Harnessing CyberFrench for Specific Purposes

INTRODUCTION

In a place like Wollongong, French is very much a foreign language. Once students leave the classroom, there is little chance of casual encounter with the written or spoken language. It is unlikely that they will see French language magazines, newspapers or books in the shops, or hear neighbours or passers-by communicating in it. Ethnic television and radio have for some years now made news bulletins, feature films and documentaries available to the broader public - but it would indeed be a remarkable thing if anyone from a non-French speaking environment were to acquire fluency in the language through the type of exposure that these media allow.

For an Australian student, the richest readily accessible source of French is undoubtedly cyberspace. In theory it would be possible for a student to turn the computer on each morning and spend the entire day wandering around over the border in French-speaking Cyberland. In reality this does not happen - for a host of reasons. The following are some of the more conspicuous:

  • When all is said and done, the study of French represents only a small part of the average student's occupations and preoccupations. They are also studying other subjects, they eat, sleep, play, party, pursue hobbies, work part-time jobs...
  • Expression in the cyber world is, at least for the time being, relatively restricted, being far more print and graphics based than audio or video.
  • Most of what is available is in the form of data to be consulted or presentations to be absorbed, rather than genuinely communicative activity requiring users to ‘talk’ to real people in real time. After the initial fascination with the medium has dissipated, such reading often becomes about as exciting as flicking through the Encyclopaedia Britannica or the telephone directory.
  • There is a lot of ‘garbage’ out there, and often a lot of trial and error for users in locating worthwhile sources.
  • In some situations computer operations may be instantaneous. With the Internet, however, one tends to have flashes of the instantaneous interspersed with numerous and often unpredictable patches of tedium associated with navigating, deciphering, sifting, typing, downloading, unanticipated ancillary software requirements, screen refreshing/redrawing, address changes, congestion/overload, or system crashes. The capacity of the time required to perform various computer-related activities to act as a disincentive for their use is widely recognised - see, for example, Debski (1997:58) and McEnery and Wilson (1997:9).
  • Even if all the French-language material in the world were instantaneously available, it would reach the student in the form of a mountain of linguistically haphazard and pedagogically unstructured data.
  • There is a financial cost associated with Web time, whether it was used productively or not, and frequently with access to in-depth information.

Effective teaching has always relied on teachers possessing and using a combination of their own mastery of the discipline or skill, their awareness of the broader picture, and their experience with a specialist kit of tools and techniques for unlocking knowledge and fostering competence. As with any other store of information, the potential of the Web as a resource in language teaching is only as great as the teacher's ability to harness it and to integrate it into the overall learning program. It is not hard to visualise wonderful outcomes from a class of students in Australia eagerly exchanging email messages with a class of native French speakers on the other side of the world, or to be inspired by the capacity for instruction of a series of assignments requiring students to draw independently and creatively on the vast, varied and ever-growing and evolving array of information sources available on the Internet. Neither is it difficult to picture serious educational casualties when, as the journalists' cliché puts it, things go “tragically wrong”. A class excursion into Cyberland can be as fruitful or as catastrophic as a field trip to France with an average group of adolescents.

The object of this paper is to demonstrate one way of imposing instructional organisation on the linguistic and operational chaos of cyberspace.

BUILDING THE LINGUISTIC CORPUS

In the context of a much broader project (see Appendix A: CALL Modules Developed at the University of Wollongong; and McCarthy, (1996)), four grammar revision/maintenance modules (Future Tense, Infinitive Government, Relative Pronouns, and Infinitives & Participles) were constructed around linguistic data gleaned from the Web1. The use of Web resources for some grammar points and not for others is not arbitrary. It stems from both pedagogical and practical considerations.

Topics such as Present Tense Verbs, Articles, and Adjectives are so fundamental to any sort of comprehension or expression that they must be covered in the very early stages of any French course. Common sense therefore usually directs teachers to optimise students' chances of mastering them by making any activity associated with them as transparent as possible in all respects, and particularly with regard to the accompanying vocabulary, syntax and grammar.

The four grammar topics mentioned above, on the other hand, all have in common one or more of the following features:

  • They fall into the “intermediate” rather than the “beginner” range of the language-learning scale.
  • Conceptually they are not particularly difficult, so they are commonly “taught” by means of an explanation, a few examples, and perhaps an exercise or two. Subsequent recognition, re-use, systematisation and refinement are left up to the learner.
  • It is not easy to find extensive sets of exercises on them.
  • They are relatively sparsely represented in everyday discourse (compared, for example, to articles or present tense verbs), so opportunities for mastery through chance encounter are correspondingly reduced.
  • They are frequently couched in a linguistic environment that is sophisticated at the level of vocabulary, syntax or construction.

It is more daunting than encouraging for any not-particularly-computer-passionate language teacher to begin the quest for data by reading search-engine claims of the type:

  • ALTAVISTA YELLOW PAGES: Alta Vista gives you access to the largest Web index: 30 million pages found on 275,600 servers, and four million articles from 14,000 Usenet news groups. It is accessed over 18 million times per weekday.
  • ANZWERS: ANZWERS is designed for Australian & New Zealand users, it provides access to over 3.5 million Australian and New Zealand Web pages and 54 million Web pages world wide. It also includes a range of powerful easy to use tools to help you refine your search.

Ordering the confusion of French-language resources available on the net is a superhuman task - not the least so because they are invisible until searched out and summoned up from their corner of cyberdarkness. And maintaining order in any register one does manage to establish is akin to co-ordinating the telephone directory of a large city single-handed, and with no system of notification from the ‘clients’ whose particulars you are endeavouring to keep track of.

One may not understand what motivates colleagues like Bob Peckham (better known as “Tennessee Bob”) and David Gatwood of the Department of Modern Foreign Languages, University of Tennessee at Martin, to create and maintain “Tennessee Bob's Famous French Links” (http://www.utm.edu/departments/french/french.html) which, in its second edition, has evolved into a “5,500-link Globe-Gate Supersite” - but their efforts must have earned them the profound gratitude of thousands of French teachers world-wide.

These listings contain 9 divisions: Finding New Francophone Sites; Books and Literature; Art, Music, Film and General Culture; History of France and the French-Speaking World; Virtual Francophone Tourism; The French Language; Press, Radio/TV, Telephone; Education in French-Speaking Schools; French across the Curriculum and in Every-day Life. In the Press, Radio/TV, Telephone category, for example, there are some 126 magazines and 45 newspapers. This alone represents an overwhelming volume of information. Decisions therefore have to be made regarding the specific sources from which data will be gathered. For this project we selected a relatively restricted set, including Agence France Presse, Radio France Internationale, Label France, Sympatico Nouvelles (Canada), Nice Matin, Les Dernières Nouvelles d'Alsace, La Une du <Monde>, France 2, France 3, TV5 (Québec).

Factors influencing this selection were:

  • Homogeneity. Although the material assembled contains a great diversity of styles, there is nevertheless a homogeneity of ‘inspiration’ that does much to unify the banks of thousands of sentences drawn from them for each module, selected only because they contained a particular feature of grammar, and with little consideration for the subject matter. Although exercises contain only semantically free-standing sentences, the genre consistency among items makes it easier for students to remain focused than if they were randomly and constantly jumping between excerpts from a wide array of sources (literature, philosophy, adolescent chitchat, specialist documents (legal, scientific, technical, economic, etc), e-mail messages, business correspondence, advertising, personal letters, etc)
  • Cultural content. Any collection of news and current affairs items creates a rich cultural backdrop containing a blend of social and political events that are of abiding significance because they had an impact on national or international affairs, and of everyday matters (sport, community, education, politics, crime, transport, religion, agriculture, town planning, industry, youth, lifestyle, human interest...) that are of perennial relevance - the names and places may vary, but the themes, newsworthiness and intrinsic cultural relevance are constant.
  • Risk reduction. To limit oneself to a single source is risky. Sources can go dead, or change address and become untraceable, or convert to a user-pay system. Individually, the styles or themes may turn out to be too repetitive or restricted. Relay or download times may make daily collection from some sources over an extended period quite impractical.

It took over three years of monkish devotion (between half an hour and one hour per day) to assemble the 9 million word corpus on which the four grammar activities are built. In addition to the familiar speed-humps to be encountered on the information superhighway (enumerated earlier), all data, once gathered, had to be reformatted to remove unwanted spaces and paragraph returns. Individual sentences were extracted from this corpus and organised for presentation to students in different ways for each of the modules.

THE 4 MODULES

Future Tense

By far the richest concentration of future tense forms was to be found in daily horoscopes2 - although second-person forms were far more common than third-person forms in this context, and first-person forms were virtually non-existent. Snippets from articles in which the journalist chose to report in the form of the subject's own words provided the bulk of the first-person items.

In order to ensure comprehensive coverage of the point, verbs were divided into 18 categories: avoir; être; aller; faire; venir, tenir and compounds; devoir; pouvoir; savoir; falloir, vouloir, devoir; regular -er; regular -ir; regular -re; semi-regular verbs of the mener or jeter types; semi-regular verbs of the é+consonant+er and -yer groups; voir and envoyer; reflexive verbs; predictable irregular verbs; cueillir, recevoir, courir, asseoir and their compounds. There is a minimum of 20 items for each person (1st, 2nd and 3rd persons singular and plural) in each category. The computer then draws on the total pool of over 4,500 items to compile 30-item tests, taking a designated number of items randomly from each of the categories. Where there are two items or more from any category, they will differ in person and/or number. The module also contains on-line grammar notes and operational help, it displays a progress score, provides feedback for incorrect responses, and stores all incorrect responses plus corrections in a field which can by studied by the user on completion of the activity.

The interface between the student and the underlying heavily-structured pool of linguistic data is a simple work screen.

Fig 1. Workscreen for Future Tense module

Infinitive Government

In French, some verbs link up directly with a following infinitive (e.g. Claude veut parler. - Claude wants to talk.), some require ‘à’ before the infinitive (e.g. Claude apprend à parler. - Claude is learning to talk.) and some require ‘de’ (Claude refuse de parler. - Claude refuses to talk.). Traditionally, the process of learning which verbs use which construction is begun by having students memorise basic lists for each category, with mastery being acquired in much the same way as it is for native speakers - i.e. by having sufficient exposure to the spoken and written word, and sufficient occasion to reuse the constructions in their own expression, that ultimately a form is used because it “sounds right”. The problem for genuine foreign language students is that this exposure is long-term and diluted. This module compensates for that situation by distilling more than 5,000 sentences from the full corpus: over 30 verbs from each of the 3 categories (as well as a small number of verbs that follow more than one pattern), and, with a few exceptions, over 20 examples for each verb. A selection of screen dumps is shown in Fig 2 below.

The computer draws random items in designated ‘doses’ from the different wells to compile banks of 40 questions. In the text of each item, the symbol [~~~] is inserted between the verb and the following infinitive, and students click on the form required (à, de, d', or “no preposition”) to complete the construction in the sentence displayed. For any incorrect answers, the correct alternative is displayed. This is followed by the full sentence incorporating the correct construction, and an “Additional Information” field containing the following information:

  • a further indication of whether the verb takes à, de (d') or no preposition before a following infinitive;
  • a list of the various English meanings conveyed by the verb;
  • a model sentence for the verb, accompanied by the English translation of that sentence;
  • For verbs with which more than one construction is possible, explanations of the various forms and meanings.

Fig 2. Screen dumps from the Infinitive Government module

As with the Future Tense module, this module also contains on-line grammar notes and operational help, displays a progress score, and stores all incorrect responses plus corrections in a field which can by studied by the user on completion of the activity.

Relative Pronouns

An interesting characteristic of relative pronouns is that it is quite possible to ‘survive' linguistically without them. Their function is to avoid the repetition of a noun when two ideas are telescoped into a single sentence. By definition, then, it is always possible to use two separate sentences to convey the message. On some occasions, for the sake of clarity, it is even desirable to avoid the use of relative clauses.

Any exercise requiring students to supply a missing relative pronoun serves a dual purpose:

  • It creates an opportunity for students to become familiar with language at a level more sophisticated than they might be inclined to use themselves, but ignorance of which could seriously restrict their comprehension of others' expression;

It requires them to have a clear understanding of who is doing what to whom in both of the ideas articulated by the missing pronoun.

Some relative pronouns can be used in quite simple constructions:

  • e.g. Quatre pompiers avaient été blessés au cours de l'incendie qui a partiellement détruit le Parlement de Bretagne dans la nuit du 4 au 5 février.

Others, however, are far more likely to occur in the articulation of rather complex ideas. For example:

  • Temps fort à Omaha Beach. La cérémonie internationale prévue cet apres-midi à Omaha Beach constitue le temps fort de cette journée au cours de laquelle le président Fran&ccdil;ois Mitterrand doit participer à plus d'une demi-douzaine de commémorations en compagnie notamment de Bill Clinton (Utah beach), de la reine Elisabeth II (Bayeux), du Premier ministre canadien Jean Chrétien (Courseulles) et du président polonais Lech Walesa (Urville-Langannerie).
  • or,
  • Derrière ces règlements de comptes saoudo-irakiens au Yemen, on retrouve les clivages de la guerre du Golfe. l'Egypte se range à côté des pétromonarchies pour soutenir Aden. La Jordanie préfère appuyer Sanaa. Si Riyad a choisi le Sud, dont le régime naguère ouvertement marxiste-léniniste fut longtemps son pire ennemi, c'est parce que le Nord lui semble à présent la menace principale.

There is, moreover, great disparity in the frequency of occurrence of the various forms.

In the light of the above considerations, it was important to organise and present the bank of approximately 2,000 “Press Clips” containing relative clauses in a pedagogically coherent way. A number of design criteria were therefore established.

  • Items were presented at two levels. Level one contained only the primary forms: qui, que and qu'. Level 2 contained the full range of relatives: qui, que, qu', dont, ou, auquel, à qui, and relatives in combination with a preposition.
  • In the case of relative constructions incorporating a preposition, students are required to do the linguistic equivalent of what is commonly referred to in mathematical problems as “showing their working”. They have firstly to identify the appropriate preposition from a set of 5 alternatives, then they must demonstrate recognition of antecedent by selecting the appropriate number/gender form from four alternatives (lequel, laquelle, lesquels, lesquelles for constructions containing a simple preposition such as ‘sur’ or ‘avec’; duquel, de laquelle, desquels, desquelles for compound prepositional forms such as ‘au bout de’ or ‘à partir de’).
  • A parallel set of simpler “Idea Link” exercises was created, of the type found in traditional textbooks (e.g. Le film que nous avons vu hier soir était formidable; Je dine souvent avec un vieux monsieur dont le fils est en Allemagne). Students are encouraged to work through these exercises as a lead-up to the ‘reality’ of the Press Clip items.
  • A comprehensive set of mini-tutorials was devised (20 in all), allowing students to reflect on the principles governing the use of each type of relative pronoun. (see Appendix C: Relative Pronouns: Sample self-help mini-tutorial).

Infinitives and Participles

The verb endings er, é, ée, és, ées, ai and ez all sound the same in French. In language programs with a heavy emphasis on listening and oral communication, there is a strong chance that students will remain confused about which ending to put where when they have to express themselves in writing. The most common sources of errors are failure to understand the grammatical concepts of ‘second-verb-infinitive’ and the uses of the past participle, and uncertainty about the rules of past participle agreement. There are very few naturally-occurring or teacher-contrived opportunities for learners to focus on accurate use of these verb endings whose sound provides such an inadequate cue to the form.

The corpus was used to compile a bank of 500 independent paragraphs containing over 1500 verb forms ending in the sound [e]. The computer moves through the passage, highlighting each of these verbs (in its infinitive form) in turn, and students are required to retype it in the appropriate form.

Sample passages:

Cue Sentences Required Responses
En Espagne, l'explosion d'un colis [piéger] dans un bureau de poste de Madrid a fait au moins un blessé grave. Le paquet a [exploser] alors qu'un employé de la poste le manipulait. Le bâtiment a été [évacuer] et des chiens policiers sont sur les lieux pour [tenter] de [trouver] d'autres bombes s'il y en a. Lundi dernier un policier a été [tuer] dans un attentat à la voiture [piéger] à Madrid. Une attaque [revendiquer] par le groupe séparatiste basque ETA. En Espagne, l'explosion d'un colis piégé dans un bureau de poste de Madrid a fait au moins un blessé grave. Le paquet a explosé alors qu'un employé de la poste le manipulait. Le bâtiment a été évacué et des chiens policiers sont sur les lieux pour tenter de trouver d'autres bombes s'il y en a. Lundi dernier un policier a été tué dans un attentat à la voiture piégée à Madrid. Une attaque revendiquée par le groupe séparatiste basque ETA.
Les négociateurs russes et tchétchènes ont [signer] un protocole d'accord sur le retour à la paix dans la république indépendantiste. Les forces tchétchènes seront [désarmer] et les troupes russes progressivement [désengager]. Le déblocage des pourparlers a été obtenu grâce à l'acceptation par les Tchétchènes de faire [arrêter] les membres du commando auteur de la prise d'otages de Boudenovsk. Les négociateurs russes et tchétchènes ont signé un protocole d'accord sur le retour à la paix dans la république indépendantiste. Les forces tchétchènes seront désarmées et les troupes russes progressivement désengagées. Le déblocage des pourparlers a été obtenu grâce à l'acceptation par les Tchétchènes de faire arrêter les membres du commando auteur de la prise d'otages de Boudenovsk.

Grammar notes, always available on the work screen, give students access to comprehensive sets of examples of the structures in which each of the different written endings can occur.

Conclusion

From a teaching perspective, the World Wide Web is a resource, in much the same way as the ideas in the head of a sage, or a library of books, or the collective experiences of a community of informants are a resource. And the computer is a tool - like a piece of chalk, or a cassette recorder, or a video camera. The value of the information to a learner in the context of institutional education is a function of the organisation imposed on it by the teacher and of the appropriate and efficient use of the media through which it is presented.

In reflecting on the task of harnessing the wealth of linguistic data available on the Web, two images come to mind (with apologies to any colleagues who might find imagery incompatible with the genre of an academic paper). The first is of the teacher as a parent bird who knows what food sources are appropriate for the nestling, and has the ability, the time and the opportunity to locate, gather, and often, to predigest it, before presenting it to his or her charges as a source of nourishment. The second is that of a knitter who, by dint of what has come to be seen an archetypal repetitive action, can produce an article that is both functional and of abiding value.

The four grammar review and maintenance activities whose elaboration has been the subject of this paper will represent only a minuscule component of the overall process of mastering French as a foreign language. And their perceived value will vary depending on the methodological inclinations of the teacher.

Several features, however, are incontestable:

  • It is highly unlikely that students would have the time, the inclination, the expertise or the resources to discover and master these language features through their own, unguided exploration of CyberFrench. Even a concordancing-based activity built around a teacher-supplied corpus could prove too time-consuming and inefficient for the average student in the light of the broader requirements of the average university foreign-language program.
  • The resulting activity has made use of the unique capabilities of the computer to produce a simple interface between the learner and a large body of relevant data organised in such a way as to provide comprehensive coverage of specific language features.
  • Students can return to any of the activities as many times as they want or need to, and be confronted by a fresh set of data on each occasion - data which is authentic, culturally relevant French.
  • The potential of the 9-million-word corpus established in the process to serve as a resource in the development of other learning materials is limited only by the imagination of the teacher.
  • The considerable investment of time, expertise and energy on the part or the teacher in producing the material is offset by the satisfaction of knowing that it was a one-off effort, and that the resource created is of abiding value for himself, for colleagues, and for future cohorts of foreign-language students.

Appendix A

CALL Modules Developed at the University of Wollongong


GRAMMAR MODULES

Present Tense Verbs Fr, It TM*
Adjectives Fr, It TM
Passé Composé Fr TM
Imperfect & Passé Composé Fr
Pronoun Objects Fr TM
Reported Speech Fr TM
Relative Pronouns Fr TM
Present Subjunctive Fr TM
Adverbs Fr TM
Articles Fr, It
Comparison Fr
Infinitives & Participles Fr TM
Passato Prossimo It TM
Imperfetto It TM
Verb Kit Fr TM
Infinitive Government Fr TM
Future Tense Fr TM

NOTIONAL/FUNCTIONAL MODULES

Dis bien bonjour! (Greeting & Leave-taking) Fr
Clockworks Fr
Date Fr
ID Kit (name, age, address, profession, nationality) Fr
Spelling Fr
Numbers Fr, It, Ja, Ge, Sp, Ind, Eng
Numbers/Alphabet Grid Games Fr, It, Ge, Sp
Basic Expressions Fr, It, Ge, Sp, Ja
Shopping Fr
Weather Fr

OTHER

Proverbs Fr, It, Eng
Bons Mots (French words/expressions in English) Fr/Eng
VocaBuilder Fr
Translation Traps Fr TM
Pot-Pourri Quiz Fr
Roadsigns Fr
ClozeMaker any roman alphabet language

* TM indicates the availability of sets of printed tests (Test Masters) for classroom use. These tests are identical in construction to the comprehensive revision tests for the respective modules - i.e. they are compiled by drawing random items from prescribed banks according to a formula ensuring that all aspects of the grammar point are covered in each test.

Appendix C: Relative Pronouns: Sample self-help mini-tutorial

step 1
step 2
step 3
step 4
step 5
step 6
step 7
step 8
step 9
step 10

Notes

  1. All materials were developed by the author in conjunction with Ray Stace (Centre for Educational Development and Interactive Resources, University of Wollongong). Without his expertise in software design and programming, sensitivity to the underlying linguistic and pedagogical principles, and commitment to the projects, none of the modules would exist.
  2. A by-product of the data-assembly process for this exercise was a set of some 4,000 one-line items of health advice (a section of which is reproduced below). Although of virtually no use for study of the future tense, it has considerable potential as a source of enlightenment on the health preoccupations of the French, and the limitations (style, imagination, legal) of the genre.

Santé: Borborygmes stomacaux.

Santé: Bourdonnements d'oreilles.

Santé: Bourdonnements dans les oreilles.

Santé: Brossez souvent vos dents.

Santé: Brulures d'estomac.

Santé: Brulures d'estomac.

Santé: Brulures d'estomac.

Santé: Brulures du tube digestif.

Santé: Brulures du tube digestif.

Santé: Brulures du tube digestif.

Santé: Buvez davantage de lait.

Santé: Buvez davantage de lait.

Santé: Buvez davantage de lait.

Santé: Buvez davantage et vous éliminerez.

Santé: Buvez des eaux minéralisées.

Santé: Buvez des eaux minéralisées.

Santé: Buvez du bon vin, mais sans excès.

Santé: Buvez en dehors des repas.

Santé: Buvez moins de café.

Santé: Buvez pour aider vos reins.

Santé: Buvez pour aider vos reins.

Santé: Buvez surtout de l'eau.

Santé: Buvez surtout entre les grands repas.

Santé: Buvez surtout entre les repas.

Santé: Buvez un verrre d'eau à jeun.

References

  • Debski, R., (1997). Support of creativity and collaboration in the language classroom: A new role for technology. In R. Debski, J. Gassin & M. Smith (Eds.), Language learning through social computing (pp. 41-65). Occasional Papers Number 16. Melbourne: Applied Linguistics Association of Australia.
  • McCarthy, B.N. (1995). Grammar drills: What CALL can and cannot do. On-CALL - The Australian Journal of Computers and Language Education, 9 (2), 30-41.
  • McCarthy, B.N. (1996). Fully integrated CALL: Mission accomplished. On-CALL - The Australian Journal of Computers and Language Education, 10(2), 15-28.
  • McEnery, T. and Wilson A. (1997). Teaching and language corpora (TALC), ReCALL 9(1), 5-14.