Stakeholders’ Perceptions over the Integration of CLIL and Museum Education and Methodological Implications

The aim of this article is to investigate stakeholders’ perceptions as regards the integration of CLIL and museum education and outline the main methodological implications. Lately, Italian museums have started offering CLIL learning programmes aimed at school groups. However, there is very little research on the affordances, issues and practical implications of integrating CLIL and museum-based pedagogies. To help fill this gap, an action research project was initiated, which involved university experts, museum staff and upper secondary teachers and students. This study focuses on the museum staff’s interview data, and reveals that successful design of CLIL museum programmes depends on different elements, such as a shared vision for CLIL and strong school-museum collaboration.


Introduction
Interdisciplinarity and the ability to capitalize upon and validate learning across different learning contexts (formal, non-formal and informal) is at the core of the present and future educational agenda of the European Union (CEDEFOP 2009). The assumption is that to really promote lifelong learning we need to create synergies among different educational providers and implement models that help teachers bridge the gap between learning at school and beyond. This is true for any type of learning, including language learning (European Council 2014Council , 2019Reinders, Benson 2017). From this perspective, it is important to underline that while Content and Language Integrated Learning (henceforth CLIL) is now widely recognized has having a positive impact on both students' learning and affective factors (Seikkula-Leino 2007;Dalton-Puffer, Nikula, Smit 2010;Coyle 2011Coyle , 2013Lasagabaster 2011Lasagabaster , 2017, research on its effects and teaching/learning frameworks has mainly concentrated on its implementation in the formal context. Thus, the questions we would like to answer here are: what happens when CLIL is implemented outside of the classroom and what methodological implications should we draw? In answering these questions we particularly focus on CLIL in the museum. The reason for this being that museums in Italy have recently emerged as alternative settings of content and language integrated learning, probably following the recent reform which made CLIL mandatory in the final year(s) of upper secondary school (Riforma degli Ordinamenti della Scuola Superiore 2009, and subsequent decrees d.P.R. 15/3/2010, n. 88-9; 1 Fazzi 2018). Several studies have already shown that engaging with museum objects has a positive impact on students' language learning and affective factors, such as increased vocabulary and positive self-concept as language users (Fazzi, Lasagabaster 2020;Díaz 2016;Parra, Di Fabio 2016;Ruanglertbutr 2016;Rohmann 2013;Wilson 2012). However, apart from our research, which we partially report on here, no study has yet fully investigated the affordances and issues of integrating CLIL and museum-based pedagogies and its methodological implications. The current article aims to fill this gap by discussing the Researcher-practitioner and museum staff' reflections, comparisons, and evaluations as regards the integration of CLIL and museum teaching during a three-year doctoral research project in the field of language education carried out in collaboration with the Civic Museums of Venice Foundation. The article starts with a brief overview of the princi-ples at the basis of CLIL and museum education. Thus, it outlines the main aspects characterising the 'CLIL in the Italian museums' phenomenon, before delving into the presentation and discussion of the results of the current study. Finally, it presents some key methodological implications for the design and implementation of CLIL learning programmes in the museum setting.

2
Theoretical Framework

CLIL
The term CLIL was first developed by the European Network of Administrators, Researchers and Practitioners (EUROCLIC) in the mid 1990s and indicates: A dual-focused educational approach in which an additional language is used for the learning and teaching of both content and language. That is, in the teaching and learning process, there is a focus not only on content, and not only on language. Each is interwoven, even if the emphasis is greater on one or the other at a given time. (Coyle, Hood, Marsh 2010, 1) As Coyle, Hood, and Marsh point out, "there is no one model for CLIL" (2010,14) as the contextual variables (i.e. teacher availability, language level on the part of both students and teachers, time, assessment processes) are so different across countries, and across schools within the same country, that CLIL can be implemented through a wide variety of curricular models. However, whatever the model, efficient language-medium instruction should be based on the integration of learning theory (content and cognition) and language learning theory (communication and culture ;Coyle 2008). This is in line with Mohan's (1986) Knowledge Framework, which has been used by Coyle (2006Coyle ( , 2008 as the starting point for her 4Cs Framework. In CLIL, this framework starts with content (such as subject matter, themes, cross-curricular approaches) and focuses on the interrelationship between content (subject matter), communication (language), cognition (thinking) and culture (awareness of self and 'otherness') (Coyle 2008, 103). Coyle, Hood, and Marsh (2010) affirm that these elements need to be integrated when planning CLIL, so as to balance the linguistic and cognitive demands. In order to accomplish this task, teachers have a variety of methods, teaching tools and strategies at their disposal, such as: -evaluation and design of materials (Coyle, Hood, Marsh 2010;Meyer 2010); -scaffolding and feedback (Bentley 2010;Ball, Kelly, Clegg 2015;Menegale 2008;Lyster 2007; European Commission 2014); -supporting listening and reading (Coonan 2012(Coonan , 2016Ball, Kelly, Clegg 2015); -pushed output (Coonan 2008;Swain 2000); -cooperative learning (Guazzieri 2009); -task-based teaching (Coonan 2008;Ellis 2003;Willis 1996), and -project-based learning (Lasagabaster, Beloqui 2015;Serragiotto 2014b).

Museum Education
On the basis of the International Council of Museums (ICOM) museum definition, 2 Vuillaume (2016, 8) claims that: The ultimate goal of museums (which may sound naive or presumptuous) is to help people to grow as individuals, become more critical of our society and more involved, learn to appreciate dialogue, feel empathy, be tolerant, become more educated and more civilised, to stand against ignorance and extremism.
In this sense, museums are much more than stiff repositories of objects aimed at experts. Instead, they act as public forums where both education and enjoyment can take place (Falk, Dierking [2000] 2018. Indeed, when investigating the educational value of school field trips to museums, researchers have indicated that student groups that visit a museum often show cognitive gain (Stronck 1983, cited in Griffin 2004, and a more positive attitude towards learning (Orion, Hofstein 1994) than those who do not. This is due to the fact that museum learning is based on engagement with objects, 3 whose realness "enables the possibility of an arousal of interest or a focus of attention that is qualitatively different from the attention given to the written word" (Hooper-Greenhill 1994a, 98). When talking about museumbased pedagogies, we thus refer to the array of approaches, strate-2 "A non-profit, permanent institution in the service of society and its development, open to the public, which acquires, conserves, researches, communicates and exhibits the tangible and intangible heritage of humanity and its environment for the purposes of education, study and enjoyment" (ICOM 2007).
3 With the term 'object' we refer to a variety of cultural expressions -both tangible, such as artefacts, made or modified by a person or persons, and specimens, natural objects, and intangible, such as oral traditions and performing arts -"collected" by museums (Dudley 2009). In the current research, most of the objects students engaged with were tangible, mainly art works (frescoes, paintings, sculptures, etc.) and animal specimens. However, intangible objects can also be used in devising foreign language learning activities in the museum context. gies, and tools that are used to learn about or through tangible and intangible museum objects (see van Veldhuizen 2017 for a more detailed description), such as: -questions (Hooper-Greenhill 1994b, 158 for a list of types of questions, and Shuh 1982 for practical examples); -activities and worksheets (see Moma online course 4 for a full list of activities and discussion of aims and procedures; Bamberg, Tal 2007 for how to design museum worksheets); -visual thinking strategies (Yenawine 2013;Housen 1997); -discovery learning and Inquiry-based Science Education (IB-SE; Pedaste et al. 2015).
It is also important to point out that 'museum learning' does not equal to 'interaction with objects'. Visiting a museum is a very complex semiotic and social experience, and what one learns from it is the result of the "never-ending interaction" of three contexts (Falk, Dierking [2000] 2018) -personal, sociocultural and physical. Indeed, according to Falk andDierking ([2000] 2018), what a visitor learns in a museum depends on a number of factors, such as his/her personal interests and previous knowledge (personal), the people with whom he/she visits the museum (sociocultural), the design elements of the halls and reinforcing events after the museum experience (physical). However, in the case of school groups, one has also to consider the instructional context (Orion, Hofstein 1994), which refers to, for example, students' participation in pre-visit activities, teachers' motivation, school-museum collaboration and communication, travel logistics, and, even, weather conditions on the day of the visit.

Why CLIL and Museum Education in Italy?
The emergence of CLIL in museums in Italy is a grassroot and multifaceted phenomenon, by which museums use and adapt CLIL principles and practices to devise educational programmes that likely complement CLIL delivery at school. Interestingly, these programmes result from the need to: -align to the internationalization of the  Education 2015Education , 2017Education , 20186 Law 107, The Good School 7 ), which encourage schools and museums to strengthen their collaboration.
As regards the first point, the need to align to the school curriculum is an aspect that characterizes museum educational provisions across the world. In fact, as Hooper-Greenhill highlights, school groups "are unlikely to make much use of museums unless their provision relates fairly closely to the areas which are being studied" (1994b, 165). This means that museums are always in search of innovative ways to support teachers' delivery of the school curriculum (Xanthoudaki 2015). Considering the effect that CLIL has had on all school stakeholders (e.g. teachers, headmasters) in Italy (Aiello, Di Martino, Di Sabato 2017), it was only a matter of time before CLIL crossed the school borders and spread to the museum field. However, while there are certain aspects that are similar across CLIL museum programmes, there are also many differences. For example, a similarity is that most of the CLIL museum learning activities on offer are grassroot experimentations, which often involve the collaboration of other local institutions (e.g. private language institutions, schools). On the other hand, the way CLIL is conceptualized differs across museums, depending on factors such as the museum mission and targets.
Further research is certainly needed to investigate this phenomenon so as to understand better its rationale, characteristics, and repercussions at both organizational and practical level (see also Fazzi 2018).

The Research Project
This study was part of a three-year doctoral research project (2015-18), whose aim was to understand the impact of integrating CLIL and museum-based pedagogies on students' learning and attitudes (Fazzi, Lasagabaster 2020), and develop a model to support teachers and museum educators co-design and implement CLIL projects across the school and museum setting. In particular, we initiated an action Moreover, we also liaised with a local liceo artistico 8 (partner school) and designed pre-and post-visit activities, 9 integrating the museum visit on Animal Classification into the science school curriculum of 1st and 2nd year classes. We thought that an action research methodology 10 could support these goals as it allows to put in practice a series of actions with an end-result of improving the situation and of learning from the consequences of those actions through collaborative, structured and reflective inquiry (Burns 1999). We collected both quantitative and qualitative data using eight research instruments, and adopted an embedded mixed methods design (Ivankova, Creswell 2009, 143-4), as well as triangulation procedures.
In this article, we only focus on the stakeholders' perceptions of integrating CLIL and museum-based pedagogies, and thus report and discuss the qualitative results collected through the Researcher-practitioner's journal, museum staff's fruitful discussions, and interviews. Moreover, we draw some methodological implications for the empirically-based model under construction.
8 Following the Reform in 2003, the Italian upper secondary school system is currently structured in: Lycée system (Licei), technical schools (istituti tecnici), and vocational schools (istituti professionali). The Licei are generally directed at students aged 14 to 19, and are divided in sub-types with different curricula and specializations: liceo artistico specializes in arts (i.e. fine arts, design, sculpture), liceo linguistico specializes in languages (i.e. foreign languages, cultures and literatures), liceo scientifico specializes in sciences (i.e. biology, chemistry, physics and maths). For a full description see Cinganotto 2016, 384 andINDIRE, MIUR 2014. 9 Pre-and post-visit activities as developed in collaboration with the partner school are described in detail in Fazzi, Lasagabaster 2020.
10 According to Burns, 'action' "involves putting deliberate practical changes or 'interventions' in place to improve, modify, or develop the situation", while 'research' "involves a systematic approach to collecting information, or data, usually using methods commonly associated with qualitative research" (2009, 114).

Outline of the CLIL Museum Programmes at the Civic Museums of Venice Foundation
All the CLIL museum programmes specifically devised for the project aimed at offering middle and upper secondary students the opportunity to use English to engage with art or science contents in an alternative and stimulating environment. At the macro-level, we developed a structure, which integrated Willis ' (1996) Task-Based Approach and Johnson's (1995Johnson's ( , 2009) Touring Strategies: 6

Figure 1: Macro-level structure of the CLIL museum visit
At the micro-level, we used scaffolding strategies and materials, such as questions, visuals, worksheets, which included different types of cooperative activities, and a glossary, with the aim of balancing linguistic and cognitive demands. In so doing, we followed Coyle, Hood, and Marsh's (2010, 43-44) CLIL Matrix (adapted from Cummins 1984), and designed a progression of different types of activities, from easier to more difficult, which could flexibly be adapted according to students' cognitive and language level. However, we also made sure that the task at Ca' Rezzonico and Ca' Pesaro followed Hooper-Greenhill's (1994) suggestions on how to engage with objects, starting from close observation, to analysis and shared discussion. On the contrary, the task at the Natural History Museum of Venice followed the stages of the scientific method: data collection, hypotheses creation and testing, and observation formulation (see Pedaste et al. 2015).

Participants
The study involved the following participants:

Museum staff (Foundation Civic Museums of Venice) and roles
-Education Director: oversaw the design, implementation, and management of all the educational programmes on offer at the Foundation Civic Museums of Venice. She had an art history background and several years of experience in museum education. -Director of the Natural History Museum (MSN Director): is the key person responsible for the running of the Natural History Museum of Venice. He has a science background and several years of experience in science communication, and museum management. -Education Manager 1 and 2: were responsible for the design of the educational provision of the Natural History Museum of Venice. They both had a background in science and several years of experience in devising non-formal science educational activities. -Museum Educator 1: was a very experienced science educator, having a strong scientific background, and a B2 level in English. He took part in two training workshops on CLIL.
He also job shadowed the Researcher-practitioner while delivering a couple of CLIL workshops on Animal classification. In the third year of the project (2017/2018) he codelivered with the Researcher-practitioner the programme on Animal Classification, and was the only educator responsible for the delivery of the programme on the Lagoon at the Natural History Museum of Venice.

Figure 1 Macro-level structure of the CLIL museum visit
At the micro-level, we used scaffolding strategies and materials, such as questions, visuals, worksheets, which included different types of cooperative activities, and a glossary, with the aim of balancing linguistic and cognitive demands. In so doing, we followed Coyle, Hood, and Marsh's (2010, 43-4) CLIL Matrix (adapted from Cummins 1984), and designed a progression of different types of activities, from easier to more difficult, which could flexibly be adapted according to students' cognitive and language level. However, we also made sure that the task at Ca' Rezzonico and Ca' Pesaro followed Hooper-Greenhill's (1994) suggestions on how to engage with objects, starting from close observation, to analysis and shared discussion. On the contrary, the task at the Natural History Museum of Venice followed the stages of the scientific method: data collection, hypotheses creation and testing, and observation formulation (Pedaste et al. 2015).

Participants
The study involved the following participants: Researcher-practitioner The Researcher-practitioner (RP) was a doctoral student in the Department of Linguistic and Comparative Cultural Studies at Ca' Foscari University of Venice, but has also been a museum educator and a teacher of English for several years. She delivered the museum visits, and also collaborated as a teacher assistant for the partner school during the project.

Research Question
Our research question was: "What are stakeholders' perceptions as regards the integration of CLIL and museum-based pedagogies?".

Research Instruments
We used four types of qualitative instruments and methods, with the aim of collecting participants' perceptions and reflections about the aspects that need to be taken into consideration when integrating CLIL and museum education. As regards the fruitful discussion, 11 the Researcher-practitioner provided the MSN staff with a questioning route, which only offered inputs for discussion.
Museum Educator 2's oral interview was semi-structured, as the Researcher-practitioner prepared a questioning route, but left the format open-ended so as to allow the participant to "elaborate on the issues raised in an exploratory manner" (Dörnyei 2007, 136).
Two written interviews were sent via e-mail to the Education Director and Museum Educator 1, because of both time constraints and the difficulty encountered in scheduling face-to-face meetings. The questioning route for the Education Director was similar to that used for the fruitful discussion between the MSN staff and the Researcher-practitioner, while that for Museum Educator 1 followed the one used for Museum Educator 2.
The Researcher-practitioner kept a journal in which she recorded both the descriptive and reflective aspects of her observations regarding the CLIL museum visits, to better focus her analysis and interpretations (Burns 1999, 90). With this introspective method, she was able to collect her on-going interpretations and reflections on the weaknesses and strengths of the CLIL museum learning programmes. The journal also enabled the Researcher-practitioner to record her emotions in relation to her practice, allowing her to draw hypotheses on the profile a museum educator involved in CLIL museum learning programmes should have.

Data Analysis Procedures
The analysis was based on Thematic Analysis (Liamputtong 2011). We first read through the participants' open-ended responses to find repeated patterns of meaning (Liamputtong 2011). Then, we read through each transcript and combined the interview data with the open-ended responses data. We performed initial and axial coding, working from a more descriptive to a more analytical perspective.
11 According to Foreman-Peck and Travers (2013), fruitful discussions are group discussions in which the moderator and the participants share equal responsibilities.

417
During the initial coding, we grouped the data and assigned labels, which we then further categorized. In so doing, we looked for the most significant and frequent initial codes. Finally, we conducted axial coding by making connections between major categories and their respective sub-categories (Liamputtong 2011). N-vivo was used to assign the codes to the focus group segments and open-ended responses, and to keep track of how the codes were organized and represented in the transcripts during both coding stages. In analysing the focus groups, we considered three levels of analysis: the individual, the group, and the group interaction.

Results
To answer the research question, we conducted an integrated analysis of the stakeholders' perceptions. Four different categories emerged from the data analysis: Vision, Context, Methodology, and the CLIL museum Educator Profile.

Vision
In explaining the reasons behind her interest in CLIL, the Education Director claimed that museums can play an important role in supporting teachers to implement changes in their delivery of the school curriculum: Education Director: The introduction of CLIL involved and is still involving a great effort on the part of teachers to respond to the curriculum changes, and not always they have been able/ are able to do so. Museums, when the appropriate pedagogical approach is used, can be incredible tools to support learning.
However, CLIL was also seen by the MSN staff as a tool to address the mission of the museum, by supporting students' knowledge and communication of the local territory from an international perspective: Education Manager 1: I've always liked the fact that students could tell about their territory also in English, given the fact that they more and more take part in exchange programmes, so to be able to tell where they live, it's nice... and I think this works... to have a vocabulary that can narrate the territory where you live is certainly an added value.
During the fruitful discussion, the MSN staff drew a comparison with another museum programme called Treasure Hunt, run by an exter- , because sometimes if you don't know the tool you don't know it exists, but when you encounter it a light goes on and you start thinking "this is an instrument that I can use to speak to a certain audience, to make them use English, to do things that I wanted to do but I couldn't, to do it with older students in a certain way".

Context
According to the MSN staff, the collaboration with the Researcherpractitioner, who had different but complementary skills, was pivotal when designing the CLIL museum programme on Animal Classification: The data analysis also revealed that there are certain differences between the museum and school context, which have an impact on the learning experience. For example, the Researcher-practitioner reported the invigilators' comments and behaviour as sometimes disruptive of the positive learning atmosphere: RP: I also found it really heavy not to be able to allow them to sit on the floor [for security reasons]. Half of my mental energy goes into keeping an eye on how students move around the objects, making sure they don't irritate the invigilators. "Don't get too close to the art works, don't lean on the walls".
Similarly, Museum Educator 2 claimed that CLIL delivery in the museum also has to consider the size, and design of the rooms, as well as the tourist flow: ME2: Yes, I'd say that all museum are appropriate if they don't have many people crammed together, but it also depends on the design of the exhibition… I'd never do it in the [name of museum], for example, because it's sort of a Wunderkammer, and there's no space to move.
As regards the school-museum collaboration, participants reported that teachers often played a passive role during the museum visit in both CLIL and non-CLIL museum learning activities:

ME1:
Most of the time, they mind their own business and leave the workshop. Then in the end, they pay you compliments for the little they've seen/experienced. They very rarely participate in an active way, supporting students' [understanding].
Another important aspect to consider is the fact that teachers often underestimated the demands of participating in a CLIL museum visit for students and failed to capitalize on its educational value by booking multiple visits to other museums and institutions on the same day: ME2: I'd never consider to do another activity after this one… I think the students already accomplished what they were supposed to… it's not [feasible] that "yes, after this we take them to the Guggenheim and then to the Biennale" [laughing in astonishment]. RP: [Laughing]. ME2: How can they do it? It's impossible, right?! From this perspective, the MSN staff pointed out that, given the oneoff nature of any museum learning experience, the CLIL museum visit should work as a supplement to CLIL delivery at school: Education Manager 2: But also because I think that here in the museum we can give a "taste" of CLIL.

Methodology
When reflecting on the positive aspects of the CLIL museum programmes, the Education Director said that she was particularly proud of the materials developed and the balance obtained between content and language: Education Director: The added value of the museum experience, the right balance we were able to establish between language and contents, and the materials [worksheets] we designed.
However, despite the efforts, the activities and worksheets created were not always exempt of problems. The stakeholders reported that some activities originally planned were later modified because they were both cognitively and linguistically too difficult. An example is the definition of homologous and analogous structures in the CLIL museum programme on Animal Classification. In fact, in her journal, the Researcher-practitioner claimed that:

RP:
In relation to the second part, building the definitions of homologous and analogous structures with the students resulted too complex and too time-consuming. Thus, Education Manager 1 and 2 suggested to use the keywords I used to scaffold the definitions and design a cloze in which to insert them.
Other activities were modified because they did not support students' engagement with the museum objects. Indeed, the Researcher-practitioner reported that the activities that were most successful across age groups and language levels were those involving a discovery element. For example, the activity which asked students to look for the corresponding shapes of Arturo Martini's sculptures in the CLIL museum programme at Ca' Pesaro:

RP:
The activities on the comparison of styles based on Casorati and Martini's artworks went better. There was much more interaction on the part of the students and I asked them to make a list of all the differences they noticed. There were two Spanish girls who were enthusiastic about the visit and kept asking questions. Same with a couple of Italian girls. They were able to identify the majority of the differences, which I later systematized in plenary and expanded.
Interestingly, what also emerges as important from the analysis is the way the worksheets and activity books were used to both facilitate students' comprehension and promote their vocabulary acquisition: For example, the Researcher-practitioner explained how she used the activity book as a first step to promote students' outputs, and how she would then systematize and expand them according to students' interest and cognitive-language level: RP: Every time we approached a new work of art, I would ask students first to work on the group activities in the worksheets, then I'd build on the outputs of the group activities to develop further discussion. I was also able to build multidisciplinary connections, which is what they needed in preparation for the final exam of upper secondary school (especially the tesina).
Indeed, what emerges from the data is that museum educators placed great effort in facilitating students' comprehension and use of vocabulary during the visit: ME1: You need to change your approach completely in comparison to a "normal" activity. You need to reduce the concepts you want to convey and continue to repeat the same words to fix the few [concepts] you've chosen to focus on.
As regards the structure of the CLIL museum visit, Museum Educator 2 highlighted that the first two stages, the Welcome stage and the Ice-breaker, were fundamental not only in lowering students' anxiety, but also in creating a positive relationship with the teachers and thus a positive learning atmosphere:

ME2:
Maybe the most difficult part is to jump in the void as you're asking them to do something and to test themselves, and it's not the most amazing thing ever "do this exercise, test yourself"… you can also sell it to them like it's a game, but they know that it's not a game [emphasis]. RP: Ha ha. ME2: So the ice-breaker at the beginning is fundamental. RP: Mmh, why do you say so? ME2: First of all because you tell them something about yourself, because I always choose three things about myself, two are true and one is false, and they think of you as a human being, and not as a teacher, and this is very important afterwards… and it's also something that encourages discussion, something fun, a bit nonsense but they like it, and it always ends with a good laugh.
She also suggested that the Final remarks stage is a step that cannot be missed to reinforce students' self-esteem and encourage them to return to the museum: ME2: Anyway, I think it's very important to thank them at the end [of the workshop], because they understand that you had fun as well, that they did something nice, and that they worked well, because it was a test for them and I thank them for the time they've dedicated to interacting with me.

The CLIL Museum Educator Profile
As regards the skills needed by a museum educator when delivering a CLIL museum visit, the Education Director claimed that he/she should have both excellent language-and content-related knowledge EL.LE e-ISSN 2280-6792 9(3), 2020, 407-436 423 and skills and should be able to actively engage students with the museum contents: Education Director: He/she should know the museum contents, so as to be able to also answer the unforeseen questions; he/she should adopt the cooperative/active methodology in a critical way, and have an excellent competence in the FL.
Interestingly, Museum Educator 2 claimed that, when acting as a CLIL museum educator, she perceives herself as a language "model": Indeed, Museum Educator 2 highlighted the differences between a traditional museum visit delivered to tourists and a CLIL museum visit: ME2: I'm less instinctive, I follow more of a structure and [pause] and I tend to observe more how they react… and I try to understand if they're following me, because… one thing is to adapt the contents on the basis of the group you have in front of you and that… maybe it's because I've been working [as a museum educator] for a long time, but I mean… with [CLIL] is totally different because I need to pay attention to what they say and how they say it, if they can understand me, I also pay attention EL.LE e-ISSN 2280-6792 9(3), 2020, 407-436 424 to how they react to what I say because I need to adapt to their English level… recently at Ca' Pesaro I had students that had already passed the C1 level and students that had never spoken in English before in their life. Flexibility in approaching students by adapting his way of speaking and scaffolding students' understanding was also acknowledged by Museum Educator 1, who drew a parallel between CLIL and good museum-based pedagogy: ME1: It's necessary to ask lots of questions or ask them to complete sentences to make sure that they understand. However, this is something that is normally done also during the non-CLIL museum workshops.
However, Museum Educator 2 also indicated other soft skills as pivotal for any museum educator delivering a CLIL museum workshop. In fact, in building a profile of the perfect CLIL museum educator, Museum Educator 2 claimed that empathy and being able to build a relationship of trust with students is also incredibly important in terms of students' engagement: ME2: Perfect knowledge of English, perfect knowledge of the museum and related to it; mmh, to be really nice, great empathy, to be an excellent listener [pause], great patience, lots of it [laughing].
Interestingly, when considering the array and multidisciplinary knowledge and skills required by a CLIL museum educator, the MSN Director claimed that museums need to consider that both designing a CLIL museum programme and training the person to deliver it takes much more time than it would for any other type of museumlearning programme: MSN Director: This is the most difficult thing to explain when you propose a CLIL museum programme to an institution that has never worked with it before... there are certain objectives, but what do I need to deliver them? eh-h [sighing] you need quite a lot of professional training, in the sense that... you need the people that are available, because [pause] what I see as a possible critical point is that it requires people that have a very complex professional expertise, which you can't develop in a quick way, also because a good scientific background needs to be developed through a degree and not a series of short courses, [then you need] the English language competence, which also requires time, and then you also need a methodological [CLIL] EL.LE e-ISSN 2280-6792 9(3), 2020, 407-436 competence... all these competences that a museum educator needs to develop need to be worth it from an economical point of view [in terms of professional opportunities], otherwise why should someone make such a life investment?

Creating a Shared Vision for CLIL in the Museum
The results of our research show that the reasons underpinning museums' decision to learn about and implement CLIL are mainly two: supporting teachers' delivery of the school curriculum, and addressing their mission of promoting lifelong learning in the 21st century.
As regards the former, the Education Director said that, besides the financial agenda, she perceived the need to introduce CLIL in her institution as fundamental to help teachers face the challenges of CLIL training and delivery. 12 The Education Director's perception is supported by Xanthoudaki's assumption that "museums can make a difference because they are able to instill a methodology which is part of their very nature, integrated in the things they do well, do for a long time, and are unique at doing" (2015, 252; italics in the original).
As regards the latter, the interest of the MSN staff towards CLIL was due to it being the right tool to engage a specific audience, and, at the same time, to focus on what is contemporary and relevant for the museum to be sustainable (Di Pietro et al. 2014).
We need to keep in mind that CLIL is not mandatory in museums and that the reasons behind why museums are interested in its implementation may differ dramatically from those of schools.

Collaboration across Institutions
According to Hooper-Greenhill (1994b, 93), "the success of museum and gallery education services is closely geared to the efficacy of relations outside the museum". Indeed, our results show that the integration of CLIL and museum-based pedagogies depends on the collaboration with other institutions. In particular, while Coyle, Hood, and Marsh advice teachers to create communities of "fellow professionals" (2010,69), what emerged as necessary in our project was to build a bottom-up partnership across contexts -the museum, the school, and the university. In fact, the MSN staff highly valued the collaboration with the Researcher-practitioner both as regards designing the programmes, and training the other museum educators on the principles of CLIL learning and teaching.
However, in reflecting on how we came to design the CLIL museum programme on Animal Classification, the MSN staff claimed that an important lesson they had learnt regarded the need to involve teachers in the designing process. Education Manager 1 underlined that if a museum is committed to support teachers' CLIL delivery in the classroom, then it has to 'learn' from the teachers themselves what they need and expect from a CLIL museum visit. In fact, according to DeWitt and Osborne, teachers' perceived needs for resources, his or her agenda or goals for the school trip, and the context in which he or she operates should be a primary consideration in the development of resources for school trips. (2007,689) However, which teachers to involve is still an open question: language teachers, content teachers, teachers trained in the CLIL methodology?
When reflecting on how to integrate CLIL and museum education, the MSN Director himself highlighted the need to collaborate with the university to experiment methodological approaches that are inherently cross-disciplinary. This is the only way they can really embark on a journey, which the MSN Director defined as challenging and requiring more effort and time than any other kind of innovation/experiment in the museum.
As regards museum-school collaboration, the museum educators perceived the role played by most teachers as passive at best. As widely reported in the literature in museum studies, this seems to be a common feature of school trips to museums. For example, Mathewson-Mitchell (2007, 7) claims that teachers often assume they have a marginal role on the realization or value of museum experiences, and this might be the reason why most of them simply focus on "following the museum guide, helping with keeping the order, and watching their students" (Falk, Dierking [2000] 2018, 932; see also Griffin 2004, 37). In fact, some of them wrongly assume that, once in the museum, "meanings will be transmitted in a naturalistic manner" (Mathewson-Mitchell 2007, 7). In our study, museum educators also reported that there were multiple occasions in which teachers did not understand the demands for students of participating in a museum visit through English, and booked several other activities in the same day. The museum educators involved in our research reported how students were sometimes exhausted before even starting the CLIL museum visit, simply because they had just finished another activity in another museum.

Learning Context
The findings in our research show that in designing and implementing CLIL museum programmes there are at least two aspects to take into consideration in relation to the learning context: -the specific characteristics of the museum setting; -CLIL museum educator profile.
In considering the first aspect, museum educators highlighted how museums have specific characteristics, which may impede learning. Indeed, being public spaces, one needs to consider aspects such as the design of the exhibition, the size of the room(s) in which the visit takes place, and the fact that the visit might be interrupted by the presence of other visitors or by the invigilators enforcing security regulations.
As regards the second aspect, while both the Italian law (see Decree D.D. n. 6 dated 16 April 2012 13 ) and the literature on CLIL (see Coonan 2012;Balboni, Coonan 2014;Cinganotto 2016;Ludbrook 2014) have clearly established the profile required by a CLIL teacher, there are currently no criteria to follow for museum educators involved in CLIL teaching. First, we need to consider that the profession of the museum educator still needs to be formally recognized (ICOM 2008). Second, our results suggest that a CLIL museum educator should have a hybrid profile, integrating both CLIL and museumbased teaching principles. In particular, in exploring the knowledge, skills, and competences required, the participants in our research affirmed that a CLIL museum educator should: -have a high competence in the FL, and be familiar with subject-specific and classroom management vocabulary in the FL; -be familiar with the museum context, and subject-specific contents in relation to the museum collection; -be familiar with the principles of museum-based pedagogy, facilitating inquiry, thinking, problem-solving, and observation skills, encouraging students' interaction, and promoting their interest and curiosity; -be able to use different strategies to scaffold students' understanding and interaction; -be able to exploit the potential of the materials devised, by flexibly adapting them to the specific audience; -be positive, emphatic, and a great communicator.

Methodology
When discussing the positive and negative aspects of the methodology adopted in the CLIL museum programmes, the museum staff claimed that finding the right balance between content and language was not an easy task. Indeed, the issue with teaching in CLIL in the museum is that the design of the visit/programme does not have a specific group of students in mind. In fact, we realized that what we needed was a standard visit format with standard activity books and worksheets that could be flexibly used and adapted on the spot by the museum educators. From this perspective, we also found that the true challenge was to use questions (Menegale 2008) and practical activities/experiments to scaffold students' learning through English, but also to capitalize upon students' engagement with museum objects and expand/integrate their learning (Hooper-Greenhill 1994a).
Only in this way we were able to face the challenge of engaging students with different language and content competences (Caon 2008), and prior knowledge. However, the results of our inquiry show that some of the tasks originally designed (e.g. definition of homologous and analogous structures; CLIL programme on Animal Classification) were too demanding, both linguistically and cognitively. The difference between CLIL at school and in the museum is that, unlike classroom learning, which is composed of linear sequences units that rely on prior knowledge and previously learned scientific concepts, museumbased learning occurs in short time units, does not require continuity, and relies on curiosity, intrinsic motivation, choice and control. (Bamberger, Tal 2007, 77) In this sense, we realised that the combination of the short duration, one-off experience of the museum visit, and the use of CLIL made it necessary to revise some of the tasks, and reduce the objectives initially set.
On the other hand, when analysing the activities of the CLIL museum programmes, while they were successful in following a "route from low linguistic and cognitive demands to high linguistic and cognitive demands" (Coyle, Hood, Marsh 2010, 68), they did not always meet the conditions of museum-based pedagogy, as outlined by Falk andDierking ([2000] 2018, 24). From this perspective, we need to highlight that for museum tasks to be successful, whatever the medium of communication, they have to promote students' close observation of objects, curiosity, and a sense of discovery.
As regards the format of the visit, our findings suggest that the introduction and learning agreement in the Welcome stage at the beginning of the visit are fundamental in both explaining the methodology and the structure of the visit, and in encouraging students' engagement and the use of English. The literature on classroom CLIL underlines the importance of sharing with students learning objectives, expectations, and responsibilities (Serragiotto 2014a, 59;Coonan 2012;Coyle, Hood, Marsh 2010). However, in the context of CLIL teaching in the museum, the extra challenge is represented by the fact that the visit is led by the museum educator, who has very little time to bond with the students. In this context, the museum educators agreed that the Ice-breaker was very important, both to give students the chance to get familiar with the museum educator, and to encourage them to interact in English by lowering their affective filter (Krashen 1982;Balboni 2015). Our findings also suggest that praising students' work and performance in English during the Final remarks at the end of the visit is important, as it promotes their positive self-concept as users of English and motivates them towards engaging in similar experiences.

Conclusion
More and more museums in Italy have become interested in CLIL and its potential to support formal education as well as to address their social agenda. However, we still know very little about the potential and issues of integrating CLIL and museum-based pedagogies.
Our article aimed at helping filling this gap, by investigating the reflections, evaluations, and comparisons of multiple actors involved in the design and implementation of CLIL learning activities at the Civic Museums of Venice Foundation. In doing so, we came to define the following methodological implications.
First of all, designing successful CLIL museum programmes requires the creation of a shared vision for CLIL across different institutions, including museums, schools, and university. In fact, it is only through long-term partnerships, which combine different competences and share aims, that professionals can exploit the potential of CLIL beyond the four-walled classroom.
Second, museums need to establish a successful collaboration with schools through providing teachers with both training and supporting documents (e.g. lesson plan of the CLIL museum visit, examples of pre-and post-visit activities). Teachers and museum educators have different agendas, and these need to be made clear beforehand the CLIL museum visit, so that teachers act in a supporting capacity during the visit.
Third, the designer(s) need to carefully think about the design and size of the museum rooms in which the activities will take place, and how to redirect students' attention and/or facilitate their comprehension in the case of interruptions caused by either other visitors or by the museum staff.
Fourth, in terms of methodology, at the macro-level, the CLIL museum visit should consist of four main moments: a Welcome stage, also including an Ice-breaker, to introduce the aims of the visit, and allow students and museum educator to get to know each another; a Task, consisting of activities that balance content and language, while at the same time promoting students' engagement with the objects and conversation with their peers; Final remarks, in which the museum educator praises students' efforts and encourage them to come back to the museum. At the micro-level, materials should be designed so as to be flexibly adapted to groups of different ages and language levels, and several scaffolding strategies should be applied (e.g. questions, glossary).
Finally, as regards the profile of the CLIL museum educator, the right person should have a great knowledge of the museum context and contents, as well as excellent language skills and knowledge of CLIL learning and teaching principles. He or she should also be a good communicator, and an emphatic educator. Unfortunately, as the MSN Director said, such a hybrid profile requires training and sacrifices, which are rarely repaid either economically or professionally in the museum sector.
Our findings are far from being generalizable. However, we think that they can be of help in unveiling a still rather unexplored field of research. Certainly, further research is needed to deepen the understanding of each of the aspects here presented.