Some Highlights from The Evidence is In: Drills are Out

Originating from the Audiolingual Method, drills have become a signature activity in many language classes. In the article titled "The Evidence is In: Drills are Out," Wong and VanPatten (2003) argue that drills are neither necessary or helpful for building an internal linguistic system, which is foundational to the abilities to use the target language. They highlight the essential role of input in building this internalized system while discussing output and interaction as playing facilitative roles. They acknowledge that developing accuracy and fluency is a separate matter built upon the internal system.

Some highlights below clarify their argument.

1. Mechanical Drills

Mechanical drills are defined by complete control of the response, where there is only one correct way to respond and students do not need to understand the meaning to be successful.

• Singular-to-Plural Transformation: A teacher reads a singular sentence, and the student provides the plural form.

  • Teacher: "He's a lawyer." → Student: "They're lawyers".

  • Teacher: "The student is tall." → Student: "The students are tall".

• Verb Conjugation (Spanish): A teacher provides an infinitive and a subject pronoun, and the student provides the conjugated form.

  • Cue: "andar (tú)" → Response: "andas".

  • Cue: "trabajar (tú)" → Response: "trabajas".

• "Slash" or Dehydrated Sentences: Students are given a string of words and must combine them into a grammatically correct sentence.

  • Cue: "Claude / nettoyer / la chambre / à Richard" → Response: "Claude fait nettoyer la chambre à Richard".

2. Meaningful Drills

These remain highly controlled with one right answer, but the learner must comprehend the stimulus to answer correctly.

• Object Pronoun Practice: A teacher asks a question about a visible object.

  • Teacher: "Where does John put his books in class?" (while looking at the books under the seat) .

  • Response: "He puts them under his seat".

3. Communicative Drills

In these drills, learners must provide information that was not previously known, meaning there is no single "right" answer as long as the response is grammatically correct.

• Information Gap/Opinion: Using the same object pronoun focus as above, but asking for a personal opinion or prediction.

  • Teacher: "Where do you think John puts his books when he gets home?".

  • Possible Responses: "He puts them on the kitchen counter" or "He leaves them on a chair".

4. Contextualized Drills

The authors argue that adding a "story" to a drill does not change its mechanical nature if it can still be completed using nonsense words.

• Apartment Scenario: Students use prompts to create a conversation about a new apartment.

  • Cue: "I / to want / the apartment / to be / near / university".

  • Response: "I want the apartment to be near the university".

  • Critique: The authors note that a student could replace "apartment" with a nonsense word like "doiter" and still complete the drill successfully without understanding it.

2. The Essential Role of Input

Input is characterized as the "single most important concept" in SLA. It is defined strictly as meaning-based language that a learner hears or sees within a communicative context.

• When a learner encounters input, their primary objective is to comprehend the message being conveyed.

• The internal linguistic system (the "mental grammar") is constructed through exposure to this communicative input, not through the study of rules or the use of a rulebook.

• To count as input, the language must involve a speaker (or writer for that matter) attempting to express a real meaning to the learner (e.g., a teacher giving directions in the target language. “Open your books to page 24. Is everyone on page 24?”).

3. The Role of Output and Interaction

While input is the primary builder of the internal system, output (producing the language) and interaction serve three specific secondary functions that support acquisition according to Swain (1998):

• Noticing: Attempting to produce language makes learners aware of what they can say or not say and pushes them to search for language forms to express their meaning.

• Hypothesis Testing: Learners use output to "try out" ways to express meaning. They then confirm or disconfirm if they are correct based on the feedback or input they receive in response.

• Heightened awareness: Producing output promotes conscious awareness of language and language use, which may promote processes responsible for acquisition.

4. Fluency and Accuracy

The author argue that development accuracy and fluency in language production is a separate matter from the development of the internal linguistic system. Much like playing chess makes one a better chess player, producing language in communicative contexts helps develop the skill of using the system quickly and correctly. However, the authors clarify that using the system is not the same as building it.

4. The Argument Against Drills

Wong and VanPatten argue that traditional drills are not only unnecessary but can occasionally be detrimental for the following reasons:

• Lack of Meaning: By definition, drills are mechanical and not meaning-based. Because they lack a communicative purpose, they do not qualify as the type of input necessary to build a mental grammar.

• Task-Specific Success: Research shows that students who excel at drills often develop "test-taking strategies" rather than actual linguistic competence. They learn how to perform the drill, not how to use the language.

• Opportunity Cost: Time spent on intensive pattern practice is often "lost time" that could have been devoted to meaning-based activities that actually promote acquisition.

The Authors' Stance on Instruction

It is important to note that the authors are not against all forms of instruction. They also support a Focus on Form. They advocate for Processing Instruction (PI), an interventionist approach that is informed by how the brain processes input, rather than the "output-first" approach found in traditional drilling.

NOTE from the author of this blog post: We are sharing a recent podcast interview with Professor Rod Ellis, who also discusses implicit learning as the main mechanism for SLA and the importance of input-based tasks in Task-based Language Teaching especially for beginners. However, he also points out the possible interface between implicit and explicit knowledge.