English curriculum: insights from a French lifelong learning context

For about three decades, Higher Education (HE) institutions across Europe have been facing the challenging task of developing policy measures on lifelong learning, which is acknowledged as one of the major responses to socioeconomic changes related to globalization, rapid technological progress, and demographic transformation in aging societies (EUA, 2008; Holford, Milana, Mohorčič, & Špolar, 2014). This challenge exposes a need to widen education access to an increasingly large range of adults with different professional and personal needs and interests with the aim of enhancing their employability, mobility, and competitiveness. In France, while universities are still struggling to adapt degree programs to adult needs and blur the boundaries between initial and continuing education (Borras & Bosse, 2017)2, the Conservatoire National des Arts et Métiers (hereafter Cnam) has been successfully tackling these particular issues since the early 1970’s (Dubar, 2008). The Cnam is actually a unique HE public institution in that it is exclusively dedicated to lifelong learning and offers a variety of training programs in the economic, technical, and social fields. Adults3 enrolled in the


Introduction
For about three decades, Higher Education (HE) institutions across Europe have been facing the challenging task of developing policy measures on lifelong learning, which is acknowledged as one of the major responses to socioeconomic changes related to globalization, rapid technological progress, and demographic transformation in aging societies (EUA, 2008;Holford, Milana, Mohorčič, & Špolar, 2014). This challenge exposes a need to widen education access to an increasingly large range of adults with different professional and personal needs and interests with the aim of enhancing their employability, mobility, and competitiveness. In France, while universities are still struggling to adapt degree programs to adult needs and blur the boundaries between initial and continuing education (Borras & Bosse, 2017) 2 , the Conservatoire National des Arts et Métiers (hereafter Cnam) has been successfully tackling these particular issues since the early 1970's (Dubar, 2008). The Cnam is actually a unique HE public institution in that it is exclusively dedicated to lifelong learning and offers a variety of training programs in the economic, technical, and social fields. Adults 3 enrolled in the Cnam are all professionals pursuing education at all levels of qualifications (undergraduate, graduate, and doctoral levels). They have the opportunity to tailor their training to meet their career aspirations at a pace that suits their personal circumstances -including the possibility to enroll in distance learning and evening or Saturday courses.
Foreign Language (FL) proficiency is a key competence that is significantly promoted in lifelong learning, as it undeniably supports social inclusion and economic growth. In France, the basic effect of the internationalization of business and industry has been the progressive adoption of English as the corporate lingua franca. Research has shown that mastering English at an advanced level is highly valued in the French labor market, and demand for overall expertise in English (including communication and intercultural skills) has been rising continuously and steadily (Chancelade et al., 2016;Taillefer, 2007;Truchot, 2015). Despite the dominant role of English and the existing opportunities to learn it in the French education system -FL 4 learning is compulsory at the primary, secondary, and tertiary levels -English proficiency outcomes remain disappointing. The European Survey on Language Competences (European Commission, 2012), a large-scale comparison of the English proficiency of pupils finishing formal secondary education (average age=16), revealed that only 14% of French students reach the threshold independent user level or better -that is, B1 or above on the Common European Framework of Reference for languages (CEFR). This actually means that 86% of French students who could potentially pursue HE studies are non-proficient speakers of English. Other studies have corroborated this finding, which (even more unfortunately) seems to have held steady since the Bologna Declaration in 1999 (see for example Bonnet, 2004;Hilton, 2002Hilton, , 2003Manoïlov, 2019;Terrier, 2011;Zoghlami, 2015). Most recently, according to the 2018 edition of the EF English Proficiency Index -which ranks adult English proficiency in 88 countries and regions all over the world -France placed 35th with a score of 55.49, indicating moderate proficiency 5 .
Researchers have tried to explain the low English proficiency of the French. The ESLC report (European Commission, 2012) identifies national protection of the French language, the lack of English exposure in everyday life, and the ineffectiveness of teaching approaches in developing communication skills in English as major challenges facing the French education system. In addition, despite revised and increasingly internationalized curricula, no clear language policy has been committed to, let alone a research-based one. Taillefer (2007) described language learning and teaching as being "paradoxical on all levels of the educational system" (p. 137), with no connection between secondary and HE. In particular, English training for non-language majors 6 in universities seems to be sorely lacking institutional structuring. Beyond the shortage of human and material resources as well as the absence of clearly defined learning outcomes and research-grounded teaching practices, English courses are generally poorly integrated into curricula and fall short of meeting the needs of the targeted public, employability needs included (Braud, Millot, Sarré, & Wozniak, 2015;Brudermann, Mattioli, Roussel, & Sarré, 2016;SAES, 2011;Taillefer, 2007). Braud and her colleagues (2015) take this line of argument further and speak of "improvisation" (p. 59) with regard to language programs and pedagogical measures since the teachers who are generally asked to give such courses lack training in what English for Specific Purposes (ESP) involves.
In HE contexts, efforts have been made to address these shortcomings. These efforts, however, are institution specific. In the Cnam, specific measures have been taken to internationalize the curriculum and overcome discrepancies between language (particularly English as an FL -ELF) programs and the labor market while at the same time responding to massification concerns. In this chapter, I report on the findings of a large-scale Needs Analysis (NA) performed to uncover the English communicative needs of Cnam adult learners and thus inform the design of a task-based Professional English (PE) syllabus.
6. In the French HE system, ESP courses offered to students specializing in disciplines other than languages are part of what is generally referred to as LANSAD (langues pour spécialistes d'autres disciplines).

Analyzing language needs
Although NA is still often overlooked by teaching professionals and curriculum designers (Chan, 2018;Iizuka, 2019), it is now well established that it is actually central to the design of Learning for Specific Purposes (LSP) programs that can bridge the gap between institutional learning and workplace requirements (Basturkmen, 2010;Brown, 2009Brown, , 2016Huhta, Vogt, Johnson, & Tulkki, 2013;Hutchinson & Waters, 1987). In a lifelong learning context, conducting a sound NA is of paramount importance, as it ensures the development of courses specifically tailored to meet the immediate and future English needs of practicing professionals. Our first concern is then to define what is meant by learner needs and the process of NA. It is beyond the scope of this chapter to review the diverse definitions and classifications existing in the literature (see for example the seminal works of Brown, 2016;Hutchinson & Waters, 1987;Richards, 2001). For the purpose of this study, I adopted a straightforward working definition of NA, which does not merely deal with the identification of the language forms to be mastered, but also takes into consideration a range of other factors, including (1) the learners, their actual competencies in English, and their perceptions of their aims for English learning; (2) the reality of the teaching context; and (3) the target workplace situation and the type of work tasks performed in English. Basturkmen (2010) stated that in NA, "the language and skills that the learners will use in their target professional or vocational workplace […] are identified and considered in relation to the present state of knowledge of the learners, their perceptions of their needs and the practical possibilities and constraints of the teaching context. The information obtained from this process is used in determining and refining the content and method of the ESP course" (p. 19).
Language needs can be probed from the perspectives of learners, teachers, or professionals from the targeted fields. A number of authors have stressed the subjective nature of NA as being a process dependent on learners' (Richards, 2001, p. 54) as well as teachers' (Hyland, 2009, p. 113) interests, values, and beliefs about what workplace needs can be but also about teaching, learning, and language. For example, Lam, Cheng, and Kong (2014) surveyed resources tailormade by government bodies and commercial publishers for a special module on learning English through business workplace communication introduced in the senior secondary English language curriculum in Hong Kong. Discrepancieswhich could have been avoided if a thorough NA had been conducted -were found regarding the most frequent spoken and written professional genres covered in both settings 7 . This state of affairs is of course neither specific to the teaching of English for professional purposes nor limited to this Asian context (Iizuka, 2019;Martin & Adrada-Rafael, 2017;Taillefer, 2014). Long (2015, pp. 147-149) also reported on several studies that identified notable differences between the type of language used in targeted situations and the language modeled for those situations in commercial teaching materials -language that is oversimplified, inauthentic, and presented in unrealistic situations.
Methodological rigor must then be observed to increase NA's reliability and validity. Pertinent guidelines on how to conduct sound NAs -particularly survey use -have been proposed in the literature (Brown, 2016;Huhta et al., 2013;Long, 2005Long, , 2015Richards, 2001;Serafini, Lake, & Long, 2015). For example, Brown (2016) provided a detailed account of questionnaire design and other qualitative methods of data collection, including interviews, observations, and focus groups. He also discussed ways of analyzing and reporting NA results.
In their comprehensive survey of the design, methods, and procedures reported in ESP NAs conducted over the past three decades, Serafini and her colleagues (2015) outlined several methodological inconsistencies, mainly in relation to interactions between the sources and the methods used to collect data and interpret findings. The review enabled the authors to offer a set of practical recommendations -an adaptable methodological checklist (p. 25) -for careful NA practice, emphasizing in particular the importance of methodological triangulation (i.e. employing several sources and methods to study the same 7. For example, there was an overemphasis on phone calls and complaint-related genres in the teaching materials, whereas formal meetings and emailing were the top genres in this globalized workplace context (Lam et al., 2014, pp. 72-73).
phenomenon from different perspectives) as well as the contribution of a taskbased approach to language NA.

2.2.
Motivating the use of a task-based approach to NA The adequacy of adopting a task-based approach to NA -and thus to LSP courses -is now empirically established (Long, 2005(Long, , 2015Serafini et al., 2015), though it is still not widely implemented, particularly in French HE institutions. Using the task as the unit of lesson organization is meaningful, and hence motivating, to professionals. Carrying out a work task (e.g. writing a business report, responding to a customer complaint) has work-related goals that call on adequate language. In addition, the basic tenets of Task-Based Language Teaching (TBLT) are consistent with second language acquisition theory and psycholinguistic research findings (Long, 2015;Nunan, 2004;Robinson, 2011), an essential grounding in theory and research that linguistic NAs have failed to account for. In fact, traditional linguistic approaches rarely go beyond the text level, and tend to produce lists of decontextualized unitstypically grammatical, lexical, notional-functional, or a combination of these -for learners to master. The task-based NA, however, acknowledges language learning as a complex sociocognitive process with a focus on meaningful units, through the identification of the different types of communication tasks that specific communities of learners need to perform in the real world in the target language (Long, 2005(Long, , 2015. Recently, a meta-analysis of TBLT programs and their long-term effect on FL learning demonstrated an overall positive and strong effect (d=0.93) of TBLT as opposed to more traditional pedagogies (Bryfonski & McKay, 2017). The meta-analysis also revealed that stakeholders held positive views regarding such programs.
Although most studies in the LSP literature have investigated English needs for business purposes, recent task-based NAs have thoroughly examined the needs of learners in other professional contexts. For example, responding to a growing demand for courses of Spanish for specific purposes in American universities, Martin and Adrada-Rafael (2017) conducted a robust multiphase NA to identify the tasks that business professionals have to perform in Spanish, as well as their perceived frequency and difficulty. One of the distinctive features of this study was the numerous sources of information and the genuine interaction of sources and methods. The researchers interviewed both business insiders and outsiders including graduates, professors, researchers, and experts (qualitative phase) to ensure that only tasks really carried out in business settings would be included in the questionnaire administered on a larger scale to students and business professionals (quantitative phase). In the curriculum design phase, the authors proposed regrouping the most frequent tasks identified and classifying them into five more superordinate task categories that constituted the course objectives. The tasks and their corresponding objectives were organized by modality in accordance with the five C-goal areas promoted in the ACTFL's World-Readiness Standards (communication, cultures, connections, comparisons, and communities). In a similar vein, Malicka, Gilabert Guerrero, and Norris (2017) relied on in-depth qualitative data obtained from observations in the workplace and semistructured interviews to explore English-use needs in the professional domain of hotel receptionists in Spain. The study made a major contribution as it focused on how NA results can be meaningfully applied to the design of genuinely relevant pedagogical activities, and more specifically, how data about task difficulty can assist in designing tasks that vary in levels of cognitive load, thus providing insights on the importance of task complexity and sequencing in a language curriculum.
In France, the paucity of NAs is striking. A search of the published literature revealed very few studies which investigated real-world workplace language needs, with English being -unsurprisingly -the one FL systematically considered in all these studies (Braud, 2008;Taillefer, 2004Taillefer, , 2007Wozniak, 2010). For instance, using data obtained from expert and novice guides, Wozniak (2010) assessed the language needs of French mountain guides.
A key finding in her study was that oral communication skills represented the most important to improve -rather than knowledge of English related to technical skills. Taillefer's (2004Taillefer's ( , 2007 projects were institutional as they took place at the University of Toulouse. Already at that time, the author drew attention to the alarming situation of FL training in France. With the objective of encouraging more coherent English training in HE contexts, as is the case in the present study, her large-scale NAs tapped into the professional needs of economics students. The needs were investigated via questionnaires administered to language teachers, economics teachers, undergraduates, and recent graduates. Overall, the researcher highlighted the learners' feeling of being linguistically ill prepared for workplace demands. She noted the mismatch between what is taught at university and what are perceived as professional requirements, in particular with regard to the frequency of FL use, the degree of importance of the four different language skills, and the level of competence necessary in each as expressed in the CEFR scale 8 . As an example, unlike economics graduates, language teachers underestimated target productive and receptive levels of English proficiency, minimizing the importance of written communication in business workplaces (Taillefer, 2007). Accordingly, the researcher provided practical recommendations for language training in the economics sector, which included guidance on raising university and professional stakeholders' awareness of the importance of NA, taking into consideration the specificity of each context, and adopting an interdisciplinary approach to institutional curriculum design by integrating disciplinary and language components.
Only one empirical study -which also took place at the University of Toulouse -explored professional needs within a task-oriented approach (Joulia, 2014). Like Taillefer (2004Taillefer ( , 2007, Joulia (2014) underscored the necessity of preparing learners to face the workplace language challenge and advocated for the adoption of a professionalizing approach. To this end, the researcher first probed into the English needs of students in computer science through the observation of programming courses and the use of a questionnaire sent to regional companies hiring programmers, which often recruited the University of Toulouse's computer science undergraduates for internships. The assessment revealed that high-level proficiency in reading comprehension was the most important need. Learners encountered several reading difficulties and applied inappropriate strategies to overcome them (e.g. word-by-word reading). This finding paved 8. It is noteworthy to mention here that the actual CEFR levels (A1, A2, B1, B2, C1, C2) were not given in the questionnaires. The researcher argued that when the studies took place between 2002 and 2004, the reference system developed in 2001 was hardly known by most language teachers and not known at all by the general public.
the way for a reading experiment in which the researcher tested the efficiency of online resources in helping learners read and understand authentic technical documentation in English and ultimately enabling them to write lines of code -as they would actually do in their programming sessions and in real work settings. This professionalizing experiment, though not conclusive as far as the role of the chosen resources in assisting reading comprehension is concerned, proved to be highly motivating for learners, as they appreciated the authenticity of the reading content and task.
In light of the above theoretical background highlighting the relevance of the task unit in professionalizing approaches to language teaching, and given that very little attention has been paid to the role of NA in course design in French HE contexts, the present study addressed these issues by exploring the English needs of learners enrolled in the Cnam. The study particularly sought to answer the below research questions.
• What are the typical tasks French adult professionals (specializing in different fields) need to perform in English at work?
• Is there a difference in learners' perception of these tasks across levels of English proficiency (A2, B1, B2) and learning modes (self-directed, blended, face-to-face)?

Context
This NA study was carried out at Cnam Paris in the first semester of 2019-2020. The researcher holds an assistant professor position in the languages 9 department, and therefore benefited from direct access to all informants. The study is actually part of a larger action research project undertaken by the 9. The Communication en Langues Étrangères department also offers training in languages other than English including French, Arabic, Russian, and Sign Language. English classes, however, have the highest rates of enrollment. department to internationalize the English curriculum while responding to massification concerns, including the implementation of blended EFL courses on a large scale. It is also worth mentioning that two methodologically distinct EFL training programs are offered. The first one is mainstream in that it involves regular group courses. The second one seeks to promote language learner autonomy via a particular self-directed learning program, which combines autonomous learning guided periodically by an English teacher taking on the role of an adviser during counseling sessions, along with oral practice in groups with a native speaker.
The study is mainly exploratory in nature, as its overall purpose is to expand our understanding of the English-use needs of professionals. It is also action oriented in that it ultimately seeks to inform the design of an English curriculum and illuminate course content. The very particular lifelong learning context of the study guided the multimethod research design adopted, including the combination of both quantitative measures (questionnaires) and qualitative measures (open-ended questions and interviews) used to gather data from EFL adult learners as well as teachers. This methodological triangulation was meant to enhance the overall validity of the NA. Prior to the study, all the respondents were informed about its objective and assured that data would be used exclusively for research and teaching purposes.

Participants
The present NA involved the collaboration and input of a large number of adult learners enrolled in different non-language programs for the academic year 2019-2020 as well as English teachers from the Cnam languages department. From a total of 564 learners who signed up for either of the EFL programs, 242 (45.5% male and 54.5% female) took part in the study by responding to the learner NA survey (supplementary materials, Appendix 1). The great majority of the respondents (202 out of 242) were French speakers. As for the minority (n=40) who reported having a different mother tongue, Arabic and Berber were the most mentioned languages (22 and 12, respectively). The age range varied considerably, as shown in Figure 1, with the group 36-40 years old (20.5%) slightly outnumbering the other lower age groups. It is also interesting to note that nearly a quarter of our participants (n=54) were over age 40.

Figure 1. Age distribution of Cnam adult learners
The largest majority of the informants (71%) were employed, while 13% were either undertaking professional retraining or actively looking for a job 10 . As active professionals, our respondents can be considered domain insiders -the most valid data source given their expertise in domain contents and tasks (Long, 2005(Long, , 2015. Accordingly, data obtained from the sample is believed to derive from their accurate knowledge of their current or projected language needs rather than a mere perception of these needs. Regarding the participants' career fields, the answers revealed quite significant diversity. The domain of Information Technology (IT), telecommunications, and interactive digital media had the largest representation (17%), while the fields of accounting and audit (7.4%), management (7%), and business, marketing, and sales (6.6%) were less represented. This finding can probably be explained by the special place held by some of the (prestigious) Cnam institutions 11 -10. About 25% of the surveyed population reported that their return to HE was motivated by their desire to evolve their careers, including being promoted. Nearly 15% indicated personal growth as their main incentive. Another 15% referred to both career evolvement and personal growth, and for 16.4% the pursuit of HE was part of a professional retraining process.
11. Examples include the CNAM Engineering School (EICNAM) and the National Institute for Economic and Accounting Techniques (INTEC), which specializes in accounting, management control, auditing, and finance. specializing in engineering, IT and digital media, and accounting and auditing -which seem to attract more lifelong learners than other schools, and in which English training is compulsory for graduation.
Eleven teachers were also included as a valuable source of information about learners' language needs. The English teaching staff of the Cnam language department willingly participated in the study. While the teacher sample had considerable overall teaching experience (M=16), their particular experience in the Cnam ranged from 0 to 14 (M=5). Two of the instructors were also researchers.

EFL proficiency measure
English proficiency was used as a variable to see if any categorization of needs by FL linguistic level would potentially emerge from the data. In Cnam Paris, learners are placed in different groups according to their level in the targeted language. The placement measure used is the standardized CEFR self-assessment grid. The grid presents the different reference descriptors of receptive and productive proficiency in correspondence to the three broad levels of basic user (A1 and A2), independent user (B1 and B2), and proficient user (C1 and C2). The validity and reliability 12 of the grid are now confirmed (see for example North, 2007, for a discussion of the validity and consistency of the CEFR levels). This type of placement is of particular relevance given the heterogeneity of the learners' professional fields.

Questionnaires
Two questionnaires were developed in order to investigate the perspectives of the different stakeholders involved in our NA: a learners' survey (supplementary materials, Appendix 1) and an EFL teachers' questionnaire (supplementary 12. As a member of the Cnam English teaching staff, I actually had the opportunity to observe the consistency of the CEFR self-assessment grid as it yielded accurate learner grouping. materials, Appendix 2). To design the survey, I reviewed relevant research on language NA (e.g. Brown, 2016;Long, 2005Long, , 2015Serafini et al., 2015), and conducted informal interviews (a year before the main study) with both learners and teachers to bring out initial data about perceived needs and teaching practices that would inform item construction. Following key guidelines on questionnaire design (Dörnyei, 2003(Dörnyei, , 2007, the preliminary pool of targeted tasks was piloted with fellow researchers and with a convenience sample from the targeted learner population (N=20), who provided useful feedback on the wording of instructions as well as item readability, redundancy, and relevance. The final version of the survey contained 24 questions spread over five sections, and it took about 15 minutes to complete. In Part 1 (Q1-Q5), the personal profile of the learners is investigated in terms of age range, gender, professional status, major, and reasons for undertaking HE studies at the Cnam. In Part 2 (Q6-Q13), I explore their linguistic profile by looking at their native language, exposure to English, attitudes toward the English language and culture, perceived proficiency in the different language skills, and perceived difficulty of developing these skills. Part 3 (Q14-Q18) investigates the learners' English training. In Part 4 (Q19-Q20), the learners' perceptions of the importance of English for work are explored. Part 5 (Q21-Q24) first taps into the learners' perception of the importance of the target English tasks via a four-point importance scale (1=not important at all, 2=slightly important, 3=important, 4=very important). The tasks are organized under five categories pertaining to the five language skills (reading, writing, listening, and oral communication) along with relevant language elements. Perceptions about learner motivation and learning modes were also investigated via a sixpoint agreement scale (1=strongly disagree, 2=disagree, 3=slightly disagree, 4=slightly agree, 5=agree, 6=strongly agree) in hopes that these would provide further information that would be of benefit in designing blended courses. The survey ends with two open-ended questions meant to elicit qualitative data about perceptions of the most efficient EFL learning activities and English practice outside the classroom. It is noteworthy that I obtained a high reliability index for the final survey (α=.93) 13 , suggesting that the items work well together and that the survey should produce consistent answers if used in similar study situations.
13. A commonly accepted coefficient of reliability using Cronbach's alpha is .7 or higher.
The purpose of the teachers' questionnaire was to obtain a more comprehensive picture of the English needs of Cnam adult learners. Its design was then inspired by the learner survey, but it contained mainly open-ended questions. For example, and to obtain accurate and valid data, each teacher was asked to report on the most frequent needs mentioned by learners at a specific level of English proficiency. They were also asked to give their opinions on the most important elements a language course should include, the most successful activities, and the difficulties they and their learners often encounter. The last open-ended question required teachers to reflect on the appropriateness of a task-based course for adult professionals. The questionnaire was also piloted with two experienced English teachers and a researcher who provided comments on its content and layout. About 10 minutes were required to complete the teachers' questionnaire.

Follow-up interviews
The preliminary analysis of the teacher questionnaires yielded interesting responses that were worth further investigation. Unstructured follow-up interviews, lasting about 20 minutes, were conducted with volunteer teachers (N=3) to allow them to reflect retrospectively on some of their answers and provide clarifications and additional information. These revolved in particular around the type of language challenges learners encounter as well as their perception of the efficient ingredients for completing a task successfully.

Procedure and analysis
I administered the questionnaires to teachers and learners concurrently between October and November 2019. The EFL teaching staff was actually already well informed about the NA study and the larger research project in which it fits. As a fellow colleague, the researcher emailed the questionnaire to all the teachers (N=12). Emailing was deemed suitable for two main reasons. First, the majority of the teachers expressed their preference for this form of data collection. It also allowed them to contact the researcher if they needed further clarification. Second, the researcher needed to identify the teachers with whom follow-up interviews were to be conducted. Although teacher data was not anonymous, confidentiality was guaranteed to all participants. The teachers emailed back the filled questionnaires. The return rate was highly satisfactory, as only one teacher did not respond. It should be mentioned that the teacher questionnaire was administered in English to guarantee question comprehension, as some of the teachers were English native speakers. However, to further ensure the overall validity of the data obtained, teachers were given the choice of responding in either French or English. A call for voluntary interview participation was also emailed, and three teachers responded positively.
For validity and reliability purposes, the learner NA survey was administered in French a week before the language courses actually started. To ensure a large number of respondents, the survey was generated on Google Forms and distributed via the Cnam Moodle learning platform to which all Cnam learners have access. The survey was withdrawn when I reached a quantitatively acceptable return rate, as mentioned earlier (N=242).
Quantitative as well as qualitative analyses were undertaken. The survey quantitative data were analyzed using SPSS (Version 25). In particular, I computed the reliability index of the items (Cronbach's alpha) as well as descriptive statistics including frequency distributions, means, and modes. Pearson chi-square statistics (with a significance value of p≤.01) were performed to account for any differences in response frequencies between learner subgroups corresponding to different learning modes (self-directed learning, blended, face-to-face) and different English proficiency levels (A2, B1, B2). Teachers' perceptions were analyzed both quantitatively and qualitatively. Their answers to the open-ended questions were coded, allowing certain categories to emerge. These related mainly to learners' needs and difficulties, teachers' difficulties, and efficient language activities. A comparison with learners' answers was conducted, and illustrative teacher comments were provided when relevant.

Learners' English profile
In this section, I draw an English language proficiency profile for Cnam adult learners based on their answers to language biography questions included in Parts 2 and 3 of the survey and covering the following aspects: perceived English proficiency and exposure to English.
Two questions investigated learners' perception of English proficiency. They were first asked to indicate their degree of agreement with a common statement, "the French are bad at English" (Q9). Like other fellow colleagues (e.g. Taillefer, 2007), I have been witnessing this negative self-image 14 , and thus sought to investigate the extent of its consistency among the present particular population.
Results were quite striking. The mode index showed that the most frequent response (40.2%) was slightly agree. However, collapsing the scale into two meaningful categories -agree and disagree -reveals that the difference between the adult learners who agreed with the statement (59%) and those who disagreed with it (41%) is not as significant as expected. Interestingly, I also obtained quite a similar tendency from language teachers (N=11), with seven teachers agreeing (5 of them slightly) and four instructors disagreeing. Our learner results differ from those of Taillefer (2007), who asked the same question to French graduates in economics who were using English at work. Nearly 93% of her respondents believed the French were bad at English. In addition, the results of the chi-square test showed that our learners' beliefs about this poor self-image are independent from their own general English level (Q17; X 2 (30, 224)=41.42, p=.08) as well as their perceived proficiency in the different skills (Q12), sketched in Figure 2 below. It is tempting here to speculate that, if the answers I obtained can be considered a reflection of the important personality trait of self-esteem, which is essential to successful (cognitive) learning activity, the French seem to feel less insecure with regard to their English ability than they used to be. However, caution is due here as our results may be an artifact of the study sample, being confirmed adult professionals who may have gained self-confidence and selfesteem via their work experience. More research is then required in the French context to provide further insights into the issue of self-esteem in FL use.

Figure 2. Cnam learners' perceived English proficiency by skill
A quick look at the figure reveals quite similar perceptions with regard to the skills of writing and listening in English, with half of the sample believing their level in both skills to be low while the other half perceived it as good. It is also interesting to note the difference in the learners' perceived proficiency with regard to reading and speaking in English. Most of the respondents considered their reading ability to be good (n=168; 69%), whereas the relative majority believed their speaking ability to be low (n=134 extended speaking; n=120 spoken interaction). These results, however, need to be interpreted with caution as the majority of the surveyed learners (62%) are actually at the B1 level of the CEFR.
Learners' English exposure was explored from two angles: previous exposure in academic contexts (Q14) and everyday exposure (Q8). The findings indicate that prior to enrolling in the Cnam, 72% of the respondents pursued English courses at an HE level, while only 18.5% reported having pursued English courses up to the end of (French) high school. Interestingly, only a small number of learners seemed to have benefited from language and study stays (Q7; 6% and 3% respectively), where language contact was probably more regular and intensive. With regard to everyday exposure to English, it appears to be context dependent, as illustrated in Figure 3. Degree of exposure to English was first examined for three personal contexts. Inperson exposure is very limited, as only 46 informants reported regular contact with the language in real in-person communicative situations. However, exposure to English seems to be noticeably more frequent in virtual situations, particularly when using media and social networks (53% often or always; 33% sometimes). Caution should be observed here as the data do not allow speculation on the type of virtual exposure -that is, whether it involved productive or receptive skills. Cnam learners' exposure to English in professional contexts was also relatively limited. In fact, approximately half of the participants were either never (n=48) or rarely (n=59) exposed to English at work. Sixty-seven informants reported being exposed to English only sometimes while the remaining minority reported frequent English exposure. The lack of contact with English is consistent among the different learners' fields of employment (X 2 =112.2, p=.27). This finding is rather unexpected given the continuously growing context of globalization and internationalization of French companies since the 1990's, where FL proficiency -in particular English -is found to bolster employability, mobility, and competitiveness (Chancelade et al., 2016;Truchot, 2015).

Perceived importance of English
Overall, positive attitudes toward the English language and culture were expressed (Q11; M=3.18), and the majority of the informants viewed English as considerably useful (Q10; M=4.55), which is interesting in light of the results presented above. When asked to indicate the extent to which being able to communicate effectively in English is important (Q19), most of the respondents reported an equally significant relevance of this skill in both professional and personal contexts. In particular, learners perceived work-related English proficiency to be equally important for both oral and written communication. In fact, frequency results showed that about 86% of the informants perceived oral and written professional communication as important or very important. Given the learners' reported low speaking proficiency, I would have expected the perceived importance to be more marked for oral communicative proficiency.
Reporting on the importance of English as a career driver (Q20), about 20% of Cnam professionals indicated that their English ability was rather an obstacle for recruitment and professional growth alike. Interestingly, however, it seems that, for the majority of the respondents, being proficient in English was not actually a decisive factor in recruitment or professional growth purposes (42.6% and 47.1%, respectively). This finding aligns with another striking finding emerging from the data, namely the reasons reported for taking the English course at the Cnam (Q18). In fact, a minority of the respondents (12.7%) indicated solely a current and/or a future need for English at work. This result could partly explain the fact that the majority (78%) responded negatively when asked whether they had previously taken PE courses (Q15). Interestingly, among the minority of respondents who seemed to have benefited from such courses, only 5% reported that the course was offered by their company.
Nevertheless, analysis of motivation items (Q22) revealed that our participants seemed to enjoy learning English (M=4.96) and were extremely motivated to learn and improve in the language (M=5.38), including by undertaking extra language work outside the classroom (M=4.77) as confirmed by the numerous examples of activities they provided in answering the last question of the survey.

English needs: identifying target tasks
Before presenting the results of the target tasks to be performed in English at work and hence perceived as important by our adult learners, I first report on the teachers' thoughts about the relevance of a task-based English course in meeting the needs of such learners (Q12). Interestingly, most of the teachers seem to believe that a task-based course is appropriate for Cnam learners regardless of their level in English. They principally advocate that such an approach is purposeful and motivating: "Learners are motivated as it makes sense to them" (Teacher 1).
"I favour task-based courses … [they] tend to suit all levels … popular with students … [who] appeared to enjoy working on specific tasks that relate to their everyday professional lives" (Teacher 2).

"[It] makes learning meaningful and gratifying" (Teacher 3).
Some of the teachers (n=3), however, argued that the relevance of taskbased language courses depends on learners' proficiency and specific needs. For them, learners need to reach a certain level in the language to be able to perform tasks autonomously. In follow-up interviews with these teachers, they clarified that the mentioned specific needs are not work related but rather of a linguistic nature.
The typical work tasks French adults are required to perform in English were investigated quantitatively in Part 5 of the learner survey as well as qualitatively in Section 2 of the teachers' questionnaire. Table 1 displays the quantitative results, mainly the descriptive statistics per item for the whole learner sample (N=242).  * Tasks for which a significant difference was found in the reported importance frequencies across levels of proficiency (A2, B1, B2) and type of learning mode (self-directed, blended, face-to-face).
In general, Table 1 shows that the most frequent modes correspond to the scale importance ratings important (3) and very important (4). The relatively high mean scores (i.e. ≥2.75) also indicate that the respondents tended to rate the majority of the tasks as important. Given these figures, I report here only on the sum of the percentages of the important and very important responses in the rightmost column of the table. For convenience, the items have been translated from French in the table.
According to the overall learner perceptions reported in Table 1, and apart from learners' assigning roughly equal prominence to the examined tasks regardless of the communicative mode -as I actually expected given their answers to Question 19 -the results show that our questionnaire design procedure was highly satisfactory. The initial identification of learner English needs that allowed item construction and which was based on information obtained from learners and teachers seems to have targeted the most frequent tasks encountered by French professionals. Nevertheless, a few striking findings were observed.
The three lowest mean and percentage figures obtained for writing abstracts and/or scientific articles (17), writing on social networks (19), and taking care of foreign visitors (35) indicate that these tasks were perceived by the learners as the least important English tasks. Concerning task 35, only 57% of the respondents perceived it as quite important. The most frequent answer obtained for the writing tasks was slightly important (mode=2). The particularity of these tasks can explain the lowest importance scores obtained. In fact, writing abstracts and/ or scientific articles may be appealing principally to learners pursuing graduate education and hence have a limited target audience in the Cnam. Writing on social networks, on the other hand, is not work related and could be considered not demanding with respect to English proficiency. Table 1 also shows that of the 36 target tasks, 17 (presented in bold) seemed to be of paramount significance for the majority of Cnam learners, obtaining the highest mean scores with percentages higher than 80%. What can be noticed is that most of these tasks (11 out of 17) pertain to oral communication skills. For example, all four listening skills -Items 20, 21, 22, and 23 -were reported to be important and very important by respectively 95%, 92.1%, 90.9%, and 81.9% of the respondents. As for the remaining reportedly essential communicative tasks, these emphasize speaking abilities and relate to situations in which our professionals would need to converse informally and socialize, discuss work matters, communicate by phone, participate in meetings, make presentations, ask for and give information, and instruct, explain, and demonstrate. This result is actually in line with learners' views of the most efficient English learning activities (Q23) as the majority mentioned activities and tasks focusing on speaking practice for the development of oral fluency.
In addition, the analysis of teacher data revealed a similar trend. In fact, when asked to report on the most frequent English needs mentioned by Cnam learners (Q5), all 11 instructors cited developing speaking fluency and listening ability as the major reported needs. The teachers confirmed the necessity to develop oral-aural skills independently of learners' English level in their answers to the question of "which skills and/or language elements should an English course concentrate on" (Q6). Two teacher comments illustrate this finding: "'Writing' is not usually a skill they seem interested in developing: they often consider they have already spent a great deal of time writing English (and having been assessed on writing) at school. I mainly focus on speaking/listening activities" (Teacher 1).
"Ideally, spoken interaction in a face-to-face course, and the other skills via the Learning Management System (so the students can pick and choose depending on their needs, and take their time to work on the content and tasks provided)" (Teacher 4).
Interestingly, apart from language tasks per se, the majority of surveyed learners also reported the high relevance (M≈3.4) of other language needs -also bolded in Table 1 -including specific vocabulary (88.8%), general vocabulary (92.2%), grammar (93.4%), and pronunciation (92.1%). The high degree of importance given to pronunciation is striking, as most of the learners also agreed with the statement "I prefer having a native English speaker as teacher" (Q22, Item 4), with 32.2% strongly agreeing and 26% agreeing. Only one (non-native) teacher, however, seemed to insist on the role of pronunciation in her answer to Q6, commenting that the English course should concentrate on "spoken English with a strong stress on pronunciation".
Chi-square tests were carried out to answer the second research questionthat is, to identify potential differences in the frequencies of the reported task importance and which might depend on the general proficiency level (A2, B1, B2) or the type of learning mode (self-directed learning, blended, face-to-face).
The items for which a significant difference (p≤.01) was found are indicated by an asterisk in Table 1. One difference was found between the different level groups for task 31 (request and provide information/clarifications), with B1 learners placing more stress on its importance than the other level groups.
It is unclear to me why adult B1 learners would perceive this specific task as significant, as the type of tasks investigated here are professional rather than linguistic and proficiency oriented. Given the results on the perceived difficulties reported in the next paragraph, we could hypothesize that they perceived this oral task as more challenging than the others. Interestingly, the comparison revealed more significant differences between groups following different learning modes. These pertain mainly to writing (Items 13,14,15,and 16) and listening (Items 20 and 21) tasks, which seem to be less important for learners enrolled in self-directed learning programs. Other differences were found for the perceived importance of language certifications (X 2 =30.791, p=.000), grammar (X 2 =24.711, p=.003), and pronunciation (X 2 =26.02, p=.002), with a noticeably high degree of prominence expressed by learners taking face-to-face group lessons versus a moderate degree of importance for learners enrolled in the blended course.
In addition to identifying the most important target tasks, I sought to explore learners' and teachers' perceptions of the difficulties they might encounter while learning/teaching English. It was deemed that such information might help further classify the tasks as 'easy' or 'difficult'. I first asked learners and teachers to rate the difficulty of skills development in English (Q13 and Q7 in the Learner Survey and Teachers' Questionnaire respectively). Learners reported that speaking skills were the most difficult to develop (M=2.59 spoken extended production; M=2.66 spoken interaction). This is interesting considering that oral communication tasks were reported to be of extreme importance for the learners. Listening and writing skills were perceived as neither difficult nor easy (M=3.04 and M=3.06, respectively), whereas reading seemed to be the easiest skill for the learners (M=3.57). English teachers' perceptions differed from those of learners only with regard to spoken interaction. In fact, six instructors rated this skill as neither difficult nor easy, and three as rather difficult. Only one of the teachers perceived developing this skill as very difficult, adding that "spoken interaction also depends on interpersonal skills". Nevertheless, some caution has to be observed when interpreting these data given the small teacher sample (N=11) and the variation in learner English levels. It is surprising to me that FL listening is viewed as only moderately demanding, since research has shown that listening is the most anxiety-provoking and hardest skill to master for language learners regardless of their proficiency (Terrier, 2011;Zoghlami, 2015). It is hard to explain this result, but it might be related to our adults' urgent need to speak English at work, thus minimizing problems posed by other skills.
Other nonlinguistic problems that have to be taken into consideration in task design emerged from teachers' answers to Questions 8 and 9, which asked them to report on recurrent difficulties. Many teachers referred to learners' lack of time to study but more importantly to metacognitive aspects of learning, including learners' frustration and lack of self-confidence, in particular regarding improving speaking abilities. Below are some of their comments for illustration: "negative representations they have about themselves as learners (low levels of self-confidence, self-efficacy, and self-esteem) and about learning English in general (including the role of the teacher)" (Teacher 5).
"I find it hard to get them to talk. Many of them are literally traumatized by their secondary school teachers and are afraid of making mistakes" (Teacher 3).
"I found that B2 learners could be quite frustrated students. … [They] sometimes stated that they felt like they had reached a plateau. … They became irritated that their receptive skills were far greater than their speaking skills and got disheartened when they couldn't express themselves thoroughly" (Teacher 2).

Discussion: informing the PE curriculum
In view of improving the existing PE curriculum in the Cnam, I carried out a sound NA to identify the tasks French professionals need to perform in English in the workplace. I believe the methodological design adopted is one of the strengths of the present study. The qualitative method (i.e. informal unstructured interviews) used prior to the administration of the quantitative instrument (i.e. online survey) allowed for the emergence of relevant work-related English-use needs that I, as a domain outsider, might have overlooked. Following methodological recommendations in recent research (e.g. Malicka et al., 2017;Martin & Adrada-Rafael, 2017;Serafini et al., 2015), my aim was to avoid perpetuating the tendency of using a top-down approach to target task identification, an approach based on the researcher-teacher's own intuitions of what students need to learn, and which can be biased by what is offered in commercial resources.
In answer to the first research question, the present study sought to obtain a more comprehensive picture of the target tasks in light of the English profile I drew for the target learner audience as well as the perceived importance of English to them and the difficulties they expressed in the language. All of these elements should be taken into consideration when constructing pedagogical tasks to be integrated in a PE curriculum. As opposed to other French HE contexts (Braud et al., 2015;Brudermann et al., 2016;SAES, 2011;Taillefer, 2007), the similarities in perceptions of English needs and difficulties among the present NA stakeholders (learners and teachers) obviously reflect the effective articulation and interface between the ongoing language measures undertaken in the Cnam and professional life.
The study identified several tasks perceived as significantly important and pertaining to the four language skills. The findings also clearly demonstrate that the learners viewed oral communication tasks as the most important tasks for the workplace, thus confirming previous findings in the French literature -though exploring English needs for specific professional domains (Taillefer, 2007;Wozniak, 2010). The level of importance can be used to decide on the order of appearance of tasks in a curriculum. Accordingly, all the tasks presented in Table 1 could constitute units in a PE course -probably at the exception of the three tasks identified as the least important ones (Items 17, 19, and 35). Training on using English for the essential (bolded) tasks should be introduced first, with priority given to oral communication tasks.
Improving spoken fluency is a major, yet problematic, need. The findings reveal that the learners believed their speaking proficiency to be the lowest, which probably also explains the fact that they found this skill to be the most difficult skill to develop. This result is in line with previous findings on the actual limited French proficiency in speaking English as reviewed in the introduction (Chancelade et al., 2016;European Commission, 2012;Manoïlov, 2019). However, this study hopefully contributes to further explaining this poor level and the oral difficulty expressed. These may be related to the relatively negative self-image of the French as English users. The reported limited exposure to English in daily life, including in professional contexts, is also a key factor. A striking piece of evidence revealed in this study is that still only a minority (approximately one out of five adults) seems to be in frequent contact with English in the French workplace even in the present internationalized economic context. The lack of exposure undoubtedly jeopardizes the acquisition of English (general and specific) vocabulary -also revealed as an important need by the quantitative results. Most importantly, the data seem to indicate that most of the learners wanted to have training on English pronunciation and preferred a native speaker model, as they reported a preference for teachers who are native English speakers. This might mean that Cnam learners merely want an opportunity to be exposed to a native English accent in the classroom. However, it is also possible that they actually target native-like pronunciation when they speak English, which adds further challenges for the development of oral fluency. All of these factors are undeniably connected and need to be accounted for when designing and implementing the identified oral communicative tasks. Some work and awareness raising on the characteristics of spoken English could be injected to improve learners' aural-oral skills, probably during the pre-and post-stages of a task. It is vital, however, that such work be research based. I believe it is critically important to raise EFL teachers' awareness on current psycholinguistic L2 comprehension and production models 15 , which highlight the complex processes (cognitive, linguistic, and pragmatic) that make oral L2 communication possible and point to the necessity of automatizing L2 declarative knowledge for L2 aural-oral fluency. Several other studies could be useful for the EFL classroom as they outline the features of English connected speech that are particularly challenging for the French learner (e.g. Hilton, 2003, on the teaching of the spoken form of French-English cognates; Terrier, 2011, on the segmental and suprasegmental features of English speech; Grosbois, 2014, on the role of metalinguistic awareness in reducing the effects of phonetic and phonological nativization).
To answer the second research question, I explored the potential differences in learners' answers with regard to task importance across different levels of English proficiency (A2, B1, B2) and learning modes (self-directed learning, blended, face-to-face). As expected and with the exception of one task, as explained earlier, no significant differences were found across levels of proficiency, a quite reassuring result since the explored tasks investigated in this study were work oriented and required to be performed in English by Cnam adult learners regardless of their real level in the language. In designing the curriculum, I would need however to be mindful of the significant differences that emerged between groups enrolled in methodologically distinct learning modes. This is particularly important in this case as the Cnam language department is currently piloting a blended EFL course in order to deploy it massively in the upcoming years. For example, given the results of this NA on the importance of grammar and language certifications for the blended group, work on these elements could be provided online (via the Moodle English platform being currently tested). Classroom time would then be dedicated to expanding on learners' knowledge of general and specific vocabulary in connection with the productive and receptive professional tasks being practiced -again with a focus on improving English oral communicative abilities. Telecollaboration sessions (pairing French learners with native English speakers) could also be considered in planning the blended course given the associated high level of motivation expressed by this group.

Conclusion and directions for future research
The present study attempted to fill the existing gap in the French NA literature by exploring the potential of a professionalizing task-based needs assessment for efficient English training. The findings and the discussion of the teachinglearning implications has shown that NA can certainly provide an accurate profile of a target learner community and reliable guidelines for task-based curricular planning. I also hope that the study has brought to light the necessity of conducting more theoretically driven research in the field of lifelong language learning in HE, another underdeveloped area, at least in France.
This study revealed a few interesting areas for future research. First, a possible weakness of the present NA is that I overlooked the potential cognitive and linguistic difficulties of the tasks themselves (Malicka et al., 2017). Learners' perceived difficulty could have produced more valid data on the complexifying factors and the order in which the pedagogical tasks would appear in the English course. Second, this study further points to the importance of metacognitive aspects -self-esteem and self-confidence -in language learning. It would be interesting to conduct further research to determine the exact nature of these factors as well as the potential teaching techniques that could be applied to raise professional adults' awareness of the impact such factors have on their learning, and ultimately help them overcome their negative self-image.
Every NA is context dependent. Constrained by a heterogeneous grouping for language courses, I have identified real-life tasks that can usually be relevant to adult professionals in different domains. Likewise, every group of learners has specific learning needs. Teachers who would like to investigate their groups' language needs might find it convenient to create and use a shorter version of the learner questionnaire I used for this study. In our particular case, the next step would be to design task-based syllabi per level of proficiency, taking into consideration the task characteristics outlined all through this chapter. To do so, I would further dive into learners' and teachers' reflections about the most efficient English activities. I believe detailed analysis of this qualitative data would undoubtedly enlighten course design -potentially the design of a database of ready-made language teaching units. Finally, great NA effectiveness would be achieved if the informed PE curriculum clearly stipulated the learning outcomes specifying what learners can do in English -another very often neglected dimension in language programs.