1. Introduction
Proficiency guidelines from both the American Council on the Teaching of Foreign Languages (ACTFL) and the Common European Framework of Reference (CEFR) include knowledge of vocabulary as a critical component of learners’ growing language competence, for both comprehension and production in their target language. Exactly which words learners should learn and when remains unstated in these guidelines, however. Previous research in vocabulary acquisition focusing on English as a second language/English as a foreign language (ESL/EFL) (Schmitt & Schmitt, Reference Schmitt and Schmitt2014; Webb & Nation, Reference Webb and Nation2021) has led to estimates that learners need a minimum of 2,000–3,000 word families for basic competence in English. This research has been highly influential in the creation of the General Service List (West, Reference West1953), the New General Service List (Brezina & Gablasova, Reference Brezina and Gablasova2013), and the Academic Word List (Coxhead, Reference Coxhead2000), which are now frequently incorporated into ESL/EFL textbooks as vocabulary targets. In other languages, much less work has been done in terms of creating standardized vocabulary lists. In this study, I compare the most frequently appearing vocabulary in a French corpus with the vocabulary taught in commonly used beginning-French textbooks. In doing so, I highlight the need for a standardized vocabulary list for our learners of French, deriving from twenty-first century oral and written French.
2. Background
Few studies have focused specifically on the acquisition of vocabulary in French as a second/foreign language, but in those that have, researchers agree that the acquisition process is laborious. Horst and Collins (Reference Horst and Collins2006) found that learners did not add significantly more low-frequency words to their lexicon over the course of instruction, but instead solidified their knowledge of the 2,000 most frequent words, using them more correctly and in more diverse ways. Milton (Reference Milton2006) noted that the best learners tended to learn about four words per contact hour, plateauing after only one year of instruction, with weaker learners acquiring even fewer words overall. Ovtcharov et al. (Reference Ovtcharov, Cobb and Halter2006) noted that anglophone learners of French in Canada relied extensively on the 1,000 most frequent vocabulary words, mirroring native speaker usage only at the highest levels of proficiency. All studies highlighted the importance of teaching learners the most frequent French vocabulary first.
3. Research questions
The following research questions were examined in this study:
1. Which are the most frequent words used by native French speakers in the authentic materials examined? How are they represented proportionally in the corpus?
2. To what extent are these words actively taught as vocabulary in beginning-level French textbooks in the United States?
4. Methods
4.1 Textbooks
Six beginning-level textbooks from three publishersFootnote 1 were chosen for analysis. Each textbook organized chapters by themes and incorporated a four-skills approach, with focus given to speaking, listening, reading, and writing in each chapter, in addition to explicit instruction of vocabulary, grammar, and culture. All words included as vocabulary items in chapter lexicons as well as in the end-of-textbook glossary were noted on an Excel spreadsheet. This approach allowed me to track vocabulary that was explicitly taught to learners in each chapter, as well as vocabulary that appeared in readings and cultural passages and that may have been learned implicitly. Items within a multi-word string (Comment tu t'appelles? “What is your name?”) were noted as individual words on the Excel spreadsheet (comment – adverb, tu – pronoun, te – pronoun, appeler – verb).
4.2 Authentic materials
An in-house corpus was created for use in this study, consisting of one million words from ten sub-corpora of twenty-first century materials evenly divided between written and spoken genres. My goal was to create a corpus focusing on the authentic language that second language (L2) French learners at the novice (ACTFL) or A1-A2 (CEFR) levels are most often exposed to in the language classroom, such as music lyrics, short stories, conversations, and blogs, in addition to the academic sources frequently included in existing corpora, such as newspaper and journal articles and radio broadcasts.
4.3 Vocabulary profile
Once all sub-genres of the corpus had been created, documents were passed through LexTutor (Cobb) for lexical profiling. The French 1–25 K Framework was chosen, which categorizes words in texts into frequency bands from the 1,000–25,000 most/least frequent words, based on a much larger corpus (23 million words) of digitized written (66%) and transcribed spoken (33%) materials. A flemma categorization was chosen, such that a verb and all its inflected forms were categorized together, a noun and its inflected forms were viewed as a separate flemma (habiter was thus separated from habitation), and so forth. Derived forms were also categorized as separate from their stems. The 2,000 most frequent flemmas (flemmas 1–1,000 are abbreviated K1 and 1001–2,000 K2) are considered here, with particular focus on the K1 flemmas.
5. Results
LexTutor identified 981 flemmas from the in-house corpus as belonging to the K1 vocabulary from the larger LexTutor corpus. (Proper nouns were classified as “off list,” in order not to skew the amount of K1 vocabulary identified.) At the same time, 963 flemmas were identified as belonging to the K2 vocabulary from the larger corpus. In the K1 vocabulary, there were 460 distinct nouns, 242 verbs, 114 adjectives, 97 adverbs, and 70 function words. Ratios by part of speech were similar at the K2 level, with 477 nouns, 252 verbs, 173 adjectives, 40 adverbs, and 19 function words. Some of the most frequent K2 flemmas in the in-house corpus were used more often than the least frequent K1 vocabulary, as identified by LexTutor. This suggests that there was large but not complete overlap between the most frequent words chosen by speakers/authors in the contemporary corpus and those used in the larger corpus, likely reflecting the degree of formality in the larger corpus, which relied more heavily on written academic materials. Some of the more frequent items in the K2 vocabulary (greetings, days of weeks, months, discourse markers) were flemmas that did not rise to the level of K1 vocabulary in the LexTutor corpus. All flemmas in the K1 vocabulary were observed to follow Zipf's law,Footnote 2 with the most frequent verb (être, to be, 30,832 tokens) occurring significantly more often than the second most frequent (avoir, to have, 22,485 tokens),Footnote 3 followed by expected proportional decreases in tokens as words decreased in frequency. Verbs 31–50 on the K1 list appeared an average of 531 times in the corpus (range: 653 for regarder, to watch/to look at – 426 for perdre, to lose), while usage of verbs 101–120 decreased by an additional 50–70%: average 196 tokens, range: 220 (aider, to help) – 180 (tirer, to pull). Similar trends were observed with all other parts of speech.
An examination of the six beginning-level French textbooks showed uneven coverage of K1 vocabulary. With rare exception, all textbooks included the 20 most frequent flemmas across all word classes in their vocabulary lists. Moving down the K1 list, however, we observed that coverage became more dissimilar. An analysis of the 100–120 most frequent nouns and verbs showed that only three of six textbooks, on average, listed these flemmas amongst vocabulary for learners. A first tendency might be to argue that this decrease is due to Zipf's law, but the criteria for inclusion was simply whether or not textbooks included the flemmas in their vocabulary lists, not overall frequency in each textbook. Instead, it was observed that some verbs (préférer, to prefer, for example) were included by all six textbooks, while others (tenter, to attempt) were included by none. When examining the 200–220 most frequent nouns and verbs on the K1 list, inclusion in textbook vocabulary lists dropped to a mean of two of six textbooks for nouns and less than one of six for verbs. It would appear, therefore, that textbook authors are not guided by overall frequency when choosing vocabulary for learners to learn, but by factors such as chapter theme and cognateness. This is further corroborated by examining the other parts of speech and flemmas beyond the 20 most frequent on the K1 list.
For adjectives, adverbs, and function words, where there were considerably fewer types on the K1 list to begin with, the last 20 flemmas were examined. Adjectives fared extremely poorly, with fewer than one of six textbooks including these least frequent K1 adjectives. It is not the case that fewer than 117 adjectives were taught as vocabulary in most textbooks, however. Rather, the textbooks tended to focus on cognate-based adjectives that were less frequently used by native speakers, but that fit the themes of the chapters and that could be easily learned by beginning-level students, despite their low frequency usage amongst native speakers. These included adjectives describing one's personality (sociable, timide), family (marié, divorcé), living quarters (spacieux, confortable). Not included were other cognate-based adjectives such as relatif, libéral, or civil that did appear on the K1 frequency list, but that may have been more difficult to “finesse” into a theme-based chapter. An average of only two of six textbooks listed the least frequent K1 adverb and function word flemmas amongst the vocabulary that they expected learners to acquire. Noticeably absent from the textbooks were adverbs not derived from adjectives. Antes (Reference Antes2022), in an examination of coverage of adverbs and conjunctions in these six textbooks, argued that these parts of speech are largely considered grammar points rather than vocabulary. All of the textbooks examined included a grammar lesson devoted to creation of adverbs from adjectives (poli – polite → poliment – politely), yet many neglected to include the most frequently used non-derived adverbs and conjunctions (aussi – also, as; car – because) from their vocabulary lists.
Among the K2 vocabulary that was identified in the in-house corpus, some of the most frequent items were well represented in the textbooks examined. These included greetings, adjectives of color, days of the week and months, and verbs such as habiter (to live) and chanter (to sing). Whether the textbooks recognized the frequency of these flemmas or were simply following a long-standing tradition of teaching learners thematic lists of vocabulary, however, remains an open question.
6. Discussion and implications
The comparison of a contemporary French language corpus and six beginning-level French textbooks demonstrated that there are serious gaps in the words that textbooks choose for learners to focus on during vocabulary instruction. While the most frequent words in the corpus were represented across all textbooks, beyond the top 50 flemmas in each part of speech (or approximately 250 words), inclusion across textbooks was highly sporadic, with no guarantee that learners would have been exposed to the majority of the K1 vocabulary at the end of instruction. Some of the most frequent K2 vocabulary was included, highlighting its importance for communication, but there appeared to be no rigorous method of vocabulary selection across textbooks examined. Based on this research, I argue for the creation of a general service list for French, based largely on word frequency in contemporary oral and written language, but also taking into account factors such as cognateness and dispersion.
Supplementary material
The supplementary material for this article can be found on the journal website.
