (receiving verbal information about the aversive features of a neutral stimulus)19. Here we follow the work of Phelps et al.20, who employed an instructed fear paradigm where fear conditioning and the corresponding physiological response were elicited by just providing verbal instructions explaining the association between the CS and US without actually experiencing the US (see also Luck and Lipp21, for extensions and review). Hence, the use of verbal instructions for fear conditioning provides a unique opportunity to explore the extent to what different language contexts can modulate fear acquisition.
In our study, we adapted the experimental setting used by Phelps et al., and manipulated the language context of the experiment (namely, the language of the task: native or foreign). Verbal instructions were given either in the foreign or native language, and participants had to complete a series of countdowns either in one or the other language during the experiment. Skin conductance and pupil dilation data were collected as measures of the psychophysiological response to the conditioned and unconditioned stimuli. It is important to notice that psychophysiological responses are also affected by cognitive load, and that arguably foreign language processing is likely to increase such load. Indeed, other studies have demonstrated that foreign language contexts elicit larger pupil dilations than native language ones even in conditions with relatively low emotional connotation10,22,23. Hence, a correct comparison of the fear correlates across languages should consider the potential effect of cognitive load associated with foreign language processing. To do so, following the recommendations on human fear conditioning paradigms24, the current experiment follows a differential protocol including a neutral condition in which no fearful stimuli is expected (CS−), and in which participants have to perform the same task as in the fear condition (CS+). The critical issue then is whether the difference between the magnitude of the psychophysiological measures between the neutral and the fear conditions is larger in the native than in the foreign language, and not so much whether pupil dilation or skin conductance levels per se are larger in one language than in the other, since this may reflect just differences in the cognitive load. We predicted a different fear conditioning effect depending on the language context, being the difference between the CS+ and the CS− items lower in the foreign language.
Learning and conditioning have a key role in pathological emotion and behavior regulation, and concretely in the acquisition of fear, linked to important mental disorders like anxiety disorders and stress-related disorders. Thus, showing a mediating role of language context in processes related with fear conditioning can open doors to applications in psychotherapy.