Translanguaging in L3 Spanish classrooms: practices and attitudes

. Research into pedagogical translanguaging in second/foreign language education has largely been focused on the interplay between two languages, one of which is usually English. Little attention has been paid to the practice of trilingual translanguaging (translanguaging between three languages), and how this can be helpful in the acquisition of a third language. This study, conducted in two Spanish as a third language classrooms in a Chinese university, aims to explore teachers’ translanguaging practices as well as the teachers’ and students’ attitudes to these practices. Analysis of data collected through audio-assisted classroom observation, interviews, and questionnaires reveals that teachers proactively and flexibly mobilize their multilingual resources in classroom talk. Students in general express positive attitudes towards teachers’ translanguaging practices, and a wish to experience a greater amount of bilingual translanguaging between English (L2) and Spanish (L3). Meanwhile, having access to the views of L3 learners on teachers’ classroom talk proves to be a crucial component in the understanding of how L3 teaching and learning can take place in the most effective way. This study calls for further research into translanguaging practice in multilingual classrooms and its impact on students’ learning.


Introduction
Teachers' choices of languages have always been a salient and contentious feature in second/foreign language classroom research. Progressive advances in our understanding of bilingual and multilingual development have given rise to studies of translanguaging, which, as a practical theory of language, serves as a constructive means to capture interlocutors' flexible and sometimes creative maneuvers between two or more languages (Li, 2018).
Unlike traditionally used terms such as code-switching or code-meshing, translanguaging refers "not simply to a shift or a shuttle between two languages, but to the speakers' construction and use of original and complex interrelated discursive practices that cannot be easily assigned to one or another traditional definition of a language, but that make up the speakers' complete language repertoire" (García & Li, 2014: 22). Studies into translanguaging have been approached from both naturalistic and pedagogical perspectives (Creese & Blackledge, 2019;Li, 2014Li, , 2016aPennycook, 2017). In second/additional language learning settings, translanguaging has frequently been employed as a pedagogy to probe into the dynamics of classroom interactions (Canagarajah, 2011;Charamba, 2019;Jones, 2017). It is mostly observed in contexts where the interplay between two languages, usually English and the learners' L1, is accentuated, or in cases where students are bilinguals in English and another language (Lin & He, 2017;Makalela, 2015;Moore, 2014). It is rare to witness scenarios in which multiple languages (three or more) are simultaneously utilized and woven effortlessly into classroom discourse for the purpose of sense-making. It is also yet to be established how students respond to such multilingual practices conducted by their teachers in L3 classrooms where both teachers and students possess shared L1 and L2.
Meanwhile, an increasingly mobile world population highlights the need for people to learn more than one language (Gao & Zheng, 2019;Jaensch, 2013). Efforts have been made to improve our understanding of multilingual speakers' abilities to use different languages in an integrative and compatible way (Canagarajah & Gao, 2019;Leonet, Cenoz, & Gorter, 2017). In the area of L3 acquisition, studies tend to mainly focus on two areas of research: cross-linguistic influences at the phonological, lexical, and syntactic levels; and the impact of bilingualism or prior linguistic knowledge on the acquisition of a third language, particularly among immigrant populations (Cenoz & Gorter, 2019). Nevertheless, in the process of proving that learners' prior linguistic knowledge and learning experiences are advantageous to the learning of a third language (Ma, Yao, & Zhang, 2018;Park & Starr, 2016), it seems that a monolingual mindset is often adopted, where learners' L3 competence is benchmarked against native-speaker standards, with little attention given to the multilingual practices of these learners and their attitudes to multilingual discourses in the classroom (Arıbaş & Cele, 2019;Jaensch, 2011).
Therefore, the researchers in the current study were prompted to explore two L3 Spanish classrooms in a Chinese higher education institution from a multilingual perspective. Teaching and learning foreign languages other than English has always remained an underresearched area in China, with insufficient attention paid to teachers in these classrooms (Guo, Sit, & Bao, 2020). This study is contextualized within the "Focus on Multilingualism" framework, which emphasizes the significance of "the multilingual speaker", "the whole linguistic repertoire", and "the context" to the study of L3 acquisition (Cenoz, 2013;Cenoz & Gorter, 2011). Teachers' pedagogical translanguaging practices, together with teachers' and students' attitudes, are examined to understand how bilingual and trilingual translanguaging between L1 Putonghua, L2 English, and L3 Spanish is conducted, and the ways in which these translanguaging practices enhance students' L3 learning progress.

Translanguaging
As a form of practice and a language theory, translanguaging has restructured our understanding of bilingualism and multilingualism. Instead of seeing bilinguals/multilinguals as those who have separate competences in different languages, it sees them as fundamentally equipped with multi-competence that can be fully used in the organization of meaning and the process of delivery. It emphasizes a person's whole linguistic repertoire, including (particularly but not exclusively) their total understanding of languages, as well as the semiotic, modal, and sensory resources they possess (Canagarajah, 2011;Li, 2018). Faced by the challenges of Post-Multilingualism, translanguaging provides an understanding that incorporates practice and process-it is "a practice that involves dynamic and functionally integrated use of different languages and language varieties, but more importantly a process of knowledge construction that goes beyond language(s)" (Li, 2018: 15).
A review of existing research suggests that translanguaging studies are to a large extent conducted in second language/bilingual classrooms where the interactional dynamic is restricted to two languages, almost always English and another language. Seldom does one encounter the hybrid and fluid use of three or more languages, including switching to languages other than English, in L3 learning settings. Although a small number of recent studies have attended to the issue of pedagogical translanguaging in multilingual classrooms, they are often associated with the agenda of raising teachers' awareness of translanguaging as a pedagogy, or only interested in analyzing teachers' and students' perspectives on translanguaging (Cenoz & Santos, 2020;Galante, 2020;Gorter & Arocena, 2020). There is still a paucity of research showing direct evidence of how bilingual and/or trilingual translanguaging practices are actually conducted, or analyzing the classroom interlocutors' attitudes to these practices in the hope of offering implications for pedagogical improvement.
For this reason, the current research focuses on the use of Spanish as a third language in classrooms in China. The aim is to observe teachers' translanguaging practices, and also obtain teachers' and students' viewpoints from the perspective of translanguaging as a pedagogical resource. It is expected that evidence of an invigorating fusion of multiple languages will emerge, along with the chance to make constructive suggestions for how to employ one's full linguistic repertoire when teaching or learning a third language.

Third Language Acquisition
Learning a second foreign language, also termed third language acquisition, is "the acquisition of a language that is different from the first and the second and is acquired after them" (Cenoz, 2013: 71). It once fell under the umbrella term of second language acquisition, but has now become an independent area of linguistic research, providing an invaluable source of information for languages and language acquisition researchers.
Research into L3 acquisition attaches great importance to two main areas. First, studies of cross-linguistic influence are apt to draw conclusions about the relationship between language distance and positive/negative transfer (Gui, 2014;Mulík, Carrasco-Ortiz, & Amengual, 2018;Zheng, 2018). Second, research into the influence of prior linguistic knowledge or bilingualism on the acquisition of a third language tends to focus on the idea that bilinguals possess certain advantages in learning a third language when compared to monolinguals (Bardel & Falk, 2007;Ma et al., 2018;Park & Starr, 2016).
In terms of this second field of interest, there are two critical issues. First, learners' L3 competence is emphasized in comparison to that of bilingual and/or monolingual groups, instead of in terms of the learners themselves (Cenoz, 2013). In particular, there is insufficient research into their beliefs, attitudes, and understanding of classroom pedagogy, as well as how improvements can be made to enhance learning. The other issue concerns the lack of research into multilingual teachers in L3 classrooms, and how they maneuver between different languages and naturally build this into a form of pedagogy to facilitate communication and make meanings in classroom discourse. The teachers' deployment of their whole linguistic repertoire, as a valuable asset at specific pedagogical moments, is largely overlooked. To this end, a holistic approach to understanding multilingual education, a "Focus on Multilingualism", has been proposed (Cenoz, 2013;Cenoz & Gorter, 2011, advocating a multilingual perspective embracing the multilingual speaker, the whole linguistic repertoire, and the context when conducting research in multilingual classrooms.
In summary, to address the issues raised in relation to existing research on pedagogical translanguaging and L3 acquisition, an empirical study which incorporates translanguaging between multiple languages and attitudes of classroom interlocutors, while adopting a multilingual perspective, would seem appropriate. With this motive in mind, the current study sets out to investigate two teachers' translanguaging practices with Spanish as the third language, as well as teachers' and students' attitudes to such practices, with the purpose of observing how multiple languages are employed simultaneously and integratively to achieve pedagogical outcomes. By gaining access to teachers' and students' attitudes, it provides opportunities to assess how translanguaging can be shaped or improved as a pedagogy. In the meantime, by situating the analysis in specific micro-contexts in multilingual classrooms, it offers a refreshing viewpoint on how multilingual speakers make full use of their whole linguistic repertoire to negotiate meaning and facilitate communication.

Research Methodology
The study explored the translanguaging practices of two teachers of Spanish as a third language, as well as their own attitudes and those of their students in relation to the use of translanguaging as a pedagogy. It aims to describe and analyze the dynamic and fluid classroom interactions in multilingual classrooms where the teachers and students have a shared L1 (Putonghua) and L2 (English), and to inquire, from the translanguaging perspective, about the extent to which one's whole linguistic repertoire can be conducive to the teaching and learning of a third language. The study hopes to answer the following three questions: (1) What translanguaging practices can be observed in Spanish as a third language classrooms? (2) What are the teachers' attitudes to their use of such translanguaging practices in these classrooms?
(3) What are the students' attitudes to teachers' translanguaging practices?
As a mixed-method study, both qualitative and quantitative data collection methods were utilized with the objective of gaining views from teachers as well as students both individually and collectively (Tashakkori & Teddlie, 1998). Audio-assisted classroom observations were employed to record classroom interactions, particularly instances where the teachers' translanguaging practices took place. In addition, questionnaires and interviews recorded the participants' views of such practices. It is considered that lesson recordings provide the most direct way of capturing the teachers' shuttling between multiple languages and perceiving the nature of the classroom community. On the other hand, questionnaires are the standard approach to comprehending respondents' thoughts and retrieving specific information from a group of learners (Dörnyei, 2007;Richards, Ross, & Seedhouse, 2012). In addition, interviews enable the researcher to further examine the reasons behind practices or the decisions participants have taken. Specifically in this instance, it was possible to investigate the beliefs that teachers hold with reference to the flexible deployment of their multilingual competence, and the students' attitudes to such deployment (Richards, 2003).

Context and Participants
The current study was conducted in two university classrooms where Spanish was taught as a third language. Institution X enjoys a nationwide reputation for delivering high-quality undergraduate courses in foreign languages and literature. Its mission is to develop multilingual talents with language expertise and subject knowledge who can excel in multilingual and multicultural professional contexts. Students are required to learn another foreign language apart from English (which they are expected to have learned for approximately 12 years by the time they enter university) as a compulsory course during their second and third years of a four-year program.
The participants in the study were two teachers and fifty students. Both teachers were relatively young practitioners with teaching experience of fewer than 3 years and 5 years respectively. Born after 1990 and thus belonging to the "post-1990s" generation (Yang, 2018), they are considered fluent multilingual speakers of Spanish, English, and Putonghua. Compared to previous generations of Spanish teachers whose English proficiency tended to deteriorate considerably after becoming Spanish majors, these two young teachers represent a new generation who are likely to retain their English language competence by constantly being in touch with resources for study, work, and entertainment in English, mostly on the internet. Consequently, it is expected that these teachers possess the ability to make full use of their linguistic repertoire as they make language choices in classroom talk.
Fifty students in total from two parallel classes formed the student body of the participants. They were all in their second year of undergraduate studies, majoring in literature, journalism, economics, and business management. They had been receiving formal English instruction since they were in primary school, and on average their English expertise was CEFR B2 level. However, none of them had any knowledge of Spanish before joining this course. Enthusiasm for Spanish language and culture was the main motivation for them to choose Spanish among other choices of second foreign language such as French, German, and Japanese.
Students received two 90-minute intensive skill-based Spanish lectures per week. The total in-class hours of formal Spanish teaching amounted to 108 hours annually. The main aim of the course is the attainment of Spanish language skills at the beginner level. Students are expected to acquire basic Spanish grammar and vocabulary, and to be able to hold elementary conversations in Spanish within two years. Teacher-led lecturing and teacher-initiated interaction account for a significant portion of the teaching.

Data Collection and Analysis
Consent from all participants was obtained before the data collection procedures were initiated. Audio-assisted classroom observation lasted for 10 weeks, with one 90-minute session recorded per week. Notes were taken during the observations. Instances of classroom discourse containing the flexible mobilization of more than one language within one sentence were transcribed and then subjected to micro-analysis in order to closely examine the process of translanguaging selection and use, during particularly interactional moments to help achieve certain pedagogical aims.
Questionnaires concerning students' attitudes to teachers' translanguaging practices were distributed and collected in Week 5. Students (n=50) were invited to answer questions in relation to their teachers' mixed use of Putonghua, Spanish, and English in classroom teaching, including their desired classroom talk by the teachers, expected changes in teacher talk if any, their views towards translanguaging in classroom teaching in general, and the impact of teachers' bilingual and trilingual translanguaging on their Spanish learning progress. The results from the questionnaires were then analyzed quantitatively.
Semi-structured interviews were conducted with two teachers, as well as with two groups of selected students from the two classes. The teachers were invited to explain their overall views on using multiple languages in a mixed manner to organize classroom discourse, and the reasons behind their multilingual choices. The interviews lasted for approximately 30 minutes each, and the guiding questions were as follows: (1) What language(s) do you believe should be used in teaching Spanish as a third language?
(2) What is your attitude to the monolingual approach to language teaching, and why? (3) Do you have preferences for each language you employ in specific classroom tasks?
Two group interviews were held with students from the two classes (three students in each group). These lasted for one hour each, during which the students expressed their views about the teachers' translanguaging practices in class, their attitudes to the current arrangement, and their expectations for future changes. They further explained some of their answers in the questionnaire at greater length. All the interviews were recorded, transcribed, and subjected to content analysis.

Results
The results from the analysis of lesson recordings, questionnaires, and interviews are presented in this section so as to answer the three Research Questions.

Teachers' Translanguaging Practices in L3 Classrooms
Since three languages are involved in classroom discourse in the present research, translanguaging practices take place in two broad forms: bilingual translanguaging where L1 Putonghua and L3 Spanish are used interchangeably; and trilingual translanguaging where L1 Putonghua, L2 English, and L3 Spanish are employed for interactional purposes. On the basis of the data collected, there was no translanguaging practice which engaged Spanish and English only. Looking at the total number of translanguaging instances recorded in all the observed sessions, trilingual translanguaging practice accounts for 33% of the total, whereas bilingual translanguaging practice consists of 67%.
Instances containing both forms of translanguaging practices were identified and closely studied via comparison with previous studies. It was revealed that teachers' translanguaging practices do not present a wide variety of patterns with regard to pedagogical strategies. Translanguaging for explanatory and managerial purposes were the two most prominent categories, illustrated by the following examples extracted from transcripts of classroom discourse.

Translanguaging for Explanatory Purposes
This type of translanguaging occurs when the teacher employs a combination of either Putonghua and Spanish, or Putonghua, English, and Spanish to explain textbook-related content. It particularly focuses on specific features of Spanish language such as grammar and vocabulary (in textbook and after-class exercises), as illustrated in Example 1 to 4. EXAMPLE 1: (TA: Teacher A; S: a student) TA: Nada, 什么都没有，Nada hay en la sala.
Nothing. Nothing. Nothing is in the room. 这里的nada用在动词之前。 Here "nada" is used before the verb.
No hay nada en la sala.
Nothing is in the room. 这里的nada用在动词之后。用在动词之后时需要注意什么？ Here "nada" is used after the verb. But what needs to be paid attention to when it is used after the verb? S: Hay前面有个no.
Very good. When "nada" is used after the verb, you need to put "no" before the verb.

Yes.Yes.
In Examples 1 and 2, the teachers interchangeably utilize Putonghua and Spanish to explain rules of grammar to students. As Spanish language beginners, the students would not be able to understand the explanation if it was given in Spanish. Hence, in these particular pedagogical moments the teachers, who apparently hold the belief that explicit instruction in Spanish grammatical rules is an integral part of the course objectives, employ a mixture of L1 and L3, while maintaining a certain amount of talk in Spanish to facilitate understanding so as to achieve communicative purposes. Examples 3 and 4 represent the teachers' proactive and flexible mobilization of their full linguistic repertoire. Associations of linguistic aspects between Spanish and English are made, possibly because English is a more similar language to Spanish than Putonghua in terms of language genealogy; English is therefore viewed as a more useful and suitable resource in those pedagogical moments when new words and phrases are introduced to the students. Through exercising their multilingual competence, despite occasionally inaccurate expressions of English, pedagogical outcomes of meaning-making and effective interaction are accomplished.

Translanguaging for Managerial Purposes
Classroom instructions can be incomprehensible to beginners if they are articulated in full Spanish, leading to confusion and discouragement. Hence, teachers tend to actively deploy bilingual translanguaging between Putonghua and Spanish in order to facilitate communication, clarify misunderstandings and hold students' attention, while still maintaining a certain amount of Spanish input. Examples 5 and 6 provide evidence for this: As can be observed in Examples 5 and 6, translanguaging between Spanish and Putonghua is employed to provide instructions regarding homework and exams. In these examples, maintaining the provision of L3 input and minimizing misunderstanding about crucial instructional information are both on the pedagogical agenda. By navigating these cross-linguistic interactions, the teachers manage to deliver classroom management messages effectively using their broad linguistic repertoire.

Teachers' Attitudes to Their Own Translanguaging Practices
Analysis of the semi-structured interviews with two teachers has revealed their attitudes to the use of one's own multilingual competence in classroom teaching. Both rejected the notion of a monolingual approach of teaching a foreign language, and expressed a positive attitude towards being resourceful with their prior knowledge of other languages and previous experiences of language learning. Although they lacked formal understanding of the notion of "translanguaging" and the various translanguaging strategies that may be used to achieve pedagogical outcomes, they spoke with confidence about the legitimacy and necessity of consistently correlating with and integrating languages other than the L3 in different classroom tasks to help with the flow of interaction and ensure understanding. Teacher A shares her view from an affective perspective: "It is important to draw on knowledge of different languages to help one learn another language… I firmly believe this because I taught myself other languages, for example, Korean, by associating it with the linguistic knowledge I learned in English and Spanish. So I think now it is right to use all my language knowledge to help my students learn Spanish… In my daily teaching tasks, I often refer to Putonghua or English to explain new words and grammar, or to assign homework to students, or inform them of upcoming tests and exams… Because they are beginners, it is important not to frighten them or discourage them in the first place." One particularly noteworthy discovery that came from the interviews was that although both teachers stated that they often deployed trilingual translanguaging involving Putonghua, English, and Spanish simultaneously in classroom teaching, and one of them (Teacher B) was even convinced that he was more prone to shifting between English and Spanish than between Putonghua and Spanish, this is not strongly evidenced by the results of the classroom observation and the students' attitudes. Teacher B claims:

"I think English is very helpful in the learning of Spanish. They are very similar, compared to Putonghua and Spanish… When I was a beginner in Spanish language, my knowledge of English used to provide me with a lot of benefits in the process of studying Spanish… For instance, when I explain the Spanish preposition 'con' to the students, I would automatically associate it with the English preposition 'with', instead of referring it to the Chinese counterpart, because the Chinese word may cause misunderstanding or confusion to students in this situation."
It is possible that when the two teachers incorporate English in their classroom talk, they are not fully aware that their existing classroom talk often already consists of Putonghua and Spanish, with Putonghua sometimes playing the role of scaffolding the sentence. By integrating English, their teacher talk effectively becomes a combination of three languages woven together simultaneously. Therefore, bilingual translanguaging involving only English and Spanish is less likely to occur.
Overall, as advanced English speakers, the teachers agreed that trilingual translanguaging is a valuable linguistic and pedagogical asset, and that one should make full use of one's linguistic repertoire to construct and negotiate meanings in an L3 learning setting. However, their views about proactively and frequently making efficient use of their English knowledge to teach Spanish are not entirely reflected by their practice, given the aforementioned result that two thirds of the translanguaging instances involved shifts between Putonghua and Spanish.

Students' Attitudes to Teachers' Translanguaging Practices
Students' attitudes to teachers' translanguaging practices are presented from two perspectives. While findings generated from questionnaires provide a holistic view of their attitudes, the results from group interviews allow individual opinions to emerge. With regard to the questionnaires exploring students' viewpoints about teachers' classroom talk, as well as their expectations for change, the findings have been gathered under the following three headings:

Desired Teachers' Classroom Talk
Of a total of 50 participants, 36% (n=18) were generally satisfied with the current pattern of the teachers' classroom discourse incorporating Putonghua, Spanish, and English. When asked what their ideal form of teachers' classroom talk would be, their views varied, with 78% (n=39) of them opting for a combination of three languages while 12% (n=6) preferred the use of Putonghua and Spanish only. Additionally, 8% (n=4) of the participants believed that the interchangeable use of Spanish and English was most conducive to learning, and the rest 2% (n=1) would have preferred a Spanish-only L3 classroom. Table 1 illustrates the results.

Desired Changes to Teachers' Classroom Talk
When questioned about whether they would like any changes to the current pattern of teachers' classroom talk, and if so, in what way, 36% (n=18) of the participants expressed general satisfaction with their teachers' classroom discourse and thus had no suggestions for changes. Among the rest of the participants, 42% (n=21) of them wished the teachers would increase their intermixed use of Spanish and English. 12% (n=6) believed that shifting between Spanish and Putonghua could be increased, while 10% (n=5) hoped for more Spanish input in classroom discourse. Table 2 summarizes these results.

Impact of Teachers' Translanguaging Practices on the Spanish Learning Process
The vast majority (n=47) of the participants held a positive view towards the use of translanguaging as a pedagogical strategy in L3 classrooms, and believed that teachers' translanguaging practices are a helpful resource in the process of learning Spanish. In particular, translanguaging between Spanish and English as a pedagogical resource was favored by 42.6% (n=20), who considered that a flexible intermixing between Spanish and English in classroom teaching has been a helpful asset in improving their learning effectiveness. This number is only slightly lower than the students (n=27) who prioritized the value of translanguaging between Spanish and Putonghua. The results from the questionnaires seem to reveal that although only a small proportion of the participants (8%) would opt for a "Spanish + English" mode of bilingual translanguaging in the classroom, quite a large proportion of the students appear to welcome the integration of English as a beneficial agent in the process of classroom learning. Therefore, it is of interest to probe their attitudes towards different types of translanguaging practices (multilingual or bilingual) and the extent of the support that teachers, and their own knowledge of Putonghua and English, have offered in the acquisition of Spanish.

Results from Students' Group Interviews
All six students participating in the two group interviews expressed a positive attitude towards the employment of multiple languages by teachers during L3 Spanish courses. Their opinions can be summarized in three points. First, a monolingual approach to teaching Spanish as a third language is not considered plausible, because "without the aid of Putonghua and Spanish, misunderstanding will increase", which will eventually slow the learning process. Second, although switching between Spanish and Putonghua facilitates the flow of classroom teaching, incorporating English into Spanish classroom talk is more helpful in delivering new vocabulary and grammar knowledge, since "analogies can easily be drawn between these two languages and more accurate and effective understanding can be achieved", whereas Putonghua, despite being their shared L1, "cannot help clarify syntactic confusion in Spanish when encountering long and complex sentences". Finally, there was a consensual expectation regarding a more "targeted" increase of shifting between Spanish and English when possible. This has confirmed the positive impact of previous English learning experience on current Spanish learning progress.

Discussion
The "Focus on Multilingualism" approach has offered a holistic framework for this study to be conceptualised in (Cenoz, 2013;Cenoz & Gorter, 2011. Through placing emphasis on the multilingual speakers, their whole linguistic repertoire, and the contexts within which they are situated, the findings from this study contribute to our understanding of pedagogical translanguaging in third language learning contexts mainly in two aspects. First, we have explored teachers' translanguaging practices incorporating three languages, which has received little attention in previous research. Second, we have investigated both teachers' and students' attitudes to translanguaging practices in the classroom, and have thus shed light on how teachers' pedagogical translanguaging can be adapted to be even more conducive to students' learning process. The rest of this section will elaborate these two aspects respectively.

Teachers' Bilingual and Trilingual Translanguaging Practices
One of the most prominent findings revealed by this study is that teachers' mobilization of their full linguistic repertoire is manifested through their active and flexible deployment of multiple languages in the classroom. We found that trilingual translanguaging involving L1 Putonghua, L2 English, and L3 Spanish was deployed by the teachers for the purpose of negotiating meaning and facilitating interaction. This type of pedagogical translanguaging has not been observed in previous research, and thus requires more investigation in various contexts to further verify its pedagogical value.
Additionally, bilingual translanguaging between L2 English and L3 Spanish, though claimed by teachers to be a frequently employed strategy, was in fact perceived by students as less effective. It could be argued that a more balanced approach towards translanguaging practice, which dynamically engages all the prior linguistic knowledge and previous language learning experiences of multilingual speakers, might be more conducive and beneficial for L3 learners. In this context a shared L2 with more lexical and syntactic similarities to the L3, compared with a shared L1, should be seen as a significantly valuable linguistic asset, and should be put to extensive use in L3 acquisition.
In comparison to previous research into teachers' translanguaging practices in second or additional language learning contexts (Makalela, 2015;Mazak & Herbas-Donoso, 2015;Wang, 2019), analysis of the current transcripts has not shown a wide variety with regard to the types of translanguaging strategies. Bilingual and trilingual translanguaging were observed in the explanation of grammar and vocabulary, as well as in classroom management. This lack of variety in translanguaging categories might be linked to the fact that course objectives emphasize the attainment of basic Spanish grammar language skills, with less focus on spoken and written competence. It may also be ascribed to insufficient teaching experience on the part of the teachers, in the sense that they might not have felt sufficiently at ease to establish a rapport with the class. This would mean that they rarely attempted to build interpersonal relationships with the students, and hence may have felt less comfortable using diverse translanguaging strategies for social or affective purposes.

Teachers' and Students' Attitudes to Translanguaging Practices
Attitudes to teachers' pedagogical translanguaging practices are to a large extent under-researched. Previous research has revealed that teachers' awareness of and willingness to adopt translanguaging as a valuable pedagogical strategy can be increased if they are provided with appropriate guidelines or formal professional development training on implementing translanguaging (Cenoz & Santos, 2020;Gorter & Arocena, 2020). Nonetheless there may still be hesitation on the part of teachers, who may worry that using translanguaging might compromise target language exposure (Galante, 2020) and preference for a monolingual approach to teaching from students (Wang, 2019). Findings from the current study have indicated that translanguaging practices are regarded by both teachers and students as a necessary and helpful pedagogical strategy, which can lead to positive impact on students' L3 learning progress. Understanding students' attitudes is crucial for the adaptation of teachers' language choices, if needed. In this case, the students' preference for a multilingual approach to L3 acquisition coincides with and serves as an approval for teachers' current practices. Meanwhile, their desire for more translanguaging between Spanish and English could prompt teachers to adapt their translanguaging practices and help to create a more mutually beneficial learning ecology.
Previous studies have suggested that students' L1 and/or L2 knowledge is an elemental resource in their learning of L3 (Bardel & Falk, 2007;Park & Starr, 2016), but few researchers have addressed this when studying teachers' translanguaging practices and students' attitudes to them, as we have done here. It can be argued that in L3 learning contexts where both teachers and students are fluent speakers of the L1 and L2, adopting a multilingual approach to teaching which fully engages the holistic sum of all the stakeholders' multiple language capacities is likely to benefit L3 acquisition. Nonetheless, the extent to which the shared L1 and L2 can be of help depends on various aspects, one of which could be how closely related the languages are. In this case, because they belong to the same Indo-European language family, L2 English and L3 Spanish clearly share more similarities than there are between L3 Spanish and L1 Putonghua, which is a member of the Sino-Tibetan language family. Unlike Zheng (2018), who suggests avoiding drawing on students' L2 English in the process of teaching L3 Spanish for fear of negative language transfer, the results of the present research demonstrate that students believe their prior linguistic knowledge of English can be reactivated and the skills and strategies obtained in previous English learning experience can be readapted in the process of learning Spanish. In this way their metalinguistic awareness is enhanced, as suggested by Gui (2014). Our results offer the view that a number of factors, including students' attitudes, language distance, and possibly the L2 proficiencies of both teachers and students, should be fully considered to maximize the potential benefits of trilingual translanguaging as a pedagogy.

Conclusion
Although limited by the small number of participants, the current study nevertheless sheds some light on pedagogical translanguaging in L3 acquisition. It presents the bilingual and trilingual translanguaging classroom practices of two teachers of Spanish as a third language, demonstrating their fluid and skillful use of their full linguistic repertoire to achieve pedagogical purposes. It also highlights the significance of understanding students' attitudes to teachers' classroom talk, since classroom input plays a critical role in students' language learning, and therefore it is crucial that students' opinions and preferences regarding teachers' choices of languages in classroom discourse are heard. More empirical evidence is needed to further buttress these arguments, preferably with a larger sample and thus more data concerning students' attitudes and their actual learning outcomes. Subsequent studies could be conducted in contexts where the interplay of a large variety of languages exists. However, an initial proposition can be drawn from the results of this study: the use of translanguaging as a pedagogical resource in L3 acquisition is not only justifiable, but also valuable.

Funding
This work was supported by the Shanghai International Studies University under Grant [2019114018].