The Trade-Off Hypothesis is Peter Skehan's account of how second-language learners allocate limited attentional resources during task performance, and what this allocation predicts about the shape of their output. Developed in A Cognitive Approach to Language Learning (1998) and refined across the next two decades, the hypothesis frames task performance as a competition among three measurable dimensions of language use: complexity, accuracy, and fluency. When one rises, another typically falls.
The hypothesis rests on a triad of performance measures that have become the standard outcome variables in task-based SLA research:
| Dimension | What it captures |
|---|---|
| Complexity | Range and sophistication of syntactic structures and lexis used |
| Accuracy | Correctness of forms produced relative to target norms |
| Fluency | Speed, smoothness, and lack of dysfluency in production |
Skehan's claim is that learners' attentional resources are bounded and that pushing more attention onto one dimension tends to leave less for the others. The cognitive architecture cannot do everything at once.
Attention and working memory are treated as a single bottleneck, in line with Skehan's broader Limited Attentional Capacity (LAC) model. Two coping strategies follow:
Task design influences which strategy dominates. Pre-task planning, task structure, familiarity, and time pressure each tip the balance.
The Trade-Off Hypothesis is the principal counterweight to Robinson's Cognition Hypothesis, and the two positions structure most cognitive research in task-based language teaching. The disagreement is sharp.
| Question | Skehan (Trade-Off) | Robinson (Cognition Hypothesis) |
|---|---|---|
| What does increased task complexity do? | Forces trade-offs across CAF | Pushes accuracy and complexity up together along resource-directing dimensions |
| Is attention pooled or specialised? | Single limited pool | Multiple resources that can be recruited in parallel |
| What sequencing principle follows? | Manage cognitive load to avoid overload | Sequence simple to complex along resource-directing dimensions |
| What happens under high load? | Fluency or accuracy collapses | Attention is mobilised toward new functional demands |
Both positions have generated substantial empirical work. The picture is genuinely mixed: trade-offs appear in many studies, but so do simultaneous gains under particular complexity manipulations. Differences in target language, proficiency level, task type, and outcome measure all moderate the result.
If the trade-off picture is correct, classroom decisions involve managing rather than eliminating competing demands.
The general advice is to know which dimension a task is targeting and not expect simultaneous gains across all three.
Reviews of task-based research find consistent evidence that planning supports complexity and accuracy at some cost to fluency, that task repetition benefits later performances, and that complexity manipulations produce interpretable but mixed CAF profiles. The hypothesis predicts the direction but not always the magnitude of trade-offs, and a substantial number of studies report results that are consistent with neither Skehan nor Robinson in their pure forms.