How does an expert teacher meet learning aims while also ‘fixing’ linguistic inequality through their pedagogical talk?

Recently (here and here), I wrote about how Jim Cummins defined “transformative pedagogy” as meeting learning aims and social justice aims at the same time, through collaborative relations of power. I continue that theme by asking the title question, which was what the late sociolinguist Alexandra Jaffe asked in a bilingual, Corsican-French school. Teachers of minoritized languages like Corsican have a challenging balancing act—to ensure that, in the wider society, the minority language is standardized/official enough to ‘fight’ for itself, and yet there is enough flexibility in teaching it to accommodate the diversity and needs of the learners. Her significant study in the late 1990s shows that what we would think of as ‘standardization’ can in fact be supportive of equal access and positioning for ALL learners… if that ‘standard’ is co-constructed by the class in as egalitarian a manner as possible. It appeared as a book chapter and as a journal article; in this post, I summarize the book chapter. At the end, I include a tribute to Jaffe’s paper and other “old” studies I summarize on the blog, explaining why I do so.

Jaffe, A. (2003). Talk around text: Literacy practices, cultural identity and authority in a Corsican bilingual classroom. In A. Creese & P. Martin (Eds.), Multilingual classroom ecologies: Inter-relationship, interactions and ideologies (pp. 42–60). Multilingual Matters.

A teacher’s goal may be to position languages equally in the classroom, but they are never equal in the wider society. This begs the question of what the teacher can and cannot control to position them equally. Abundance or lack of resources for teaching each language is beyond teacher control. So are top-down assessments (i.e., standardized tests in K–12 education) that add legitimacy to the powerful languages and marginalize the minoritized ones. But what CAN teachers control? To answer this question, we need to look at the practices of an expert teacher, who recognizes things they can control more readily than less expert teachers, and seizes that opportunity.

In a bilingual classroom, the aim is to learn two languages (majoritized and minoritized, e.g. English and Welsh; Spanish and Quechua), and both are given equal attention in the curriculum. However, due to their inequality in the wider society, differences and inequities in social functions for each language become apparent even as people try to follow this curriculum (e.g., Hamman, 2018). This is why Jaffe begins her paper with a discussion of cultural/linguistic identity and power. Every day, people naturally construct “sameness” and “difference” with regard to dominant and minoritized cultures and languages. The message of Jaffe’s paper is that BOTH injustice/inequality AND equality/equity arise from such “management” of sameness and difference. Through the critical educator’s deliberate actions, equality arises from marking sameness when it is good/just to do so, and equity arises from marking difference when it is good/just to do so.

Jaffe next turns to the topic of literacies. (See this post for a definition of literacies, and how they differ from language.) Literacy practices don’t just “belong” to different groups—rather, they create said groups, bringing them into being. Jaffe draws on Gee (1991), who argued that literacy practices create both knowledge and relationships. Minority language education cannot help copying, to some extent, some models of majority language education such as “standard” grammars and conventional forms of literacy (e.g., dictionaries, mainstream news outlets, standardized tests to earn language qualifications) because without these strategies, minoritized languages will lack social authority: “Models of minority language literacy are invariably defined in reference to dominant language literate histories and practices” (Jaffe, 2003, p. 43). Teachers of minoritized languages therefore have a challenging balancing act—to ensure that, in the wider society, the minority language has this kind of leverage, and, in the classroom, that there is enough flexibility to accommodate the diversity and needs of the learners. Interestingly, Jaffe’s study shows that having an authoritative ‘standard’ can support equal access and positioning for ALL learners, and maximize learning… if that ‘standard’ is co-constructed by the class in as egalitarian a manner as possible. The paper goes into how this is done.

The research context

Jaffe’s career as a sociolinguist was built in Corsica, an island in the Mediterranean that is part of France. It’s where Napoleon was from, and the island has a minoritized language/dialect called Corsican. Corsicans have fought hard for their right to education in their own language. In the 1970s, schools in Corsica did not teach Corsican language classes even though they taught foreign language classes, but fortunately, in the 1980s, a small number of experimental bilingual programmes were founded called “le corse intégré” (a form of content and language integrated learning or CLIL, teaching subject matter in Corsican). Unfortunately, government support was not extended beyond the three-year experimental period. Yet le corse intégré created knowledgeable teachers who would re-emerge as educational leaders a decade later.

The main challenge to Corsican education was that although Corsicans were proud of their heritage, there was a lack of grass-roots support for compulsory Corsican learning in schools (that is, people were proud of their heritage language, but didn’t see why it needed to be taught at school, where French was seen as appropriate; they were so used to the current domains of Corsican). In the mid-1990s, however, Corsican-French bilingual schools made a comeback in parts of Corsica where there was parental support, and some of their staff were former practitioners of le corse intégré. Jaffe describes these teachers as “highly motivated Corsican-speaking teachers who were already using a substantial amount of Corsican in their classes” (p. 44), and they were effective at partnering with the supportive parents.

Jaffe’s study took place at a small village primary school that was one of the first three schools to begin bilingual instruction in 1996. By 2003 (the time of the study’s publication), there were 17 bilingual sites. Teachers and administrators were desperate to fix the language shift from Corsican to French that had accelerated after WWII. By the 1990s, the first language (L1) of almost all Corsican children and youth was French, so virtually all the children were learning their heritage language (HL) as an additional language. At the school, at least 50% of instruction was Corsican-medium, and “[i]f policy had permitted, the teachers would have made the first three years of school (for children aged 3 to 5) total Corsican immersion” (p. 44). All subjects were taught 50-50 in French and Corsican. However, there was one class—grammar and spelling—which was offered in French only. This reflected a lack of textbooks of any sort in Corsican, which Jaffe discusses further on.

Another major goal the bilingual school had was to foster students’ appreciation of Corsican language and culture, their pride in their heritage, and their identity as Corsicans. One major barrier to this, of course, was that many barely spoke Corsican, and were constantly reminded of this in class: “the shortfalls of linguistic competence addressed by the language-learning activities implicitly undermine children’s claims to authentic ownership of the minority language” (p. 45). To be clear, teachers did not position students as incompetent—quite the contrary, they were very encouraging. However, objectively speaking, students’ lack of Corsican proficiency could get in the way of their deeper appreciation of cultural material, pride when speaking Corsican, and seeing themselves as Corsican.

Another goal was to assert political and symbolic equality between Corsican and French. This was naturally challenging because it didn’t reflect the reality beyond the school walls. But that didn’t matter—in school, things would be different. Teachers strove to build reading and writing skills in both languages “by introducing reading and writing in Corsican simultaneously with French, by having the children produce the same text genres (summaries, reports, letters, scripts) in the two languages, and by translating French pedagogical materials into Corsican” (p. 45). The languages were to be as equal as possible, even though standardized tests would only be in French (the government did not administer tests in Corsican), which meant parents would emphasize French more. But not every upper hand was held by French. There was a special heritage status attached to Corsican that bound the school community. Thus, some instructional goals were shared across both languages; other goals were specific to only one language. The children Jaffe studied, now in their 30s, were “the first generation of Corsicans to go through a school system in which French is not the only language of authority” (p. 45).

The first of two lessons

Jaffe made original contributions to bilingual education scholarship by analyzing only two lessons. The class was working on adapting an Indian story, written in French, titled The Tiger, the Brahman, and the Jackal, into Corsican and changing the cultural content to Corsican cultural content (i.e., The Bear, the Monk, and the Fox). In the story, the Brahman takes pity on a tiger in a cage and releases it after making it promise not to eat him, only for the tiger to come out and declare that it will eat him. To resolve their dispute, they agree to ask the first five creatures they meet, in a trial by jury, whether the tiger has any right to eat the Brahman. “The story was also earmarked as the source for the end-of-year play, which meant that it was ultimately to undergo a transformation of genre” (p. 45).

The activities in the unit were important literacy activities that occurred in both Corsican and French classes: “intensive oral preparation for writing in which [the teacher] guided the children in the crafting of a sentence that, once ‘perfected,’ was written on the board (either by her or a student secretary),” “the day’s work on a particular text… [was] copied by all the students into their notebooks,” “finished texts (or important segments of texts) were either neatly handwritten or typed on a computer, photocopied, and distributed back to the whole group,” and “the created texts became scripts that were passed out to all the children and then memorized” (p. 46). Jaffe’s research question was how authority was ascribed to texts and individuals in classroom interaction.

The first session was a week into the unit. The teacher had written down a summary of the story, but had forgotten to bring the photocopies to class, so everyone was trying to remember the story by heart and recite it. This created some interesting data. The teacher said, “We must revisit the story a bit because we have forgotten it,” and indeed, “there was no written record of that work before either the teacher’s or the children’s eyes, so they were all in the same position of recalling it from memory” (p. 46).

Below are the transcripts of dialogue. Jaffe analyzes three key parts, which I’ve highlighted in different colours.

What is interesting here is that when Vanina begins “long ago,” the teacher immediately chimes in “so, you remember, ‘long ago,’” which means she is immediately evaluating Vanina’s recollection of a text from memory. Vanina couldn’t have begun the story by, “Well, at the start, the monk was blah blah…” She had to begin it in the “proper” way, which frames the activity as NOT just a recalling of the story itself, but a verbal memory test of sorts. Vanina then continues, “when the animals knew how to speak” (l. 6), and the teacher repeats the sentence after her (l. 7), which is equivalent to saying, “That’s correct, Vanina!” and the others repeat the sentence after the teacher (l. 8), which suggests that they are also trying to memorize exactly how it went, word for word. From lines 9 to 20, you see people debating insignificant details of wording—was the monk passing? traveling? going?—and finally they settle on “was going from one village to the other” as the “correct” way to say it (ll. 19-20). It’s the teacher who decides this, saying, “No no no… ‘Who went [from one village to the other],’ it was a proposal, but we corrected it after, eh? ‘Who was going from one village to the other.’”

This may seem ridiculously pedantic, even oppressive, but notice the teacher saying, “we corrected it,” positioning the class, and not her, as the editors of their own work. Also consider what she is trying to do: expand their linguistic range: “from village to village” —> “from one village to the other.” Note that the teacher’s choice here is also more literary: it’s a language arts class that teaches how to do oral storytelling in Corsican.

The verbatim searches for what it was they wrote on the missing page means that “classroom talk in this sequence was effectively defined as reciting the written text. This definition of talk as oral performance of a remembered text excluded a number of other possibilities, including paraphrasing and rephrasing of previously made choices” (p. 48). In other words, some things were learned at the expense of other things—but if open/flexible paraphrasing and rephrasing were allowed, a different set of things would be learned and not what was learned here. This highlights that teachers’ decisions must be based on learning aims. In addition, the story was going to be a play, and lines needed to be fixed in students’ minds. Jaffe writes:

The teacher’s comment in lines 19-20, ‘era una pruposta, ma emu fattu una curreziò dopu’ [that was a proposal, but we changed it afterwards] … also reveals something else: that one of the sources of authority of the prior text was the fact that it was a product of collaborative negotiation and revision. The teacher presented the text almost as if it were a contract, that once drafted, could not subsequently be challenged without undermining the authority of the group. (p. 48)

While one might be apt to judge the teacher, there is a good reason for why she does what she does, which becomes clear when you focus on the strongest student (Vanina), who also liked to show off. She was one of the oldest students in a multi-grade class, with higher Corsican proficiency than the others.

Vanina wants to add a concrete detail to make the story more interesting: the bear complains that the shepherds didn’t put water in his cage. If the teacher had accepted this, the story would be better, but then it would start to belong more to Vanina, and then everyone would have to memorize her new addition, which would encourage her to make others, and then everyone would have to keep up. Instead, the teacher asks, “Did we write that?” and the other students laugh. “…we didn’t mention it the other day, we didn’t even say it at all.” In rejecting Vanina’s contribution, the teacher emphasizes our story, our story. In fact, to co-construct knowledge with the students, the teacher often used the pronoun “we.”

On another occasion, the children were debating whether one of the five jury animals would be a cow or a donkey, and she said, “I don’t know why, you liked, you preferred the donkey,” (p. 49), reminding them of their earlier choice. That is, it didn’t really matter what the animal was, but she reminded them that they were creating things together—as the directors making a group judgment. When they forgot their decision, she reminded them what it was, not necessarily because they ought to stick with it, but because they needed to remember they were directors making a group judgment. Another time, the teacher even included Jaffe in the collective: “And so we also chose, Alexandra, do you remember, we also chose…” (p. 49).

Vanina, who was very vocal, did not always agree with the collective text:

In one case, she corrected Ghjuvan’s use of the adjective ‘tamantu’ [huge] to describe the bear, telling him: ‘emu dettu, tremendu’ [we said ‘tremendous’]. The teacher corrected her, saying that it was in fact ‘huge.’ (p. 49)

Here is another instance:

In line 1, the teacher is tired of Vanina making all the contributions, and explicitly tells the class this has to stop. Vanina doesn’t stop, and contests (paraphrases) Francè’s contribution. The teacher announces that Francè is right, and claims that there is “a big difference” between the two contributions even though there is not any substantive difference in meaning (ll. 11-12). Jaffe suggests: “[Vanina’s] frequent, dominant voice in the classroom, while welcome as a creative force, also challenged the ideal of democratic, cooperative work among peers and perhaps, the teacher’s authority. … This exchange raises questions about the distribution of power in the practice of collaborative textual production” (pp. 49-50).

The second lesson

The next day, the teacher remembered to bring the handout:

The teacher’s delivery as she read the dialogue was animated and dramatic. It was the kind of crafted performance that could be done on a finished text, a known text. … In this respect, we can view the performance of the text as constitutive of the text’s status as finished and authoritative. The teacher’s reading also anticipated the text’s intended use as the basis for a play: with it, the teacher began a practice of modelling of theatrical delivery… Additionally, specific aspects of her delivery—for example, the use of pauses before the bits of dialogue—called to mind a style associated with the reading of well-known fables and fairytales. In this respect, it framed the text as being as well-known, and as shared as other stories in this genre. (p. 51; Jaffe’s emphases)

This is a major point: that the students’ and teachers’ collaborative work has been given the stamp of authority, as ‘standard.’ What they made together can be a standard! In other words, they are joint creators of standards which must be followed, to which people are answerable. (In this post, I mention bilingual storybooks jointly authored by students and their families and borrowable from the school library alongside books by published authors as an example of this “giving the stamp of real authority.”)

Next Jaffe focuses on one of the weaker students, Ghjuvan. She found that he tended to echo what the teacher and others said to demonstrate his competence in the language and perform being a “good student,” as this was the best he could do given the circumstances. Once, it distracted another child and the teacher enough that they commented on it, switching into French (in bold).

In this session, Jaffe also noticed Vanina being corrected (i.e., learning something), in this case the word “man” as meaning “all humankind,” not a specific group of people. Vanina states that the bear has a problem with the shepherds who put him in a cage. The teacher corrects her to say that this is a bigger issue—animals and plants versus humankind who exploit them. This is why the “jurors” (e.g., a bush) side with the bear. And so in effect, the teacher is not just teaching students a new word, the universal “man,” but a new concept, and an abstract one at that, i.e., humankind, and a new discourse and grown-up philosophical topic to go along with this concept, about humankind’s relationship with nature.

Jaffe writes:

[W]ith respect to some other instances where the teacher exercised authority, I would argue that they are all in keeping with the teacher’s pedagogical agenda, and criteria of effective narrative and dialogue established throughout this learning event. For example, she twice insisted that the children refer to the fruits of the bush by the phrase ‘baghi rossi’ [red berries] rather than generically as ‘fruits.’ In another discussion, of how men exploit these very fruits to make a variety of products, she disqualified one child’s suggestion of ‘honey’ as a product on the grounds that honey was never made exclusively from this particular plant. (p. 55)

The teacher’s pedagogical aims can be seen in her explanations, for example: “We are going to put, not ‘men’ but ‘man’ which is what the other characters in the story said, eh? ‘Man’ means man in general eh?” (p. 56). The message here is that teacher authority (i.e., the right to tell students if their contribution is right or wrong, and to what extent) has a purpose, and isn’t always bad. Often, such authority has a necessary pedagogical function… or even a function related to inclusion, ethics, or critical thinking:

[I]n the process of textual production, the teacher played an authoritative role, both in the way that she defined a good narrative, and in the way she defined appropriate participation as shared, relatively equally distributed and democratic. This definition of participation mitigated her authoritative role, and made room for joint ownership that was the product of negotiation between the teacher and the students and between students and their peers. (p. 56)

Before concluding the paper, Jaffe briefly discusses how these practices are similar or different from literacy work in the dominant national language, French.

Comparison with French

The following were the ways literacy activities were similar across both languages (p. 56-57)—luckily for these Corsican heritage speakers:

  • “a great deal of collective oral preparation for writing”
  • “teacher’s attention to lexical precision, and to form and delivery in children’s spoken language”
  • “The same attention was applied to student writing in Corsican, with spelling corrections made on the spot (as the student secretary wrote on the board) as well as when the texts were recopied by hand or at the typewriter”
  • “The way these corrections were made took for granted both a high level of competence in reading (knowledge of the ‘rules’ of Corsican spelling) and a high level of oral competence. They were indistinguishable, in this respect, from the way that the teacher corrected children when they misspelled French words.”

These were the differences:

  • “in the children’s class notebooks… the volume of French grammar worksheets, exercises and tables was three times greater than similar exercises in Corsican.”
  • “almost all Corsican texts used in the class were presented to the children orally. … That is, the children never encountered a Corsican text first as a written product, and they did almost no silent or individual reading of such texts. In contrast, they were regularly assigned French language books. These were read at home, and they had to write book reports on them in French. French texts and worksheets were also given out in class for individual, rather than collective work.” [Indeed, Jaffe notes in this book chapter and in the journal article published the same year (Jaffe, 2003) that individual French literacy practices stood in contrast to collective Corsican literacy practices. Which brings us to the next point.]
  • “Moreover, there was a difference in the way that individual competence was judged in French as compared to Corsican. Despite the fact that the teacher did a great deal of correcting of Corsican oral and written production at all stages of the process, this correction was almost always done on children [like Vanina] who volunteered to speak up during group work. Student secretaries (who sometimes wrote the texts-in-elaboration on the board and thus put their spelling skills in Corsican on display) were also self-selected. Children with less competence or confidence in Corsican were only forced to repeat a Corsican sentence that had either just been spoken, or which had been worked and reworked many times. In the case of writing, they could passively copy what had been written on the board. They were never forced to volunteer.”

What it means is that language acquisition-wise, there may be an even greater Matthew effect for heritage language learning than for dominant language learning. Because everything is voluntary and optional, those who volunteer tend to be the stronger students, who then get far more instruction. Jaffe concedes that yes, stronger students would also self-select more in French class, so there would be a Matthew effect there as well, but it would not be as pronounced. Why? Because

in French, they were regularly forced to demonstrate their individual knowledge. Moreover, at the end of the year, they had to pass written tests in French, but were assigned a grade in Corsican based on the teacher’s global evaluation of their ability [likely a generous one]. Thus it would be fair to say that Corsican literacy work [collective, allowing weaker students to hide, to parrot, or to merely recite what other students had authored] was a form of safe talk that backgrounded problems of competence in the language of lesser proficiency (Arthur, 2001; Hornberger & Chick, 2001; as cited in Jaffe, 2003, p. 57).

So what does it all mean?

There’s a reason why Jaffe focused first on Vanina, and then Ghjuvan. The point is that their teacher, who knew what she was doing, worked with the circumstances in the best possible way. She focused on two goals. First, functional parity would be the same between the languages, which is equality: treating people/things the same when necessary (in this case, against the grain of the wider society). Second, she used the strongest students as RESOURCES to put words into the mouths of others who needed them, but framed the achievement as everyone’s, which is equity: treating people/things differently when necessary, also to achieve fairness. Jaffe concludes:

As I have shown, language practices surrounding the production and recitation of Corsican texts treat children as competent writers/authors and speakers who are able to propose, debate and sometimes judge appropriate or good literary usage in Corsican. It is in fact the case that active participation in both of these activity structures demands a fairly high level of Corsican competence… In fact, even the brief excerpts from the transcripts presented here show that discussion was dominated by a handful of children and many children did not talk at all during this class period. … That is, the teacher’s curriculum revolves around meeting a variety of academic goals in Corsican, rather than teaching Corsican as a second language. This approach to the curriculum has symbolic and political importance, since it creates functional parity between Corsican and French [i.e., Corsican, too, is used for CLIL]. At the same time… in the secondary activity [i.e., the weaker students were not primary generators of linguistic output] of recalling the original text or reading completed texts out loud, some symbolic ownership and competence is transferred to the children who aren’t strong participants in the composition process. The collective, oral summarizing and recitation of the jointly created text illustrated in both class periods puts teacher-approved language into the mouth of even the weakest student as authentically ‘theirs.’ (p. 58)

This means that safe talk (reciting what more proficient classmates composed as lines in a play, speech, or even classroom dialogue) creates the shared cultural ownership of the minority language. While safe talk is often seen as bad (face-saving but not the best form of education), Jaffe points out that in a heritage language learning setting where ownership of the language is collective, we do not need to have such a negative view of it:

Nor are the texts merely ‘safe.’ They are also authoritative… From this perspective, the teacher’s repeated use of the first person plural [“we”] not only emphasises collaborative practice in the class, but also functions to culturally and linguistically authenticate the children’s ownership of the minority language. (p. 58)

Jaffe cites O’Connor and Michaels (1997), who coined the term “revoicing”: a form of scaffolding in which the teacher improves but also recontextualizes students’ spoken utterances so that they meet the instructional goals and yet give students the credit. (In the abstract of their article, O’Connor and Michaels state that revoicing is defined as “reformulations… that credit students with teachers’ [or more expert peers’] warranted inferences). Yes, the teacher exercises authority in revoicing. Nevertheless, agency and legitimacy are ascribed to the novices: they OWN the phrasings they have just learned. Nobody attributes authorship to the teacher or to the more proficient classmate/peer tutor. This is important when functional equality must be created between the dominant and minority languages, which means experts must share utterances with novices without making it seem like that at all, so that the novices fully “own” the words and so does the whole collective. They can also get language resources from texts: for example, past simple (“apri” = “opened” the cage in Corsican). In French, past simple only exists in written literary texts:

The choice of this tense is significant with respect to parity because it is almost exclusively used in writing, and anything more than a passing familiarity with this tense in either language is not acquired in the home or on the streets: it is only acquired in the academic context. To teach it is to position Corsican as a language with the same range of literary genres as French. (p. 59)

Popular discourse tends to define “authentic” forms of minoritized language competence purely in terms of the oral traditions of past speakers. Here, the teacher is initiating students into these traditions while showing them that there is written language to be mastered by current speakers. Minoritized languages can be “modern”; minoritized languages can be “institutionalized.” Not just related to the past and tradition, not just for family or community purposes, not just “valuable only for communication within linguistically minoritized homes and social arenas” (Beiler, 2021, p. 133). If “standard” material is created in class, and framed as intimate and collectively owned, it may still not compensate for the majority language’s dominance in the wider society, as shown by the things the teacher cannot change about lack of print resources in Corsican and official assessments in French only. On the positive side, here’s the article’s last sentence:

But it does create, within the classroom, the potential for [ALL] students to acquire an identification with Corsican [or any other minoritized language] as a privileged language of identity and heritage at the same time as they come to view it as a legitimate, authoritative code. (p. 59)


Arthur, J. (2001). Codeswitching and collusion: Classroom interaction in Botswana primary schools. In M. Heller and M. Martin-Jones (Eds.), Voices of authority: Education and linguistic difference (pp. 57–76). Ablex Publishers.

Beiler, I. R. (2021). Marked or unmarked translanguaging in accelerated, mainstream, and sheltered English classrooms. Multilingua, 4o(1), 107-138.

Gee, J. (1991). Social linguistics and literacies. Falmer Press.

Hamman, L. (2018). Translanguaging and positioning in two-way dual language classrooms: A case for criticality. Language and Education32(1), 21–42.

Hornberger, N. and Chick, J.K. (2001). Co-constructing school safetime: Safetalk practices in Peruvian and South African classrooms. In M. Heller and M. Martin-Jones (Eds.), Voices of authority: Education and linguistic difference (pp. 31–56). Ablex.

Jaffe, A. (2003). Talk around text: Literacy practices, cultural identity and authority in a Corsican bilingual classroom. International Journal of Bilingual Education and Bilingualism, 6(3–4), 202–220.

O’Connor, M. C., & Michaels, S. (1993). Aligning academic task and participation status through revoicing: Analysis of a classroom discourse strategy. Anthropology & Education Quarterly, 24(4), 318–335.


Published by annamend

Assistant Professor in the Faculty of Education, University of Hong Kong

%d bloggers like this: