"Learn French" is not a useful instruction. It is a category, not a target. French is spoken by 300 million people across five continents, in dozens of regional and national varieties that differ significantly in phonology, lexicon, and register. A person learning to live and work in Montréal is not learning the same language as a person learning to live and work in Paris — even though the writing is mostly identical.
Every major language learning platform teaches standard metropolitan French. Duolingo teaches Parisian French. Babbel teaches Parisian French. Rosetta Stone teaches Parisian French. If your goal involves Quebec, this is a problem.
The Phoneme Gap
The phonological differences between Québécois French and Metropolitan French are significant and systematic. The high vowels /i/, /y/, and /u/ are affricated in Québécois before high consonants — producing sounds that simply do not exist in Metropolitan French. The vowels /ɑ/ and /ɔ/ are maintained as distinct phonemes in Québécois but have merged in Metropolitan French. The phoneme /ʁ/ has a different realization. Elisions and contractions that are optional in Metropolitan French are obligatory in informal Québécois.
A learner trained exclusively on Metropolitan French has not been trained to perceive these sounds. They have been trained to perceive a different phoneme set. When they arrive in Montreal and encounter fast, informal Québécois speech, the signal does not match the phoneme inventory their training has built. Comprehension fails not because the vocabulary is unknown but because the sounds are being misparsed.
A learner trained on Metropolitan French has built a phoneme inventory for a different language. Dialect is not an accent — it is a different acoustic target.
What Happens in Practice
This is not a theoretical problem. It is one of the most common frustrations reported by intermediate French learners who travel to Quebec after learning from standard sources. They can read Le Monde. They can watch French films with subtitles. They can converse with Metropolitan French speakers with moderate effort.
Then they arrive in Montreal, walk into a dépanneur, and the person behind the counter says something at native speed that is completely opaque. The learner's French has not prepared them for this. Their vocabulary is wrong. Their phoneme expectations are wrong. The contracted, informal, affricated speech of everyday Québécois is a different acoustic experience from what they trained on.
This is not a failure of the learner. It is a failure of the training material to specify its target.
Why All Major Apps Teach Standard
The reasons are structural. First, creating dialect-specific content requires expertise in that dialect — native speakers, phonetically accurate recordings, vocabulary databases specific to the regional variety. This is expensive and requires local knowledge that global platforms do not have incentive to develop for markets smaller than their default audience.
Second, the assumption that "standard is good enough" is widespread but false. Standard language instruction is a production shortcut, not a pedagogical decision. The learner who says they want to "learn French" and specifies Quebec as their context would benefit from Québécois training — but platforms have not historically made that distinction available.
Third, the phoneme assessment problem: providing pronunciation feedback for Québécois French requires acoustic models trained on Québécois French, not standard French. Building and licensing these models requires investment that most language platforms have not made.
Dialect-Specific Training as the Actual Differentiator
Voicely's dialect-specificity is not a marketing feature. It is a direct consequence of taking the CEFR behavioral descriptor seriously as a target. The descriptor for B1 French in a Quebec context is operationally different from the descriptor for B1 French in a Paris context — because the communication situations, the phoneme set, and the lexical register are different.
HEXI selects content, pronunciation targets, and audio references calibrated to the specific dialect the learner has specified. A Québécois French learner trains on Québécois French phonemes, encounters Québécois French vocabulary, and is assessed against Québécois French acoustic targets. A Colombian Spanish learner trains on Colombian Spanish — not Castilian, not Rioplatense.
This matters at every level of the system. The UUS™ phoneme scoring compares your production against native Québécois French — not a generalized "French" acoustic model. The Island vocabulary surfaces words in use in Montreal, not words in use in Paris. The Pulse conversation scenarios are set in contexts relevant to life in Quebec.
Specifying your dialect is not optional in Voicely. It is the first question, because "which French?" determines the entire training architecture that follows. Language fitness requires knowing what you are training for. "Learn French" is not specific enough. "Québécois French to navigate daily life in Montréal at a B1 level within seven months" is a training target.
A Note on Mutual Intelligibility
Dialect-specific training does not mean provincial. A learner trained on Québécois French will understand Metropolitan French with minimal accommodation — the reverse is more difficult. Dialect-specific training builds a superset of the standard competency: everything the standard learner has, plus the phoneme inventory and register knowledge specific to the regional variety. This is not a narrowing of scope. It is a raising of the floor.