Language Choice in Psychotherapy of Multilingual Clients: Perspectives from Multilingual Therapists

Introduction

Language is crucial to psychotherapeutic practice. It is by means of language that good rapport and effective communication are established between therapist and client. Indeed, psychotherapy is often called “talk therapy”. However, as Costa (2020) points out, there is insufficient “talking about the talk” in training for psychotherapists, who may end up unprepared when facing clients – and possibly interpreters – with unusual linguistic profiles, which may change the dynamics in the session. The likelihood of psychotherapists encountering such situations is increasing these days, with large numbers of refugees fleeing their home country, often under traumatic conditions, before settling down in host societies.

Addressing language issues in psychotherapy is vital as they can impact patient participation (Schinkel et al., 2018) and the understanding of patient's complaints as well as doctor's advice (Twilt et al., 2019). Since therapy involves emotion communication (i.e., the expressing and interpreting of emotions), it is vital to gain a better understanding on the issue in a wide variety of contexts. However, attention in the profession for linguistic and cultural diversity of therapists and clients is limited. To the best of our knowledge, even if there is a focus on language in psychotherapy research, accounts predominantly examine the role of the client's multilingualism, but rarely the therapist's multilingualism, which may shape encounters with

multilingual/multicultural clients. Research has shown that awareness about multilingualism is crucial for therapists working with clients from different linguistic and cultural backgrounds (Bager-Charleson et al., 2017; Costa & Dewaele, 2019). The current qualitative study adds to this avenue of research by focusing on how multilingual and multicultural therapists in the Netherlands use their languages. Our research focuses on the importance of making explicit the Dutch therapists' implicit, experiential knowledge of allowing multiple languages in the room.

Theoretical Framework

Following Dewaele and Li (2013, p. 231), we define multilinguals as “people with at least partial mastery in a number of languages”. It is not uncommon in literature on multilingualism to see the terms “multilingual” and “multicultural” used next to each other. However, being multilingual does not necessarily mean being multicultural. Comparing these notions, Grosjean (2015, p. 575) speaks about multiculturalism as “cultural competence or knowledge” and “interacting in two or more cultures”, and we adopt this definition. Therefore, in this paper the term “multilingual/multicultural” means speaking more than one language and identifying with more than one culture. This does not imply that the people involved have equal skill and familiarity with all languages and cultures in their repertoire, but simply that they have had exposure to and interaction in more than one language and culture. Dewaele and Li Wei (2014) pointed out that code-switching, “the alternation and mixing of languages within a conversational utterance” (p. 235) is a common phenomenon in exchanges between bi- or multilinguals. Although code-switching is typically linked to higher levels of linguistic mastery (Dewaele, 2013), is has sometimes been associated with a lack of control or a poor understanding of the languages. Dewaele and Li Wei (2014) found that attitudes toward code-switching were rather mixed, even within bi- and multilingual populations. In their sample of 2070 mono- and multilinguals, they found that participants knowing more languages did not have significantly more positive attitudes toward code-switching. Personality played a role, with participants scoring high on Tolerance of Ambiguity, Cognitive Empathy and Emotional Stability reporting more positive attitudes toward code-switching. Participants' upbringing, experience of living abroad, and working environment also had an effect: early bilinguals, those who had lived abroad, and those working in an ethnically diverse environment reported more positive attitudes towards code-switching. Gender, age, and education levels also played a role, with female participants, older participants, and participants at both the low and the high end of education appreciating code-switching the most. The implication of this finding for the current study is that multilingual therapists might hold different attitudes towards code-switching with their patients, and they might be unaware of the potential therapeutic benefits of tapping into their own and their clients' languages. In addition, different attitudes may be held towards different varieties of the same language. This is particularly important when the client's L1 is a low status variety of the language in question, as such speakers may fear negative judgments because of the way they speak L1.

Communicating Emotions in Multiple Languages

One of the striking findings in multilingualism research is that multilinguals' languages can have different functions and can be used in different discourse domains. Grosjean (2012) named this the “Complementarity principle”, the fact that multilinguals' language preferences differ according to situations, interlocutors, and purposes. Somebody who is fluent in two languages may struggle to discuss something in a language that that person does not usually use for that topic. Being forced to talk about emotions in a foreign language (LX), for example, may trigger a sense of constraint, of frustration, and even of alienation (Panicacci & Dewaele, 2017).

Multilinguals' languages may also “feel” different because of variation in levels of emotional resonance: LXs typically have less emotional power than first languages (L1s), and LXs are typically dispreferred for communicating emotions (Dewaele, 2013). Even films and news bulletins in the LX have been found to elicit weaker emotional reactions among LX users compared to L1 users (Dewaele, Lorette, Rolland & Mavrou, 2021). The authors discovered that 553 English LX users reported lower frequency of feeling emotional when watching the news or a film, lower frequency of laughter when watching a funny film, and a lower degree of trust in news reports in English compared to 271 L1 users.

The reason is that L1(s) are more strongly embodied, having been acquired from birth and used in childhood through a process of intense affective socialization, when linguistic knowledge develops with social and cultural knowledge, as well as emotion regulation systems and autobiographical memory. In contrast, LXs are acquired later in life, less intensely, and typically in the relatively artificial classroom environment; they are characterized by more restricted linguistic input, less authentic interaction, and resulting in words that are not linked to a complete conceptual representation (Pavlenko, 2008), which makes them feel uncalibrated and “detached”. Intense secondary affective LX socialization may change this, but LX users may occasionally struggle with the exact meaning and power of emotion words and emotion-laden words, and may be at a loss on how to fit them in an appropriate script (Pavlenko, 2008).

The detachment effect of the LX can have both negative and positive psychological consequences. The 468 participants in Panicacci and Dewaele (2017) were Italian migrants in English-speaking countries for whom Italian had typically remained the language of the heart, with English typically used in professional environments. As a result, their English LX did not allow them to express their intimate feelings with sufficient detail and sophistication, which created a sense of frustration when they had no choice but expressing their emotions in English, especially with less familiar interlocutors (Panicacci & Dewaele, 2018).

Similarly, Dewaele and Salomidou (2017), in a study involving 429 participants who were in multilingual/multicultural relationships showed that some multilinguals complained about a lack of emotional resonance in their LX. This was rather characteristic of early stages in a relationship, when participants reported feeling inauthentic, more distant, and occasionally unable to express their emotions in the LX. With time, though, partners became more familiar with each other's conceptual understanding of emotions, and found their way in common language(s).

In another study, De Leersnyder et al. (2011) demonstrated how knowledge of the grammar and vocabulary of the partner's language was no guarantee for smooth emotion communication in multilingual/multicultural families. Their findings on the emotional experiences of Korean immigrants in the USA and Turkish immigrants in Belgium highlight the role of conceptual understanding and the joint effort in meaning-making. The necessity to speak a LX in a newly formed multilingual family could be a challenge, especially at the beginning, when the partners get to know each other and learn to express their own and interpret the other's emotional states, or when children are involved. When such multilingual/multicultural families communicate, their language choice for expressing and interpreting emotions is highly dependent on each member's language proficiency, their emotional relation to the language, and cultural specifics of emotions.

Expressing and perceiving emotion may already be a challenge in a common language, and even more difficult when more than one is involved. In addition, if the partners decide to

undergo therapy in the local language, foreign to the multilingual family setting, it may be an additional problem (Ruiz-Adams, 2019). Since making meaning of the narrative, talking about sensitive topics, and becoming closer to one's emotions are integral parts of psychotherapy, informing therapists of relevant findings about the linguistic aspects of emotion expression in an accessible way, specifically about both the opportunities and the barriers linguistic identities can trigger, may enrich the therapeutic process and facilitate treatment and healing of the client.

Languages and Emotion in Psychotherapy

Establishing adequate rapport and continuously striving for understanding serves as the basis for a mutually beneficial therapeutic process and good outcome (Hill, 2008). Language plays a crucial role in this process; pre-existing linguistic forms used to express emotion may be misunderstood or misinterpreted, conjure up biases and wrong assumptions, and disrupt the process of giving and receiving adequate treatment. For example, if the client's L1 has an emotional concept that is absent in the language of therapy, the therapist may misjudge the true emotion described by a client. For example, Polish has two equivalents for the English word “angry”, one with a positive and another with a negative connotation, hence causing potential ambiguity for a Polish person speaking English.

An investigation by Costa and Dewaele (2012) into the beliefs, attitudes and practices of 18 monolingual and 83 multilingual therapists in their interactions with their multilingual patients revealed significant differences between both groups. The multilingual therapists were more attuned to their clients (they felt better able to connect) compared to their monolingual colleagues, who were more likely to collude with them. Next, Dewaele and Costa (2013) focused on 182 multilingual clients. The clients underlined the importance of using L1s and LXs with their therapist, and enjoyed the possibility of code-switching when needed. This typically happened when retelling traumatic episodes, allowing them to zoom in or out of the event depending on the need. This finding confirmed Kokaliari et al.'s (2013) observation that therapists who know their client's languages can allow them to use their L1 to increase, or their LX to decrease the emotional load of a traumatic event being discussed. Their qualitative study suggests that multilingual clients tend to switch to their L1 when talking about highly affective topics, dreams, or trauma.

Following a similar avenue of research, Rolland (2019) and Rolland, Dewaele, and Costa (2017) analysed data collected from 109 multilingual clients from around the world, most of whom reported code-switching linked to emotion regulation and to the need to verbalise language-specific concepts, autobiographical memories, complex social relationships, and perception of the self. Clients reported that the lower emotional resonance of the LX could be an obstacle, a much appreciated analytical ‘third’ space or a way to express painful emotions (Rolland, 2019). When the language of retelling a traumatic event was congruent with the language of the experience, it allowed participants to express themselves with more detail and clarity. Byford (2015) suggested that the use of an LX may allow clients to detach themselves from the trauma but that this strategy is not purely defensive as it also enables them to access those traumatic memories.

The option to code-switch in therapy is not equally appreciated by all bilinguals. Pérez-Rojas, Gelso, and Bhatia (2014) found that among the 63 Spanish-English bilingual university students who listened to a simulated psychotherapy session with a client and a bilingual Latina therapist who invited the client to switch from English to Spanish when the client had trouble expressing emotions, and a recording with the same therapist who did not invite the client to switch, only “participants who expressed a greater sense of belonging to US culture rated the

therapist who invited the client to switch as being more multiculturally competent” (p. 55). A recent study by Pérez-Rojas et al. (2019) focused on the role of Spanish and English in psychotherapy with bilingual Latinx clients in the US. Phenomenological analysis of eight interviews with such clients demonstrated that it was important for them to have the possibility to use both languages in therapy according to their linguistic and psychological needs, and to foster a sense of agency. Using a similar approach, Cook (2019) analysed data collected through interviews with 15 refugees from around the world who had suffered from rape or torture and who had settled in London in a therapeutic setting. Using an Interpretive Phenomenological Analysis, the author found that the use of the L1 or LX had its unique advantages and limitations. While the use of limited English LX could be draining and could lead to a feeling of frustration at being too blunt and not articulate enough, it could also empower clients and allow them to bypass the pain and the shame engrained in the L1 words and expressions surrounding the events. English LX made it possible to recount the traumatic experience which occurred in their L1 without being overwhelmed by despair. As one participant put it: “The English language enables me to visit my pain” (Cook & Dewaele, 2021). English LX allowed them to enhance their self-esteem and confidence. Dewaele et al. (2020) and Cook and Dewaele (2021) pointed out that the use of an LX may be just one strategy in the healing process, allowing the processing and integration of the trauma at an embodied and cognitive level.

Finally, Bager-Charleson et al.'s study (2017) with 88 therapy trainees and qualified therapists showed the effects of training sessions about multilingualism. A survey showed that they significantly raised awareness about the danger of making assumptions, and interviews revealed an increased awareness about the role and the importance of code-switching in therapy.

In summary, empirical findings suggest that multilingual clients need to be able to use their various languages in therapy according to their needs and that they are thankful of therapists who allow other languages in the room. However, the majority of research has been carried out among clients, rather than therapists, and in English-speaking contexts. It is important to find out whether similar patterns emerge when the language configuration is different in order to establish clearer roles of linguistic, cultural, and psychological factors in multilingual therapy.

Based on ten individual interviews with Dutch multilingual/multicultural therapists, the following research question was formulated: What role does the therapist's multilingualism play in the therapeutic process with multilingual/multicultural clients, in particular in emotion communication?

Methodology

In order to investigate the role of the therapist's language(s)in the therapeutic process and its outcome, as well as therapists' views of employing their multilingualism in their practice, a qualitative study was designed. Because we were interested in individual stories and subjective opinions, a qualitative design was the most appropriate. The empirical context of the study was the provision of psychological help and support to multilingual/multicultural clients in the Netherlands.

We opted for semi-structured interviews in order to collect multilingual therapists' stories about their own experience (Potter & Hepburn, 2005), and to follow the uniform format of a prepared questionnaire while still allowing enough room for the conversation to go beyond the

question-answer paradigm, which is crucial when obtaining detailed information about participants' experience (Cohen & Crabtree, 2006). The interview audio files were transcribed and analysed using thematic analysis. Compared to other methods of analysis, thematic analysis is common practice in qualitative studies (e.g., Bager-Charleson et al., 2017; Schweitzer et al., 2015) since it can be applied to a wide range of research questions and allows the researcher to work with the data in a flexible yet consistent way. Before conducting the interviews, ethical clearance was obtained from the first author's research institution (Project #2020/142; obtained on 9^th October, 2020).

Participants

In total, ten therapists (9 female, 1 male) participated in this study. Eight participants were recruited via three centres for psychological help in the Netherlands and two individually via the international database of psychologists/psychotherapists. The criteria for choosing the centres were that 1) they offered psychological help to multilingual/multicultural clients, 2) they had multilingual/multicultural therapists available, and 3) they allowed the possibility to hold interviews in English. The inclusion criteria for therapists were that 1) they worked as a recognised psychologist, psychotherapist or counsellor in the Netherlands, 2) they had experience working with multilingual/multicultural clients, 3) they spoke more than one language, and 4) they spoke English as one of their languages. Since the aim was to interview participants from a variety of linguistic and cultural backgrounds, the sample was heterogeneous.

The average age of the participants was 37.1 years (range 25-50). Seven therapists reported practising an integrative approach, i.e., adapting their therapy to the client's needs. Out of the remaining three, one therapist offers trauma counselling, another one hypnotherapy, and the third one mainly practices EMDR (Eye Movement Desensitization and Reprocessing). Eight different nationalities were reported: Dutch (n = 2), Moroccan-Dutch (n = 1), Greek (n = 1), Ukrainian (n = 1), French (n = 1), Venezuelan (n = 1), Belgian (n = 1), Italian (n = 1), and Iranian (n = 1). The most frequent L1 was Dutch (n = 4) and the most frequent LXs were English (n = 7), followed by French, Arabic, Berber, Greek, Italian, Russian, Ukrainian, and Spanish.

Procedure

The interviews followed the interview guide, in accordance with the recommendations of Kvale and Brinkmann (2015). The interview guide was designed based on the literature and research aims of this study. In addition to the demographic questions, the topics covered during the interviews included the therapists' linguistic and cultural backgrounds and development, their experiences when working with multilingual/multicultural clients, and the ways they utilize their own and clients' languages in therapy. Before the main questions, each participant was asked to self-assess their frequency of use and their fluency of every language they speak. The assessment form was adapted from Grosjean (2015).

The data obtained in the course of the interviews were transcribed and coded using the software tool Atlas.ti, and then analysed on the basis of reflexive thematic analysis suggested by Braun and Clarke (2020) and Kvale and Brinkmann (2015). The main aim in choosing this methodology was the need to find patterns in the data within the theoretical commitments staked out by the theoretical framework and the research question. Since thematic analysis helps establishing and organizing patterns in the data, the end result comprised a theory-based

complete narrative, allowing an interpretation in the context of multilingual/multicultural therapists' experience with multilingual/multicultural clients.

Coding

As defined by Braun and Clarke (2006, p. 79), thematic analysis (TA) is “a method for identifying, analysing, and reporting patterns (themes) within data”. In particular, the reflexive TA approach was applied in this study, according to the typology given by Braun and Clarke (2020). Since theme development depended on the researcher's analytic and reflexive abilities, and the themes were generated through close engagement with the data, this type of TA was the most suitable. Each interview was transcribed immediately after its completion, adding to the previous one(s), and coded and analysed in order to generate themes. This contributed to revisiting and re-evaluating of themes constantly, which is crucial in this type of data analysis (Braun & Clarke, 2020).

After conducting the interviews, to further develop and analyse the emerging themes, the following steps of reflexive TA were taken. First, familiarisation with the data was done by reading through the transcripts. Next, the transcripts were coded in order to find all emerging themes, yielding codes. These codes from all ten interviews formed the basis of the code book. Next, the overarching themes and subthemes were developed. Only the themes and subthemes pertinent to the research question were included. These themes were reviewed across the entire data set to ensure consistency, and the complete code book was compiled. Finally, a complete narrative was composed and a critical analysis was written up.

In order to ensure coding validity, two important steps were undertaken – themes saturation and group review of the overarching themes and subthemes. As some studies demonstrate, sufficient data saturation may increase research validity (e.g., Aldiabat & Le Navenec, 2018; Ando et al., 2014). Since this study employed thematic analysis, the goal was saturation of the themes rather than the entire data. This means that data was to be collected until no new themes could be developed. Continuous review of the information from new interviews showed that by the end of the data collection (by interview 8), no new codes or concepts appeared that were pertinent to the research question. In the final step, consensus was reached concerning the main themes and subthemes. The finalized themes and subthemes were consequently interpreted and analysed, and are presented below.

Results

During the interviews, participants referred to several topics as the most important aspects of their work with multilingual/multicultural clients in the context of expressing emotion (e.g., discussing trauma or behavioural issues). In accordance with the research aims, three main themes and six subthemes were distinguished, as illustrated in Table 1. The table also shows how many participants contributed to each theme and subtheme.

The next section includes quotes from the interviews that illustrate the themes and subthemes. To ensure privacy, all quotes were anonymized. In the cases when more than five participants expressed a similar view, the word “most” is used, whereas for an opinion that was voiced by one to five participants, the word “a few” or “some” is used. The L1 is referred to as L1 (acquired before age 3) and a foreign language as LX throughout.

Availability of languages. Each participant talked about the role(s) their own and their clients' languages play in the course of therapy. All therapists inform their clients in advance which languages are available in therapy, either via their website or during the intake, and most of the therapists consider it important to do so. One of the therapists said, “I do so…because that's kind of symbolic of a certain flexibility across cultures that I have and that I offer gladly in therapy” (Participant 2). Therefore, clients can already become aware of the possibility to use their L1 in sessions, should they wish to do so.

Language switching can serve various functions. One recurring subtheme integral to multilingualism in therapy, which all participants experienced in their practice was language switching and its various functions. For instance, participants described shifts from LX to L1 when clients tried to recall certain memories, to use a more socially acceptable language for the purpose (e.g., for swearing or taboo words), or to use words with a different connotation (e.g., less ambiguous). Participant 3 said that it might seem worse to swear in Russian than in English, possibly because in Russian, swearing is much more personal and therefore considered less socially acceptable. Another participant spoke about how her own language choice depended on the conceptual meaning she wanted to convey to the client:

“…sometimes I too switch when I can't find the concept, which happens mostly when I provide therapy in French, because some concepts, I think, are more…some words have different connotations in French and how some word can have, like, different dynamic meaning to it. And if I don't want my client to understand it this way, I'm going to use the English words” (Participant 6).

In addition, language switching is sometimes used to exclude a third person from the conversation. This function is of particular interest for the dynamics in systemic or family therapy since here language is used as a tool to wield power. While this allows one family member to express something that is important to the client, it also creates an ethical dilemma for the therapist. This dynamic was illustrated by the example from one participant's session with a mother and a daughter, when the daughter switched from L1 to LX in order for her mother not to understand her, thus trying to gain more control of the situation:

“She [daughter] thought, ‘Okay, I'll say it in Dutch,’ and the mother was ‘What? What's going on?’ So that was for her [daughter] the thing – to change the language, like we have a secret language in Dutch” (Participant 9).

Another important function that was reported for language switching was to counter the lack of certain words or concepts in one of the languages. For instance, one therapist working with LGBTQI clients remarked that particular words denoting gender categories are sometimes missing in her clients' L1 (Participant 3), while another one talked about her clients' explaining to her concepts that did not have a good translation in the LX (Participant 4). Yet another therapist made the following remark:

“In Dutch, there are no words to express anxiousness. You have fear and you have restlessness, but there's no anxiousness, no specific word for anxiousness. So sometimes you just, you know, you also use your language palette to express even more” (Participant 2).

These instances of language switching are mentioned as part of therapists' experiential knowledge, their unique encounters with multilingual/multicultural clients. Their own multilingualism seems to have provided them with a higher sensitivity towards their clients' language and culture. However, this does not appear to be something they had explicitly known about before they started working as therapists. In the words of one participant, “this whole letting your client say something in their L1 is something that I had to learn throughout my years of work, that I figured out myself” (Participant 1).

Low proficiency can be an obstacle or a facilitator of the therapeutic process. Sometimes code-switching can occur due to limited language proficiency, i.e., when the client's or therapist's LX is not sufficient to be able to express thoughts and feelings. All participants stated their willingness to go above and beyond in case language proficiency was an issue, setting their client's benefit as a priority. Only two participants said that it might not be possible for them to agree to conduct therapy with clients whose language skills were too low: “… talking is key in therapy, so if you're not able to communicate, then you're just not able to work” (Participant 4). Many participants noted the necessity to be patient, to repeat, to take more time with the client, and to ask more questions. Interestingly, having to ask more questions was also referred to as an advantage since the therapist's lack of knowledge of the client's L1 seems to “legitimize” posing more questions and on topics that might have been omitted in a monolingual setting. One of the participants explained: “I found that it's an opportunity for me to get to know what they particularly mean by the phrases they're using, so yeah, I've turned it into an advantage, in a way” (Participant 4). Another participant added that not speaking the same language provides an opportunity to ask more questions than one could in a monolingual setting:

“… maybe it's also a plus if you don't understand the language very well, because I know Arabic, Moroccan Arabic, but not that well, so I have to ask a lot of questions also: ‘What do you mean about it? I don't understand. What do you mean?’” (Participant 10).

Another result of lower proficiency that was seen as an advantage was that clients use simpler words and expressions than native English speakers would. Not having many words at one's disposal and thus expressing oneself in LX in a simplified way can be beneficial in therapy since the person has to get to the point more quickly by means of those words they already know. According to one of the participants, this puts the therapist and the client in the same boat: “I'm not speaking my L1, they're not speaking their L1, so they can connect with that” (Participant 8). Overall, most of the participants consider the client's LX vocabulary or grammar a minor obstacle, which they overcome with the help of dictionaries, drawings, and

toys, and by resorting to professional translators. According to one participant, “language is a tool to connect, and how much more words you have, you have more tools to connect” (Participant 7). While lack of linguistic resources (e.g., grammar or vocabulary) can hardly be viewed as an advantage on its own, in the context of multilingual therapy, the necessity to think out of the box can serve to the benefit of both sides. Therefore, most of the participants think that because it triggers extra creativity and openness, therapist's and client's low language proficiency does not have to be a hindrance and actually opens up new opportunities.

Language and emotion. Nine out of the ten participants remarked on the connection between language and emotion in their practice. Their view of this connection is based on their experience working as a psychotherapist. Therapists are normally highly attuned to their client's emotion expression and the way language may interfere in it. Their professional knowledge, that is, again seems to be mixed with intuitive understanding and learning “from experience, mainly from experience, because I don't know whether there is some kind of education like that” (Participant 8). This is an important finding since the evidence-based information from linguistic studies does not seem to reach psychotherapists, leading to their reliance on primarily intuitive and implicit knowledge.

Emotion encoding and decoding. Most therapists noticed that it is sometimes more difficult for their clients to connect with their feelings when speaking in LX, which can be related to the language in which the initial emotion was encoded. For instance, in some past experiences (e.g., childhood events), when only L1 was available or other languages were not sufficiently known yet, an emotion was encoded in L1. Later on, when the emotion is being decoded, and other languages are available, it often does not find an immediate way out in LX. As one participant described it:

“… when you're in therapy, you get in contact with emotion of your inner child. The inner child doesn't speak the foreign language that you learnt as an adult. So that's when sometimes people really get stuck – in that child. I had such experience…I had to connect to my inner child and I realized my inner child doesn't understand all this Dutch speaking…I could communicate but the child… speaks Italian” (Participant 2).

Viewing emotion as a concept, not just as a word, was also expressed by many participants. They find it important to look behind the words and make meaning together with the client. Asking questions about the emotion being discussed, and trying to understand what exactly the client means seem to be commonplace among the therapists in the sample. This may also be related to the issues the participants mostly deal with, such as trauma, depression, working with children, etc. One participant gave the following example of conceptual probing from her practice:

“…I had to pause and say, ‘Okay, so you're saying that causes you grief. How do you experience that? How is that grief?’ And then she [client] would say something completely different. […] Maybe in her native language saying something like that is like grief could refer to what she was explaining, but in English it's different” (Participant 4).

This quote shows how important it is for the therapist to be sensitive to such differences in conceptual understanding of emotion. In a hospital, if a patient uses a certain word to describe their symptoms, and the doctor misunderstands it and underestimates the severity of the patient's state, it may lead to serious mistakes in treatment. If a psychotherapist only focuses on their own understanding of the word “grief”, they may also miss some crucial information about the client's emotional state.

In the context of encoding and decoding emotion, some therapists spoke about their client's language switch when trying to resolve a highly traumatic experience. Even when the clients have adequate LX proficiency, speaking about the trauma in LX may not bring any release, as some participants showed in their examples. Only after the therapist had invited the client to say the same things in the client's L1 did the release occur:

“… then she tried to say it first in Dutch [LX] and then she switched to Arabic [L1], and I gave her some words, and then when I gave her that word, it was just like [an explosion] breaking out, because it was exactly what it was. […] When I gave it to her, it was just the key” (Participant 7).

Language also became a focus in another participant's practice when treating trauma. Her personal and professional intuition told her how to help the client achieve the emotional release she had longed for – by inviting her to speak in her L1:

“… it was a kind of a cork stopping the expression because she [client] believed if she didn't know the translation, she could not express herself. The moment that she could express herself – and the rest flew” (Participant 2).

This again was not based on explicit knowledge of how language functions in case of intense emotions, but on the participant's experience and her primary focus on helping the client connect with their true feelings. She said that she offered therapy in multiple languages “as a personal development project also” because she had learned nothing about working across cultures and languages in her formal education in the Netherlands.

Emotional distance in L1 and LX. Talking about trauma in L1 may be too emotionally charged, which one of the participants experienced when she had been in therapy herself (a required practice for some types of psychotherapy). She talked about having therapy in her L1first but not feeling entirely comfortable. Later on, she had therapy in LX and it actually worked much better. Afterwards, she received therapy in her L1 again, feeling more “able to talk about herself in both languages” (Participant 6). The change from L1 to LX in therapy did not seem to be a conscious choice, but it did work. Looking back, the therapist explained this effect by the possibility to put some emotional distance between her experiences and her words. Most participants feel emotionally closer to their L1. It has been noted by some participants that the emotional distance of an LX may sometimes guide the client's language choice in therapy. For instance, when their L1 conjures up a strong sense of stigma or rejection, it may be easier for clients to receive therapy in LX. One participant described such cases:

“…when it comes to emotional expression, there're a sort of two kinds of clients, I would say – some for whom emotional expression in their own language is a lot easier because they grew up speaking it and it's pretty straightforward. And then others, for whom it is actually easier to express their emotions in a different language because their own language stigmatizes those emotions or because a certain trauma happened to them in that language” (Participant 3).

Becoming more aware of language and emotional distance may enable the therapist to enhance the process of treatment and healing by, for instance, inviting the use of a client's L1 for emotion expression or the other way around: encouraging the client to speak about the emotionally rich event in LX. Some participants say that the client does not even have to use a word that they know because the emotion expressed may be so strong that it is clear what it is (e.g., it can be derived from non-verbal signals), which is usually confirmed by further probing from the therapist. From one therapist's practice:

“… once they [clients] want to share a feeling, a very deep feeling, they really need to use the word in their own language. And sometimes that word only can show me how heavy, how disgusting or something like that the situation is” (Participant 7).

When the therapist is attuned to clients and their needs, his/her linguistic awareness may contribute to a deeper understanding of emotion communication during sessions than when the process is guided by the therapist's fear of not being able to follow a client's every word. The advantage of giving the client a path to express their emotion fully by inviting them to use their L1can be illustrated by the following quote:

“If I hear also word in a language that I absolutely do not understand, I think I would still feel the energy behind that word, and again, it's more important that the client feels that – oomph, I am in that emotion, that's what's playing here. It's much more relevant to healing than me understanding everything” (Participant 2).

As these quotes show, learning more about emotional distance in L1 and LX may help create a better therapeutic alliance and a safe emotional environment for the client.

Language, culture, and therapeutic relationship. All participants see language as an inseparable part and reflection of culture. It is not always straightforward in their multilingual/multicultural practice whether a difficulty expressing or understanding a concept is due to language or to culture: “the difficulty is different cultural references and the style of communication that comes through the language” (Participant 3). These two notions are especially closely intertwined when it comes to making sense of what is being expressed by the client.

Cultural differences in making meaning. When a certain emotion or state has a different conceptual meaning in the client's L1, the therapist may come to a wrong interpretation of the value the client gives to the word. For instance, the word “discipline” may evoke something different in a person from a military family, or, as Participant 7 described, in a person from a very traditional reformed religious background. Such (mis)understanding of the values clients give to certain words may significantly affect the therapeutic relationship between therapist and client. Describing her own experience as a therapist, another participant found it fascinating “to see how the same complaints are experienced sometimes very differently because of the cultural influences” (Participant 4). Which is also true about language: acknowledging that “every culture is different from my culture”, another participant spoke about learning gradually what the same words or expressions mean in different cultures (Participant 5). He illustrated it by talking about a Syrian client for whom saying that he “needs” something implicitly puts a time stamp on it, as the need presupposes getting it right now, whereas a Dutch therapist might assume that the client realises that time and effort are required to satisfy that need. Naturally, this does not automatically generalise to all representatives of a certain culture, but it creates the understanding that the same words can mean different things. The following quote from another therapist illustrates figuratively how such understanding works:

“If I ask a Turkish guy, ‘Go and take a bread for me,’ he will bring something fluffy and round. When I ask a Dutch man, ‘Go and take a bread for me,’ it's something long and brown” (Participant 7).

A telling example of understanding the value of words and putting them in the cultural and linguistic context of the client was given by one participant who had experienced difficulty connecting with the client due to such lack of understanding: “It was a girl who I had in

treatment and she was telling me that she was there because she was stealing a lot” (Participant 10). If one stops right there and looks at the word “stealing”, one may come to various conclusions, such as kleptomania (compulsive stealing) or repetitive criminal behaviour. However, examining the rest of the story, the word “stealing” assumes a different cultural dimension:

“And she told me that that was something what was high-seen in the culture, because if you are a good thief, then you find a good partner, then you marry very quickly because everyone wants you in their family. And that was for me then very difficult to understand: so you have to steal to get accepted” (Participant 10).

Tracing one's own assumptions and biases. All accounts mentioned the ways in which linguistic and cultural differences shape therapeutic processes. The participants underscored the special importance of being aware of one's own stereotypes and assumptions in the multilingual/multicultural setting. Unconscious bias and assumptions may become more salient when the therapeutic setting changes from monolingual/monocultural into multilingual/multilingual. Therapists are recommended to stay as impartial as possible and to not “take sides”. When not everything in therapy can be translated, various assumptions may be activated. As an example, one participant spoke about the frustration among Dutch people towards other cultures when it comes to presuming that people who came to the Netherlands would automatically conform to all cultural norms. In reality, this process is rather complex, effortful, and time-consuming. Acquiring the local language and following social customs does not automatically entail leaving one's original language and culture behind. This participant emphasized that if a client speaks Dutch, it does not necessarily mean they understand what she means, because “they're listening to you through their own culture” (Participant 7). That is why she strives to understand the culture of the words and the unique meaning her client is conveying through them. This is when understanding goes beyond words, and this is echoed in most interviews.

The value of tracing one's own assumptions and biases and of using linguistic and cultural differences to their advantage is a common thread running through most accounts. The existence of stereotypes and assumptions does not have to mean adhering to them:

“It is very important to question yourself when it comes to thinking about someone's culture or about someone's language, thinking that even though they're part of that culture that they will necessarily subscribe to that reference, for example, or that they would condone the use of a certain word” (Participant 3).

Two participants spoke about the “Dutch way” of their work, which they consider to apply a one-size-fits-all too quickly when it comes to clients who speak different languages and come from different cultures. The Dutch way is a country-specific way of following a clear protocol with a limited amount of time allocated to each step. These two participants are not Dutch by origin, which may be the reason why this way of working was salient for them. One participant said that she “noticed that Dutch people literally work as they teach – you really follow certain steps” (Participant 2). The other gave the following description of the Dutch way and compared it to her own:

“It's about the protocol, it's about the program that you have and you make, and you're going from step one to three, and the main purpose is to bring, to send the person to society as soon as possible. […] My way of working, it's not very Dutch, and I'm not working really through protocols, I really work with people, with the person” (Participant 7).

According to the participants, when treating multilingual/multicultural people, it is not always feasible to stay within these timeframes, so they have to be more flexible and open-minded, focusing on the client's needs. Customizing the pre-designed protocol to every client brings about changes in every therapeutic relationship. As one participant says, “…you walk around, you take the time, but as long as you don't let the inner critic [an internal voice that judges and criticizes us] take over the session as a client or as a therapist, that's not a problem” (Participant 2). Additionally, becoming more aware of language and culture in therapy may also contribute to the therapist's openness and understanding of linguistic and cultural distinctions within the same language and culture as theirs. Indeed, every language has dialects and regional peculiarities, and every culture contains various subcultures, and this realisation comes with a better understanding of the role of language and culture in the therapeutic interaction. To quote one participant:

“…with a Dutch person, still there are cultures inside, subcultures, you know, inside the big culture, that I don't know. So when I have to do with that subculture, I really need sometimes to understand, okay, what are you talking about?” (Participant 7).

As most participants said, such cultural adaptations that they make to their approach to multilingual/multicultural clients seem to come not from formal education but from experience and privately sought-out trainings. As the quotes above illustrate, keeping the client's needs a priority the therapists become more alert and attentive, and adapt their protocols to serve the client's benefit.

Being a multilingual/multicultural therapist means having more frequent encounters with the languages and cultures that are often not present in the therapist's social environment. Therefore, being more attentive and alert to what is being said and how it is being said from both linguistic and cultural perspectives is considered to be of utmost importance for a smooth and beneficial therapeutic relationship.

Discussion

The current study investigated the role the therapist's multilingualism plays in emotion communication with multilingual/multicultural clients. The interview data showed that the therapists who practise in the Netherlands see a clear connection between language, emotion, and culture, based on their own work experience. The results revealed therapists' views on the various functions of language switching in therapy, the advantages and disadvantages of not speaking the same language with the client, the connection between language and emotion, and the way multilingualism may shape the therapeutic process and its outcome. The participants spoke about their clients' strategic language switching for particular acts (e.g., swearing), more fitting for talking about past events (e.g., childhood memories), better able to reflect certain concepts in the L1 (e.g., gender pronouns), or useful in temporarily excluding a third person from the conversation. The therapists also talked about their own language switching, for instance, when looking for a word or aiming for a particular connotation. It highlighted its crucial role in shaping the therapeutic process in terms of conceptual understanding and meaning-making (cf. Cook, 2019; Costa & Dewaele, 2012, 2019; Bager-Charleson et al., 2017; Rolland, 2019; Rolland et al. 2017).

One of the key findings was the commonly shared experience that an insufficient level of common language in therapy can be both an advantage and a hindrance to effective therapy. Though having low proficiency is a constraint, in the context of seeking and offering psychological help, it is not always viewed as a major obstacle. Having to ask more questions and to use other tools (e.g., toys, cards, etc.) sometimes offered an opportunity for the

participants to connect more with their client. Karamat Ali (2004) made a similar point in an overview of multilingualism in systemic and family therapy, stating that the “potential gain of being a bilingual therapist” is to sometimes have to ask the obvious (p. 347). Karamat Ali speaks about it in the context of family therapy, and the current study confirms this observation for therapists who practice other approaches (e.g., systemic therapy, schema therapy). That this notion is present in the implicit, experiential knowledge of the participants is encouraging, as it brings up the importance of questioning possibly misguided assumptions, and simply asking clients for more explanation.

The results of the interviews also demonstrated the essential role that therapists' understanding of their own and their clients' languages plays in discussing emotionally sensitive topics (e.g., family relations, memories, etc.). The participants' observations of closer emotional involvement in L1 than in LX confirms earlier work on multilingualism and emotion (Cook & Dewaele, 2021; Dewaele, 2013; Dewaele & Costa, 2013; Kokaliari et al., 2013; Moate & Ruohotie-Lyhty, 2017; Panicacci & Dewaele, 2017, 2018). Our study demonstrated that the sensitivity of emotional encoding and decoding to the language in which it is expressed seems to be particularly relevant when discussing traumatic events (Cook, 2019). The findings add to the previous research by showing that the therapist may not have to understand every word the client uses as long as the language use promotes the client's therapeutic benefit. Additionally, according to the participants, the link between language and emotion (especially regarding trauma) is not something they learnt about during their formal training, but rather from private courses they personally sought out, and/or through experience. Another key finding was related to the linguistic and cultural dynamics of the therapeutic relationships with multilingual/multicultural clients (e.g., cultural differences in conceptual understanding of “grief”). Therapists in the sample emphasized the importance of being aware of the value and meaning clients give to certain words and concepts depending on their language and culture. This is especially the case when it comes to emotional concepts which cannot be easily translated from L1 into LX, thus potentially causing misunderstanding and wrong interpretations by the therapist (cf. Costa, 2020; ; Softas-Nall et al., 2015).

Regarding the therapeutic process, therapists emphasized the importance of tracing and resolving their own biases in communication with their multilingual/multicultural clients. The description of the “Dutch way”, of strictly adhering to set protocols and timeframes, echoes findings in some studies on healthcare communication (e.g., Geerlings et al., 2018; ). Given the therapists' special attention to this topic, their arguments for possible adaptation of the local way of psychological guidance and treatment should be taken seriously. Psychotherapy of multilingual/multicultural clients requires a certain degree of adjustment due to cultural, but also linguistic, differences (cf. Costa, 2020). Therapists need to be aware of their own biases. Unlike in monolingual settings, therapists in multilingual practice face more challenges, and biases may only become salient when two or more languages and cultures are involved (Costa & Dewaele, 2012).

Finally, all participants reported that they were more than willing to go the extra mile in trying to understand and accept their clients, and to use their differences as a way to connect. However, some answers suggested that language was not always on the participants' radar. For instance, Participant 2 mentioned having the most difficulty with the client who spoke her L1, ascribing this to the differences in culture and way of thinking. Based on literature and the current findings, it seems to be the case that multilingual/multicultural psychotherapy in the Netherlands tends to place the main focus on culture, leaving language as an implicit and taken-for-granted part of culture.

Limitations and Future Research

The current study is not without some limitations. First, the therapists who agreed to participate were mainly raised multiculturally with early exposure to foreign languages. Thus, they may have already been inclined to pay special attention to language in their practice and have positive attitudes toward code-switching (Dewaele & Li Wei, 2014). The choice of English for the interviews, a LX for all participants, may have influenced the way certain topics were discussed. It would be interesting to compare the answers to the same questions in participants' L1. It is also important to mention the size and nature of the sample. Even though ten participants overall for a study of this scale is considered to be sufficient (Francis et al., 2010), it means that the language profiles are not necessarily representative of all therapists in the Netherlands.

Finally, we should mention the researchers' interest in the topic under investigation and its potential impact on data analysis. Being a linguist may make one more aware of language patterns and its use. All of us are multilingual and our research interests are focused on human communication in different contexts, which may have impacted our perspective on the data. However, our team was diverse enough to ensure objectivity and a valid analysis of the interview data. We believe that group discussions and reviews of the themes provided us with a maximally objective analysis presented in this article.

Future research could dig deeper into intentional as well as accidental language switching in therapy and how it is related to emotion expression and interpretation. Knowing how, when, and why to invite a client to freely switch between languages may enhance the therapeutic relationship. Finally, we agree with Bager-Charleson et al. (2017) and Costa and Dewaele (2019) that it is crucial to investigate ways of training multilingual/multicultural psychotherapists to raise their awareness about multilingualism so they can transfer it to their practice and become better practitioners.

Conclusion

This study has highlighted the importance of multilingualism in psychotherapy, in particular when it comes to emotion expression and interpretation. The interviews demonstrated that language switching occurs when clients try to connect to deeper feelings and when the topic is emotionally intense. Instances when the therapists encouraged such switches in their sessions had a profound and beneficial effect on the therapeutic process and outcome. These findings are in line with previous research in different language environments. It suggests that the status of the language does not alter the mechanisms, challenges and strategies of multilingual psychotherapists and their clients. Whether a Syrian client uses Dutch or English with their psychotherapist, code-switching will occur for the same reasons in The Netherlands as in the UK, and the therapist will be confronted with the same issues in both contexts.

A number of recommendations can be drawn from the present study. Firstly, it is essential that training courses for psychotherapists pay attention to the clients' and the therapists' language profiles and focus on ways to bridge the linguistic and cultural gap. Realizing the fact that speaking the same language does not guarantee a better understanding in therapy is also key to being a more efficient therapist. Trainers could highlight the variety of creative ways to bridge linguistic and cultural gaps which may otherwise endanger the therapeutic relationship.

To conclude, language plays an essential role in psychotherapy, influencing rapport, attunement, needs assessment, and treatment. Working with clients of the same linguistic and

cultural origin can be a challenge but when psychotherapists are dealing with multilingual/multicultural clients, they need to be even more aware of their own and their client's languages in order to serve their clients to the best of their abilities.