Praxis French: World Language study materials
Praxis French: World Language study resources

« Ce n'est pas du français »: Assessing a Laurentian Recording

Praxis French: World Language · Put your knowledge to work in context.

All case studies

« Ce n'est pas du français »: Assessing a Laurentian Recording

Your French 3 class listens to a 75-second interview recorded in a Montreal bookstore. The bookseller says « Bienvenue ! » when a customer thanks her, « on jase un peu », « mon char est dans le stationnement », « c'est correct », and gives a price as « dix piasses ». You wrote five comprehension items; the class average is 38%. After the second playing, a student says, "This isn't French, it's a different language," and another asks why you are "testing Canadian." Two students who did a summer programme in Quebec score 5/5. One student who has family in Dakar scores 2/5 and is visibly annoyed.

1. Diagnose the failure item by item. Which are lexical, which are phonological, which are pragmatic?

Read the worked answer

Pragmatic (the one students will misread as vocabulary):

  • « Bienvenue ! » in response to merci = "you're welcome," a well-documented calque. Students will hear it as "welcome to my shop" and build the wrong scene.

Lexical - regional but ordinary:

  • « jaser » = to chat, informal.
  • « char » = car, informal.
  • « stationnement » = parking lot (where a French speaker would more often say parking).
  • « c'est correct » = that's fine / no problem.
  • « piasse » = dollar, slang from piastre, highly colloquial.

Phonological, and the silent killer: affrication of /t/ and /d/ before high front vowels (tu → "tsu", dix → "dzis"), plus diphthongised long vowels. This is systematic and applies to every word, so it degrades comprehension far more than the five lexical items above - but students cannot name it, so they blame the vocabulary.

Diagnosis: the class did not fail on five words. They failed because an unfamiliar accent plus a colloquial register plus one false friend (bienvenue) exceeded their processing capacity, and the items probably depended on the words they lost. That is a task-difficulty problem, not a "this isn't French" problem.

2. Is the assessment valid as administered? Decide, and defend the decision.

Read the worked answer

Decision: the items are valid in construct - francophone listening - but invalid as administered, because the students were never taught to hear this variety. Fix the instruction and the scoring, not the construct.

Reasoning both ways:

  • For keeping it: if the course goal is "understand spoken French in the francophone world," a Quebec recording is straightforwardly in scope, and dropping it because it is hard tells students that only one accent counts. That message is worse than a low average.
  • Against the current administration: validity depends on whether the construct was taught. If every listening exercise this year has been Hexagonal, standard-register studio audio, then this test measures exposure, not proficiency. A 38% average is a measurement of your syllabus.

Concrete actions:

  1. Do not simply curve. Curving hides the gap and teaches nothing.
  2. Re-teach the five items plus the affrication pattern in one 20-minute lesson, then re-administer a parallel short task for a replacement grade.
  3. Keep the original scores on record as a diagnostic, not as a summative result.
  4. Next unit: build in one non-Hexagonal recording per week from the start, and include at least one in the practice items before it appears in an assessment.

3. How do you answer the two students - "this isn't French" and "why are we testing Canadian?" - and what do you do about the student with family in Dakar, who scored 2/5 and is annoyed?

Read the worked answer

"This isn't French." Short and factual, no lecture: French is spoken as a first language in Quebec, Belgium, Switzerland, Senegal, Côte d'Ivoire and elsewhere; each has its own accent, slang and some vocabulary. Quebec French differs from Parisian French roughly the way American English differs from British English - the comparison is the fastest way to make it land, and it is accurate at the level of register and lexicon, though not at the level of state institutions. Then name the mechanic: "The accent made it harder to segment the speech stream. That is a listening skill, and we're going to practise it."

"Why are we testing Canadian?" Reframe: "We're testing whether you can understand a francophone you didn't expect. That's what the exam's listening section does too." Show one item from a past paper with a non-Hexagonal speaker if you have one.

The Dakar student - handle this separately and privately. Her annoyance is legitimate and diagnostic. Two possibilities, and you should find out which rather than assume: (a) she expected African francophone varieties to be treated as "real French" and read the Quebec lesson as the class finally acknowledging pluricentric French while still ignoring hers; or (b) she assumed her own variety would make this easy and is embarrassed by the score. Ask. Then act: add a West African recording to the next unit, and - importantly - do not ask her to be the class's informant on it. Invite her to compare, as a speaker, not to validate.