The Aspect Hypothesis, why preterite versus imperfect is harder than we admit, in a predictable way.

Roger Andersen, Yasuhiro Shirai, and Kathleen Bardovi-Harlig built one of the most replicable findings in second language acquisition. The order in which learners pick up preterite and imperfect is not random. It is structured by the kind of verb the marker lands on.

The Vendler categories first

Zeno Vendler (Verbs and Times, Philosophical Review, 1957) proposed a four-way classification of verb meanings that philosophers had been gesturing at for decades. The categories are defined by two binary features: telicity (does the action have an inherent endpoint) and duration (is it punctual or extended).

States

Atelic and durative. Saber, ser, tener, querer, creer. No endpoint built in. Extended in time. They simply hold.

Activities

Atelic and durative. Correr, caminar, leer, nadar. No endpoint built in. Extended. They unfold.

Accomplishments

Telic and durative. Escribir una carta, leer un libro, construir una casa. They have a built-in endpoint and they take time.

Achievements

Telic and punctual. Llegar, encontrar, romper, nacer, morir. Built-in endpoint. Instantaneous in their core meaning.

The Aspect Hypothesis in plain language

Roger Andersen formulated the hypothesis (Developmental Sequences: The Emergence of Aspect Marking in Second Language Acquisition, 1991; revised with Yasuhiro Shirai in 1996). Across many L2 learner populations and target languages, perfective markers (in Spanish: preterite) first attach to achievement and accomplishment verbs (telic verbs). Imperfective markers (in Spanish: imperfect) first attach to state and activity verbs (atelic verbs).

In the early stages, the verb's inherent lexical aspect determines the tense marker the learner picks. The learner is not really choosing between preterite and imperfect. They are reading the verb's semantic class and producing the marker that lines up with it.

The hypothesis predicts that learners produce llegó and se rompió early and easily, because llegar and romperse are telic punctuals. They produce sabía and era early and easily, because saber and ser are prototypical states. They struggle with the cases that violate the prototype: llegaba (achievement in imperfect, meaning “was arriving” or habitual arrival), supe (state in preterite, meaning “found out”).

Why this matters for teaching

Most Spanish curricula teach preterite versus imperfect as a binary contrast, usually with rules like “use preterite for completed actions, imperfect for ongoing or habitual.” The rule is not wrong. It is also not what the learner is doing in their head when they pick a form.

The Aspect Hypothesis predicts that the rule fails in exactly the places we see it fail. Students consistently produce sabía when they mean supe. They produce llegaba when they mean llegó. They are not breaking the binary rule. They are following their prototype mapping. The verbs they have heard most in each tense are the prototype examples, and the non-prototype combinations have to be learned separately.

Learners do not begin with a clean grammatical contrast. They begin with a prototype mapping between lexical aspect and grammatical aspect, and they extend it slowly.

The Bardovi-Harlig longitudinal evidence

Kathleen Bardovi-Harlig spent the nineteen nineties collecting longitudinal data from L2 learners of English and Spanish (Tense and Aspect in Second Language Acquisition: Form, Meaning, and Use, Language Learning Monograph Series, 2000). Her data set is the most comprehensive empirical defense of the Aspect Hypothesis in print. The pattern holds across languages, across proficiency levels, and across instructed and naturalistic learners.

The data also show what learners need to break the prototype. They need explicit, repeated exposure to the non-prototype combinations in meaningful contexts. Forms like supe, conoció, quiso, tuvo que, llegaba have to be encountered many times in stories, dialogues, and texts where their meaning is clear and necessary. Drills do not get you there. Reading and listening to native speakers using these forms in context does.

The Spanish-specific complications

Spanish has a wrinkle that English does not. Several verbs in Spanish change their core meaning depending on whether they appear in preterite or imperfect.

Supe means “I found out.” Sabía means “I knew.”

Conoció means “he met for the first time.” Conocía means “he was acquainted with.”

Quiso means “tried to / refused to.” Quería means “wanted to.”

Tuvo que means “had to and did.” Tenía que means “had to (no information on whether it happened).”

These verbs are state-class lexically. The prototype mapping says imperfect. The preterite forms break the prototype and carry a specialized meaning. Students need the meaning differences taught explicitly, not just the endings.

How to teach preterite and imperfect aligned to the research

Lead with the prototypes

First exposure to preterite should be heavy with achievement verbs. Llegué, fui, encontré, nací, se rompió, se cayó, pasó. The prototype gives the student a strong initial form-meaning anchor.

First exposure to imperfect should be heavy with states and activities. Era, tenía, sabía, vivía, quería, trabajaba, jugaba. Same logic.

Use the prototypes inside a single narrative

A story that contains imperfect background and preterite foreground is the natural environment for this. Vivía en una ciudad pequeña con mi abuela. Era pobre, pero era feliz. Un día, mi abuela se cayó. Llegamos al hospital en quince minutos. Imperfect for the setting, preterite for the events. The prototype mapping is doing the work.

Teach the special verbs separately and explicitly

Pull the seven or eight state verbs that flip meaning in preterite out as their own unit. Teach the meaning difference, not just the form. Pair examples in context. Show students stories where supe and sabía both appear, with the difference visible. This is where explicit teaching beats input flooding.

What this means for your Monday

Pull out your last unit on preterite and imperfect. Look at the verb lists you used to introduce each tense. If the preterite verbs were mixed across all four Vendler categories and the imperfect verbs were too, you ran against the Aspect Hypothesis. Students will have absorbed the markers as random.

For the next time you teach this contrast, restructure. Lead with the prototypes. Build the contrast inside stories where the lexical aspect is doing the meaning work. Pull the meaning-flipping verbs into a separate explicit lesson once the prototypes are solid. The order is doing what your students' brains are already trying to do anyway.