Kernerman Dictionary News • Number 11 • July 2003

Some Lexicographic Concepts Stemming from a
French Training in Lexicology

Jean Pruvost

Jean Provost is the Vice-President of Université de Cergy-Pontoise, where he teaches linguistics, lexicology and lexicography, directs the laboratory of the CNRS [National Centre for Sceintific Research] (Metadif, UMR 8127) that is devoted to dictionaries and their history, and organizes the annual international conference, La Journée des dictionnaires, a meeting point for lexicologists, lexicographers and dictionary-makers. He is the author of Dictionnaires et nouvelles technologies [Dictionaries and New Technologies], for which he received the international linguistics prize Logos in 2000, and of two titles in the Que sais-je? [What Do I Know?] series, one on French language dictionaries, the other on neologisms (with Jean-François Sablayrolles), as well as of numerous articles. Professor Pruvost co-directs, along with Bernard Quemada, two collections at Éditions Honoré Champion (Études de lexicologie, lexicographie et dictionnairique; Lexica), serves on several editorial boards (Cahiers de lexicologie; Études de linguistique appliquée; International Journal of Lexicography), and presides the Association des Sciences du Langage.

Sections 2 and 3 of this contribution will be published in issue Number 12, July 2004.

All general references to lexicographers and users using he/him/his, apply to both male and female persons.

This article has been translated from the French language. The original French manuscript is available online:

I am most obliged to Anthony Cowie for his dedicated and competent assistance with the English language version. Inefficencies that might concern any aspect of the translation are my own. IJK


During the Euralex Congress held in Copenhagen in August 2002, three key ideas in particular occurred to me as a French lexicologist and lexicographer making contact with English-speaking colleagues:

  • first, there was obviously a great deal for a French specialist to learn from his non-French-speaking colleagues concerning their specific approaches to our discipline;

  • then, it seemed to me from the outset that to date there had not been much interraction between “Francophone” approaches and “Anglophone” ones (the latter may of course involve speakers of Italian, German, Russian, etc., who choose to publish in English, and who may be influenced by “English” work in the strict sense);

  • finally, I realized that while I was learning a lot from these colleagues and friends, I might also have a number of points of view and methods that were peculiar to my training, and which could effectively contribute to a meeting of minds.

Then, too, I was encouraged by Ilan Kernerman, whose paper at the Congress was particularly stimulating, as well as by Tony Cowie, whose benevolent dynamism I admire, to try to set out some of the ideas which form part of my credo and my background as a contribution to this newsletter. Three perspectives in particular seemed to me to be worth exploring.

The first corresponds to the distinction to be drawn between “lexicographie” (roughly, “theoretical lexicography”) and “dictionnairique” (or “practical dictionary-making”). The notion of “dictionnairique” has in fact been quite recently introduced by Bernard Quemada, editor-in-chief of the Trésor de la langue française, and has been adopted fruitfully by numerous French lexicologists. This distinction seems fundamental to me.

The second perspective is the one developed by Robert Galisson under the heading of ‘lexiculture’. Galisson is one of our most original and productive lexicologists in the field of French as a foreign language. Yet, in fact, lexiculture is probably one of the most neglected of the components involved in the editing of entries in French or English dictionaries, and is sometimes even completely overlooked.

The third perspective is what I call the ‘triple investigation in dictionary-making’. Various lectures I have given on the subject have convinced me that this particular approach could very likely play an important part in the improvement of our dictionaries.

1. The helpful distinction between theoretical lexicography and practical dictionary-making

In order to understand fully the difference between these concepts and to appreciate their essential complementarity, it is necessary to place them within the recent history of French dictionaries which is, more or less, not all that remote from the history of lexicography in other Western countries. One can actually distinguish four successive stages during the second half of the 20th century.

1.1 Lexicology disassociated from lexicography, in the traditional sense of the term

From 1950 to 1965, we can separate off an initial period in which one needs to draw a distinction between “lexicology”, the scientific study of words, and “lexicography” in the traditional sense of the term – that is, the actual compilation of dictionaries. We know that, in fact, lexicology as the study of words did not really attain the level of a scientific discipline until the second half of the 20th century. In France, a particular date has become symbolic in this respect: it was not actually until 1959 that the first issue of the Cahiers de lexicologie appeared; this scientific journal, established and edited by Bernard Quemada, entered the 21st century having appeared in no less than 78 issues and with an unchallenged scientific reputation.

During this first period, lexicology and lexicography in the classical sense of the term have each redefined themselves and been redefined in relation to each other, lexicology becoming in the fullest sense a scientific discipline, and lexicography clearly coming to meet at one and the same time the requirements of a craft and a science.

Lexicologists, while fully accepting the continuity of their discipline with philology, had then to take the measure of the emerging theory of structuralism and of the most recent technologies of the day, technologies represented at the time by punched-card machines. The study of the lexicon and of specialized vocabularies by reference to large corpora, and with the aid of punched cards – such was the pioneering issue of this period. It was notably at Besançon, in its laboratories equipped with punched-card machines, that lexicologists from all over Europe were trained. Thus, in June 1961, a symposium on the mechanization of lexicological research, which today has symbolic value in my eyes, was organized by Quemada at the University of Besançon. It was a meeting altogether in keeping with the new state of mind which had set in. What in fact does one of the participants, the Reverend Father Busa, director of the ‘Centro per l’automazione dell’analisi letteraria’, from Gallarate, near Torino, have to say? “One is aware that all of us taking part in this conference are pioneers in the automation of lexical analysis. We are playing a necessary part in the continuing evolution of the book […]. Today, alongside printed journals and books, the electronic book is in the process of finding a place for itself.” Such a statement, made in 1961, deserves to be recognized as visionary! It shows in any event that lexicology is acquiring a further dimension derived from the new technologies that are now emerging.

As for the lexicography of this period, it is distinguished in France by the awareness that dictionaries should be based on a closer analysis of the criteria by which they are defined. Dictionaries are no longer considered simply as working tools, but are examined in an altogether new light.

Interest was shown most particularly in the ideas underlying the Dictionnaire du français fondamental (1958), with a line of descent from “Basic English”. This dictionary of basic French derived from a scientific programme with didactic perspectives, being based on an analysis of vocabulary frequency. Then, while we awaited a new large-scale dictionary of the French language, we witnessed the republication of a large dictionary that was symbolic of the 19th century, the Dictionnaire de la langue française by Littré (first edition, 1873; reprinted in 1956). During the same period, in the Grand Larousse encyclopédique in ten volumes (1960-1964), the first defining steps were being taken towards the use of the new technologies – those then developing and specifically the 400,000 punched cards assembled in advance of the compilation of this paper dictionary. An encyclopedic dictionary, but one that was effective also in its treatment of the language and of technical varieties, the Grand Larousse encyclopédique deserves to be remembered as one of the landmark dictionaries of the period. We had not yet reached the computer era in the precise sense of that term, but very rigorous methods, based on algorithmic analysis, were already in evidence.

However, for the moment, “lexicography” could still keep its traditional meaning: it approximated in fact to the compilation of dictionaries, with recourse as circumstances required to the most suitable technologies, and the use of editorial teams that were becoming more and more professionalized.

1.2 The birth of metalexicography and the new distinction between theoretical lexicography and practical dictionary-making

The second period, broadly speaking, ran from 1965 to 1980, and demarcated a stage when the dictionary enjoyed a new status, being widely recognized as an object of scientific research. A French thesis entitled Les dictionnaires du français moderne (1539-1863) (Didier, 1968), submitted by Bernard Quemada, later to become editor-in-chief of the Trésor de la langue française, established itself as a benchmark for numerous studies that would come to fruition concerning one dictionary of the past or another. A new discipline was thus born: “metalexicography”. Lexicography, which until then, had been mainly concerned with meeting everyday needs, with the dictionary regarded above all as a tool, henceforth became linked up with corpora that were studied in order to improve understanding of the history of the genre and the functioning of the language. In the process, dictionaries ceased to be simply the creation of philologists and skilled craftsmen; they became a matter for linguists as well.

The second period coincided with a time of intense commercialization of dictionaries intended for the general public, and a veritable revolution in information technology—in advance of actual dictionary production—in the classification of data and its interpretation. Actually, the field of lexical research assumed a new breadth, while it became easier to produce dictionaries based on different computerized databases, adapting them for different categories of reader. At that stage, Quemada introduced a new dichotomy, between “lexicographie”, to which he gave a new meaning in relation to its traditional sense, and “dictionnairique”, both concepts forming a helpful dichotomy while being perfectly complementary.

In the new opposition established between “theoretical lexicography”, i.e. in its new definition, and “practical dictionary-making”, lexicography went well beyond the activity of editing a dictionary to include actual scientific research, carried out into words and their categorization, with all the related work of definition.

“Practical dictionary-making”, represented, by contrast, everything that has to do with the concrete aspects of production, and presentation, for a given readership, with all the commercial imperatives that are called for in order to satisfy that public.

With “theoretical lexicography”, one is situated squarely within the domain of research, without being concerned about high-lighting details for the non-specialist reader, or worrying about adapting the content for users who are simply buying a product. One is indeed at some distance from the dictionary as something that is shaped in order to be sold; one is engaged in pure research. There can even be a form of lexicography that, in contrast to the common definition of lexicography as “the compilation of dictionaries”, does not necessarily lead to a dictionary that can be sold. Corresponding to this or that piece of research into groups of words, or their definitions, it may very well not leave the laboratory but correspond, for example, to the computerized databases designed uniquely for researchers. Here one is not at all concerned with grading information for an attractive product to be marketed in a purchasable form.

As for “dictionnairique”—a word that Charles Nodier had already used in the 19th century, but which fell into oblivion until Quemada resurrected it—this denotes by contrast the activity of developing a dictionary as a product, which is then offered for sale—with all the constraints and problems relating to each edition—as a reference tool, a cultural product deliberately designed for a specific group of potential buyers. Thus, one must never forget that the dictionary represents a technical-commercial product whose content is defined as a function of the means granted to it for a defined clientele, within the context of a precise piece of market research.

Let us take, for example, two dictionaries that are very widely diffused in the French-speaking world and which are considered to be of high quality, in this case the Petit Robert or the Petit Larousse. (And let us remind ourselves that on average 200,000 copies of the Petit Robert are sold each year, and 800,000 of the Petit Larousse, thus over one million copies in 2001.) When a new edition is being considered (and this is every year, since they bear a date) and it is necessary to add a new word to a particular page, it is out of the question to reset the entire printed dictionary at the beginning of each school year. This or that example is simply removed from a neighbouring entry on the same page—or one or another shade of meaning—to free up the few lines which will allow us to insert this new word without touching the beginning of the page and its end, and thus without having to modify the preceding pages and those that follow. Here one is plainly in the domain of “dictionnarique”: these are the practical constraints that prevail over definition quality and precision.

One should emphasize, to illustrate yet more clearly the difference between “lexicographie” and “dictionnairique”, that one may be an excellent lexicographer, that is to say carry out effective research on groups of words, and their definitions, and yet turn out be a dreadful dictionary-maker, that is one who fails to meet the production deadlines and the material restrictions that are inevitably imposed. One knows of major dictionaries that in their earlier volumes boast of enormous entries, making them almost illegible, and in which, as one goes along, because of a lack of space and because it has already been necessary to increase the number of volumes that were initially planned, the entries get thinner, and one can even find oneself at the end of the alphabet with very poor entries indeed.

Publishers should not be confused with researchers. Publishers must sell a product of a given size to a readership that can be won over at a particular price, and during a given period of sale: practical lexicography concerns them above all else. There is no mystery about the rules: if the product is inappropriate, excessively large, inconsistent in the density of information provided, the dictionary as a product will not be successful, it will not sell, and the publishing house will be in mortal danger.

Be that as it may, theoretical lexicography and practical dictionary-making are complementary: there is indeed no dictionary-making of interest that does not rest on a solid basis of lexicography, and lexicography is sometimes more effective if it has been able to take account of the practical constraints of time and space which, in a sense, contain it and lead it towards greater homogeneity in the description of a large body of words.

1.3 A revealing distinction between basic principles

It is possible to learn a number of lessons from the necessary distinction between theoretical lexicography and practical dictionary-making.

First of all, it is important to separate the two perspectives, ‘lexicographiques’ and “dictionnairiques”. A dictionary, that is to say, a product, in which the two perspectives are blurred, runs the risk of failing to meet the user’s expections. In general, the user wishes to have information that is precise, but not overwhelming. If he buys, for instance, a thousand-page book, he would rather have useful and clear information than information that tends towards the exhaustive and transforms each entry into an exercise in compression, into a dreadful digest. To aim always at providing the maximum of information in the minimum of space, is to condemn the user to reading with a magnifying glass, to the “intellectualized” reading of the researcher. After all, one might reflect, the dictionary represents a genre in which the editor in principle avoids stylistic verbosity, which is considered out of place. Superfluous details are hunted down; the object is scientific and for this reason, one needs, it is assumed, to be austere.

Note, however, that the first monolingual French dictionaries, those of the 17th and 18th centuries, which possess great charm, do not seem restricted to this degree by scientific rules, of an almost monastic nature, governing the entire work. Constantly saving space, by condensing as much as possible in order to add new information, is not sound practice. Outside the “dictionary” genre, in works of a didactic nature, redundancy is indeed very much present, not to say essential, for the explanations. It allows one to space out the information, to make it accessible and digestible; it also affords one the possibility of putting forward different approaches. Too much dense information definitely frustrates the effective conveying of information, while taking away all the pleasure from consultation. Without a practical approach which allows for flexibility of editing from the start, and without a theoretical side that can keep the “lexicography” involved within certain limits, so as to allow for the “dictionary-making” that is required—the kind that will make the reading of the dictionary a pleasant experience—you will no doubt sacrifice one of the primary functions of the dictionary: to make the information clear but also pleasant and legible. It is easy to add a mass of information in the name of “lexicography”; it is difficult to curb oneself and to choose as a good dictionary-maker should the discourse which is best suited to the task.

Then, to compare the findings of research to the editing of an entry that should account for them in their entirety, is to confuse the two stages. There is a time to conduct research, to be engaged in lexicography, resulting in an entry intended for the researcher alone; there is another time to adapt the results to the user’s needs, to move into practical dictionary-making. Here, one does not want necessarily to hand on everything that has been found at the research stage. The entry which is then compiled is aimed at a reader who is not a linguist, nor willing to read and reread definitions that are too dense. The information to be given to the reader should not be confused with the straightforward recapture of the dry, scientific language expected by the linguist. Thus, the absolute meticulousness and the concern for exhaustiveness which are uppermost in research are no longer necessarily the primary criteria: one must adapt in order to explain better. The lexicographer-researcher can write for his peers when he is in the domain of research, but when he becomes a dictionary-maker, he no longer writes for his peers; he writes for all his readers and especially for those who are not linguists. The dictionary has a didactic function as a reference tool for everybody.

Finally, have we reflected sufficiently on the fact that though it is desirable that the researcher knows everything possible about the use of a word in the language, he should, as a dictionary-maker, not necessarily summarize it in as little space as possible, but on the contrary take account, as much as he can, of the particular questions that the reader asks concerning that word? Yet, the often systematic treatment of information that is characteristic of the way we proceed in linguistics, sometimes falls short of answering many of the specific questions that dictionary users ask themselves about such and such a word.

As a matter of fact, of course, the term “word” is used in various senses: “word as part of the language system”, “word as part of discourse”, “literary word”, “word as referring to something (having a referent)”, and so on. Now, let us not forget that the word as treated in the dictionary, which often corresponds to a more or less successful synthesis of some or all of these senses hidden in a single word, is not the dictionary word itself (i.e. the item in bold print that heads an entry and is often referred to in English as the ‘headword’ or ‘entry word’). The word described in the entry, then, is not simply that which has been analysed with reference to language and discourse; it may first be perceived as the “consulted” word (the headword). And, in this sense—the word as something consulted—has to some extent its own difficulties which often escape the uniform rules of description, devised for the whole body of words in the dictionary.

I have drawn up a summary list of difficulties relating to certain French words, and it will be noted that each word is almost always consulted in the dictionary to resolve the same difficulty. It is a curious fact that (in France) no-one has brought together groups of linguistically naïve observers, or of dictionary users, noting systematically the questions that they turn to the dictionary to resolve. A large-scale investigation of this type would be quite revealing. There has been such a proposal concerning online electronic dictionaries, connected with the organization of automatic logging of questions posed, but studies of the needs that were revealed are lacking.

Thus, how is the French verb rejeter (reject) written in the future tense: rejetera, or rejettera? Hardly any dictionary considers including this as an example, although 80% of look-ups of the entry focus on this question. For the commonly-used abbreviation pro (a professional), the plural is never given, and people hesitate – can they write pros? There too, those looking up this word are concerned for the most part with this question of spelling. Now, for the linguist, the problem does not arise. He has in fact thought in terms of very general rules which apply throughout the entire work, and he considers that, if he does not provide this or that piece of information, it is because in his eyes it goes without saying. If he does not provide a special note, it is because in this case one follows the general rule. Obsessed as we are, as lexicographers, with the need to save space, any saving of space is welcome, general rules enable us to gain typographic space: too bad for the reader who is not aware of our obsession and who looks up a word just because he is not aware of the general rule! This is to treat with disdain the concerns of the anxious dictionary user. The truth is that, statistically, the user never reads introductions and wishes for a direct reply to his queries. Thus, as regards the particuliar and complicated rules of pronominal agreement in French according to which “ils se sont développé” (they have developed (themselves)) does not take an ‘s’, here too, it is quite rare that the dictionaries offer the illustration hoped for in the examples. The same is true regarding semantics, where one does not expect necessarily and systematically an exhaustive description of all the semantic components of the word, but sometimes examples clarifying the referent, as it was yesterday and is today. Hardly any French person looks up the entry chaise (chair) for its spelling or its use in the language: here, it is the referent that is crucial. It is therefore treatment of the referent that should be further expanded in sound practical dictionary-making, yet a French “lexical” dictionary in principle does not provide illustrations. It is therefore necessary to know beforehand the meaning of chaise haute (highchair), chaise longue (deckchair), chaise à porteur (sedan chair), chaise percée (commode), chaise roulante (wheelchair), to gain any advantage from these words, presented in lists in lexical dictionaries, married (not always) to the briefest of definitions.

The distinction which to us seems fundamental between the notions theoretical lexicography and practical dictionary-making, the latter being concerned with providing a pleasant and efficient tool for the user and not for the researcher, does not disappear during the two subsequent periods of modern dictionary history. On the contrary, it has become all the more important. The second period we have called to mind saw the birth, in fact, of very great dictionaries such as the Dictionnaire du français contemporain (1966), of a distributionalist nature; the Trésor de la langue française (1971-1994), of a philological nature, based on a computerized textual documentation, unrivalled for the French language; the Grand Larousse de la langue française (1971-1978), likewise of a distributionalist approach; the Dictionnaire alphabétique et analogique de la langue française (1964; Supplément in 1970) by Paul Robert, in a renewed continuation of Littré. In truth, then the third period, which extends from 1980 till about 1995, only further reinforces the useful distinction we have drawn between “lexicographie” and “dictionnairique”.

We are witnessing then a connection between, on the one hand, the domains appropriate to dictionaries designed for human consultation, and, on the other hand, computational lexicography, a discipline reserved until now for computer scientists. This last discipline brings together all that constitutes the basis of lexical knowledge and all that has to do with machine-dictionaries for the computerized treatment of languages and the language industries. There has been a meteoric rise in research in this field, and computerized resources are making possible very broad-ranging studies. Lexicography in the Quémadien sense of the term is in full swing.

Elsewhere, information technology, even before the birth of the first CD-ROMs and Internet, made possible the gearing-down of dictionaries intended for the general public by working from well-nourished databases. Many small dictionaries thus appeared, diversified according to the ages of their potential users; and practical dictionary-making could henceforth even dispense with the data provided by research, and sometimes enjoy an autonomy outside of lexicography. Here was no longer research, but the adaptation of data, with as many “cocktails” of the data as there were potential readerships. Diversify to achieve better sales! And this is sometimes made possible by adapting, with talent and efficiency, existing knowledge. This was the case with the Dictionnaire historique de la langue française (Le Robert, 1992), for example, which set out in a pleasant way etymological information provided by researchers, those of the CNRS, most particularly. Sometimes, by contrast, we are provided with nothing more than a fairly dull mixture of information, selected and targeted so as to secure a well-calculated commercial profit, in the manner of a well-packaged product.

If we wish to ensure that lexicography does not turn in on itself, and that practical dictionary-making does not self-reproduce, these are the goals that should not be lost sight of. Both perspectives should remain united and complementary. Without research, there is in fact no interesting future for dictionaries. And without sound dictionary-making, lexicography might shrivel and be of benefit to very few, without really attracting new competence and skill.

As for the last period, the very end of the 20th century and the beginning of the 21st, this has been marked by the development of the Internet, which distinguishes itself primarily by the revival of editorial strategies, extended and adapted to the new virtual spaces, infinite spaces of information accessible in real time. It is defined also by a profound metamorphosis of our look-up habits.

A problem remains: for the time being, we are confronted chiefly by electronic adaptations of products, published not long ago in paper form, which are in the process of being redeveloped or offered on the market. This is “redictionarization”, moving from paper on to computer, adding to it all the proper tags for the richest and most reciprocal look-up possible, and matching to it internet links. We are nevertheless left with the task of developing dictionaries conceived from the outset for the computerized medium, with no doubt real openings-up for the hypertextual means between the encyclopedia and the language, between the synchronic and the diachronic, between the general vocabulary and the specialized vocabulary, between textual examples and visual and sound examples, all in all synesthesic. Added to all this is lexiculture, which we will introduce later.

New approaches to lexicography and practical dictionary-making remain to be developed: the potential field of activity is immense. Many users start off by preferring disorganized searching on the Internet, which is no doubt rich but also unpredictable, to the consultation of a real “dictionary” based on this breaking down of barriers. Already, some works are taking up the challenge, especially in the field of learners’ dictionaries. Working together, on a global scale, we will not be cautious.


Books and articles

Galisson R. et J. Pruvost (dr.), 1999. Vocabulaire et dictionnaires en français langue maternelle et en français langue étrangère. Étude de linguistique appliquée, n° 116 (octobre-décembre). Paris: Didier Érudition.

Pruvost J.,  2002. "À la recherche de la norme: sa représentation lexicographique et dictionnairique chez Larousse et Robert et la triple investigation". In La représentation de la norme dans les pratiques terminologiques et lexicographiques, Langues et sociétés n° 39. Montréal: Office de la langue française, Québec, pp. 139-170.

Pruvost J., 2002. Les dictionnaires de langue française. Collection "Que sais-je ?" n° 3622. Paris: Presses Universitaires de France.

Quemada B., 1961. "Actes du Colloque sur la mécanisation des recherches lexicologiques". In Cahiers de lexicologie, n°3. Paris: Didier Érudition.

Quemada B., 1987. "Notes sur lexicographie et dictionnairique". In Cahiers de lexicologie, n° 51. Paris: Didier Érudition, pp. 235-245.


Dictionnaire alphabétique et analogique de la langue française, 5 vol. and 1 Supplément, 1964-1970. Paul Robert. Paris: Société du Nouveau Littré.

Dictionnaire de la langue française, 4 vol., 1863-1873. Littré É. Paris: Librairie Hachette.

Dictionnaire du français contemporain, 1966. Dubois J. (dir.). Paris: Larousse.

Dictionnaire du français fondamental, 1958. Gougenheim G. Paris: Didier.

Dictionnaire historique de la langue française, 2 vol., 1992. Rey A. (dir.). Paris: Le Robert.

Grand Larousse encyclopédique, 10 vol., 1960-1964. Dubois Cl. (dir.). Paris: Larousse.

Le Petit Larousse illustré, since 1905, revised every year. Paris: Larousse.

Le Petit Robert, Dictionnaire alphabétique et analogique de la langue française, since 1967, revised every year. Paris: Société du Nouveau Littré, Société Le Robert.

Trésor de la langue française, Dictionnaire de la langue du XIXe et du XXe siècle (1789-1960), 16 vol., 1971-1994. Paris: Klincksieck, vol. 1-10. Paris: Gallimard, vol. 11-16. Directors: Imbs P. (vol. 1-7), Quemada B. (vol. 8-16).


K Dictionaries Ltd
10 Nahum Street, Tel Aviv 63503 Israel
tel: 972-3-5468102 • fax: 972-3-5468103