The explicit teaching of vocabulary in French L2: theoretical models, teaching practices and experimental itineraries

This contribution presents the results of a recent experimental research aimed at investigating the potential of explicit teaching of French L2 vocabulary in the university context of Langues pour Spécialistes d’Autres Disciplines (LANSAD) in Italy. After introducing the research background, the tools, methodologies and teaching materials are outlined. The qualitative and quantitative evaluation of the results obtained from an experimental test (ET) is then presented. The test consisted of four tâches concerning the identification and use of paraphrastic rules, the analysis of lexical relations through the use of LF and a written production on one of the specialised topics addressed during the course (with the use of LU and PU) in order to measure the students’ acquisition of metalessical knowledge, as well as their ability to spontaneously and correctly use PU in specialised communication. Finally, the results that emerged during the experimental course are outlined, with reference to the significance of the tools and methodologies tested and the promotion of teaching activities focusing on the explicit teaching of vocabulary even in areas such as LANSAD.


Introduction: theoretical premises and research aims
In recent years, the development of lexical competence in L2 learning processes has gained considerable importance in the field of linguistic research, fostering a multi-voiced dialogue between different fields of investigation, such as lexicology, terminology, educational linguistics, corpus linguistics and foreign language teaching. According to the descriptors of the Common European Framework of Reference for Languages (CEFRL), a B2 level learner should have "a good lexical repertoire related to his field and to many general topics" (CoE 2020: 143). He/she should also be able to "vary formulation to avoid excessive repetition [...]; understand and use a large proportion of the vocabulary related to his or her specialist field; [...] insert appropriate words/signs in most contexts in a fairly systematic way" (CoE 2020: 143). However, recent studies on lexical competence and its combinatorial properties (better defined as "phraseological competence" 1 ) have shown that learners at both intermediate and advanced levels have significant lexical deficits, especially with regard to the use of locutional and idiomatic phraseological units (PU) of low and/or medium frequency (Hamel et al. 2007;Tsedryk et al. 2019). In order to avoid linguistic errors, learners often resort to the mechanisms of substitution and repetition, relying on a very high frequency general lexicon (use of the hypernym → poisson, instead of the hyponym → requin, loup de mer) and making excessive use of simple locutions, sometimes generated by L1 linguistic interference (i.e. avec tout le coeur rather than de tout mon coeur; profit manqué instead of manque à gagner). In addition to the possible interferences with the mother tongue, further elements come into play in the process of learning the lexicon of an L2: the time needed to store new LU, the ability to reuse them spontaneously and correctly according to the various communicative contexts, motivation and, lastly, the use and experimentation of didactic tools, methodologies and materials produced ad hoc according to the learners' linguistic and educational needs. One need only think of the approaches adopted for the teaching of Français sur Objectifs Spécifiques (FOS) at university level (Raus et al. 2020) or of the methodologies employed in the area of "Langues pour Spécialistes D'Autres Disciplines" (LANSAD) (Chaplier et al. 2019). In these contexts, a purely thematic approach (Binon et al. 2004) is usually preferred for teaching vocabulary through activities based on the creation and mnemonic learning of lists of names, inventories and specialised glossaries taken from textbooks, paper and electronic dictionaries, thesauri, digital corpora, etc. The result of this approach is undoubtedly the expansion of encyclopaedic knowledge relating to a domain, but the development of knowledge and lexical skills associated with the function of LU within a specialised discourse is compromised. Therefore, there is the risk of moving from the specificity of a lexicon in its technical-scientific and professional use to the likely aseptic labelling of world objects in specialised communicative situations. According to the attentional model suggested by Picoche (1992) and the theoretical framework of Meaning-Text Theory (MTT) and Explanatory Combinatorial Dictionary (ECD) (Mel'čuk et al. 1995), the aforementioned problems can be solved, at least in part, through theoretical-applicative didactic paths centred on the explicit teaching of the lexicon. More specifically, this approach emphasises the explanation of theoretical notions, the description of lexical structures and reasoned exercises in reformulation and written production, with the help of two purely linguistic tools: lexical functions (LF) 2 and paraphrase rules.
In the light of the above-mentioned theoretical linguistic models, which have possible didactic implications, this contribution presents the results of a recent experimental research aimed at investigating the potential of the explicit teaching of French L2 vocabulary in the university context of LANSAD in Italy. More specifically, the experimentation took place in a French language course aimed at Italian-speaking students of Bachelor's Degree Programmes in Statistics and Business Management Sciences at the University of Naples Parthenope. The objective of the experimentation was dual: on the one hand, it fosteredthe development of basic lexical and metalexical skills through theoretical-applicative, reasoned and structured training of the main concepts of ECD and through reformulation, paraphrase andwritten production excercises; on the other hand, it taught students the effective use of dictionaries and other lexical resources, such as corpora or other lexical databases for the study of specialised lexicons related to the dominant disciplines of their study paths. After introducing the research background and the minimum language requirements ( § 2.1.), the tools, methodologies and teaching materials characterising the operational protocol of the theoretical-applicative training are outlined ( § 2.2The qualitative and quantitative evaluation of the results emerging from an experimental test (ET) is then presented ( § 2.3.). The test consisted of four tâches concerning the identification and use of the paraphrastic rules, the analysis of paradigmatic lexical relations and collocations through the use of LF, and a written production on one of the specialized macro-themes tackled during the course (with the compulsory use of the LU and PU studied), in order to measure the students' acquisition of metalexical knowledge, as well as their ability to spontaneously and correctly use PU in specialized communicative contexts. Lastly, some of the main themes dealt with during the experimental course are outlined, with particular reference to the significance of the tools and methodologies tested and the promotion of teaching activities focused on the explicit teaching of vocabulary even in areas such as LANSAD.

The context: target group and language requirements
The experimentation was carried out during the second semester of the academic year 2021/2022, as part of the course in French Language Skills. The course was divided into two modules of 36 hours each for a total of 72 hours and aimed at Italian-speaking students of the Bachelor's Degree Programmes in International Business Management and Statistics of the University of Naples Parthenope. These are two professional training courses that aim, on the one hand, to train specialists capable of governing the internationalisation processes of companies and marketing policies and, on the other, to qualify future statisticians and analysts in the management of big data in a corporate context, through mathematicalcomputational methods and new information technologies. Within the curriculum of the two above-mentioned degree programmes, particular attention is paid to the teaching of foreign languages, whose macro-objectives include the attainment of at least a B2+ level of the CEFRL and a methodological basis for learning the specialised languages of the various disciplines studied over the three-year period. With regard to the above-mentioned teaching programme, the contents and objectives were designed for a particularly numerous (n= 300 students in total, among which 196 attending lectures) and markedly heterogeneous class, as fair as their French language knowledge and skills are concerned. An analysis of the numerical data shows that 20% of the students hold a C1 level DALF language certificate; 55% hold the B2 level EsaBac and/or EsaBac Techno Italian-French upper secondary school dual diploma; 15% claim to hold a B1 level DELF certificate and, lastly, 10% claim to have studied French at lower and upper secondary school but do not hold any language certificate.

The operational protocol: methodologies, tools and teaching materials
As outlined above, the course was divided into two modules of 36 hours each. During the first module entitled "The Language System of the French Language", basic knowledge was provided on the phonetic, morphosyntactic and lexical aspects of a B2+ level of French according to the CERFL. The topics covered ranged from the revision of various grammatical aspects of the analysis of the main syntactic, lexical and phrasal structures, and the study of the major writing and specialised communication strategies representative of three professional sectors, including international trade, blue economy and environmental sustainability in the marine environment.
In the second module entitled "Lexicology, specialised lexicography and digital literacy", a detailed, albeit introductory, path was provided on the various word formation processes from a contrastive perspective (Italian/French), thus directing learners to the study of specialised languages as well as the traits differentiating them from the general language in their lexical, terminological, pragmatic and translation aspects, with the aid of digital tools such as corpora, online databases and term extractors. A series of introductory seminars aimed at learning the basic metalexical notions, such as lexeme, vocabulary, LU, PU, as well as their interconnection with terminological studies (L'Homme 2020; Bonadonna 2020) was also proposed. Referring to the actantial model proposed by Picoche and to the overall theoretical background of ECD, a clear examination of the main syntagmatic and paradigmatic lexical relations through the use of LF as well as of the concepts of semantic actant, collocation (base+collocate) and locution was presented. Through a brief summary of the various types of LF, a wide range of criteria to classify them (idiomaticity, non-compositionality, semanticconceptual transparency/opacity, polylexicality and fixity) was illustrated and, in addition, the difficulties faced by those studying them were outlined. Depending on the perspective one takes, it is possible to encounter terms that differ in the various languages and traditions of phraseological studies. An attempt has therefore been made to offer as schematic a view of the main PU as possible, based on the classification proposed by Mel'čuk (2013). With regard to the basic theoretical aspects concerning reformulation operations, the definitions of sense, semantic equivalence and paraphrase were commented upon, preferring the criteria of lexico-syntactic paraphrase and focusing especially on the lexical criteria of synonym, antonym and conversive substitution. Once the seminar training was completed, a practical activity was designed to link the aforementioned paraphrastic processes to the more general operations of text composition, in order to encourage students to use the vocabulary studied in their written productions. Starting from two short expository texts on two very topical subjects, such as the internationalisation and business organisation sector and the marine bioeconomy for the correct business management of marine protected areas, the students were asked to reformulate some parts and fill in the blanks, using appropriate collocations and/or phrases with the help of dictionaries and open access digital terminology databases (GDT, TERMIUM Plus, FranceTerme, TLFI, DES, IATE and LOTERRE). In addition, they were shown the importance and usefulness of lexical resources such as Spiderlex (RL-fr) and IdiomSearch, as well as some linguistic corpora (frTenTen20, TermiTH and ScienQuest) to identify PU through the use of the "co-occurrences" or "word combination" entry, as well as to analyse their frequency and linguistic properties from context. At this stage, ways of searching for and modelling lexical relations through the use of LF were defined. Particular interest was paid to certain paradigmatic relations (hypernymy, hyponymy, synonymy, antonymy, conversion, syntactic derivation and semantic derivation). As far as syntagmatic LF (or collocations) are concerned, one adjectival FL expressing intensity, two adverbial LF representing place and medium and, lastly, two verbal LF composed of a supporting verb and a realisation verb were selected 3 . The choice of the aforementioned LF is to be correlated to their potential didactic value in the class context under investigation, assuming that it would have been undoubtedly easier for learners to understand the above formalisms through examples relating to nouns, actions, agents and places related to the specialised domains of their interest. Although brief, this theoretical-applicative examination proved particularly useful from a didactic point of view: on the one hand, it contributed to the students' mastering of metalinguistic language; on the other hand, the paraphrasing and reformulation exercises enabled them to learn how to organise the syntactic structures of a sentence (becoming aware of their errors and self-correcting), to explore some of the paradigmatic and syntagmatic relations between several LU belonging to the lexicons studied and, lastly, to develop a good metalexical competence.

The experimental test (ET): activity design and results
In order for the learners to familiarise with both the theoretical and methodological concepts outlined so far, an experimental test (ET) lasting 4 hours (2+2) was proposed, consisting of four tests concerning the identification (ET1) and use (ET2) of the paraphrastic rules examined. The aim of the test was to verify the acquisition of lexical knowledge and skills; the analysis of paradigmatic lexical relations and collocations through the use of LF (ET3) and a written production (ET4) involving the use of the LU and PU studied, with the aim of proving their correct use in written text production. The first test included an exercise consisting in the identification of the rules of paraphrasing (syntactic and/or lexico-syntactic) in four pairs of sentences, which were given a maximum point of 4 4 . By way of example, for the pair of sentences (1a. Les températures de l'eau de mer montent considérablement; 1b. Les températures de l'eau de mer connaissent une forte hausse), the paraphrase is based on the equivalence between the verb monter and the equivalent lexico-syntactic structure involving the noun of monter, i.e. its nominalisation -S0(monter)=hausse -preceded by the corresponding supporting verb: connaître. In this case, it is a lexico-syntactic paraphrase since there is not only a lexical substitution, but also a modification of the syntactic construction of the sentence. In contrast, in the pair of sentences (2a. L'économie bleue garantit la sécurite alimentaire sur le marché mondial; 2b. La sécurité alimentaire sur le marché Mondial est garantie par l'économie bleue), only a syntactic transformation from active to passive sentence was made, without any lexical change. Students were then asked to rephrase six sentences of one point each, using the criteria of synonymous, antonymic and conversive substitution (ET2) 5 . The results of the ET1 total averages show that the concept of paraphrasing was well understood by the majority of the students with a minimal percentage difference between correct identification of syntactic paraphrasing criteria (75%) and lexico-syntactic paraphrasing criteria (73%), while approximately 27% failed the test due to lack of adequate metalexical knowledge and skills. The scores obtained for ET2 confirm almost the same success rates as ET1: around 74% of the students were able to rephrase the sentences in the test, except for 27% of the participants who experienced problems. From a more detailed analysis of the data collected, one could hypothesise that these shortcomings can be traced back to elements such as non attendance of classes (most of the students, in fact, work), a lower language level than the other members of the class and difficulties in metalinguistic thinking starting from one's mother tongue. As regards usage, the total number of reformulations produced by the students is 19, among which 6 are synonymic substitutions, 3 antonymic substitutions, 2 collocations, 1 locution and 7 cases of non-total semantic equivalence which, instead of being evaluated as errors, we preferred to consider as quasi-synonyms. In the second part of ET, we focused on the recognition of paradigmatic lexical relations, collocations and/or locutions (ET3). After having submitted a list 6 of LU and PU to the students they were asked to identify the paradigmatic relations by means of LF and, in the case of LU composed of more than one element, to indicate the type of phraseme (between locution or collocation). We were not asked to carry out a detailed linguistic analysis, but rather to make a distinction between locution and collocation, illustrating, in summary, that -by way of example -PU payer quelqu'un en monnaie de singe is a locution because its meaning is not compositional; while chimie verte (AB) is a collocation because it is composed of the base chimie (A) freely chosen by the speaker for its meaning and the collocate verte (B) which expresses the meaning of the expression, i.e. a concept of chemistry that aims to promote sustainable approaches to chemical industry by eliminating the use of hazardous procedures and substances. Test ET3 was undoubtedly the most successful compared to all the others: 80% of the class was able to identify the lexical relations and classify the proposed PU, except for 20% of the participants who misinterpreted the above concepts in several examples and only answered correctly in four cases (économie bleue, chimie verte, pollution sonore, aire marine protégée). Lastly, the last activity consisted of a free production (ET4) on one of the macro-themes tackled during the course, with the necessary usage of LU and PU correctly, among those belonging to the specialized lexicosemantic fields studied in the class. The ET4 results were also particularly interesting. The average total score for written production was 78%, which corresponds more or less to the average of the ET3 results on metalexical knowledge. 17 out of the 30 points awarded for correct answers were for examples of LU and PU studied in class and reused, consisting mostly of collocations -adjectival type (Bon: énergie verte, chimie verte; économie bleue; AntiBon: marché noir), verbal type (Real: defender/sauver/preserver/détruire/polluer/nuire à l'environnement; Oper: pratiquer une politique soutenable, mener un audit, etc.) -and a few locutions (n=4) (col blanc, voie maritime, marché réglementé, agrobusiness mondial, obstacle commercial, etc.).

Conclusions
The ET results suggest that students have acquired good metalexical skills. More than 70% of the participants benefited from the explicit teaching of the lexicon and from the exercises to identify paraphrastic rules (ET1) and lexical relations with the help of LF (ET3), as well as from the resources made available, through which the learners learnt to search for information on the combinatorial properties of the lexicon. As far as the use of LU and PU is concerned, it was high in both tâches (ET2 and ET4): more than 70% of the correct answers contained examples of LU and PU. The percentages were on average higher in written production (ET4: 78%) than in paraphrasing (ET2: 74%). This result can be explained as a consequence of the fact that the reformulation exercise presented more constraints due to the search for synonymic expressions and patterns to follow, unlike ET4 whose only constraint was the use of LU and/or PU studied in class. With regard to collocation patterns, it was found that the "supporting verb+noun" pattern was more frequent in the rephrasing exercise, while the "noun+adjective" pattern was more frequently used in written production. As for the locutions, the percentage of use remained very low in both tests, a result that was expected for two reasons: due to their non-compositionality, learners found it difficult to translate and use them; moreover, more attention was given to collocations than to locutions in class. Even for the weaker learners and/or those with a scarce knowledge of lexical resources, cognitive effort was noted during the paraphrase and production exercises: for example, rephrasing or the use of compound collocations with inexact collocates caught our attention. The fact that a learner used an incorrect support verb highlights the need to teach broader syntactic structures in order to include the corresponding support verb in a noun phrase, especially if its use is limited to one or two verbs or well-defined syntactic patterns.
The theoretical and methodological reflections and empirical data outlined here allow us to state that the explicit and structured teaching of L2 vocabulary, combined with numerous reformulation and lexical network identification exercises, can contribute to the enhancement of the various lexico-syntactic constructions -and, more generally, to the development of metalexical competence -albeit with varying success rates. Although the choice of linguistic modelling tools tested may, at first sight, seem ambitious for a university context such as ours, we believe that even an audience of non-specialists in language sciences can benefit greatly from this approach. Promoting explicit and reasoned vocabulary training in French L2 aimed at specialists in other disciplines at university level appears, therefore, necessary both for the maintenance of the French language in the specialist fields and the quality of linguistic exchanges -oral and/or written -between technicians and experts in the field. However, it is necessary to ensure the creation of manuals and pedagogical-didactic tools for teaching/learning the lexicon in the LANSAD sector.