The concept explainer
Why Does Your English Get Worse When You Have More to Think About?
The English you can use in an easy conversation may become harder to access when the task demands more reasoning, planning, and monitoring. Here is why.
An idea, brought into focus.
07 sources, open to exploreIn this article 15 sections
Your English can seem to get worse when a speaking task becomes mentally harder because speaking and thinking draw on limited attention at the same time.
You are not only choosing English words. You may also be comparing options, remembering instructions, organizing an argument, checking grammar, watching the listener, and planning what comes next.
When those demands rise together, performance can change even if your underlying English knowledge has not changed.
This is the useful distinction:
knowledge is what is available in your language system; performance is what you can coordinate under the demands of this task, right now.
The same English can produce very different speech
Answer this:
What did you do last weekend?
Now answer this:
Your city can fund either cheaper public transport or more affordable housing, but not both this year. Which should it choose? Give two reasons, address one objection, and explain what evidence could change your mind.
The second prompt does not merely require “better English.”
Before you speak, you have more conceptual work to do. You need a position. You need reasons. You need to keep the instructions active. You may need to compare consequences while you are already talking.
Then English has to carry all of that.
A learner who speaks smoothly about Saturday can therefore hesitate during the policy question without suddenly losing vocabulary or grammar.
The language task changed because the thinking task changed.
Watch what happens to one ordinary sentence
Suppose you want to say:
I would choose public transport because it helps more people immediately.
On an easy topic, the idea may already be clear. Most of your attention can go to putting it into English.
On a harder task, the sentence may be competing with several unfinished thoughts:
Is transport actually the better choice?
What about low-income renters?
I need two reasons.
Did the prompt ask for an objection?
Should I say would choose or would have chosen?
What was my second point?
The grammar did not become harder.
The word transport did not become rarer.
But the sentence is now being produced inside a busier system.
Speaking is a coordination problem
Models of speech production separate several operations that have to work together.
You form a message. You select words. You build grammatical and phonological plans. You articulate them. You also monitor what you are saying and adjust when something goes wrong.
In a second language, some of these operations can demand more conscious attention because they are less automatic. Kormos's model of bilingual speech production describes how L2 production can place distinctive demands on lexical access, formulation, monitoring, and control. Kormos, 2006
That gives us a better model than:
difficult question → bad English
Think instead:
conceptual planning + language formulation + monitoring + time pressure → one shared performance problem
As language processes become more automatic, they can become less expensive to run. But no speaker has unlimited processing capacity.
Task-complexity research shows trade-offs, not a simple collapse
Second-language researchers often study complexity, accuracy, and fluency as different dimensions of performance.
A harder task does not necessarily push all three in the same direction.
Peter Skehan's limited-attentional-capacity account predicts that when tasks place greater demands on learners, attention may be distributed unevenly across these dimensions. A speaker can prioritize meaning and flow, accuracy, or more complex language, but maximizing everything at once is difficult. Skehan, 2009
Robinson's Cognition Hypothesis offers an important counterpoint. Increasing some kinds of cognitive complexity can direct attention toward language that is needed to express more complex meanings, potentially encouraging greater linguistic complexity or accuracy rather than simply making performance worse. Robinson, 2001
That disagreement matters.
The evidence does not justify saying:
more thinking always makes your English worse.
A harder task can change which parts of performance receive resources.
“Worse English” can hide several different changes
| What you notice | What may be happening | What it does not prove |
|---|---|---|
| More pauses | More online planning or lexical search | You forgot the language |
| Simpler sentences | Attention is protecting meaning or flow | Your grammar knowledge disappeared |
| More grammar errors | Monitoring/formulation is under pressure | You never learned the rule |
| Slower answers | Conceptual planning takes longer | Your speaking skill suddenly declined |
| More self-correction | Monitoring is active during production | Every repair is a fluency failure |
| More elaborate language but slower delivery | Resources shifted toward complexity | The task made all dimensions worse |
These are diagnostic possibilities, not one-to-one rules. The same visible hesitation can have different causes.
Harder tasks can sometimes produce more complex language
This is the part learners often miss.
If a task forces you to explain causes, compare alternatives, justify a decision, or describe relationships, it may create a reason to use more elaborate language.
Research has repeatedly tested this idea by manipulating task complexity and measuring changes in complexity, accuracy, and fluency. Results vary, but several studies have found that cognitively more demanding tasks can increase aspects of linguistic complexity while imposing costs elsewhere.
For example, Gilabert studied narrative tasks manipulated along reasoning demands and found effects consistent with the idea that increased cognitive demands can influence complexity, accuracy, and fluency differently rather than causing a uniform decline. Gilabert, 2007
A later meta-analysis of task complexity research also found that effects differed across performance dimensions and task manipulations. The field does not support one universal “hard task penalty.” Jackson & Suethanapornkul, 2013
So when your speech slows during a difficult answer, ask what else changed.
You may be trying to express a more demanding idea.
Myth: if you really knew English, thinking harder would not affect it
Myth: Strong English should remain equally fast and accurate no matter how difficult the question becomes.
Reality: Language production happens while other cognitive work is happening. Even skilled speakers take longer when they must plan unfamiliar content, hold constraints in mind, or choose between competing ideas.
L2 learners can feel this more sharply because language formulation itself may still consume substantial attention.
That does not make every error a harmless “cognitive load” effect. A missing word can still be a real vocabulary gap. A recurring tense error can still reflect unstable knowledge.
The useful point is narrower:
one difficult performance is not a clean measurement of everything you know.
Planning time changes the budget
Give yourself 30 seconds before answering a difficult question and something important happens.
You can do part of the conceptual work before speech begins.
You can choose your position. Build a rough structure. Retrieve useful vocabulary. Pre-plan a difficult phrase.
That leaves fewer decisions to solve while your mouth is already moving.
Research on task planning has found that pre-task planning can affect L2 oral performance, often benefiting fluency and sometimes complexity, although effects on accuracy are less consistent and depend on how planning and tasks are designed. A meta-analysis by Johnson and Abdi Tabari found an overall positive effect of planning on oral task performance while also showing variation across performance measures and study conditions. Johnson & Abdi Tabari, 2022
Planning does not give you new English in 30 seconds.
It changes when you pay some of the processing cost.
That is why prepared and unprepared speaking can feel like two different versions of you.
Do not use “cognitive load” as an explanation for every mistake
The phrase is tempting because it can explain almost anything after the fact.
You forgot a word? Cognitive load.
Wrong tense? Cognitive load.
Bad pronunciation? Cognitive load.
That is not useful diagnosis.
To decide whether task demand is part of the problem, compare conditions.
Can you answer the same question much better after 30 seconds of planning?
Can you explain the idea clearly in your first language but struggle to formulate it in English?
Does your English improve when the reasoning stays the same but the response becomes shorter?
Does the same grammar error appear even in an easy, prepared answer?
Patterns matter more than one bad attempt.
If the problem appears mainly when another person is evaluating you, social pressure may be the stronger mechanism. That separate problem is explored in Is Your English Worse When Someone Is Listening?.
Automatic language leaves more room for the idea
Imagine that producing this phrase requires almost no deliberate construction:
On the other hand...
You can use your attention to decide what the other hand actually is.
Now imagine that you must consciously retrieve each part:
Is it in the other hand? No. On. What comes next?
A small language problem has consumed attention that could have gone to the argument.
This is one reason automaticity matters for spoken fluency. Familiar words, constructions, and useful multiword sequences can become easier to access, reducing the amount of conscious control needed for routine formulation.
The broader relationship between retrieval, pausing, and fluent speech is covered in What Does It Mean to Be Fluent in English?.
The goal is not to speak without thinking.
It is to spend less thinking power on language operations that no longer need to be expensive.
Test whether the bottleneck is English or the task
- Answer an easy version. Speak for 45 seconds about a familiar topic with no special constraints.
- Add reasoning. Compare two options or defend a position on the same topic.
- Repeat with planning. Take 30 seconds to outline the difficult answer, then try again.
- Compare the recordings. Listen for pauses, abandoned sentences, grammar errors, lexical searches, and idea organization.
- Change one demand at a time. Keep the topic but remove time pressure, shorten the answer, or remove one instruction.
- Find what recovers. If planning restores much of the performance, online coordination is probably part of the bottleneck. If the same language gaps remain everywhere, they deserve direct language practice.
This is a diagnostic practice routine derived from the research logic. It is not a standardized cognitive-load test.
Practice should sometimes make the thinking harder — on purpose
If you only practice English with easy questions, you can become good at easy questions.
Real conversations ask more.
You explain why. Compare options. Remember what someone said two minutes ago. Change your mind halfway through. Listen while preparing your response.
So practice should eventually include those demands.
But adding everything at once is a poor way to learn.
A useful progression is:
- first make the language for a topic accessible;
- then add one reasoning demand;
- then reduce preparation;
- then add interaction or another constraint.
This lets you train coordination without confusing overload with productive difficulty.
Your English did not necessarily disappear. The task got more expensive.
A hard question can make familiar English slower, simpler, or less accurate because the language system is sharing attention with more conceptual work.
But the effect is not a universal collapse.
More demanding tasks can also push speakers toward more complex language. Planning can move some of the work out of real-time production. Automatic language can leave more capacity for the idea itself.
So after a difficult conversation, do not ask only:
Why was my English so bad?
Ask:
What else was I trying to do while I was speaking?
That question separates a language gap from a coordination problem.
And once you know which one failed, you know much more about what to practice.
Sources and further reading
- Skehan (2009) — Modelling second language performance: Integrating complexity, accuracy, fluency, and lexis
- Robinson (2001) — Task complexity, task difficulty, and task production
- Jackson & Suethanapornkul (2013) — The cognition hypothesis: A synthesis and meta-analysis of research on second language task complexity
- Gilabert (2007) — The simultaneous manipulation of task complexity along planning time and reasoning demands
- Kormos (2006) — Speech Production and Second Language Acquisition
- Johnson & Abdi Tabari (2022) — The effects of planning on L2 oral task performance: A meta-analysis
- Housen & Kuiken (2009) — Complexity, Accuracy, and Fluency in Second Language Acquisition