Aspect in the Past Tense: Heritage Speaker's Narrative Draft
Ms. Ferreira teaches Portuguese II at a suburban high school. In a 450-word personal narrative draft, Daniela, a second-generation student who speaks Portuguese at home, writes: "Todos os sábados, nós fomos à feira com minha mãe." "A casa foi grande e teve um quintal enorme." "De repente, começava a chover e nós corremos para o carro." In conference, Daniela reads the draft aloud fluently and self-corrects two agreement slips without prompting, but when Ms. Ferreira asks about the three sentences, Daniela shrugs: "Sounds fine to me." An oral recording of the same story, made a week earlier, uses morava, era, tinha, and começou correctly and naturally. The class is mixed: nine heritage speakers and fourteen second-language learners. Ms. Ferreira has one 20-minute conference slot per student before the draft is due, and department policy requires that major writing tasks be graded with a grammar-and-mechanics category worth 30% of the score.
1. Is Daniela's past-tense usage a random error pattern or a systematic one? Support your diagnosis with the three sentences and with the oral/written contrast, and state what the diagnosis implies for instruction.
Read the worked answer
The pattern is systematic, not random, and it is aspectual rather than morphological. In all three sentences Daniela uses the preterite where the imperfect is required by the sentence's function in the narrative:
- "Todos os sábados, nós fomos à feira" is habitual/repeated background → íamos.
- "A casa foi grande e teve um quintal enorme" is description of state, not an event → era grande e tinha um quintal enorme.
- "De repente, começava a chover" reverses the foreground/background logic: de repente marks the punctual event that interrupts → começou.
Two facts point to the same explanation. First, the errors cluster exactly where English has no obligatory choice: English simple past covers both perfective event and imperfective background, so Daniela's L1 supplies one default form and nothing in her sentence-level intuition flags it. Second, her oral version is correct, which rules out "she doesn't know the forms" and rules in "the knowledge is not yet available under writing load."
The instructional implication is decisive: do not reteach the full paradigm or drill conjugation. Teach the discourse function - foreground event versus background/description/habit - and make the oral knowledge the reference point she can consult while writing. The pattern also gives Ms. Ferreira a legitimate focus for the 30% grammar category, since one high-frequency system, not forty scattered errors, is at stake.
2. Design the 20-minute conference. What sequence of moves, and which traps should Ms. Ferreira avoid given the mixed class and the time limit?
Read the worked answer
A workable sequence:
- Warm read (3 min). Daniela reads her three sentences aloud. Almost always she will hear at least one problem; let her find it before any explanation.
- Oral retelling (4 min). Ask her to retell the fair scene and the rain scene as if telling a friend. Transcribe two of her correct oral sentences next to her written ones. The contrast is the lesson: "You already do this - here is what your mouth knows."
- Two-question test, not a rule list (5 min). (a) Did this happen once, or every Saturday? (b) Is this the event that moves the story forward, or the scenery behind it? Apply to the three sentences; she fixes them herself.
- Transfer hunt (5 min). She scans her draft for two more background sentences and marks them. Transfer, not compliance, is the goal.
- Exit ticket (3 min). She writes the two questions on an index card and chooses one paragraph to revise overnight.
Traps to avoid: correcting all forty errors in the draft (it converts a teachable pattern into noise); lecturing through the imperfect/preterite charts (she demonstrably possesses the forms); and using heavy English metalanguage, which also shuts out the second-language learners if this mini-lesson is later recycled for the whole class. For the mixed class, the same two questions work for everyone, with L2 learners getting additional form-focused practice.
3. How should Ms. Ferreira apply the 30% grammar-and-mechanics category to this draft without penalizing Daniela for a system she has not yet been taught, and without lowering the standard?
Read the worked answer
The category should be scored on demonstrated control of the targeted system plus overall accuracy, not on a raw error count. A defensible approach:
- Score the revised draft, not the first draft, for the grammar category, with the first draft counted as formative. The department's 30% weighting does not require that the first draft be the graded artifact.
- Use a small number of named criteria, e.g., "past-tense aspect in narration," "subject-verb and noun-adjective agreement," "accentuation of taught forms." Daniela's aspect errors then count against one criterion that she has just been taught and can revise, rather than sinking the whole category.
- Do not double-count. If the same aspect slip appears eight times, it is one pattern, not eight independent failures; error-count grading systematically punishes fluency and risk-taking.
- Keep dialect features out of the category entirely. Here the issue is aspect, which is a genuine standard-Portuguese requirement, not a regional variant; the distinction matters because Daniela's family speech may contain forms that are legitimate in speech and only proscribed in formal writing.
- Document the feedback so a parent or administrator can see that the grade reflects a taught criterion and a revision opportunity. This also protects Ms. Ferreira if the grade is appealed.