Learning strategies are the specific actions, behaviours, steps, or techniques learners use to enhance their own language learning (Oxford, 1990). They represent the learner's active contribution to the learning process: conscious or semi-conscious choices about how to approach, process, store, retrieve, and use L2 knowledge.
Two influential classification systems emerged in the same year:
Oxford divided strategies into two major classes with six subcategories:
| Class | Strategy type | Description | Examples |
|---|---|---|---|
| Direct | Memory | Storing and retrieving information | Grouping, imagery, keyword method, reviewing |
| Cognitive | Manipulating the language directly | Practising, analysing, note-taking, summarising | |
| Compensation | Overcoming gaps in knowledge | Guessing from context, using synonyms, gestures | |
| Indirect | Metacognitive | Managing the learning process | Planning, monitoring, evaluating, self-assessment |
| Affective | Managing emotions | Anxiety reduction, self-encouragement, emotional temperature | |
| Social | Learning through interaction | Asking questions, cooperating, empathising |
The SILL questionnaire (50 items) became the most widely used instrument in strategy research, with translations into dozens of languages.
O'Malley and Chamot drew on cognitive psychology (Anderson's ACT theory) and proposed a simpler three-part taxonomy:
| Category | Description | Examples |
|---|---|---|
| Metacognitive Strategies | Executive processes: planning, monitoring, evaluating | Advance organising, selective attention, self-monitoring |
| Cognitive Strategies | Operating directly on incoming information | Repetition, note-taking, deduction, imagery, transfer |
| Socio-affective | Interaction and affect regulation | Cooperation, questioning for clarification, self-talk |
O'Malley and Chamot embedded their taxonomy within information-processing theory, arguing that strategies are complex cognitive skills that move from declarative knowledge (knowing what the strategy is) to procedural knowledge (using it automatically).
The learning strategies research tradition originated in studies of the good language learner (Rubin, 1975; Stern, 1975; Naiman et al., 1978). Key findings:
These findings motivated the development of learner training programs designed to teach less effective learners the strategies used by successful ones.
From the mid-2000s, the learning strategies construct came under critique:
Dörnyei (2005) argued that strategy research suffered from definitional fuzziness, overlapping categories, and reliance on self-report questionnaires. He proposed replacing "learning strategies" with self-regulation, a broader construct from educational psychology (Zimmerman, 2000) that encompasses goal-setting, strategy use, monitoring, and motivational regulation.
Macaro (2006) partially agreed, defining strategies as "conscious mental activity" that must contain a goal and a learning situation, tightening the definition and excluding automatic processes.
Oxford (2011) responded by updating her framework as Strategic Self-Regulation (S²R), integrating strategy research with self-regulation theory while retaining the practical taxonomy that teachers find useful.
Cohen & Macaro (2007) argued that regardless of theoretical framing, the practical question remains: what do effective learners do, and can it be taught?
Green & Oxford (1995): Surveyed 374 university EFL learners in Puerto Rico and found that higher proficiency correlated with more frequent use of all strategy categories, with the largest differences in cognitive and metacognitive strategies.
Chamot (2004): Reviewed two decades of strategy instruction research and concluded that explicit strategy instruction generally improves learning outcomes, especially when embedded in regular classroom activities rather than taught in isolation.
Plonsky (2011): Meta-analysis of 61 strategy instruction studies: overall effect size d = 0.49 (medium). Instruction was most effective for younger and lower-proficiency learners, and when strategies were taught explicitly with practice and feedback.
Griffiths (2013): Proposed that the most important variable is not which strategies learners use but how orchestrated their strategy use is, combining multiple strategies in task-appropriate ways.