Similarly to the tension between the political will and the actual practices mentioned above, a discrepancy can be identified between the multilingual turn in the domain of second language acquisition (May, 2013) and the assessment practices and teachers’ beliefs, which remain strongly monolingual – and monoglossic (i.e., promoting the separation of languages). This paper aims to provide evidence for the possibility of aligning plurilingual teaching and assessment. It will analyze concrete ways of assessing plurilingual competence as a teaching and learning goal, and will show how plurilingual education can integrate plurilingual assessment practices.

The pedagogical interest in plurilingual education is a long-standing phenomenon, with evidence of its existence dating back to a considerable period in the past. In his PhD thesis, Fornel (2023) presents a comprehensive analysis of various publications from the 18th, 19th and 20th centuries that address the topic of Romance languages and promote plurilingual learning to varying degrees. A substantial number of projects and publications from the past 40 to 50 years address the topic of plurilingual education. A reference framework has even been developed by Candelier et al. (2012) for the pluralistic approaches that have been extensively investigated in recent years, including the intercultural approach (Abdallah-Pretceille, 2011; Byram, 2003; Dervin and Liddicoat, 2013), intercomprehension (Caddéo and Jamet, 2013; Escudé and Janin, 2010; Garbarino and Degache, 2017; Ollivier and Strasser, 2013), integrated didactics (Hufeisen and Lindemann, 1998; Hufeisen and Neuner, 2003), and Awakening to Languages (Candelier et al., 2003). A recent publication (Candelier et al., 2023) provides an overview of the latest developments.

Despite the fact that all of these approaches are designed to support learners in enhancing their schooling integration and developing their plurilingual competence, there has been a notable lack of interest for plurilingual assessment within the academic and pedagogic community over a considerable period of time. Indeed, while “from a social justice perspective, discrimination of multilingual practices as well as marginalization of (certain) multilingual speakers are to be contested by multilingual assessment” (Vogt and Antia, 2024, p.11), assessment tends to stay monolingual, perpetuating linguistic and cognitive injustices. This is true even in contexts where teachers implement more or less systematic practices. This means that “assessment practices that make use of more than just the target language or the full repertoire of the students to access their content knowledge” (Stathopoulou et al., 2024, p. 236) are seen with mistrust and are rarely used.

One of the challenges to mainstream plurilingual assessment (practices) might be the ambiguity of the concept itself. To start with a simple yet operative definition, Vogt and Antia, (2024, p.11) declare that “multilingual assessment can be understood as incorporating multilingual elements into assessments, whether they are content-related or language-related”. A search on academic search engines for the term “multilingual (or plurilingual) assessment” will return publications addressing the assessment of non-linguistic competences of plurilingual individuals, as well as papers and books on the assessment of plurilingual competences (for more information, see below in section 3). Both underscore that plurilingual assessment is broadly about evaluating individuals who are fluent, in various degrees, in multiple languages.

Table 1 presents some of the differences and commonalities between two main conceptual strands, in terms of: (a) characteristics of the individual, (b) the skills and competencies being evaluated, (c) the context where the evaluation occurs and, finally, (d) the purpose of the evaluation process and outcomes.

With assessment of non-linguistic competences of plurilingual individuals, we refer to the evaluation of skills or competencies such as cognitive or problem-solving abilities and subject content. Assessment of plurilingual competences is understood as the evaluation of language-related abilities, evaluating how individuals use their linguistic repertoire holistically.

This dichotomy shows that both strands share some similarities but also have distinct features, based on what and how they assess, and the borders can be blurry. Evaluation of the plurilingual individual tends to be conceived in contexts of welcoming students with a mobility history into a new educational system, to diagnose their command of or to strengthen their acquisition of school subjects’ content. This strand is concerned with issues of linguistic, social, and cognitive justice. Evaluation of the plurilingual competence is a construct developed to assess the command and use of different languages, either cumulatively or in more holistic and integrated terms. In the next section, we delve deeper into these differences.

A high number of publications focus on the assessment of knowledge and competences in non-linguistic subjects (Canagarajah, 2006, 2012; Schissel, 2019; Schissel et al., 2018; Shohamy, 2011, 2006, 2022, 2001; Shohamy et al., 2017). Most of them show that monolingual practices are creating cruel inequity and do not allow a fair and valid evaluation of what is supposed to be assessed.

Melo-Pfeifer and Ollivier (2023b, pp. 8-9) highlight that one form of discrimination in education is the practice of testing individuals from diverse linguistic and cultural backgrounds using a uniform, standardized format. The lack of consideration for the familiarity of test-takers with the language and culture of the test can have adverse effects on the results (Altakhaineh and Melo-Pfeifer, 2022). For example, Shohamy (2011) observed that Russian immigrants in Israel who were tested in Hebrew on their mathematical competencies achieved lower scores than native speakers. The research also indicated that the scores of this population were higher when the test was conducted in their first language. Furthermore, the study demonstrated that the disparate scores persisted over time, with Russian immigrants requiring 9 to 11 years to attain the same scores as native speakers.

Wright and Li (2008) demonstrated that poor proficiency in the language used in national tests in the USA creates “a source of construct irrelevance” (p. 8). They showed that the vocabulary and syntactic complexity of the task, as well as the cultural content of the tests, made them difficult for non-native speakers to understand.

This understanding of the plurilingual assessment is concerned with social justice and equity in education, through the development of “fair and equitable forms of evaluation for all students, regardless of prior language background, educational context and geographical location” (De Angelis, 2021, p. 1). Under the practices developed to reach fairer formats of evaluation, we could name testing accommodations (De Backer et al., 2019; Shohamy and Menken, 2015) including, for instance, providing students with more time to answer, reducing the number of questions or allowing students to use external resources (for example, dictionaries and online translation tools). Other strategies, more responsive in terms of students’ communicative repertoires, include presenting content in the student’s own language, allowing answers in multiple languages, and the use of home languages and other semiotic resources (Shohamy and Pennycook, 2019; Vogt and Antia, 2024).

In the domain of plurilingual competence assessment, research has particularly documented and analyzed practices since the end of the previous century. The methods of assessment can be grouped into two principal categories: hybrid and holistic/integrated (Melo-Pfeifer and Ollivier, 2023b).

The EVAL-IC project (https://www.evalic.eu) is a European project co-funded by the European Union within the Erasmus+ program. It aimed at describing and assessing plurilingual competence, in particular Romance intercomprehension competence. Intercomprehension is understood as a form of plurilingual competence in which speakers of different languages can understand each other using different languages of the same group - in this case Romance languages.

Descriptors have been designed for oral and written comprehension of narrow languages, for production in a familiar language so that speakers/writers can make themselves understood by people who have never learnt the language they use, and for plurilingual interaction. These descriptors address the different dimensions of the communicative competence: linguistic, sociocultural, pragmatic, etc., and establish six levels of proficiency (Ollivier et al., 2019).

On the basis of these descriptors and the above-mentioned definition of intercomprehension as a plurilingual form of communication involving four different language activities, an assessment protocol was designed and a pilot test carried out. The project team decided to apply the TBLA-approach (task-based language assessment) (Bachman, 1990; Bachman and Palmer, 1996; Ellis et al., 2020; Norris, 2016). Therefore, a global scenario has been chosen including different subtasks, each contributing to a coherent project.

This test was intended to be plurilingual, not only because several languages had to be activated by the students at different stages, but also because different languages were used in the subtasks.

The pilot scenario for Romance languages designed by the EVAL-IC team for university students reflects the application and selection process for the participation in an international conference on sustainable development in higher education. In the initial stage, which addresses written reception, students are required to complete an application form in a Romance language of their choice or in the official language of their university. The questions are posed in a variety of romance languages. Thereafter, the students are provided with written and video materials pertaining to sustainable solutions for universities in five Romance languages (French, Italian, Portuguese, Romanian, and Spanish). Based on their comprehension of these materials, the students are required to produce a PowerPoint presentation and an oral presentation for a plurilingual jury, who will engage in questioning in different languages. The final phase is dedicated to the participation in an online forum with other students writing in different languages in order to select a concrete action for sustainable development that can be implemented during the conference.

The METLA project (www.ecml.at/mediation), running from 2020 to 2022 and coordinated by Maria Stathopoulou, was rooted in educational methodologies that embrace diversity, particularly focusing on the significance of plurilingual and pluricultural competence. The project highlights the importance of students becoming proficient in navigating multiple languages and cultures, preparing them for effective communication across languages instead of in monolingualized communicative contexts. A key objective of the project was to provide teachers with the tools and knowledge to help them cultivate, nurture, and assess their students’ mediation skills (Stathopoulou et al., 2023). Crosslinguistic mediation involves facilitating understanding and communication between individuals or groups from different linguistic and cultural backgrounds.

The project’s outcomes strongly resonate with the core values and principles championed by the Council of Europe (see Introduction to this paper). These include a firm commitment to upholding (linguistic) human rights, fostering mutual understanding, and encouraging social cohesion across communities. The METLA project also places a strong emphasis on language inclusion, promoting the idea that multiple languages should be embraced rather than excluded in educational settings, in general, and in the language classroom more specifically. Indeed, since the introduction of the direct method - derived from foreign language learning theories based on L1 acquisition concepts such as exposure to input and immersion - the language classroom has tended to banish other languages from students’ repertoires, whether or not acquired in formal contexts. By introducing the skill of crosslinguistic mediation into the list of skills to be developed in the language classroom, the CEFR challenges this hegemony of monolingualism and the monolingualization of students in the classroom. As promoted by the CEFR Companion Volume, “in mediation, the user/learner acts as a social agent who creates bridges and helps to construct or convey meaning, sometimes within the same language, sometimes from one language to another (crosslinguistic mediation)” (2020, p. 90). This means that crosslinguistic mediation (also called interlinguistic mediation in the above-mentioned Companion Volume), as a competence integrated in language curriculum, legitimized the use of different languages in the classroom, providing realistic scenarios to understand the concomitant use of languages in communication, instead of aiming at monolingualization of communication through the use of a common language only.

As put forward by Stathopoulou et al. (2024), “Mediation in METLA tasks entails the purposeful selection of information by the mediator from a source text in one language and the relaying of this information into another language, with the intention of bringing closer interlocutors who do not share the same language. Cross-linguistic mediation can thus be taught and assessed through METLA tasks which ask for the use of different languages (i.e., passing on information from one language to another), thus softening linguistic and cultural gaps in the process” (p. 247).

Figure 1 reproduces one task of the METLA project, which shows the incorporation of students’ plurilingual C the task. The students are also required, after task completion, to evaluate their learning process.

In the METLA project, the cross-linguistic mediation tasks are designed to be inherently plurilingual, encouraging learners to recognize and utilize their additional or foreign languages, grounding these activities in their linguistic biographies and sociolinguistic realities. These tasks specifically prompt students to purposely build linguistic connections, identify similarities and differences between languages, and combine various languages with non-verbal resources (such as gestures, facial expressions, body language, and visual aids) for diverse communicative purposes (sending an email, communicating in online applications, producing flyers, etc.). By doing so, the tasks promote students’ engagement in and appreciation of language negotiation, encouraging them to navigate across languages and collaboratively construct meaning in plurilingual contexts (Stathopoulou et al., 2023, 2024). Privileged formats of assessment in this project are formative and based in self-reflection (such as Portfolios), encouraging students to return to their production and to the learning process, so that they develop agency over the learning processes.

Both projects were designed for language education programs and focus on the plurilingual competence of plurilingual subjects. The main goal was not to assess the student’ proficiency and knowledge in subjects other than languages but to evaluate their capacity to use their entire linguistic and semiotic repertoires for effective plurilingual communication. EVAL-IC and METLA opened space to languages beyond the language of instruction and the target language in the language classroom, creating - among other - spaces for translanguaging in a holistic approach to plurilingual communication and plurilingual competence. Table 2 reflects the main key-points of the two projects.

In the EVAL-IC and METLA projects, because they targeted mostly language learners, plurilingual assessment formats were not specifically drawn for newcomers to a new school system. Nevertheless, such formats can be used to show the unique challenges newcomer students face and the potential they bring with. Indeed, when integrating a new school system, especially in the European context where the learning of two foreign languages is mandatory, newcomer students are concomitantly learning the language of schooling and potentially two other languages. Even if students do not have a mobility background, they already developed skills in the language of schooling and are developing competences in two additional languages. In some cases, they also have some or full command in a heritage language. Such consideration for a plurilingual competence also calls for the development of plurilingual pedagogies and the implementation of multilingual assessment formats.

As shown in Table 2, both projects emphasized the importance of valuing plurilingual education for all students, highlighting how such formats not only support language learning but also enhance the development of transversal competences, such as cognitive flexibility, critical thinking, and intercultural understanding in diverse educational contexts. By encouraging students to draw on their entire linguistic repertoire, both projects fostered deeper engagement with the learning process and contributed to a more inclusive, learner-centered environment. By legitimizing multilingual learning pedagogies for all students, plurilingual assessment formats as promoted in the two projects also address some potential otherization occurring when these pedagogies and assessment formats are exclusively tailored to plurilingual students’ (supposed) needs.

Additionally, EVAL-IC and METLA aimed at promoting an appreciation for and recognition of individual plurilingualism, not only in the classroom but across various social and cultural settings, ensuring that students’ plurilingual abilities are acknowledged and nurtured as valuable assets, inside and outside the school.

Our study shows that new formats for assessing plurilingual competence are possible, especially procedures that integrate different languages without separating them, but rather promoting the holistic activation of all the students’ semiotic resources. Both approaches are different from those suggested by Jamet (2010), Mueller Gathercole et al. (2013) or Cenoz et al. (2013). They show that it is realistic to conceptualize and successfully implement types of assessment that do not measure competences in different languages in order to establish a plurilingual profile or calculate an index of plurilingualism. Instead, the projects consider plurilingualism as a holistic competence which makes use of all semiotic resources in order to accomplish complex communicative tasks. This study therefore provides evidence that the multilingual turn in education can include assessment and that alignment between teaching and assessment is achievable.